IS-95 CDMA and cdma2000

Cellular/PCS Systems Implementation

Prentice Hall Communications Engineering and Emerging Technologies Series

Theodore S. Rappaport, Series Editor

GARG IS-95 CDMA and cdma2000: Cellular/PCS Systems Implementation

GARG & WILKES Principles and Applications of GSM

KIM Handbook of CDMA System Design, Engineering and Optimization

LIBERTI & RAPPAPORT Smart Antennas for Wireless Communications: IS-95 and Third
Generation CDMA Applications

RAPPAPORT Wireless Communications: Principles and Practice

RAZAVI RF Microelectronics

STARR, CIOFFI & SILVERMAN Understanding Digital Subscriber Line Technology

FORTHCOMING

CIMINI & LI Orthogonal Frequency Division Multiplexing for Wireless Communication

POOR & WANG Wireless Communication Systems: Advanced Techniques for Signal Reception

REED & WOERNER Software Radio: A Modern Approach to Radio Engineering

TRANTER, KOSBAR, RAPPAPORT & SHANMUGAN Simulation of Modern Communications

Systems with Wireless Applications
IS-95 CDMA and cdma2000
Cellular/PCS Systems Implementation

Vijay K. Garg, PhD, PE

Department of Electrical Engineering and
Computer Science
University of Illinois at Chicago
Chicago, Illinois U.S.A.

Prentice Hall PTR

Upper Saddle River, NJ 07458
www.phptr.com
Library of Congress Cataloging-in-Publication Data
Garg, Vijay Kumar, 1938–
IS-95 CDMA and cdma2000 : cellular/PCS systems
implementation / Vijay K. Garg.
p. cm. — (Prentice Hall communications engineering and
emerging technologies series)
Includes bibliographical references and index.
ISBN 0-13-087112-5 (case)
1. Code division multiple access. 2. Cellular telephone
systems. 3. Personal communication service systems. I. Title.
II Series.

TK5103.452 G37 1999

621.3845—dc21 99-049860
CIP

Editorial/production supervision: BooksCraft, Inc., Indianapolis, IN

Acquisitions editor: Bernard Goodwin
Editorial assistant: Diane Spina
Marketing manager: Lisa Konzelmann
Manufacturing manager: Alexis Heydt
Cover design director: Jerry Votta
Cover designer: Talar Agasyan
Project coordinator: Anne Trowbridge

© 2000 by Prentice Hall PTR

Prentice-Hall, Inc.
Upper Saddle River, NJ 07458

Prentice Hall books are widely used by corporations and government agencies for
training, marketing, and resale.

The publisher offers discounts on this book when ordered in bulk quantities.
For more information, contact:
Corporate Sales Department
Prentice Hall PTR
One Lake Street
Upper Saddle River, NJ 07458
Phone: 800-382-3419 Fax: 201-236-7141
E-mail: [email protected].

All rights reserved. No part of this book may be reproduced, in any form
or by any means, without permission in writing from the publisher.

All product names mentioned herein are the trademarks of their respective owners.

Printed in the United States of America

10 9 8 7 6 5 4 3 2 1

ISBN: 0-13-087112-5

Prentice-Hall International (UK) Limited, London

Prentice-Hall of Australia Pty. Limited, Sydney
Prentice-Hall Canada Inc., Toronto
Prentice-Hall Hispanoamericana, S.A., Mexico
Prentice-Hall of India Private Limited, New Delhi
Prentice-Hall of Japan, Inc., Tokyo
Pearson Education Asia Pte. Ltd., Singapore
Editora Prentice-Hall do Brasil, Ltda., Rio de Janeiro
 I dedicate this book to my lovely grandchildren:
Adam Dorr, Renu Dorr, Monica Taneja, and Nevin Taneja.
This page intentionally left blank
Contents

Preface xvii

1 Introduction to Access Technologies 1

1.1 Introduction 1
1.2 Major Attributes of CDMA Systems 2
1.3 Third-Generation Systems 3
1.4 Multiple Access Technologies 5
1.5 Modes of Operation in Wireless Communications 7
1.6 Summary 8
1.7 References 8

2 Direct Sequence Spread Spectrum and Spreading Codes 9

2.1 Introduction 9
2.2 Types of Techniques Used for Spread Spectrum 9
2.3 The Concept of Spread Spectrum System 11
2.3.1 System Processing Gain 12
2.4 The Performance of DSSS 14
2.4.1 The DSSS System 14
2.4.2 Coherent Binary Phase-Shift Keying 14
2.4.3 Quadrature Phase-Shift Keying 15

vii
viii Contents

2.5 Bit Scrambling 18

2.6 The Performance of a CDMA System 20
2.7 Pseudorandom Noise Sequences 24
2.7.1 Properties of a Maximal-Length Pseudorandom
Sequence 27
2.7.2 Autocorrelation 29
2.7.3 Cross-Correlation 30
2.7.4 Orthogonal Functions 31
2.8 Summary 37
2.9 Problems 38
2.10 References 39

3 Speech and Channel Coding 41

3.1 Introduction 41
3.2 Speech Coding 41
3.2.1 Pulse Code Modulation 41
3.2.2 Adaptive Pulse Code Modulation 43
3.2.3 Code-Excited Linear Prediction 45
3.2.4 Enhanced Variable-Rate Codec 49
3.3 Channel Coding 50
3.3.1 Convolutional Code 52
3.4 Summary 54
3.5 Problems 54
3.6 References 55

4 Diversity, Combining, and Antennas 57

4.1 Introduction 57
4.2 Diversity Reception 57
4.3 Types of Diversity 58
4.3.1 Macroscopic Diversity 58
4.3.2 Microscopic Diversity 59
4.3.3 RAKE Receiver 60
4.4 Basic Combining Methods 63
4.4.1 Selection Combiner 63
4.4.2 Maximal-Ratio Combiner 65
4.4.3 Equal-Gain Combining 67
Contents ix

4.5 BPSK Modulation and Diversity 71

4.6 Examples of Base Station and Mobile Antennas 72
4.6.1 Quarter-Wave Vertical Antenna 73
4.6.2 Stacked Dipoles 75
4.6.3 Corner Reflectors 76
4.6.4 Smart Antenna 76
4.7 Summary 78
4.8 References 78

5 IS-95 System Architecture 79

5.1 Introduction 79
5.2 TR-45/TR-46 Reference Model 79
5.3 Functional Model Based on Reference Model 83
5.4 Wireless Intelligent Network 85
5.5 Summary 89
5.6 References 89

6 IS-95 CDMA Air Interface 91

6.1 Introduction 91
6.2 TIA IS-95 CDMA System 91
6.2.1 Forward Link 93
6.2.2 Reverse Link 96
6.3 Summary 100
6.4 References 101

7 Physical and Logical Channels of IS-95 CDMA 103

7.1 Introduction 103

7.2 Physical Channels 103
7.3 Modulation 109
7.4 Bit Repetition 109
7.5 Block Interleaving 110
7.6 Channel Coding 111
7.7 Logical Channels 111
7.7.1 Pilot Channel 112
7.7.2 Sync Channel 113
x Contents

7.7.3 Paging Channel 115

7.7.4 Access Channel 118
7.7.5 Forward Traffic Channels 122
7.7.6 Reverse Traffic Channels 128
7.8 Summary 132
7.9 References 132

8 IS-95 CDMA Call Processing 133

8.1 Introduction 133

8.2 CDMA Call Processing State 133
8.2.1 System Initialization State 135
8.2.2 Idle State 137
8.2.3 System Access State 142
8.2.4 Mobile Control on Traffic Channel State 143
8.3 CDMA Registration 145
8.4 Authentication 149
8.5 Summary 151
8.6 References 151

9 Signaling Applications in IS-95 CDMA 153

9.1 Introduction 153

9.2 Layered Structure 153
9.3 A-Interface 156
9.3.1 Supported Architectural Configurations 158
9.3.2 Call Processing and Supplementary Services 160
9.3.3 Radio Resource Management 162
9.3.4 Mobility Management 162
9.3.5 Transmission Facilities Management 162
9.3.6 Use of the SCCP 164
9.4 Roaming 167
9.4.1 Call Waiting 169
9.4.2 Handoffs 171
9.4.3 Over-the-Air Service Provisioning (OTASP) 176
9.5 Summary 178
9.6 References 179
Contents xi

10 Soft Handoff and Power Control in IS-95 CDMA 181

10.1 Introduction 181

10.2 Types of Handoff 182
10.2.1 Soft Handoff (Forward Link) 184
10.2.2 Soft Handoff (Reverse Link) 184
10.2.3 Softer Handoff (Reverse Link) 184
10.2.4 Benefit of Soft Handoff 184
10.3 Pilot Sets 184
10.4 Search Windows 185
10.4.1 SRCH_WIN_A 186
10.4.2 SRCH_WIN_N 186
10.4.3 SRCH_WIN_R 187
10.5 Handoff Parameters 187
10.5.1 Pilot Detection Threshold (T_ADD) 187
10.5.2 Comparison Threshold (T_COMP) 187
10.5.3 Pilot Drop Threshold (T_DROP) 188
10.5.4 Drop Timer Threshold (T_TDROP) 188
10.6 Handoff Messages 188
10.7 Handoff Procedures 190
10.7.1 Mobile-Assisted Soft-Handoff (MASHO) Procedures 190
10.7.2 Dynamic Soft-Handoff Thresholds 191
10.8 Setup and End of Soft Handoff 192
10.8.1 Setup 192
10.8.2 End of Soft Handoff 193
10.9 Maintenance of Pilot Sets 195
10.9.1 Active Set Maintenance 195
10.9.2 Candidate Set Maintenance 195
10.9.3 Neighbor Set Maintenance 195
10.10 The Need for Power Control 196
10.11 Reverse Link Power Control 198
10.11.1 Reverse Link Open-Loop Power Control 198
10.11.2 Reverse Link Closed-Loop Power Control 202
10.12 Forward Link Power Control 206
10.13 Summary 210
10.14 References 211
xii Contents

11 Security and Identification in IS-95 CDMA 213

11.1 Introduction 213

11.2 Mobile Identification Parameters 213
11.2.1 Mobile Station Identification Number 213
11.2.2 Mobile Directory Number 214
11.2.3 Electronic Serial Number (ESN) 214
11.2.4 Station Class Mark 215
11.2.5 Registration Memory 215
11.2.6 Access Overload Class 216
11.2.7 Home System and Network Identification 216
11.2.8 Local Control Option 216
11.2.9 Preferred Operation Selection 217
11.2.10 Discontinuous Reception 217
11.3 Authentication Procedures 217
11.3.1 Authentication of MS Registrations 219
11.3.2 Unique Challenge Response Procedure 219
11.3.3 Authentication of MS Originations 220
11.3.4 Authentication of MS Terminations 221
11.3.5 Authentication of MS Data Bursts 222
11.3.6 Authentication of TMSI Assignment 222
11.4 Shared Secret Data 223
11.5 Parameter Update 225
11.6 Voice Privacy 225
11.7 Summary 227
11.8 References 228

12 RF Engineering and Network Planning 229

12.1 Introduction 229

12.2 Radio Design for a Cellular/PCS Network 229
12.3 Radio Network Planning 229
12.4 Radio Link Design 230
12.5 Estimation of Cell Count 230
12.6 Radio Coverage Planning 231
12.7 Propagation Models 232
12.7.1 Modeling for the Outside Environment 233
Contents xiii

12.7.2 Models for Indoor Environment 238

12.7.3 IMT-2000 Models 242
12.8 Delay Spread 243
12.8.1 Coherence Bandwidth 245
12.9 Doppler Spread 247
12.10 Intersymbol Interference 249
12.11 Link Budget and Cell Coverage 249
12.12 Dual-Mode CDMA Mobiles 253
12.13 The Transition from an Analog System to a Digital
System 255
12.13.1 Overlay Design 255
12.13.2 Integrated Design 256
12.13.3 Partial CDMA Coverage, Integrated System 257
12.14 Facilities Engineering 268
12.15 Design Considerations at the Boundary of a CDMA
System 280
12.16 Interfrequency Handoff 280
12.17 Summary 281
12.18 References 281

13 Reverse and Forward Link Capacity of IS-95 CDMA System 283

13.1 Introduction 283

13.2 Reverse Link Capacity 284
13.3 Multicell Network 290
13.4 Intercell Interference 291
13.5 Erlang Capacity of a Single Cell 292
13.6 Forward Link Capacity 294
13.6.1 Pilot Channel 295
13.6.2 Traffic Channel 296
13.7 CDMA Cell Size 300
13.8 Forward and Reverse Link Balance 302
13.9 Forward Link Budget 305
13.10 Summary 311
13.11 References 311
13.12 Problems 312
xiv Contents

14 Wireless Data 313

14.1 Introduction 313

14.2 Data Communication Services 314
14.3 OSI Upper Layers 315
14.4 Wireless Data Systems 316
14.4.1 Wide-Area Wireless Data Systems 316
14.4.2 High-Speed Wireless Local Area Networks 317
14.5 WLAN Standards 321
14.5.1 IEEE 802.11 321
14.5.2 Wireless Information Networks Forum 324
14.5.3 High-Performance Radio Local Area Network 324
14.5.4 U.S. Advanced Research Project Agency 325
14.6 Access Methods 325
14.6.1 Fixed-Assignment Access Methods 325
14.6.2 Random-Access Methods 326
14.7 Error Control Schemes 330
14.8 Data Services in IS-95 333
14.9 Asynchronous Data and Group-3 Facsimile 334
14.10 Short Message Service 338
14.11 Packet Data Services for CDMA Cellular/PCS
Systems 342
14.11.1 Network Reference Model 342
14.11.2 Network Element 342
14.11.3 Network Reference Points 343
14.11.4 Protocol Options 343
14.11.5 Applicable Mobile Type 345
14.11.6 Packet Data Protocol States 345
14.11.7 Packet Mode Data Service Features 348
14.12 Summary 351
14.13 References 352

15 cdma2000 System 353

15.1 Introduction 353

15.2 cdma2000 Layering Structure 356
15.2.1 Upper Layer 356
Contents xv

15.2.2 Link Layer 357

15.3 cdma2000 Channels 362
15.3.1 Channel-Naming Convention 362
15.4 Logical Channels Used by PLICF 362
15.4.1 Dedicated Traffic Channel (f/r-dtch) 362
15.4.2 Common Traffic Channel (f/r-ctch) 363
15.4.3 Dedicated MAC Channel (f/r-dmch_control) 363
15.4.4 Reverse Common MAC Channel (r-cmch_control) 363
15.4.5 Forward Common MAC Channel (f-cmch_control) 364
15.4.6 Dedicated Signaling Channel (dsch) 364
15.4.7 Common Signaling Channel (csch) 364
15.5 Physical Layer 364
15.6 Forward Link Physical Channels 364
15.6.1 Forward Pilot Channel (F-PICH) 365
15.6.2 Forward Sync Channel (F-SYNC) 366
15.6.3 Forward Paging Channel (F-PCH) 366
15.6.4 Forward Common Control Channel (F-CCCH) 367
15.6.5 Forward Common Auxiliary Pilot Channel
(F-CAPICH) 367
15.6.6 Forward Broadcast Common Channel (F-BCCH) 368
15.6.7 Forward Quick Paging Channel (F-QPCH) 368
15.6.8 Forward Dedicated Auxiliary Pilot Channel
(F-DAPICH) 369
15.6.9 Forward Fundamental Channel (F-FCH) 369
15.6.10 Forward Supplemental Channel (F-SCH) 370
15.6.11 Forward Dedicated Control Channel (F-DCCH) 372
15.7 Forward Link Features 373
15.7.1 Transmit Diversity 374
15.7.2 Orthogonal Modulation 375
15.7.3 Power Control 375
15.7.4 Walsh Code Administration 375
15.7.5 Modulation and Spreading 376
15.7.6 Key Characteristics of Forward Link 378
15.8 Reverse Physical Channels 379
15.8.1 The Reverse Access Channel (R-ACH) and the
Reverse Common Control Channel (R-CCCH) 380
15.8.2 Reverse Pilot Channel (R-PICH) 380
xvi Contents

15.8.3 Reverse Dedicated Control Channel (R-DCCH) 381

15.8.4 Reverse Fundamental Channel (R-FCH) 382
15.8.5 Reverse Supplementary Channel (R-SCH) 382
15.8.6 FEC on Reverse Link 383
15.8.7 Reverse Link Physical Layer Characteristics 385
15.8.8 Reverse Link Modulation and Coding 386
15.8.9 Key Characteristics of Reverse Link 387
15.9 Data Services in cdma2000 387
15.10 Mapping of Logical Channels to Physical Channels 388
15.10.1 Forward Link 388
15.10.2 Reverse Link 391
15.11 Evolution of cdmaOne (IS-95) to cdma2000 393
15.12 Major Technical Differences between cdma2000 and
W-CDMA396
15.13 Summary 398
15.14 References 398

A Traffic Tables 399

B Abbreviations 403

C Additional References 413

Index 417

About the Author 423

Preface

The global mobile communications market is booming. There are almost 250 million users
worldwide and should be nearly 1 billion by early next century. Code Division Multiple Access
(CDMA) is the fastest-growing digital wireless technology, tripling its worldwide subscriber
base between 1997 and 1998. There are already 30 million CDMA customers and, at the current
growth rate, there will be 50 million by the millennium. The major markets for CDMA are North
America, Latin America, and Asia (particularly Japan and Korea). In total, CDMA has been
adopted by almost 50 countries around the world.
It is not hard to see the reasons for the success of CDMA. CDMA is an advanced digital
technology that can offer about 7 to10 times the capacity of analog technologies and up to 6
times the capacity of digital technologies such as Time Division Multiple Access (TDMA). The
speech quality provided by CDMA systems is far superior to any other digital cellular technol-
ogy, particularly in difficult radio environments such as dense urban areas and mountainous
regions. In both initial deployment and long-term operation, CDMA provides the most cost-
effective solution for cellular operators. After an 18-month of market rollout, Personal Commu-
nications Services (PCS) providers have adequately demonstrated the power of CDMA technol-
ogy to support a marketing strategy based on low prices and superior performance in key areas
such as voice quality, system reliability, and handset battery life.
CDMA service providers have a strong advantage when pursuing the market to the min-
utes-of-use model, given the longevity of CDMA handset battery life and the higher quality of
the voice signal. A recent analysis of wireless platform performance by the Telecommunications
Research and Action Center (TRAC) found that CDMA outperformed other digital and analog
technologies on every front, including signal quality, security, power consumption, and reliabil-
ity. Although analog technology came out ahead in availability, all three digital services (GSM,
IS-136 TDMA, and IS-95 CDMA) were rated equally over analog with respect to availability of

xvii
xviii Preface

enhanced service features. The TRAC study found CDMA to be superior in signal security and
voice quality over the other digital air interface standards. According to TRAC, CDMA has sev-
eral advantages for consumers. Lower power consumption enables CDMA handsets to support
up to 4 hours of talk time or 48 hours of standby time on a single battery charge. It has also been
found that the soft-handoff characteristics of CDMA lead to fewer dropped calls than with GSM
and IS-136 TDMA. One possible drawback for some CDMA customers is that there are some
limitations on roaming capabilities. Some PCS operators with cellular affiliates are supporting
dual-mode handsets to allow roaming between CDMA and analog platforms.
CDMA technology is constantly evolving to offer customers new, advanced services. The
mobile data speeds offered through CDMA phones are increasing, and new voice codecs pro-
vide speech quality close to wireline. Internet access is now available through CDMA terminals.
The time will soon be at hand when CDMA service providers can further exploit the enhanced
service potential of their platforms. There has been much talk of so-called third-generation (3G)
data capabilities, where PCS providers will be able to compete with wireline service providers at
high access speeds. PCS providers are looking ahead toward providing a range of service cate-
gories such as Internet and intranet access, multimedia applications, high-speed business trans-
actions, and telemetry. The CDMA network offers operators a smooth evolutionary path to 3G
mobile systems.
The IS-95B standard is quite flexible, enabling service providers to allocate data in incre-
ments of 8 kilobits per second (kbps) within the 1.25-megahertz (MHz) CDMA channel band-
width based on how service providers configure software download to already-installed network
controllers. This means operators can implement return data speeds at rates much lower than 64
kbps, ensuring much lower power consumption in handsets than would be the case at a full 64-
kbps return rate. While operators in GSM and IS-136 TDMA sectors are making efforts to
ensure they won’t be left behind as data becomes a factor, CDMA appears to have a clear edge in
its ability to go to relatively high speeds over the existing infrastructure.
The opportunity to use the CDMA platform to add a fixed wireless service feature repre-
sents an added advantage for operators. Because CDMA has ample spectrum to provide a fixed
service on top of mobile, several operators are exploring using terminals that would be able to
shift the handset between fixed and mobile service, depending on where the user is. The univer-
sal handset would serve as a cordless phone in the home and as a mobile handset outside the
home. The evolution to 3G will open the wireless local loop (WLL) with Public-Switched Tele-
phone Network (PSTN) and Public Data Network (PDN) access, while providing more conve-
nient control of applications and network resources. It will also open the door to convenient
global roaming, service portability, zone-based ID and billing, and global directory access. The
3G technology is even expected to support seamless satellite interworking.
With the cornucopia of benefits surrounding CDMA, it is evident that operators using this
platform will have every opportunity to grow the business once the community-based strategy
begins to unfold. The question is, when will they get serious about bringing these new capabili-
ties to market?
Preface xix

Recently an enhanced hybrid technology combining the CDMA air interface with the
GSM network has been built, tested, and evaluated. GSM operators can save over 60% in cumu-
lative capital costs using a GSM-CDMA overlay for network expansion of the GSM network
using IS-95 CDMA radio access in addition to, or as a substitute for, TDMA radio access. This
combines the spectral efficiency of CDMA with all GSM features, including seamless roaming
and network services. The GSM-CDMA technology provides operators with a way to serve mul-
tiple market segments economically and to offer various services on one network platform. In
addition to being a cost-effective network expansion solution, GSM-CDMA also paves an evolu-
tionary path to 3G services including high-speed data, multimedia, and mobile/fixed conver-
gence services.
CDMA is the selected approach for the 3G system, as evidenced by the proposals submitted
by the European Telecommunications Standards Institute (ETSI), the Association Radio Industry
Business (ARIB), and the Telecommunications Industry Association (TIA). The 3G cdma2000
uses a CDMA air interface based on the existing IS-95B standard to provide wireline-quality
voice service and high-speed data services, ranging from 144 kbps for mobile users to 2 megabits
per second (Mbps) for stationary users. It is important to note that cdma2000 is a core proposal of
the TIA for International Mobile Telecommunications-2000 (IMT-2000). Moreover, support for
cdma2000 is not limited to North America; Korean carriers have a great opportunity to provide
3G-like service with today’s existing CDMA technology. Mobile data rates of up to 114 kbps and
fixed peak rates beyond 1.5 Mbps are within reach before the end of the decade with today’s
CDMA technology. These capabilities will be provided without degrading the systems’ voice
transmission capabilities or requiring additional spectrum. This will have tremendous implica-
tions for the majority of operators that are spectrum constrained. A doubling of capacity and a
1.5-Mbps data rate capability within a 1.25-MHz channel structure look very appealing.
This book is an extension of the book Applications of CDMA in Wireless Communications
(Garg, Smolik, and Wilkes, Prentice Hall, 1997). In that book, the primary focus was on the
CDMA systems standardized by TIA and American Telecommunications Industries Standards
(ATIS) as standards IS-95 and IS-665. Since the publication of that book, CDMA technology
has undergone major changes and has become a viable technology for 3G systems. In this book,
I discuss those aspects of CDMA that are essential to understanding system capacity. I also pro-
vide guidelines for system parameters of a CDMA network. The book outlines a migration path
for CDMA to a 3G cdma2000 system.
In writing this book, I addressed the needs of practicing engineers and engineering manag-
ers by explaining CDMA concepts, system capacity, radio frequency (RF) engineering, and
other important aspects of the CDMA network. Students studying courses in telecommunica-
tions will also find this book useful as they prepare for careers in the wireless industry. I
included a sufficient amount of mathematics so that you can understand the operation of the
CDMA network, but I tried not to overwhelm you with very complex mathematical derivations.
This book can be used by practicing telecommunications engineers involved in the design
and operation of CDMA-based cellular/PCS networks as well as by senior or graduate students
xx Preface

in electrical engineering, telecommunications engineering, and computer engineering curricula.

I assume that you have some basic background in mobile communications and CDMA technol-
ogy. If you don’t, the book mentioned above by Garg, Smolik, and Wilkes can provide that
understanding. By selectively reading pertinent chapters of that book, telecommunications man-
agers who are engaged in managing CDMA networks but who have little or no technical back-
ground can gain enough of an understanding of CDMA systems to read this book.
This book can be divided into four segments: Chapters 1 through 4 provide a foundation
for understanding the material in subsequent chapters. Chapters 5 through 11 deal with IS-95
CDMA standards, and chapters 12 and 13 provide design aspects of a CDMA system. Chapters
14 and 15 focus on the data applications in CDMA and the evolution of IS-95 (2G system) to
cdma2000 (3G system) in order to satisfy ITU IMT-2000 specifications. The following is a syn-
opsis of the subjects covered in each chapter.
• Chapter 1. Major attributes of CDMA and the access technologies used for cellular/
PCS systems.
• Chapter 2. The different types of Spread Spectrum (SS) systems that are used. The
main focus is on the Direct Sequence Spread Spectrum (DSSS) techniques that are
employed in CDMA. I provide a relationship to calculate the performance of a CDMA
system.
• Chapter 3. Speech and channel coding applications in the IS-95 CDMA system.
• Chapter 4. The concepts of diversity reception used to improve signal-to-noise ratio
(SNR) of the system; various combining schemes used to combine the signals; some
practical antennas used in the cellular telephone industry.
• Chapter 5. Functional entities of the wireless network and the TIA-standardized inter-
faces between the entities. I examine the activities of the International Telecommunica-
tion Union (ITU) to add Intelligent Network (IN) to wireless systems.
• Chapter 6. A high-level description of the IS-95 CDMA air interface, including impor-
tant aspects of the forward link (base station to mobile) and reverse link (mobile to base
station) and modulation parameters for the channels.
• Chapter 7. Modulation schemes, bit repetition, block interleaving, and channel coding;
these are used in processing logical channels on the IS-95 CDMA forward and reverse
links. Details about information processing, message types, and message framing are
presented for the pilot, sync, paging, and traffic channels on the forward link. Similar
details are provided for the access and traffic channels on the reverse link
• Chapter 8. IS-95 CDMA call processing states that a mobile station (MS) goes
through in getting to a traffic channel; idle handoff, slotted paging operation, CDMA
registration, and authentication procedures; call flows for CDMA call origination, call
termination, call release, and authentication.
• Chapter 9. The layering concept used to develop the protocols for IS-95 CDMA; the
standardized interfaces between the functional entities, mainly the A-Interface and TIA
IS-634-defined MSC-BS messages, message sequencing, and mandatory timers at the
Preface xxi

BS and the MSC. The chapter also provides call flow diagrams for typical supplemen-
tary services, handoff scenarios, and Over-The-Air Service Provisioning (OTASP).
• Chapter 10. Handoff strategy used in IS-95 CDMA; power control schemes for the
reverse and forward links.
• Chapter 11. Various parameters used to identify an MS including International Mobile
Station Identity (IMSI), Mobile Station Number (MDN), Electronic Serial Number
(ESN), and station class mark. I focus on authentication procedures, including the
authentication of MS registration, MS originations, MS terminations, MS data bursts,
and Temporary Mobile Station Identity (TMSI) assignment. Also discussed are unique
challenge response procedures.
• Chapter 12. Basic guidelines for engineering a CDMA system, including a discussion
of propagation models, link budgets, the transition from analog operation to CDMA
operation, radio link capacity, facility engineering, border cells on a boundary between
two service providers, and interfrequency handoff.
• Chapter 13. Procedures for calculating the capacity of the reverse and forward link of
a CDMA system; a procedure to develop a link safety margin parameter for each of the
forward link channels.
• Chapter 14. Standards for data services supported by CDMA cellular/PCS systems;
highlights of the TIA IS-99, TIA IS-637, and TIA IS-657 standards. I describe the
architecture for each of the four data services (e.g., packet data, asynchronous data, fac-
simile, and short message services) and the protocol stacks supported by these services.
• Chapter 15. The cdma2000, 3G evolution of IS-95. The cdma2000 Radio Transmis-
sion Technology (RTT) is a wideband, SS radio interface that uses CDMA technology
to satisfy the needs of 3G wireless communication systems.
Appendix A presents traffic tables for a variety of blocking probabilities and channel
numbers. Appendix B comprises a list of abbreviations I introduce in the text and that are com-
mon to the industry. The references cited in Appendix C are papers and texts that I have found
useful and, when considered in addition to those cited in the text, provide a rich background for
readers interested in looking into digital wireless technology in greater depth.
I suggest chapters 1–11 for those who are interested in IS-95 standards but who do not
have much background in digital communications. Those who have adequate background in dig-
ital communications may skip chapters 1–4.
I recommend chapters 1, 2, 4–10, 12, and 13 for those who are involved with the design of
a CDMA system. The engineering managers should use chapters 1 and 5–12 to achieve adequate
knowledge of IS-95 CDMA.
I suggest chapters 1–8, 10–12, 13, and 15 for a one-semester graduate course in IS-95
CDMA and its evolution to cdma2000.

I would like to thank the many people who helped me prepare the material in this book. Bernard
Goodwin provided his encouragement in motivating me to write the book. Professor Ted Rappa-
xxii Preface

port of Virginia Tech took me under the banner of his new series. I acknowledge the many help-
ful suggestions I received from my many friends.
Finally, I acknowledge the assistance of my wife, Pushpa Garg, and the staff of BooksCraft,
Inc. during the production of this book.
 C H A P T E R 1

Introduction to Access
Technologies

1.1 Introduction
After 18 months of market rollouts, digital cellular and Personal Communications Services
(PCS) providers have adequately demonstrated the power of Code Division Multiple Access
(CDMA) technology to support a marketing strategy based on low prices and superior perfor-
mance in key areas such as voice quality, system reliability, and handset battery life. According
to the CDMA Development Group (CDG), the market base for IS-95 platform providers has
grown from only 200,000 customers in 1997 to about 3.0 million in 1999.
CDMA service providers have a strong advantage in pursuing the market to the minutes-
of-use model, given the longevity of CDMA handset battery life and the higher quality of the
voice signal. A recent analysis of wireless platform performance by Telecommunications
Research and Action Center (TRAC) found that CDMA beat other digital and analog technolo-
gies on every front, including signal quality, security, power consumption, and reliability.
Although analog was more readily available, all three digital systems (GSM, IS-136 TDMA, and
IS-95 CDMA) were rated equally over analog with respect to enhanced service features. The
TRAC study found CDMA to be superior in signal security and voice quality over the other dig-
ital air interface standards. According to TRAC, CDMA has several advantages for consumers.
Lower power consumption enables CDMA handsets to support up to 4 hours of talk time or 48
hours of standby time on a single battery charge. It has also been found that the soft handoff
characteristics of CDMA lead to fewer dropped calls than GSM and IS-136 TDMA. One possi-
ble drawback for some CDMA customers is that there are some limitations on roaming capabili-
ties. Some PCS operators with cellular affiliates are supporting dual-mode handsets to allow
roaming between the CDMA and analog platforms.
The IS-95 standard is quite flexible, enabling service providers to allocate data in incre-
ments of 8 kilobits per second (kbps) within the 1.25-megahertz (MHz) CDMA channel band-

1
2 Ch. 1 • Introduction to Access Technologies

width based on how service providers configure software download to already-installed

network controllers. This means operators can implement return data speeds at rates much
lower than 64 kbps, ensuring much lower power consumption in handsets than would be the
case at a full 64-kbps return rate. While operators in GSM and IS-136 TDMA sectors are mak-
ing efforts to ensure they won’t be left behind as data becomes a factor, CDMA appears to have
a clear edge in its ability to go to relatively high speeds over the existing infrastructure.
The opportunity to use the CDMA platform to add a fixed wireless service feature repre-
sents an added advantage for operators. With ample spectrum to provide a fixed service on top of
mobile, several operators are exploring using terminals that would be able to shift the handset
between fixed and mobile service, depending on where the user is. The universal handset would
serve as a cordless phone in the home and as a mobile handset outside the home.
CDMA offers about eight times the capacity of analog, which implies that there’s enough
capacity in 30 MHz of spectrum alone to serve the entire mobile market now and into the future
for a country like Canada. With the cornucopia of benefits surrounding CDMA, it is evident that
operators using this platform have every opportunity to grow the business once the community-
based strategy begins to unfold. The question is, when will they get serious about bringing these
new capabilities to market?
Recently, an enhanced hybrid technology combining the CDMA air interface with the
GSM network was built, tested, and evaluated. GSM operators can save over 60% in cumulative
capital costs using a GSM-CDMA overlay for network expansion of the GSM network using IS-
95 CDMA radio access in addition to, or as a substitute for, TDMA radio access. This combines
the spectral efficiency of CDMA with all GSM features, including seamless roaming and net-
work services. The GSM-CDMA technology provides operators the flexibility to serve multiple
market segments economically and to offer various services on one network platform. In addi-
tion to being a cost-effective network expansion solution, GSM-CDMA also paves an evolution
path to third-generation services including high-speed data, multimedia, and mobile/fixed con-
vergence services.
In this chapter, we discuss major attributes of CDMA and describe the access technologies
used for cellular/PCS systems, outlining their advantages and disadvantages. We also discuss the
use of CDMA for future third-generation systems and outline modes of operation for cellular/
PCS systems.

1.2 Major Attributes of CDMA Systems

The major attributes of IS-95A/J-STD-008 CDMA systems [5] are as follows:
• System Capacity. The projected capacity of CDMA systems is much higher than that
of existing analog/digital systems. The increased system capacity is due to an improved
coding gain/modulation scheme, voice activity, three-part sectorization, and reuse of
the same spectrum in every cell and all sectors.
• Quality of Service. CDMA improves the quality of service by providing robust opera-
tion in fading environments and transparent (soft) handoffs. CDMA takes advantage of
Third-Generation Systems 3

multipath fading to enhance communications and voice quality. By using RAKE

receivers and other improved signal-processing techniques, each mobile station selects
the three strongest multipath signals and coherently combines them to produce an
enhanced signal. Thus, the multipath fading of the radio channel is used to an advan-
tage in CDMA whereas, in narrowband systems, fading causes a substantial degrada-
tion of signal quality. By using soft handoff, CDMA eliminates the ping-pong effect
that occurs when the mobile is near a border between cells and the call is rapidly
switched between the two cells. The ping-pong effect results in handoff noise, increases
the load on switching equipment, and increases the chance of a dropped call. In
CDMA’s soft handoff, a connection is made to the target cell while maintaining the
connection with the serving cell, both operating on the same carrier frequency. This
procedure ensures a smooth transition between cells, one that is undetectable to the
subscriber. In comparison, many analog and other digital (TDMA) systems use a break-
before-make connection and require a change in mobile frequency that increases hand-
off noise and the chance of a dropped call.
• Economies. CDMA is a cost-effective technology that requires fewer cell sites and no
costly frequency reuse pattern. The average power transmitted by CDMA mobile sta-
tions averages 6 to 7 megawatts (mW), which is significantly lower than the average
power typically required by FM and TDMA phones. Transmitting less power means
that average battery life will be longer.

1.3 Third-Generation Systems

The time will soon be at hand when CDMA service providers can further exploit the enhanced
service potential of their platforms. There has been much talk of so-called third-generation (3G)
data capabilities, where PCS providers will be able to compete with wireline service providers at
high access speeds. PCS providers are looking ahead to providing a range of service categories,
such as Internet and intranet access, multimedia applications, high-speed business transactions,
and telemetry.
Taking into account the limitations imposed by the finite amount of radio spectrum avail-
able, the focus of third-generation mobile systems is on economy of network and radio transmis-
sion design to provide seamless service from the customer’s perspective. Third-generation
mobile systems have to provide their users with seamless access to the fixed data network. They
are perceived as the wireless extension of future fixed networks, as well as an integrated part of
the fixed network infrastructure.
In Europe, three related network platforms are currently the subject of intensive
research—Future Land Public Mobile Telephone Systems (FLPMTS, now known as IMT-2000),
Mobile Broadband Systems (MBS), and Wireless Local Area Network (WLAN). One major dis-
tinction of IMT-2000 relative to second-generation systems is its hierarchical cell structure,
which is designed to support a wide range of multimedia broadband services by using advanced
transmission and protocol technologies. Second-generation systems mainly use a one-layer cell
4 Ch. 1 • Introduction to Access Technologies

structure and employ frequency reuse within adjacent cells in such a way that each single cell
manages its own radio zone and radio circuit control within the mobile network, including traffic
management and handoff procedures. The amount of traffic supported in each cell is fixed
because of frequency limitations and also because of limited flexibility of radio transmission
which is optimized mainly for voice and low-data-rate transmissions. Increasing traffic leads to
costly cellular reconfigurations such as cell splitting and cell sectorization.
The multilayer cell structure in IMT-2000 aims to overcome these problems by overlay-
ing—discontinuously—picocells and microcells on the macrocell structure with wide area cov-
erage. Global/satellite cells can be used in the same sense to provide area coverage where
macrocell constellations are not economical to deploy and/or to support long distance traffic.
With low mobility and small delay spread profiles in picocells, high bit rates and high traf-
fic densities can be supported with low complexity as opposed to low bit rates and low traffic
load in macrocells that support high mobility. Users expect service to be selected in a uniform
manner with consistent procedures, irrespective of whether access to these services is fixed or
mobile. Freedom of location and means of access will be facilitated by smart cards, which allow
customers to register on different terminals with varying capabilities (speech, multimedia, data,
short messaging).
The choice of a radio interface parameter set corresponding to a multiple access scheme is
a critical issue in terms of spectral efficiency, taking into account the ever increasing market
demand for mobile communications and the fact that radio spectrum is a very expensive and
scarce resource. A comparative assessment of several different schemes has been carried in the
framework of the Research in Advanced Communications Equipment (RACE) program. One
possible solution is to use a hybrid CDMA/TDMA/FDMA technique, integrating the advantages
of each and meeting the varying requirements on channel capacity, traffic load, and transmission
quality in different cellular/PCS layouts. Disadvantages of such hybrid access schemes are the
high complexity and difficulties in achieving simplified low-power, low-cost transceiver design
as well as efficient flexibility management in the several cell layers.
CDMA is the selected approach for third-generation systems, as evidenced by the propos-
als in ETSI, ARIB (Japan), and the TIA. In Europe and Japan, Wideband CDMA (W-CDMA)
has been proposed to avoid IS-95 intellectual property rights (IPR). In North America,
cdma2000 uses a CDMA air-interface based on the existing IS-95 standard to provide wireline-
quality voice service and high-speed data services ranging from 144 kbps for mobile users to 2
megabits per second (Mbps) for stationary ones. The 64-kbps data capability of IS-95 will pro-
vide high-speed Internet access in a mobile environment, a capability that cannot be matched by
other narrowband digital technologies.
Mobile data rates of up to 114 kbps and fixed peak rates beyond 1.5 Mbps are within reach
before the end of the decade using wideband CDMA technologies. These services will be pro-
vided without degrading the systems’ voice transmission capabilities or requiring additional
spectrum. This will have tremendous implications for the majority of operators that are spectrum
constrained.
Multiple Access Technologies 5

1.4 Multiple Access Technologies

There are two types of access technologies [6]—narrowband and wideband. The narrowband
access technologies are Frequency Division Multiple Access (FDMA) and Time Division Multi-
ple Access (TDMA).
In FDMA, signals from various users are assigned different frequencies (see Fig. 1-1).
Guard bands are used between adjacent signal spectra to minimize crosstalk between channels.
The advantages of FDMA are that
• Capacity increases can be obtained by reducing the information bit rate and using effi-
cient digital codes.
• Technological advances required for implementation are simple. A system can be con-
figured so that improvements in terms of speech coder bit-rate reduction could be
readily incorporated.
The main disadvantages of FDMA are that
• It does not differ significantly from analog systems; capacity improvement depends on
reducing signal-to-interference ratio, or signal-to-noise ratio (SNR).
• FDMA involves narrowband filters, and, because these are not realized in very-large-
scale integrated (VLSI) digital circuits, this may set a high cost floor for terminals even
under volume production conditions.

Frequency 1 circuit

Frequency 2 circuit Downlink

Path

Frequency n circuit

Frequency
Domain

Frequency 1 circuit

Frequency 2 circuit Uplink

Path

Frequency n circuit

Time
Domain

Figure 1-1 Frequency Division Multiple Access (FDMA) with Frequency Division
Duplex (FDD).
6 Ch. 1 • Introduction to Access Technologies

• The maximum bit rate per channel is fixed and small, inhibiting the flexibility in bit-rate
capability that may be a requirement for computer file transfer in some applications.

In a TDMA system, information from each user is conveyed in time intervals called time
slots (see Fig. 1-2). Several time slots make up a time frame. Each time slot consists of a pream-
ble plus information bits. The functions of the preamble are to provide identification and inci-
dental information and to allow synchronization of the time slot at the intended receiver. Guard
times are provided between each user’s transmission to minimize crosstalk between channels.
Most TDMA systems time divide a frame into multiple slots used by different transmitters. This
approach is called Time Division Multiplex (TDM). The information is transmitted via a radio
carrier from a base station (BS) to several active mobile stations (MSs) in the downlink. In the
uplink, transmissions from mobile stations to a BS are time-sequence synchronized on a com-
mon frequency for TDMA. The advantages of TDMA are that it
• Allows a flexible bit rate, not only for multiples of a basic single channel but also sub-
multiples for low-bit-rate broadcast-type traffic.
• Potentially integrates in VLSI without narrowband filters, giving a low cost floor in vol-
ume production.
• Offers the opportunity for frame-by-frame monitoring of signal strength/bit error rates
to enable either mobiles or base stations to initiate and execute handoffs.

Slot 1 Slot 2 Slot m

Frequency 1 circuit circuit circuit

Frequency 2 circuit circuit circuit Downlink

Path

Frequency n circuit circuit circuit

Frequency
Domain

Frequency 1 circuit circuit circuit

Frequency 2 circuit circuit circuit Uplink

Path

Frequency n circuit circuit circuit

Time
Domain

Figure 1-2 Time Division Multiple Access (TDMA) with FDD

Modes of Operation in Wireless Communications 7

• Uses bandwidth more efficiently because no frequency guard band is needed between
channels.
• Transmits each signal with sufficient guard time between time slots and accommodates
time inaccuracies caused by clock instability, delay spread, transmission time delay
because of propagation distance, and the “tails” of signal pulses in TDMA because of
transient responses.

The main disadvantage of TDMA are:

• Particularly for mobile hand sets, on uplink TDMA demands high peak power in trans-
mit mode. This reduces battery life.
• TDMA requires a substantial amount of signal processing for matched filtering and
correlation detection for synchronizing with a time slot.
In wideband technologies, the entire bandwidth is made available to each mobile user; this
bandwidth is many times larger than the bandwidth required to transmit information. Such sys-
tems are called spread spectrum (SS) systems. The primary advantage of wideband technology
is its ability to tolerate a fair amount of signal interference compared to FDMA and TDMA,
which typically cannot tolerate any such interference. Because of the interference tolerance of
wideband systems, the problems of frequency band assignment and adjacent cell interference
are greatly simplified. Flexibility in system design and deployment are improved since interfer-
ence to others is not a problem. FDMA and TDMA radios must be carefully assigned a fre-
quency or time slot to assure that there is no interference with other similar radios. In wideband
systems, adjacent cells share the same frequencies; with TDMA/FDMA it is not feasible for
adjacent cells to share same frequencies because of interference. There is no need for frequency
planning in wideband systems.

1.5 Modes of Operation in Wireless Communications

There are two operational modes in wireless communications: Frequency Division Duplex
(FDD) and Time Division Duplex (TDD). In FDD, the transmitter and receiver operate simulta-
neously on different frequencies. Separation is provided between the downlink and uplink chan-
nels to avoid interference between the transmitter and the receiver (see Fig. 1-2). In TDD, a
bidirectional flow of information is achieved using a simplex-type scheme by automatically
changing in time the direction of transmission on a single frequency. The basic advantages of
TDD are that

• There is no need for a dedicated duplex stage (duplexer). The only requirements are
that a fast switching synthesizer, RF filter paths, and fast antenna switch be available.
• It increases battery life or reduces battery weight.
• It provides phones with better quality and at a lower cost.
• It requires a single frequency band instead of the two in FDD.
8 Ch. 1 • Introduction to Access Technologies

1.6 Summary
In this chapter, we focused on the growth of CDMA technology in wireless communications. We
discussed the suitability of CDMA for future 3G systems to provide higher data rate access to
Internet/intranet services because it is the most suitable digital technology to handle future wire-
less communications demand.
We briefly described narrowband channelized and wideband nonchannelized technologies
and listed their advantages and disadvantages. Finally, we described modes of operations in
wireless communications.

1.7 References
1. Balston, D. M., “The Pan-European Cellular Technology,” IEE Conference Publication,
1988.
2. Balston, D. M., and Macario, R. C. V., Cellular Radio Systems, Artech House, Boston, 1993.
3. Dasilva, J. S., Ikonomou, D., and Erben, H., “European R&D Programs on Third-Genera-
tion Mobile Communications Systems,” IEEE Personal Communications 4(1), February
1997, pp. 46–52.
4. “The European Path Towards UMTS,” IEEE Personal Communications, special issue, Feb-
ruary 1995.
5. Garg, V. K., Smolik, K. F., and Wilkes, J. E., Applications of CDMA in Wireless/Personal
Communications, Prentice Hall, Upper Saddle River, NJ, 1997.
6. Garg, V. K., and Wilkes, J. E., Wireless and Personal Communications Systems, Prentice
Hall, Upper Saddle River, NJ, 1996.
7. Marley, N., “GSM and PCN Systems and Equipment,” JRC Conference, Harrogate, 1991.
8. Rapeli, J., “UMTS: Targets, System Concepts, and Standardization in a Global Framework,”
IEEE Personal Communications, special issue, February 1995.
9. Salmasi, A., “An Overview of Code-Division Multiple Access (CDMA) Applied to the
Design of Personal Communications Networks,” in Third Generation Wireless Information
Network, Nanda, S., and Goodman, D. J., ed., Kluwer Academic Publishers, Boston, 1992,
pp. 277–98.
 C H A P T E R 2

Direct Sequence Spread

Spectrum and Spreading
Codes

2.1 Introduction
In this chapter we describe the different types of Spread Spectrum (SS) systems that are used
and then focus on the Direct Sequence Spread Spectrum (DSSS) technique that is employed in
CDMA IS-95 systems. We develop the necessary relationships to evaluate the performance of a
DSSS system with Binary Phase-Shift Keying (BPSK) and Quadrature Phase-Shift Keying
(QPSK) modulation schemes and provide a relationship to calculate the performance of a
CDMA system.

2.2 Types of Techniques Used for Spread Spectrum

Since the late 1940s, SS techniques have been used for military applications in which clandes-
tine operation is a major objective. SS techniques provide excellent immunity to interference—
possibly the result of intentional jamming—and allow transmission to be hidden within back-
ground noise. Recently, SS systems have been adopted for civilian applications in wireless tele-
phony systems.
There are three general approaches to implementing SS systems.
1. Direct Sequence Spread Spectrum (DSSS) where a carrier is modulated by a digital
code in which the code bit rate is much larger than the information signal bit rate (see
Fig. 2-1). These systems are also called pseudonoise (PN) systems.
2. Frequency Hopping Spread Spectrum (FHSS) where the carrier frequency is shifted
in discrete increments in a pattern generated by a code sequence (see Fig. 2-2). Some-
times, the codes are chosen to avoid interference to or from other non-SS systems. In an
FHSS system, the signal frequency remains constant for a specified time duration,
referred to as a time chip, Tc. The FHSS system can be either a fast-hop or a slow-hop
system. In a fast-hop system, the frequency hopping occurs at a rate that is faster than

9
10 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

TO = (2 n -1 ) TC
TO =
Period of Output TC = time chip

Waveform
+1

-1

Figure 2-1 Direct Sequence Spread Spectrum Approach

Frequency

fn
fn -1
fn -2

f3
f2
f1

0 Tc 2Tc
Figure 2-2 Frequency Hopping Spread Spectrum Approach

the message bit rate. In a slow-hop system, the hop rate is slower than the message bit
rate. There is, of course, an intermediate situation in which the hop rate and message bit
rate are the same.
FHSS radio systems experience occasional strong bursty errors, while DSSS
radio systems experience continuous but lower-level random errors. With DSSS radio
systems, single errors are dispersed randomly over time, whereas, with FHSS radio sys-
tems, errors are distributed in clusters. Bursty errors are attributable to fading or single
frequency interference, which is time and frequency dependent. DSSS spreads the
information in both the time and frequency domains, thus providing time and frequency
diversity and minimizing the effects of fading and interference.
3. Time-Hopped (TH) system, where the transmission time is divided into intervals
called frames (see Fig. 2-3). Each frame is divided into time slots. During each frame,
The Concept of Spread Spectrum System 11

Transmitted time
slots (k bits) M = time slots in each
frame; t = Tf / M
One Frame

0 t Tf 2Tf 3Tf

Figure 2-3 Time Hopping Spread Spectrum Approach

one and only one time slot is modulated with a message. All of the message bits accu-
mulated in previous frames are transmitted.

2.3 The Concept of Spread Spectrum System

The theoretical capacity of any communications channel is defined by C. E. Shannon’s channel
capacity formula [7], Eq. (2.1):

S
C = B w log 2 1 + ---- (2.1)
N

where Bw = bandwidth in Hertz,

C = channel capacity in bits per second,
S = signal power, and
N = noise power.
Eq. (2.1) gives the relationship between the theoretical ability of a channel to transmit informa-
tion without errors for a given SNR and a given bandwidth of a channel. Channel capacity is
increased by increasing the channel bandwidth, the transmitted power, or a combination of both.
Shannon modeled the channel at baseband. However, Eq. (2.1) is applicable to a Radio
frequency (RF) channel by assuming that the Intermediate Frequency (IF) filter has an ideal
(flat) band-pass response with a bandwidth that is at least 2 × Bw . This bound assumes that chan-
nel noise is Additive White Gaussian Noise (AWGN). AWGN is often adopted in the modeling
of an RF channel. This assumption is justified since the total noise is generated by random elec-
tron fluctuations. The central limit theorem provides us with the assumption that the output of an
IF filter has a Gaussian distribution and is frequency independent. For most communications
systems that are limited by thermal noise, this assumption is true. For interference-limited sys-
tems, this assumption is not true and the results may be different. The Shannon equation does
not provide a method to achieve the bound. Approaching the bound requires complex channel
coding and modulation techniques. In many cases, achieving an implementation that provides
performance near this bound is impractical due to the resulting complexity.
12 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

An analog cellular system is typically engineered to have an SNR of 17 decibels (dB)* or

more. CDMA systems can be engineered to operate at much lower SNRs since the channel
bandwidth can be traded for the SNR to achieve good performance at a very low SNR.
Next we rewrite Eq. (2.1) as Eq. (2.2).

C S
------ = 1.44log e 1 + ---- (2.2)
Bw N

Since

S 1 S 2 1 S 3 1 S 4
log e  1 + ---- = ---- – ---  ---- + ---  ---- – ---  ---- + …
S
 N N 2  N 3  N 4  N

we use the logarithmic expansion and assume that the SNR is small, e.g., SNR ≤ 0.1, so we can
neglect the higher-order terms to rewrite Eq. (2.2) as

C N
B w ≈ ---------- × ---- (2.3)
1.44 S

For any given SNR we can have a low information error rate by increasing the bandwidth
used to transmit the information. As an example, if we want a system to operate on a link in
which the information rate is 10 kbps and SNR is 0.01, we must use a bandwidth of

3
10 × 10 6
B w = --------------------------- = 0.69 × 10 Hz or 690 kHz
1.44 × 0.01

Information can be modulated into the spread spectrum signal by several methods. The
most common method is to add the information to the spectrum-spreading code before it is used
for modulating the carrier frequency (Fig. 2-4). This technique applies to any SS system that
uses a code sequence to determine RF bandwidth. If the signal being sent is analog (voice, for
example), the signal must be digitized before being added to the spreading code.

2.3.1 System Processing Gain

One of the major advantages with an SS system is its robustness to interference. The sys-
tem processing gain (Gp) quantifies the degree of interference rejection. The system processing
gain is the ratio of RF bandwidth to the information rate and is given as

B
G p = ------w (2.4)
R

* This assumes a fading radio environment, which is typical for analog cellular systems that use Frequency Modulation.
In the absence of fading, good FM performance is achievable at lower SNR.
The Concept of Spread Spectrum System 13

Chip Rate = R c = 1/TC

PN Sequence, ci (t )
cos (2π fc t )

xi (t )
Modulo-2 Baseband
Baseband d i (t )
Sum Filter

Information Rate, R i = 1/T Power

Amplifier
Figure 2-4 Basic DSSS System Transmitter

Typical processing gains for SS systems lie between 20 and 60 dB. With an SS system, the noise
level is determined both by the thermal noise and by interference. For a given user, the interfer-
ence is processed as noise. The input and output SNRs are related as

 ---
S
- = G p  ----
S
(2.5)
 N o  N i

It is instructive to relate the SNR to the Eb /N0* ratio where Eb is the energy per bit and N0 is
the noise power spectral density.

 ---
S Eb × R Eb 1
- = ---------------- = ------ × ------ (2.6)
 N i N0 × B N0 G p

From Eq. (2.6) we can express Eb /N0 as

E
-----b- = G p ×  ---- =  ----
S S
(2.7)
N0  N i  N o

EXAMPLE 2.1
Calculate the processing gain for a DSSS system that has a 10 Megachips per second (Mcps)
code clock rate and 4.8-kbps information rate. How much improvement in the processing gain
will be achieved if the code generation rate is changed to 50 Mcps? Is there an advantage in
going to a higher code generation rate with a 4.8-kbps information rate?

* The noise power spectral density actually consists of both the thermal noise and interference. Unless stated explicitly,
N0 represents the thermal noise. However, common usage of this ratio assumes that N0 includes both the thermal noise
and interference. With SS systems, interference is transformed into noise.
14 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

We assume that the DSSS waveform has a voltage distribution of (sinx)/x. The power distribu-
tion has a form of [(sinx)/x]2. The bandwidth of the main lobe is equal to the spreading code clock
rate.
7
1.0 × 10 3
G p = ---------------------3- = 2.1 × 10 = 33.1 dB*
4.8 × 10
With 50 Mcps
7
5 × 10 4
G p = ---------------------3- = 1.04 × 10 = 40.2 dB
4.8 × 10
By increasing the code generation rate from 10 to 50 Mcps, we get only 7dB improvement in
the processing gain. The effort required to get five times the operating speed of a circuit may be
much more demanding compared to an improvement of 7 dB in the processing gain.

2.4 The Performance of DSSS

2.4.1 The DSSS System
The DSSS system is a wideband system in which the entire bandwidth of the system is
available to each user. A system is defined as a DSSS system if it satisfies the following require-
ments:
1. The spreading signal has a bandwidth much larger than the minimum bandwidth
required to transmit the desired information, which for a digital system is the baseband
data.
2. The spreading of the data is performed by means of a spreading signal, often called a
code signal. The code signal is independent of the data and of a much higher rate than
the data signal.
3. At the receiver, despreading is accomplished by the cross-correlation of the received
spread signal with a synchronized replica of the same signal used to spread the data.

2.4.2 Coherent Binary Phase-Shift Keying

The simplest form of a DSSS communications system employs coherent BPSK for both
the data modulation and the spreading modulation. But the most common form uses BPSK for
data modulation and Quadrature Phase-Shift Keying (QPSK) for spreading modulation. We first
consider the simplest case.
The encoded DSSS BPSK signal is given by

x ( t ) = c ( t )s ( t ) = c ( t )d ( t ) 2S cos ω c t (2.8)

where s(t) = d ( t ) 2S cos ω c t ,

d(t) = the baseband signal at the transmitter input and receiver output,

* dB = 10 log G
The Performance of DSSS 15

c(t) = the spreading signal,

S = the signal power, and
ωc = the carrier frequency.
In Eq. (2.8), we represent the modulo-2 addition of c(t) and d(t) as a multiplication because the
binary signals 0 and 1 represent values of 1 and –1 into the modulator.
The signal s(t) has a [(sinx)/x]2 spectrum of bandwidth roughly 1/T (where T is the period-
icity at baseband), while the SS signal x(t) has a similar spectrum but with a bandwidth of
approximately 1/Tc (where Tc is the periodicity of the spreading signal). From Eq. (2.4), the pro-
cessing gain of the system is Gp = (Bw /R) = T/Tc. If the interfering signal is represented by I(t),
then in the absence of noise (assuming that the interferer limits the system performance—in
other words, that the interferer’s power level exceeds the thermal noise power), the signal at the
receiver is given as
[ r ( t ) ]∗ = x ( t ) + I ( t ) (2.9)
The receiver multiplies this by the PN waveform to obtain the signal

r ( t ) = c ( t ) [ x ( t ) + I ( t ) ] = c ( t ) [ c ( t )s ( t ) ] + c ( t )I ( t ) = s ( t ) + c ( t )I ( t ) (2.10)
since c(t)2 = 1. c(t)I(t) is the effective noise waveform due to interference.
The conventional BPSK detector output is given as

r = d Eb + n (2.11)

where d = the data bit for the T second interval,

Eb = the bit energy, and
n = the equivalent noise component.
The spreading-despreading operation does not affect the signal, nor does it affect the spec-
tral and probability density function of the noise. For this reason the bit error probability Pb
associated with the coherent BPSK SS signal is the same as with the BPSK [8] signal and is
given as

1  E 
P b = --- erfc  -----b- (2.12)
2  N 0

2.4.3 Quadrature Phase-Shift Keying

For QPSK modulation, we denote the in-phase and quadrature data waveforms as dc(t) and
ds(t), respectively, and the corresponding PN binary waveform as cc(t) and cs(t). We can repre-
sent a QPSK signal as (see Fig. 2-5)

x ( t ) = c c ( t )d c ( t ) S cos ω c t + c s ( t )d s ( t ) S sin ω c t (2.13)

where each QPSK pulse is of duration Ts = 2T.

16 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

Chip Rate = R c = 1/TC

PN Sequence, cc (t ) cos (2π fc t )

Modulo-2 I Baseband xi (t )
Information Rate, R i = 1/T Sum Filter

Baseband d i (t ) 90° phase shift φ

Modulo-2 Q Baseband
Sum Filter
Power
Amplifier
PN Sequence, cs (t )

Figure 2-5 DSSS System with QPSK Transmitter

The in-phase output component is

rc = d c Eb + nc (2.14)
Tc
2
where n c = ----- ∫ c c ( t )I ( t ) cos ω c t dt ,
Ts
0

and the quadrature component is

rs = d s Eb + ns (2.15)

Tc
2
where n s = ----- ∫ c s ( t )I ( t ) sin ω c t dt .
Ts
0

QPSK modulation can be viewed as two independent BPSK modulations. Thus the net
data rate is doubled. We consider a special case of QPSK modulation where cc(t) and cs(t) are
equal and have a value of c. The QPSK symbol energy is also the bit energy (one bit per QPSK
signal).
For this case Eqs. (2.14) and (2.15) have the form

E
r c = d -----b- + n c (2.16)
2
and
The Performance of DSSS 17

E
r s = d -----b- + n s (2.17)
2
where nc and ns are 0 mean independent with conditional variances
2
Var 〈n c|θ〉 = ( IT c ) ( cos θ ) (2.18)

and

2
Var 〈n s|θ〉 = ( IT c ) ( sin θ ) (2.19)

Next we use

( rc + rs ) E ( nc + ns )
r = ------------------- = d -----b- + --------------------
- (2.20)
2 2 2

as the statistic for decision rule; then

1 2 2 IT
Var 〈( n c + n s ) ⁄ 2|θ〉 = --- [ IT c ( cos θ ) + IT c ( sin θ ) ] = --------c (2.21)
4 4

The final expression for narrowband interference, Gj (f ), at the demodulator baseband out-
put is given as

IT I
G j ( f ) = --------c = --------- (2.22)
4 4R c

For baseband systems, we define the baseband interference, I(f ), as

I ( f ) = 2G j ( f ) ; 0 ≤ f ≤ Rc (2.23)

The bit error probability for AWGN is given [8] as

1  E   2E 
P b = --- erfc  -----b- = Q  --------b- (2.24)
2  N 0  N0 

2
–u ⁄ 2
e
where Q ( u ) ≈ -------------- u»1
2πu
We assume the demodulated baseband interference,* I, is represented by AWGN. For
coherent PSK demodulation we have

* This is not strictly true since the noise is known. It is sufficient for the purposes of this discussion.
18 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

 E   Eb  S Rc
2  ---  ------
1 1 1
P b = --- erfc  -----b- = --- erfc  ---------------------------
- = --- erfc (2.25)
2  N 0 2  ( 2I ) ⁄ ( 4R c ) 2  I   R b

I
I eff = ------------------------ is referred to as the effective interference power.
2 ( Rc ⁄ Rb )
The effective interference power, in comparison with the signal power, determines the bit
error rate probability Pb of the SS system. Note that the effective interference power is reduced
by the ratio of the bandwidth expansion between the baseband signal and the transmitted signal,
(Rc /Rb).

2.5 Bit Scrambling

Referring to Table 2-1, we consider the following activities at a given transmitter location (see
Fig. 2-6).
1. An arbitrary data sequence si(t) is generated by a digital source.
2. An arbitrary code sequence ci(t) is generated by a direct spread (DS) generator.
3. Two sequences are modulo-2 added and transmitted to a distant receiver assuming there
is no propagation delay.
4. At the distant location, the resulting sequence (assuming no propagation delay) is
picked up by the receiver (see Fig. 2-7).
5. The code ci(t) used at the transmitter is also available at the receiver.
6. The original data sequence is recovered by modulo-2 adding the received sequence
with the locally available code ci(t).

Table 2-1 Operations with Modulo-2 Addition

Transmitter 1 si(t) 1 1 0 1 0 0 1 1 1 1
2 ci(t) 1 0 0 1 1 1 0 1 0 0
3 si(t) ci(t) 0 1 0 0 1 1 1 0 1 1
Receiver 4 si(t) ci(t) 0 1 0 0 1 1 1 0 1 1
5 ci(t) 1 0 0 1 1 1 0 1 0 0
6 si(t) ci(t) ci(t) = si(t) 1 1 0 1 0 0 1 1 1 1

si (t ) s i (t ) • ci (t )

ci (t )
Figure 2-6 Mobile Receiver
Bit Scrambling 19

Table 2-2 Operations without Modulo-2 Addition

Transmitter 1 si(t) –1 –1 1 –1 1 1 –1 –1 –1 –1
2 ci(t) –1 1 1 –1 –1 –1 1 –1 1 1
3 si(t) ci(t) 1 –1 1 1 –1 –1 –1 1 –1 –1
Receiver 4 si(t) ci(t) 1 –1 1 1 –1 –1 –1 1 –1 –1
5 ci(t) –1 1 1 –1 –1 –1 1 –1 1 1
6 si(t) ci(t) ci(t) = si(t) –1 –1 1 –1 1 1 –1 –1 –1 –1

Next, referring to Table 2-2, we consider the following set of activities at the given trans-
mitter location.
1. An arbitrary data sequence si(t) is generated by a digital source. In this case, we use +1s
and –1s to represent 0s and 1s (Fig. 2-6).
2. An arbitrary code sequence ci(t) is generated by a DS generator.
3. We multiply si(t) and ci(t). The output of the multiplier is transmitted to a distant
receiver.
4. At the distant location, the resulting sequence (again assuming no propagation delay) is
picked up by the receiver (Fig. 2-7).
5. The code c i (t) used at the transmitting location is assumed to be available at the
receiver.
6. The original data sequence is recovered by multiplying the received sequence by the
locally available code ci(t).
From Tables 2-1 and 2-2 we conclude that the modulo-2 addition using 1s and 0s binary
data is equivalent to multiplication using –1 and 1 binary data as long as we remain consistent in
mapping 0s to +1s and 1s to –1s as shown in Table 2-2. (For circuit implementation, the modulo-
2 addition is preferred since exclusive OR gates are cheaper than multiplication circuits. How-
ever, for modeling purposes, the multiplication method is usually easier to formulate and under-
stand than the modulo-2 approach).
We notice that, for the output of the receiver to be identical to the original data, the follow-
ing relationship must be satisfied:
si ( t ) ⋅ ci ( t ) ⋅ ci ( t ) = si ( t ) (2.26)

si (t ) • ci (t ) si (t ) • ci (t ) • ci (t ) si (t )

ci (t )
Figure 2-7 Mobile Receiver
20 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

In other words, ci(t) · ci(t) must be equal to unity. Note that ci(t) is a binary sequence made
up of 1s and –1s; therefore
if ci(t) is 1, ci(t) · ci(t) = 1 (2.27)
if ci(t) = –1, ci(t) · ci(t) = 1 (2.28)
In our previous discussion we assume that no propagation delay and no other processing
delay occurs between the transmitter and receiver input. Thus the code copy used at the
receiver is perfectly lined up with the initial code used at the transmitter. The two codes are said
to be in phase or in synchronization (synch). In practice, however, a propagation delay and
other processing delay occur between the transmitter and the receiver input. Therefore, the
receiver may be time shifted relative to the initial code at the transmitter, and the two codes are
no longer in synch. As a result, the output of the receiver will no longer be identical to the orig-
inal data, si(t).
In order to recover the original data si(t), we must tune the receiver code sequence to that of
the incoming code from the transmitter. In other words, we must time shift the receiver code in
order to line it up with the incoming code. It should be noted that, by synchronizing or tuning the
receiver code to the phase of the incoming code, the original data (shifted by propagation delay)
can now be recovered at the output of the receiver. In these examples, the data sequence and code
sequence have the same length (one code bit for each data bit) and are used for encrypting the
data bits. This is referred to as bit scrambling and does not result in spectrum spreading.

2.6 The Performance of a CDMA System

A traditional narrowband system based on FDMA or TDMA is a dimension-limited system. The
number of dimensions is determined by the number of nonoverlapping frequencies for FDMA or
by the number of time slots for TDMA. In a TDMA system, once all time slots are assigned, no
additional users can be added. Thus, it is not possible to increase the number of users beyond the
dimension limit without causing an intolerable amount of interference to a mobile station’s
reception at the cell-site receiver.
Spread spectrum systems can tolerate some interference, so the introduction of each addi-
tional active radio increases the overall level of interference to the cell-site receivers receiving
CDMA signals from mobile station transmitters. Each mobile station introduces a unique level of
interference that depends on its received power level at the cell site, its timing synchronization rel-
ative to other signals at the cell site, and its specific cross-correlation with other CDMA signals.
The number of CDMA channels in the network depends on the level of total interference
that can be tolerated in the system. Thus, the CDMA system is limited by interference, and the
quality of system design plays an important role in its overall capacity. A well-designed system
will have a required bit error probability with a higher level of interference than a poorly
designed system. Forward Error Correction (FEC) coding techniques improve tolerance for
interference and increase overall CDMA system capacity.
We assume that, at the cell site, the received signal level of each mobile user is the same and
that the interference seen by each receiver is modeled as Gaussian noise. Each modulation
The Performance of a CDMA System 21

method has a relationship that defines the bit error rate as a function of the Eb /N0 ratio. If we know
the performance of the coding methods used on the signals and the tolerance of the digitized
voice and the data-to-errors ratio, we can define the minimum Eb /N0 ratio for proper system oper-
ation. If we maintain operation at this minimum Eb /N0, we can obtain the best performance of the
system. The relationship between the number of mobile users, M, the processing gain, Gp, and the
Eb /N0 ratio is therefore given as

Gp
M ≈ --------------------- (2.29)
( Eb ⁄ N 0 )

For a given bit error probability, the actual Eb /N0 ratio depends on the radio system design
and error correction code. It may approach but never equal the theoretical calculations.
The best performance that can be obtained is defined by the Shannon limit* in AWGN. In
Eq. (2.2), if we note that

S 1 S 2 1 S 3 1 S 4
log e  1 + ---- = ---- – ---  ---- + ---  ---- – ---  ---- + … < ----
S S
 N  N 2  N 3  N 4  N N

then, from Eq. (2.2)

------ < -------------  ----

C 1 S
B w log e 2  N

and

1 E
------ < -------------  -----b-  ------
C C
B w log e 2  N 0  B w

Thus

E
-----b- ≥ log e 2 = 0.69 = – 1.59 dB (2.30)
N0

provides error-free communications.

For the Shannon limit, the number of users we can have is

Gp
M = ---------
- = 1.45G p (2.31)
0.69
This theoretical Shannon limit shows that CDMA systems can have more users per cell than tra-
ditional narrowband systems that are limited by number of dimensions. This limit is theoretical;
in practice a wireless system is typically engineered such that Eb /N0 ≈ 6 dB. However, due to
practical limitations on CDMA radio design, it is difficult to accommodate as many users in a

* This limit is a lower bound [7]. It is assumed that the channel coding has an infinite length to achieve this bound.
22 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

single cell as given by Eq. (2.29). The CDMA cell capacity depends upon many factors. As seen
by Eq. (2.29), the upper bound theoretical capacity of an ideal noise-free CDMA channel is lim-
ited by the processing gain Gp. In an actual system, the CDMA cell capacity is much lower than
the theoretical upper bound value. The CDMA cell capacity is affected by the receiver modula-
tion performance, power control accuracy, interference from other non-CDMA systems sharing
the same frequency band, and other effects.
CDMA transmissions in neighboring cells use the same carrier frequency and therefore
cause interference that we account for by introducing a factor β. This reduces the number of
users in a cell since the interference from users in other cells must be added to the interference
generated by the other mobiles in the user’s cell. The practical range for β is 0.4 to 0.55. The
power control accuracy is represented by a factor α. The practical range for α is 0.5 to 0.9. We
designate the reduction in the interference due to voice activity by a factor υ. The practical range
for υ is 0.45 to 1. If directional antennas are used rather than omnidirectional antennas at the
base station, the cell is sectorized with A sectors. Each of the antennas used at the cell radiates
into a sector of 360/A degrees and we have an interference improvement factor of λ. For a three-
sector cell, the practical value of the improvement factor λ is 2.55. The average values for β, α,
υ, and λ are 0.5, 0.85, 0.6, and 2.55, respectively.
Introducing β, α, υ, and λ into Eq. (2.29) we get

Gp 1 1
M ≈ ---------------- × ------------ × α × --- × λ (2.32)
Eb ⁄ N 0 1 + β υ

EXAMPLE 2.2
Estimate the number of mobile users that can be supported by a CDMA system using an RF band-
width of 1.25 MHz to transmit data at 9.6 kbps. Assume: Eb /N0 = 6 dB; the interference from neigh-
boring cells β = 60%; the voice activity factor υ = 50%; the power control accuracy factor α = 0.8.
6
1.25 × 10
G p = ------------------------
3
- = 130
9.6 × 10

E
-----b- = 6 dB = 3.98
N0

130 1 1
M = ---------- × ---------------- × ------- × 0.8 = 32.64 ≈ 33 mobile users per sector
3.98 1 + 0.6 0.5

The results of this example can be compared with the capacity of an analog FM system with the
same frequency allocation, i.e., 41 FM channels. Typically, an analog system is engineered with
a frequency reuse pattern equal to 7. With a three-sector configuration and a reuse factor of 7, the
number of channels per sector equals 41/(7 × 3) ≈ 2. This comparison suggests that a DSSS sys-
tem offers a greater than tenfold improvement in the channel capacity. It is interesting to note
The Performance of a CDMA System 23

that the processing gain of a DSSS system is directly proportional to spectrum expansion while
the processing gain of an FM system is proportional to the square of the frequency expansion.*
This would seem to imply that the FM system should perform better than the CDMA system; yet
it doesn’t. There are several reasons for this CDMA performance:
• DSSS techniques take advantage of the voice activity.
• DSSS techniques use the concept of orthogonality to multiple users on a common fre-
quency channel. This concept is applicable across different base stations and sectors.
• DSSS techniques synchronize transmission for all base stations so that soft handoffs
(see chapter 10) can be implemented. This approach reduces the level of interference.

EXAMPLE 2.3
For the CDMA system (TIA IS-95), a chip rate† of 1.2288 Mcps is specified for the data rate of 9.6
kbps. Eb /N0 is taken as 6.8 dB. Estimate the average number of subscribers that can be supported by
a sector of the 3-sector cell. Assume: interference from neighboring cells β = 50%; the voice activity
factor υ = 60%; the power control accuracy factor α = 0.85; and the improvement from sectorization
λ = 2.55.
Gp 1 1
M ≈ ---------------
- × ------------ × α × --- × λ
Eb ⁄ N 0 1 + β υ

( 1 ⁄ 9.6 )
G p = ------------------------------------------
3
= 128 , E b ⁄ N 0 = 6.8 dB = 4.7863
( 1 ⁄ 1.2288 × 10 )
128 1 1
M = ---------------- × ------- × ------- × 0.85 × 2.55 = 64.4
4.7863 1.5 0.6
64.4
Subscriber/Sector = ---------- = 21.46 ≈ 21
3

EXAMPLE 2.4
A total of 40 equal-power mobile stations are to share a frequency band through a CDMA system.
Each mobile station transmits information at 9.6 kbps with a DSSS BPSK modulated signal. Calcu-
late the minimum chip rate of the psudorandom noise (PN) code in order to maintain a bit error
probability of 10–3. Assume: the interference factor β from the other base stations = 60%; voice
activity υ = 50%; and power control accuracy factor α = 0.8. What will chip rate be if the probability
of error is 10–4?
–E ⁄ N
2E e b 0
P b = Q  --------b- ≈ --------------------------------
 N0  2 π( E ⁄ N )
b 0

* For an FM system, the frequency expansion is specified by the deviation ratio.

† The chip rate is the frequency of the code clock.
24 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

–E ⁄ N
e b 0 –3
-------------------------------- = 10
2 π ( Eb ⁄ N 0 )

E
-----b- ≈ 4.8 = 6.8 dB
N0

Gp 1 1
- × ------------ × --- × α
M = --------------------
( Eb ⁄ N 0 ) 1 + β υ

G 1 1
------p- × ------- × ------- × 0.8 = 40
4.8 1.6 0.5
∴G p = 192
3
∴ R c = 192 × 9.6 × 10 = 1.843 Mcps

–4 E
For P b = 10 , -----b- = 8.43 dB = 6.9663
N0

∴G p = 278.6 and R c = 2.675 Mcps

2.7 Pseudorandom Noise Sequences

In CDMA systems, PN sequences are used to
• Spread the bandwidth of the modulated signal to the larger transmission bandwidth.
• Distinguish between the different user signals by utilizing the same transmission band-
width in the multiple access scheme.
PN sequences are not random; they are deterministic, periodic sequences. The following
are the three key properties of an ideal PN sequence [2]:
1. The relative frequencies of 0 and 1 are each 1/2.
2. The run lengths (of 0s or 1s) are: 1/2 of all run lengths are of length 1; 1/4 are of length
2; 1/8 are of length 3; and so on.
3. If a PN sequence is shifted by any nonzero number of elements, the resulting sequence
will have an equal number of agreements and disagreements with respect to the original
sequence.
PN sequences are generated by combining the outputs of feedback shift registers. A feed-
back shift register consists of consecutive two-stage memory or storage stages and feedback
logic. Binary sequences are shifted through the shift register in response to clock pulses. The
contents of the stages are logically combined to produce the input to the first stage. The initial
contents of the stages and feedback logic determine the successive contents of the stages. A
feedback shift register and its output are called linear when the feedback logic consists entirely
of modulo-2 adders.
Pseudorandom Noise Sequences 25

To demonstrate the properties of a PN a binary sequence, we consider a linear feedback

shift register (see Fig. 2-8) that has a four-stage register for storage and shifting, a modulo-2
adder, and a feedback path from adder to the input of the register. The operation of the shift
register is controlled by a sequence of clock pulses. At each clock pulse the contents of each
stage in the register is shifted by one stage to the right. Also, at each clock pulse the contents of
stages X3 and X4 are modulo-2 added, and the result is fed back to stage X1. The shift register
sequence is defined to be the output of stage X4. We assume that stage X1 is initially filled with
a 0 and the other remaining stages are filled with 0, 0, and 1; i.e., the initial state of the register
is 0 0 0 1. Next, we perform the shifting, adding, and feeding operations, where we obtain the
results after each cycle that is shown in Table 2-3.
We notice that the contents of the registers repeat after 24 – 1 = 15 cycles. The output
sequence is given as 0 0 0 1 0 0 1 1 0 1 0 1 1 1 1 (see Fig. 2-9) where the left-most bit is the ear-
liest bit. In the output sequence, the total number of 0s is 7 and total number of 1s is 8; the num-
bers differ by 1.
If a linear feedback shift register reached the 0 state at some time, it would always remain in
the 0 state and the output sequence would subsequently be all 0s. Since there are exactly 2n – 1
nonzero states, the period of a linear n-stage shift register output sequence cannot exceed 2n – 1.
The output sequences are classified as either maximal length or nonmaximal length. Maxi-
mal-length sequences are the longest sequences that can be generated by a given shift register of
a given length. In the binary shift register sequence generators, the maximal length sequence is
2n – 1 chips, where n is the number of stages in the shift registers. Maximal-length sequences
have this property for an n-stage linear feedback shift register: the sequence repetition period in

+
Modulo-2 Adder

X1 X2 X3 X4 Output

Clock
Figure 2-8 Four-Stage Linear Feedback Shift Register
26 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

Table 2-3 Results of Shifting after Each Cycle

Shift Stage X1 Stage X2 Stage X3 Stage X4 Output Sequence
0 0 0 0 1 1
1 1 0 0 0 0
2 0 1 0 0 0
3 0 0 1 0 0
4 1 0 0 1 1
5 1 1 0 0 0
6 0 1 1 0 0
7 1 0 1 1 1
8 0 1 0 1 1
9 1 0 1 0 0
10 1 1 0 1 1
11 1 1 1 0 0
12 1 1 1 1 1
13 0 1 1 1 1
14 0 0 1 1 1
15 0 0 0 1 1
16 1 0 0 0 0

Period of Output Period of Output

Waveform Waveform

TC = Time Chip

+1

-1

T 0 = (2n -1) TC T0 = (2n -1) TC

Figure 2-9 Output Waveform for Four-Stage Linear Feedback Shift Register
Pseudorandom Noise Sequences 27

clock pulses is T0 = 2n – 1. If a linear feedback shift register generates a maximal sequence, then
all of its nonzero output sequences are maximal, regardless of the initial stage. A maximal
sequence contains (2n – 1 – 1) 0s and (2n – 1) 1s per period.

2.7.1 Properties of a Maximal-Length Pseudorandom Sequence

When an n-stage shift register (see Fig. 2-10) is configured to generate a maximal-length
sequence, the sequence has the following properties:
• The number of binary 0s differs from number of 1s by one chip at most. The number of
binary 1s is 2n – 1 and the number of 0s is 2n – 1 – 1, where n is the number of stages in
the code generator, and the code length is 2n – 1 chips.
• A run is defined as a sequence of a single type of binary digits. The appearance of the
alternate digit in a sequence starts a new run. The length of the run is the number of dig-
its in the run. The statistical distribution of 1s and 0s is well defined and always the
same. Relative positions of the runs vary from code sequence to code sequence, but the
number of each run length does not.
• A modulo-2 addition of a maximal linear code with a phase-shifted replica of itself
results in another replica with a phase shift different from either of the originals.
• If a period of the sequence is compared term by term with any cyclic shift itself, it is
best if the number of agreements differs from the number of disagreements by not more
than one count.
• If we transform the binary (0,1) sequence of the shift register output to a binary (+1, –1)
sequence by replacing each 0 by +1, and each 1 by –1, then the periodic correlation
function of the sequence is given by

+ + + +

h1 h2 hn -2 hn -1

x n -1 x n -2 x1 x0

Note: h i = 1 represents a closed circuit

h i = 0 represents an open circuit

Figure 2-10 n-Stage Linear Feedback Shift Register

28 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

n
2 – 1, τ = 0
θ(τ) =
–1 , τ ≠ 0

where τ = the shift in increments of one chip (see Figure 2-11 as an example of n =
3), and
n = the number of stages in the shift register.
In the region between τ = 0 and τ = ±1, the correlation function decreases linearly
from 2n – 1 to –1 so that the autocorrelation function for a maximal-length pseudoran-
dom sequence is triangular with a maximum value at τ = 0 (see Example 2.5). With
this property, two or more communicators can operate independently if their codes are
phase shifted more than one chip. For other code sequences, the autocorrelation proper-
ties may be markedly different than the properties of the maximal-length sequences.
• Every possible state of a given n-stage generator exists at some time during the genera-
tion of a complete code cycle. Each state exists for one and only one clock interval. The
exception is that the all-0s state does not normally occur and is not allowed to occur.
It has been shown [11] that there are exactly 2n – (p + 2) runs of length p for both 1s and 0s in
every maximal sequence (except that there is only one run containing n 1s and one containing
(n – 1) 0s; there are no runs of 0s of length n or 1s of length (n – 1). The distribution of runs for
(24 – 1) chip sequence is given in Table 2-4.
Whether an n-stage linear feedback shift register generates only one sequence with period
n
2 – 1 depends upon its connection vector (see Fig. 2-10). Let h(x) be the nth-order polynomial
given by
2 n
h ( x ) = h0 + h1 x + h2 x + … + hn x (2.33)

We refer to h(x) as the associated polynomial of the shift register with feedback coefficient (h0,
h1, h2, … hn). Here h0 = hn = 1 and other feedback coefficients take values 0 and 1. Thus, the
polynomial for the four-stage linear feedback shift register as shown in Fig. 2-8 is given by
3 4
h( x ) = 1 + x + x (2.34)

Table 2-4 Distribution of Runs for a 24 – 1 Chip Sequence

Run Length 1s 0s Number of Chips Included
1 2 2 1×2+1×2=4
2 1 1 1×2+1×2=4
3 0 1 0×3+1×3=3
4 1 0 1×4+0×4=4
Total No. of Chips 15
Pseudorandom Noise Sequences 29

When h(x) is an irreducible (not factorable) primitive polynomial of degree n, then all
sequences generated by h(x) have a maximum period of 2n – 1. For an n-stage register, there are
Np(n) maximal sequences that can be generated [11]. Np(n) is the number of primitive polyno-
mial of degree n.

n k
2 –1 P –1
N p ( n ) = --------------
n ∏ -------------
i
Pi
- (2.35)
i=1

where Pi is the prime decomposition of 2n – 1.

Table 2-5 gives the number of maximal sequences available from register lengths 2
through 10 and provides an example of a primitive polynomial of degree n.

2.7.2 Autocorrelation
The autocorrelation function for a signal x(t) is defined as
∞

Rx(τ) = ∫ x ( t )x ( t + τ ) dt (2.36)
–∞

Table 2-5 Number of Maximal Sequences Available from Register Lengths 2 Through 10
Prime Example of Primitive
No. of Decomposition No. of n-sequence Polynomial of degree
Stage n 2n – 1 of 2n – 1 Np(n) n h(x)
3 2 2
2 3 3 --- ⋅ --- = 1 1+x+x
2 3
7 6 3
3 7 7 --- ⋅ --- = 2 1+x+x
3 7
3×5 15 2 4 1+x+x
4
4 15 ------ ⋅ --- ⋅ --- = 2
4 3 5
31 30 2 5
5 31 31 ------ ⋅ ------ = 6 1+x +x
5 31
3×3×7 63 2 6 6
6 63 ------ ⋅ --- ⋅ --- = 6 1+x+x
6 3 7
127 126 3 7
7 127 127 --------- ⋅ --------- = 18 1+x +x
7 127
3 × 5 × 17 255 2 4 16 2 3 4 8
8 255 --------- ⋅ --- ⋅ --- ⋅ ------ = 16 1+x +x +x +x
8 3 5 17
7 × 73 511 6 72 4 9
9 511 --------- ⋅ --- ⋅ ------ = 48 1+x +x
9 7 73
3 × 11 × 31 1023 2 10 30 3 10
10 1023 ------------ ⋅ --- ⋅ ------ ⋅ ------ = 60 1+x +x
10 3 11 31
30 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

Autocorrelation refers to the degree of correspondence between a sequence and a phase-

shifted replica of itself. An autocorrelation plot shows the number of agreements minus dis-
agreements for the overall length of the two sequences being compared, as the sequences assume
every shift in the field of interest. If x(t) is a periodic pulse waveform representing a PN
sequence, we refer to each fundamental pulse as a PN sequence symbol or a chip. For such a PN
waveform of unit chip duration and period T0 = 2n – 1 chips, the normalized autocorrelation
function is expressed as
1
R x ( τ ) = ----- [number of agreements – number of disagreements in a comparison of one full
T0
period of sequence with a τ position cyclic shift of the sequence]
The normalized autocorrelation function Rx(τ) of a periodic waveform x(t) with period T0
is given as
T0 ⁄ 2
1 1
R x ( τ ) = -------------- -----
Rx(0) T 0 ∫ x ( t )x ( t + τ ) dt for – ∞ < τ < ∞ (2.37)
–T 0 ⁄ 2

T0 ⁄ 2
1
∫
2
where R x ( 0 ) = ----- x ( t ) dt
T0
–T 0 ⁄ 2

2.7.3 Cross-Correlation
The cross-correlation function between two signals, x(t) and y(t), is defined as the correla-
tion between two different signals x(t) and y(t) and is given as

T0 ⁄ 2

Rc ( τ ) = ∫ x ( t )y ( t + τ ) dt for – ∞ < τ < ∞ (2.38)

–T 0 ⁄ 2

EXAMPLE 2.5
Consider a three-stage shift register generator that is generating a seven-chip maximal linear code. The
reference sequence is 1 1 1 0 0 1 0. Sketch the autocorrelation function if the chip rate is 10 Mcps.
Table 2-6 provides the sequence after each shift and shows the corresponding agreements (A)
and disagreements (D) with the reference sequence.
It can be noted that the net correlation A – D is –1 for all shifts except for the 0-shift or syn-
chronous condition. This is typical of all n-sequences. In the region between 0 and plus or minus one
chip shift (τ = ±1/106 seconds), the correlation increases linearly so that the autocorrelation function
for an n-sequence is triangular as shown in Fig. 2-11. This characteristic of autocorrelation is used to
great advantage in communication systems. A channel can simultaneously support multiple users if
the corresponding codes are phase shifted more than one chip.
Pseudorandom Noise Sequences 31

Table 2-6 Agreements and Disagreements with Reference Sequence

Shift Sequence Agreement (A) Disagreement (D) A–D
1 0111001 3 4 –1
2 1011100 3 4 –1
3 0101110 3 4 –1
4 0010111 3 4 –1
5 1001011 3 4 –1
6 1100101 3 4 –1
0 1110010 7 0 7

R (t )

+1

0
t
- 1/7

Figure 2-11 Autocorrelation of a Three-Stage Linear Feedback Shift Register

2.7.4 Orthogonal Functions

Orthogonal functions are employed to improve the bandwidth efficiency of an SS system.
Each mobile user uses one member of a set of orthogonal functions representing the set of sym-
bols used for transmission. While there are many different sequences that can be used to generate
an orthogonal set of functions, the Walsh and Hadamard sequences make useful sets for CDMA.
Two different methods can be used to modulate the orthogonal functions into the informa-
tion stream of the CDMA signal. The orthogonal set of functions can be used as the spreading
code or to form modulation symbols that are orthogonal.
32 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

With orthogonal symbol modulation, the information bit stream can be divided into
blocks, with each block representing a nonbinary information symbol that is associated with a
particular transmitted code sequence. If there are b bits per block, one of the set of K = 2b func-
tions is transmitted in each symbol interval. The signal at the receiver is correlated with the set
of K matched filters, each matched to the code function of one symbol. The outputs from corela-
tors are compared and the symbol with the largest output is taken as the transmitted symbol.
If we assume a simple one-path channel with perfect power control and negligible additive
noise, and if we include the interference due to multipath, multiple users, and the decision pro-
cess of the correlators, the Eb /N0 ratio can be given as [8]

E Gp
-----b- ≈ ------------------------------------------
- (2.39)
N0 ( M – 1) + (K – 1)

where M = number of mobile users,

Gp = processing gain of the system, and
K – 1 = noise from the outputs of correlators other than one corresponding to the correct
symbol.
We rewrite Eq. (2.39) as

Gp
M = ---------------- –K+2 (2.40)
Eb ⁄ N 0

Next we introduce factors β, α, υ, and λ (see Section 2.6) in Eq. (2.40) to get

Gp 1 1
M ≈ ---------------- × ------------ × α × --- × λ – K + 2 (2.41)
Eb ⁄ N 0 1 + β υ

MR M ⋅ log 2 K R s M ( log 2 K )
η = --------- = -----------------------------
- = ------------------------
- (2.42)
Bw Bw Gp

where η = bandwidth efficiency, and

Rs = symbol transmission rate.

EXAMPLE 2.6
Calculate the bandwidth efficiency of the system using the data in Example 2.2 and assuming an
orthogonal code with K = 2 symbols. If an orthogonal code with K = 16 symbols is used for the sys-
tem, how many simultaneous mobile users can be supported and what is the bandwidth efficiency of
the system?
• K = 2 symbols
6
1.25 × 10 E
- = 130 , -----b- = 6 dB = 3.98
G p = ------------------------
3
9.6 × 10 N0
Pseudorandom Noise Sequences 33

130 1 1
M ≈ ---------- × ---------------- × ------- × 0.8 – 2 + 2 = 32.64 ≈ 33 users
3.98 1 + 0.6 0.5

M ( log 2 K ) 33 ( log 2 2 )
η = ------------------------
- = -----------------------
- = 25.4 %
Gp 130

• K = 16 symbols
130 1 1
M ≈ ---------- × ---------------- × ------- × 0.8 – 16 + 2 = 18.64 ≈ 19 users
3.98 1 + 0.6 0.5

M ( log 2 K ) 19 ( log 2 16 )
η = ------------------------
- = --------------------------
- = 58.5%
Gp 130

The bandwidth efficiency of the system is improved by 33.1%. The disadvantage of the
orthogonal signaling scheme is the complexity of the receiver design. In this example, we need 16
receiver correlators per user channel instead of only one required in the simplest design.

The TIA IS-95 CDMA system uses orthogonal functions for the spreading code on the
forward channel and orthogonal functions for the modulation on the reverse channel.* One of 64
possible modulation symbols is transmitted for each group of 6 code symbols. The modulation
symbol is one member of the set of 64 mutually orthogonal functions. The orthogonal functions
have the following characteristic:

M–1

∑ φi ( kτ )φ j ( kτ ) = 0 i≠ j (2.43)
k=0

where φi (kτ) and φj (kτ) are the ith and jth orthogonal members of an orthogonal set,
M is the length of the set, and
τ is the symbol duration.
Walsh functions are generated by code-word rows of special square matrices called Had-
amard matrices. These matrices contain one row of all 0s, with the remaining rows each having
an equal number of 1s and 0s. Walsh functions can be constructed for block length N = 2j, where
j is an integer.
The TIA IS-95 CDMA system uses a set of 64 orthogonal functions generated by using
Walsh functions. The modulated symbols are numbered from 0 through 63.
The 64 × 64 matrix can be generated by using the following recursive procedure:

H1 = 0 H2 = 0 0 (2.44)
01

* The IS-665 wideband CDMA system uses the orthogonal codes for spreading in both directions; see chapter 8.
34 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

0 0 0 0
HN HN
H4 = 0 1 0 1 H2 N = (2.45)
0 0 1 1 HN HN
0 1 1 0

where N is a power of 2 and H N is the negative of H N .

The period of time needed to transmit a single modulation symbol is called a Walsh sym-
bol interval and is equal to 1/4800 second (208.33µs). The period of time associated with 1/64 of
the modulation symbol is referred to as a Walsh chip and is equal to 1/307,200 second (3.255µs).
Within a Walsh symbol, Walsh chips are transmitted in the order 0, 1, 2, …, 63.
For the forward channel, Walsh functions (Figs. 2-12 and 2-13) are used to eliminate multiple
access interference among users in the same cell. On downlink, all Walsh functions are synchro-
nized in the same cell and have zero correlation between each other. The following steps are used:
• The input user data (e.g., digital speech) is multiplied by an orthogonal Walsh function
(TIA IS-95 standard uses the first 64 orthogonal Walsh functions).
• The user data is then spread by the BS pilot PN code and transmitted on the carrier.
• At the receiver, after removing the coherent carrier, the mobile receiver multiplies the
signal by the synchronized PN code (associated with the base station).

Walsh 1 PN Code for Base Station

User # 1, s 1 (t )
X X

.
.
. Σ X
Walsh N PN Code for Base Station
cos (2 π fc t )
User # N, sN (t )
X X
Figure 2-12 Applications of Walsh Functions and Offset Code at the Base Station
Pseudorandom Noise Sequences 35

PN

X X X
Sample at
t=T
Coherent Carrier Walsh i
cos (2 π fc t )

Figure 2-13 Applications of Walsh Functions and Offset Code in the Mobile Station

• The multiplication by the synchronized Walsh function for the ith user eliminates the
interference due to transmission from the BS to other users.
The Walsh functions form an ordered set of rectangular waveforms taking only two ampli-
tudes: +1 and –1. They are defined over a limited time interval TL, known as the time base. If φi
represents the ith Walsh function and TL is the time base, then
TL
1
TL ∫ i
------ φ ( t )φ j ( t ) dt = 0 for i ≠ j (2.46)
0

and
TL
1
TL ∫ i
2
------ φ ( t ) dt = 1 for all is (2.47)
0

To correlate the Walsh codes at the receiver requires that the receiver be synchronized with
the transmitter. In the forward direction the base station can transmit a pilot signal to enable the
receiver to recover synchronization. Walsh symbol modulation is used from the mobile station to
the base station.

EXAMPLE 2.7
We consider a case where 8 chips are used per bit to generate the Walsh functions. Specify these
functions, sketch them, and show that they are orthogonal to each other.
36 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

φ
1
0 0 0 0 0 0 0 0 φ2
0 1 0 1 0 1 0 1 φ3
0 0 1 1 0 0 1 1
H4 H4
0 1 1 0 0 1 1 0 φ4
H8 = = =
H4 H4 0 0 0 0 1 1 1 1 φ5
0 1 0 1 1 0 1 0
φ6
0 0 1 1 1 1 0 0
0 1 1 0 1 0 0 1 ϕ7
φ8

Figure 2-14 shows the sketches of the eight Walsh functions. We consider φ2 and φ4 to show
orthogonality.
1 1
TL 2∫
------- φ ( t )φ ( t ) dt = ------- [ – 1 × – 1 + 1 × 1 + 1 × – 1 + 1 × ( – 1 ) + ( – 1 ) × ( – 1 ) + 1 × 1 + 1 × – 1 + 1 × – 1 ] = 0
4 TL

T/4 T/2 3T/4 T T/4 T/2 3T/4 T

+1 +1

φ1
φ5
-1 -1

T/4 T/2 3T/4 T T/4 T/2 3T/4 T

+1 +1

φ2
φ6
-1 -1

T/4 T/2 3T/4 T T/4 T/2 3T/4 T

+1 +1

φ3
φ7
-1 -1

T/4 T/2 3T/4 T T/4 T/2 3T/4 T

+1 +1

φ4
φ8
-1 -1

Figure 2-14 Plots of Walsh Functions

Summary 37

and
1 2 1
T L∫ 1
------ φ ( t ) dt = ------ [ T L ] = 1
TL
Similarly, we can show that all eight Walsh functions are orthogonal to each other.

EXAMPLE 2.8
We consider a case where 8 chips are used per bit to generate the Walsh functions. Stations A, B, C,
and D are assigned the chip sequence 0 1 0 1 0 1 0 1, 0 0 1 1 0 0 1 1, 0 1 1 0 0 1 1 0, and 0 0 0 0 1 1 1
1, respectively. The stations use the chip sequence to send a 1 bit and negative chip sequences to send
a 0 bit (e.g., station A uses 1 0 1 0 1 0 1 0 to send the 0 bit and so on). All chip sequences are pairwise
orthogonal. This implies that the normalized correlation of any two distinct chip sequences is 0 and
the normalized correlation of any chip sequence with itself is 1. We assume that all stations are syn-
chronized in time, so all chip sequences begin at the same instant. When two or more stations transmit
simultaneously, their bipolar signals add linearly. For example, if in one chip period three stations
output +1 and one station outputs –1, the net result is +2. We consider five different cases when one or
more stations transmit. We want to show that the receiver recovers the bit stream of station C by com-
puting the normalized inner products of the received sequences with the chip sequence of station C.
Chip Sequence Binary Values of Chip Sequence
A: 0 1 0 1 0 1 0 1 A: (–1 +1 –1 +1 –1 +1 –1 +1)
B: 0 0 1 1 0 0 1 1 B: (–1 –1 +1 +1 –1 –1 +1 +1)
C: 0 1 1 0 0 1 1 0 C: (–1 +1 +1 –1 –1 +1 +1 –1)
D: 0 0 0 0 1 1 1 1 D: (–1 –1 –1 –1 +1 +1 +1 +1)
Normalized inner products are
( S1 ⋅ C ) ⁄ 8 = ( 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 ) ⁄ 8 = 1

( S2 ⋅ C ) ⁄ 8 = ( 2 + 0 + 0 + 2 + 0 + 2 + 2 + 0 ) ⁄ 8 = 1

( S3 ⋅ C ) ⁄ 8 = ( 3 + 1 + 1 – 1 + 3 + 1 + 1 – 1 ) ⁄ 8 = 1

( S4 ⋅ C ) ⁄ 8 = ( 2 + 0 + 0 – 2 + 2 + 0 + 0 – 2 ) ⁄ 8 = 0

( S5 ⋅ C ) ⁄ 8 = ( 1 – 1 – 1 – 3 + 1 – 1 – 1 – 3 ) ⁄ 8 = –1
Thus, the receiver recovers 1 1 1 - 0 bit sequence for station C.
We assumed that all the chips are synchronized in time. In a real situation, this is impossible.
The sender and receiver are synchronized by having the sender transmit a known chip sequence long
enough for the receiver to lock onto. All other (unsynchronized) transmissions are then seen as ran-
dom noise.

2.8 Summary
In this chapter we considered the concept of spread spectrum systems and provided the main
features of the direct sequence spread spectrum system used in the IS-95 system. A key compo-
nent of spread spectrum performance is the calculation of processing gain of the system, which
38 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

Table 2-7 Five Cases of One or More Stations Transmitting

Stationa (A, B, C, D) Transmitting Received Chip Sequence

– – 1– C S1 = (–1+1+1–1–1+1+1–1)
––11 C+D S2 = (–2 0 0 –2 0 +2 +2 0)
111– A+B+C S3 = (–3 +1 +1 +1 –3 + 1+ 1+1)
11–– A+B S4 = (–2 0 0 +2 –2 0 0 +2)
110– A+B+C S5 = (–1 –1 –1 +3 –1 –1 –1 +3)
a
Note: a dash (–) means no transmission by that station.

is the relationship between the input and output SNR of a spread spectrum receiver. We used the
relationship to present some examples that evaluate the performance of a CDMA spread spec-
trum system.
We presented the Shannon equation for error-free communications and used it to show
that error-free communication is possible (with high delays) for an energy-per-bit-to-noise-
density ratio, Eb /N0 = –1.59 dB. SS systems trade bandwidth for processing gain, and code divi-
sion systems use a variety of orthogonal or almost orthogonal codes to allow multiple users in
the same bandwidth. Thus, CDMA systems can have a higher capacity than either analog or
TDMA digital systems. However, because of practical constraints on CDMA systems, it is not
possible to achieve the Shannon bound in system design. The upper bound of the capacity of a
CDMA system is limited by the processing gain of the system. In an actual system, the capacity
is lower than the theoretical upper bound. CDMA capacity is affected by receiver modulation
performance, power control accuracy, interference from other cells, voice activity, cell sector-
ization, and the ability to maintain synchronization of the systems. Practical CDMA systems
are designed for a value of Eb /N0 = 6–7 dB.

2.9 Problems
1. A total of 18 equal-power mobile users per cell are to share a frequency band through a
CDMA system. Each mobile user transmits data at 19.2 kbps with a DSSS QPSK-
modulated signal. Calculate the minimum chip rate of the PN code in order to maintain
a bit error probability of 10–4. Assume that the interference factor β from the other base
stations = 0.60; power control accuracy α = 0.8; gain from 3-sector antenna λ = 2.55.
What will the chip rate be if the probability of the bit error is 10–6?
2. Consider a case where 16 chips per bit are used to generate the Walsh functions. Spec-
ify the functions and sketch W0, W8, W12, and W15. Show that these Walsh functions are
orthogonal.
3. For a three-stage linear shift register generator, how many maximal-length PN
sequences are generated? What is the location of the modulo-2 adder for each
sequence? What is the period of the maximal-length sequence?
References 39

4. Consider a four-stage linear shift register generator with the initial state of the register
1001 (see Fig. 2-15). Show that it generates a maximal-length sequence. Demonstrate
the properties of maximal-length sequence. Sketch the autocorrelation function. What
is the location of the modulo-2 adder for the other maximal-length sequence?

Modulo-2 Adder

X1 X2 X3 X4 output

Figure 2-15 Four-Stage Register (Problem 4)

5. Estimate the number of mobile users that can be supported by a CDMA three-sector
cell using an RF bandwidth of 1.23 MHz and a chip rate of 1.2288 Mcps to transmit
Rate Set 2(RS2) information at 14.4 kbps. Assume (Eb /N0)reqd = 7 dB; the interference
from neighboring cells β = 0.67; the voice activity factor υ = 0.6; power control accu-
racy factor α = 0.8; gain from three-sector antenna λ = 2.55.
6. We consider a case where 8 chips per bit are used to generate the Walsh functions.
Mobile stations A, B, C, and D are assigned W2 , W5 , W6 , and W7 , respectively. The sta-
tions use the Walsh sequence to send a 1 binary bit and a negative Walsh to transmit a 0
binary bit. Assuming all stations are synchronized in time, the chip sequences begin at
the same instant. When two or more stations transmit simultaneously, their bipolar sig-
nals are added linearly. Considering the following four cases when one or more stations
transmit, show that the receiver recovers the bit stream of stations B and C.

Station (A, B, C, D)
(–) means no transmission by that station Transmitting Stations
––10 C+D
111– A+B+C
11–– A+B
110– A+B+C

2.10 References
1. Bhargava, V., Haccoum, D., Matyas, R., and Nuspl, P., Digital Communications by Satellite,
John Wiley & Sons, New York, 1981.
2. Dixon, R. C., Spread Spectrum Systems, Ed., John Wiley & Sons, New York, 1984.
40 Ch. 2 • Direct Sequence Spread Spectrum and Spreading Codes

3. Feher, K., Wireless Digital Communications Modulation and Spread Spectrum Applications,
Prentice Hall PTR, Upper Saddle River, NJ, 1995.
4. Garg, V. K., and Wilkes, J. E., Wireless and Personal Communications Systems, Prentice
Hall, Upper Saddle River, NJ, 1996.
5. Lee, W. C. Y., Mobile Cellular Telecommunication Systems, McGraw-Hill, New York, 1989.
6. Pahlwan, K., and Levesque, A. H., Wireless Information Networks, John Wiley & Sons, New
York, 1995.
7. Shannon, C. E., “Communications in the Presence of Noise,” Proceedings of the IRE, no.
37, 1949, pp. 10–21.
8. Skalar, B., Digital Communications—Fundamental & Applications, Prentice Hall, Engle-
wood Cliffs, NJ, 1988.
9. Steele, R., Mobile Radio Communications, IEEE Press, New York, 1992.
10. Torrien, D., Principle of Secure Communication Systems, Artech House, Boston, 1992.
11. Virterbi, A. J., CDMA, Addison-Wesley Publishing Company, Reading, MA, 1995.
12. Viterbi, A. J., and Padovani, Roberto, “Implications of Mobile Cellular CDMA,” IEEE
Communication Magazine 30(12), 1992, pp. 38–41.
 C H A P T E R 3

Speech and Channel

Coding

3.1 Introduction
In this chapter, we consider the speech and channel coding applications in the IS-95 CDMA sys-
tem (the data application is discussed in chapter 14). Since speech encoding is critical to digital
transmission and since it is the primary application needed for most wireless phone uses, we will
focus on the speech coding algorithms used by CDMA. CDMA uses an 8-kbps (or a 13.3-kbps)
data rate for voice transmission.
The wireline network is based on voice transmission using digital Pulse Code Modulation
(PCM) at 64 kbps and data transmission at rates of 64 kbps or multiples of 64 kbps. Many older
analog facilities still exist, especially in residential areas, and these use voice band modems at
rates of up to 28.8 kbps for data and analog electrical signals for voice. At the central office the
analog voice and analog data are converted to digital signals using PCM or, optionally, using
modem pools for data.
It would be optimal if identical systems could be used for all wireless communications.
Unfortunately error rates on radio channels are many orders of magnitude higher than those of
copper or fiber-optic cables. In addition, PCM is inefficient for use over scarce and expensive
radio channels.
CDMA systems use an efficient method of speech coding and extensive error recovery
techniques to overcome the harsh nature of the radio channel. The CDMA system uses a Code-
Excited Linear Predictor (CELP) speech coding system at 9.6 kbps (optionally at 13.3 kbps).

3.2 Speech Coding

3.2.1 Pulse Code Modulation
The simplest form of waveform coding scheme is linear PCM, in which the speech signal
is band limited, compressed, sampled, quantized, and encoded (see Fig. 3-1). This approach is
widely used for analog-to-digital conversion of a signal. In radio and telephone communications,

41
42 Ch. 3 • Speech and Channel Coding

Input Output
Analog 64-kbps
Speech PCM

Compressor Sampling Digital-to-

Band-
µ-law 8000 Analog
Limited
or samples Converter
Filter
A-law per second (8 bits)

Figure 3-1 PCM Encoder

it is not necessary to send the entire 20-Hz to 20,000-Hz signal normally used for high-fidelity
music. Intelligible speech communications can occur with a much narrower and therefore more
efficient range of frequencies. For telephone communications, the speech signal is band limited
to a frequency range of 300 to 3,300 Hz. To achieve telephone-quality speech, 12 bits per sample
are required at a sampling rate of 8,000 samples per second. However, by using a logarithmic
sampling system, 8 bits per sample are sufficient. Each sample is then quantized into one of 256
levels. Telephone speech uses two widely different variations of PCM to achieve quality speech
(µ-law and A-law PCM). Both are based on a nonuniform quantization of the signal amplitude
according to a logarithmic scale rather than a linear scale. Such coders utilize the static character-
istics of amplitude nonstationary in speech to achieve good quality at a bit rate of 64 kbps. This is
the basis of PCM.
The decoder for PCM (Fig. 3-2) inverts the stages of the encoding process. PCM encoding
and decoding are inherently simple systems. However, they require a high bit rate for transmission.
For PCM, North America and Japan use µ-law encoding where the output digital signal,
s(t), is related to the input signal, i(t), by

ln ( 1 + µ i ( t ) )
s ( t ) = sgn [ i ( t ) ] -----------------------------------, –1 ≤ i ( t ) ≤ 1 (3.1)
ln ( 1 + µ )
where a typical value for µ = 255 is used in the United States. In Eq. (3.1), the input signal is
normalized to a range of ±1. It can be noted that, for small i(t), s(t) approaches a linear function;
for large i(t), s(t) approaches a logarithmic function. The purpose of µ-law encoding is to
improve the SNR for weak speech signals. The overall data rate is 64 kbps, with sampling at 8
kbps and 8 bits per sample.

Input Output
64-kbps Analog
PCM Speech

Analog-to- Compandor
Band- Band-
Digital µ-law
Limited Limited
Converter or
Filter Filter
(8 bits) A-law

Figure 3-2 PCM Decoder

Speech Coding 43

In Europe, PCM uses A-law encoding where the output digital signal, s(t), is related to the
input signal, i(t), by

1 + ln ( A i ( t ) ) 1
s ( t ) = sgn [ i ( t ) ] -----------------------------------, --- ≤ i ( t ) ≤ 1
1 + ln A A (3.2)
1 + ( A i(t ) ) 1
s ( t ) = sgn [ i ( t ) ] ------------------------------, 0 ≤ i ( t ) ≤ ---
1 + ln A A
where a typical value of A = 87.6 is used in Europe. In Eq. (3.2), the input signal is also normal-
ized to a range of ±1.
Note that s(t) is logarithmic for |i(t)| < 1/A and linear for |i(t)| > 1/A. Thus, A-law provides
a somewhat flatter signal-to-distortion performance compared to µ-law when the signal is
greater than 1/A, at the expense of poorer performance at low signal levels.
Telephone communications that cross borders of continents must have conversion routines
in their transmission paths if the two continents use different encoding laws.

3.2.2 Adaptive Pulse Code Modulation

High bit rates are not desirable for wireless systems since the capacity of the system is
low. Higher system capacities are obtained with differential coders where compression can be
applied dynamically, such as Adaptive Predictive Coding (APC) and Adaptive Differential Pulse
Code Modulation (ADPCM). The reason for these coders is to achieve a better signal-to-quanti-
zation noise performance and a lower coding rate over PCM.
Differential coders generate error signals, as the difference between the input speech sam-
ples and corresponding prediction estimates. The error signals are quantized and transmitted.
ADPCM and APC differential coders are often used for an intermediate bit rate—16 to 32 kbps.
ADPCM employs a short-term predictor that models the speech spectral envelope.
ADPCM achieves network-quality speech (Mean Opinion Score [MOS] of 4.1 or better) at 32
kbps. This is a low-complexity coder of reasonable robustness with channel bit error rates in the
range of 10 –3 to 10 –2. The ADPCM coder is well suited for wireless access applications.
In an ADPCM encoder (Fig. 3-3), first analog speech is converted to PCM. If the signal is
already PCM—from the network, for example—then the analog-to-PCM step is not needed. The
A-law- or µ-law-encoded signal is then converted to a uniform PCM level (i.e., equal steps
between levels) signal. The encoder generates a difference signal between the converted signal
and an estimated signal and encodes the estimated signal using 15 levels. The resultant signal is
transmitted at 32 kbps (half the rate for PCM). In the encoder, the signal estimator is generated
by an inverse quantizer and an adaptive predictor. The use of differential signals and proper
design of the predictor enables an overall coding efficiency improvement over PCM.
In the ADPCM decoder (Fig. 3-4), the input 32-kbps signal is processed by an inverse
adaptive quantizer and an adaptive predictor. The output of the quantizer and the output of the
predictor are combined to generate a reconstructed signal that is converted back to PCM. The
regenerated PCM signal is then processed (in the synchronous coding adjustment stage) with sig-
nals from the input, the quantizer output, and the predictor output to generate the A-law or µ-law
44 Ch. 3 • Speech and Channel Coding

Input 64-kbps Output

Analog A-law or 32-kbps
Speech µ-law ADPCM
PCM

Nonuniform-
+ Adaptive
PCM
Encoder
to-Uniform
PCM + 15-level
Quantizer
Converter

Inverse
Adaptive Adaptive
Predicator 15-level
Quantizer

Figure 3-3 ADPCM Encoder

32-kbps Quantized Reconstructed 64-kbps Output

ADPCM Difference Signal A-law or Analog
Input Signal µ-law Speech
PCM

Inverse Uniform-to-
+ Synchronous
Adaptive
15-level
Quantizer
+ Nonuniform
PCM
Converter
Coding
Adjustment
PCM
Decoder

+
Signal
Estimate

Adaptive
Predicator

Figure 3-4 ADPCM Decoder

PCM signal. The processing in the synchronous coding adjustment stage ensures that the PCM
signal is modeled correctly by converting it to uniform PCM and comparing the resulting error
signals with the actual received signal. If an error occurs, it is corrected before the output PCM
signal is generated. The ADPCM signal conforms to International Telecommunications Union
(ITU) recommendation G.721. Finally, if analog speech is needed, the signal is processed by a
PCM decoder.
The IS-665 standard supports a modified version of ADPCM that has been optimized for
the PCS environment. The functional operation is the same as ADPCM; the coding algorithm
has been modified. (Refer to the IS-665 standard for details.)
Speech Coding 45

3.2.3 Code-Excited Linear Prediction

PCM, ADPCM, and APC operate in the time domain. No attempt is made to understand or
analyze the information that is being sent. To achieve lower coding rates, the industry has suc-
cessfully used redundancy removal techniques, operating in the frequency domain. Frequency
domain waveform coding algorithms decompose the input speech signal into sinusoidal compo-
nents with varying amplitudes and frequencies. Thus, the speech is modeled as a time-varying
line spectrum. Frequency domain coders are systems of moderate complexity and operate well
at a medium bit rate (16 kbps). When designed to operate in the range of 4.8 to 9.6 kbps, the
complexity of the approach used to model the speech spectrum increases considerably.
The other class of speech coding techniques consists of algorithms called vocoders which
attempt to describe the speech production mechanism in terms of a few independent parameters
serving as the information-bearing signals. These parameters attempt to model the creation of
the voice by the vocal tract, decompose the information, and send it to the receiver. The receiver
attempts to model an electronic vocal tract to produce the speech output.
The model operates this way: vocoders consider that speech is produced from a source-fil-
ter arrangement. Voiced speech results from exciting the filter with a periodic pulse train (simu-
lating the opening and closing of the vocal cords). Unvoiced speech results from exciting the
filter with random noise (simulating air rushing past a constriction in the vocal tract). Vocoders
operate on the input signal using an analysis process based on a particular speech production
model and extract a set of source-filter parameters that are encoded and transmitted. At the
receiver, they are decoded and used to control a speech synthesizer, which corresponds to the
model used in the analysis process. Provided that all the perceptually significant parameters are
extracted, the synthesized signal, as perceived by the human ear, resembles the original speech
signal. Nonspeech signals are often not modeled well, so this method works poorly for analog
modems.
Vocoders are medium-complexity systems and operate at low bit rates, typically 2.4 kbps,
with synthetic-quality speech. Their poor-quality speech is due to two factors: the oversimplified
source model used to drive the filter, and the assumption that the source and filter are linearly
independent.
For bit rates of about 5 kbps to 16 kbps, hybrid coders that use suitable combinations of
waveform coding techniques and vocoder techniques produce the best speech quality. Resid-
ual-Excited Linear Prediction (RELP) coding is a simple hybrid coding scheme for tele-
phone-quality speech with a few integrated digital signal processors. RELP belongs to a class
of coders known as analysis-synthesis coders based on Linear Predictive Coding (LPC).
RELP systems employ short-term (and, in certain cases, long-term) linear prediction to
formulate a difference signal (residual) in a feed-forward manner. RELP systems are capable of
producing communication-quality speech at 8 kbps. These systems utilize either pitch-aligned
high-frequency regeneration procedures or full-band pitch prediction in time domain to remove
the pitch information from the residual signal prior to band limitation/decimation. At bit rates of
less than 9.6 kbps, the quality of the recovered speech signal can be improved significantly by
46 Ch. 3 • Speech and Channel Coding

using an Analysis-by-Synthesis (AbS) optimization procedure to define the excitation signal. In

these systems both the filter and the excitation are defined on a short-term basis using a closed-
loop optimization process that minimizes a perceptually weighted error measure formed
between the input and decoded speech signals.
CDMA uses a variation of RELP called Code-Excited Linear Prediction (CELP). With
this technique, the CELP decoder (Fig. 3-5) uses a codebook to generate inputs to a synthesis fil-
ter. The codebook is characterized by its codebook index (I) and gain (G). The spectral filter is
characterized by three sets of parameters: the pitch spectral lines (a), the lag generated by the
AbS process (L), and the pitch gain (b). The output of the filter is processed by a postfilter and
gain adjustment.
CDMA implements a rate 1 encoder at 8.55 kbps and supports rates of 4, 2, and 0.8 kbps
(rates 1/2, 1/4, and 1/8, respectively). Each of the rates uses successively fewer bits for encoding
the values of I, G, L, b, and a. At rate 1/8 (Fig. 3-6), insufficient bits are available to send the
codebook index (I), and a pseudorandom code generator (synchronized at both ends) is used and
seeded by a random seed of value CBSEED (codebook seed).
The basic frame for CDMA is 20 milliseconds (ms). At rate set 1, 160 bits, plus an 11-bit
parity check field, are sent for encoding the data. Lower numbers of bits are used at lower data
rates (see Table 3-1).
Implementation of the CELP speech encoder requires three steps. First the Line Spectral
Pairs (LSP), i, values are determined. Then the LSP values are used in an AbS process to deter-
mine the values for the pitch lag (L) and pitch gain (b). Finally, the values of i, L, and b are used
in a second AbS step to determine the codebook indices (I) and gains (G). These steps are
described in more detail in the following paragraphs.

Output
Speech

Synthesis Spectral
Codebook
X Filter—
Pitch &
Format
Postfilter
and
Gain Control

Codebook Codebook Pitch Lag (L),

Index Gain Pitch Gain (b),
(I ) (G) Pitch Spectral Lines (a)

Inputs to CELP Decoder

Figure 3-5 CELP Decoder for Rates 1, 1/2, and 1/4
Speech Coding 47

Output
Speech

Pseudo- Synthesis Spectral

random
Vector
Generator
X Filter—
Pitch &
Format
Postfilter
and
Gain Control

Codebook Codebook Pitch Spectral Lines (a)

Seed Gain
(CBSEED) (G)

Inputs to CELP Decoder

Figure 3-6 CELP Decoder for Rate 1/8

Compute
Determine
DC Offset Hamming Auto-
LPC
Removal Window correlation
Coefficients
Coefficients

Scale LPC
Coefficients
Analog- µ-law PCM-
to-Uniform to-
PCM Uniform PCM
Converter Converter
Transform
LPC
Coefficients
into LSPs

Analog µ-law PCM

Speech Speech
Convert LSP
Input Input
Frequencies
into
LSP Codes Output
LSP Codes
(a)

Figure 3-7 CELP Encoder for LSP Codes

48 Ch. 3 • Speech and Channel Coding

Table 3-1 CELP Parameters for Various Coding Rates

Rate Rate 1 Rate 1/2 Rate 1/4 Rate 1/8
CELP Parameters
Line spectral pairs (i) bits 40 20 10 10
i updates per frame 1 1 1 1
Total i bits per frame 40 20 10 10
Pitch lag (L) bits 7 7 7 0
L updates per frame 4 2 1 0
Total L bits per frame 28 14 7 0
Pitch gain (b) bits 3 3 3 0
b updates per frame 4 2 1 0
Total b bits per frame 12 6 3 0
Codebook index (I) bits 7 7 7 0
I updates per frame 8 4 2 —
Total I bits per frame 56 28 14 0
Codebook gain (G) bits 3 3 3 2
G updates per frame 8 4 2 1
Total G bits per frame 24 12 6 2
Codebook seed (CBSEED) bits 0 0 0 4
CBSEED updates per frame — — — 1
Total CBSEED bits — — — 4
Parity check bits per frame 11 0 0 0
Total number of bits per frame 171 80 40 16

LSP determination (Fig. 3-7). The encoder for the LSP codes first converts the speech to
uniform PCM with at least 14 bits. If the encoder is in a base station, then the received
speech is most likely µ-law PCM; if the encoder is in a mobile station, then the received
speech is analog. After the speech is converted to PCM, it is processed to remove the DC
component and filtered by a Hamming window. The autocorrelation of the sampled output
is then computed and used to determine the coefficients for the LPC. The LPC coefficients
are then scaled, transformed into the frequency components, and converted into the values
for the i bits of the coder output.
The pitch lag and gain bits (Fig. 3-8). These are computed by a recursive process where
the output of the PCM encoder is combined with the LSP codes previously calculated
Speech Coding 49

and with all possible values of pitch and gain. For each value of pitch and gain, an error
function is computed and the transmitted values for pitch and gain are chosen to mini-
mize the error.
The codebook index and gain (Fig. 3-9). These are computed in a recursive process sim-
ilar to the pitch lag and gain bits using the uniform PCM signal; the computed values for
frequency, pitch, and gain; and all possible codebook values and gains.

For the rate 1/8 system, codebook indices are not computed—this type of system uses a
random vector generated at both sides.
For every frame or subframe the excitation waveform is selected from a codebook consist-
ing of a large number of candidate waveform vectors. The codebook vector chosen to excite the
speech coder filters minimizes the weighted error between the original and synthesized speech.

3.2.4 Enhanced Variable-Rate Codec

The technique used by the Enhanced Variable-Rate Codec (EVRC) [5] to reduce the num-
ber of bits required for linear predictor coefficients and pitch synthesis enables the algebraic
codebook to generate excitation. As a result, EVRC has higher voice quality.

Choose
Compute
L & b to
Uniform PCM Error
Minimize
Function
Error
Output
Lag (L) &
Gain (b)
LSP Codes (a)

All Possible
Values of L & b

Figure 3-8 CELP Encoder for Pitch Parameters

Choose
Compute
G & I to
Uniform PCM Error
Minimize
Function
Error
Output
Codebook (I ) &
Gain (G)
LSP Codes (a)
Pitch Lag and Gain (L & b)

All Possible
Values of I & G
Figure 3-9 CELP Encoder for Codebook Values
50 Ch. 3 • Speech and Channel Coding

Unlike conventional CELP encoders, EVRC does not attempt to match the original speech
signal exactly. Instead, EVRC matches a time-wrapped version of the residual that conforms to a
simplified pitch contour. The contour is obtained by estimating the pitch delay in each frame and
linearly interpolating the pitch from frame to frame. While this adds to computational complex-
ity, the result is higher voice quality per each bit transmitted.
The simplified pitch representation also leaves more bits available in each packet for sto-
chastic excitation and the channel impairment protection than would be possible if a traditional
fractional pitch approach were used. The result is enhanced error performance without degraded
speech quality at the small cost of added processing requirements.
EVRC also enhances call quality by suppressing background noise. The IS-127 [5] stan-
dard recommends a noise suppressor algorithm, but allows system designers to define their own.
This is an important factor in choosing a processing platform, making programmable DSP a
desirable choice.
The EVRC algorithm is based upon the CELP algorithm. It uses the Relaxed Code-Excited
Linear Prediction (RCELP) algorithm and thus does not match the original residual signal but
rather a time-wrapped version of the original residual signal that conforms to a simplified pitch
contour. This approach reduces the number of bits per frame that are dedicated to pitch represen-
tation, allowing additional bits to be dedicated to stochastic excitation and to channel impairment
protection. The EVRC algorithm categorizes speech into full-rate (8.55-kbps), 1/2-rate (4-kbps),
and 1/8-rate (0.8-kbps) frames that are formed every 20 ms. The EVRC algorithm offers a signif-
icant performance improvement over the IS-96A speech codec. Table 3-2 shows the performance
of the CDMA Development Group’s 13-kbps (CDG-13kbps) speech codec along with IS-96A
and EVRC codecs. The CDG-13kbps offers high voice quality but results in a decrease in chan-
nel capacity of about 40%. Fig. 3-10 shows a functional diagram of the EVRC algorithm.
Table 3-3 provides the bit allocations by packet type.

3.3 Channel Coding

Channel coding is used to provide the coding gain that is defined as reduction in the
required Eb/N0 (in dB) to achieve a specified error probability of the encoded system over
an uncoded system with the same modulation and channel characteristics.The channel
coding process usually falls into two classes: block codes and convolutional codes. There

Table 3-2 Comparisons of CDG-13kbps, IS-96A, and EVRC Vocoders in MOS

FERa % CDG-13 kbps IS-96A EVRC

0 4.00 3.29 3.95
1 3.95 3.17 3.83
2 3.88 2.77 3.66
3 3.67 2.55 3.50
a
FER = frame error rate.
Channel Coding 51

Rate 1/8
Encoding

Signal Model
Packet
Pre- Parameter
Formatting
processing Estimation
Rate 1/2
Sampled or Rate 1 Formatted
Speech Encoding Packet

Rate
Determination
External Rate
Rate Decision
Command

Figure 3-10 Functional Diagram of EVRC

Table 3-3 Bit Allocations by Packet Type in EVRC

Packet Type
Field Rate 1 Rate 1/2 Rate 1/8 Blank
Spectral Transition Indicator 1
LSPa 28 22 8
Pitch delay 7 7
Delta delay 5
b
ACB gain 9 9
FCB shapec 105 30
FCB gainc 15 12
Frame energy 8
Unused 1
Total encoded bits 171 80 16
Mixed mode (MM) bit 1
d
Frame quality indicator (CRC) (F) 12 8
Encoder tail bits (T) 8 8 8 8
Total bits 192 96 24 8
Rate (kbps) 9.6 4.8 1.2 0.4
a
A representation of digital filter coefficient in a pseudofrequency domain; this representation has good
quantization and interpolation properties.
b
Adaptive codebook.
c
Fixed codebook.
d
Cyclic redundancy check.
52 Ch. 3 • Speech and Channel Coding

are many subclasses of block codes, including linear block codes, binary cyclic codes, and
Bose-Chadhusi-Hocquenghem (BCH) codes. Binary cyclic codes are also called Cyclic
Redundancy Check (CRC) codes. A BCH code is a special CRC code. BCH codes are repre-
sented by (n,k,q) where k bits are mapped into n output bits (n > k), and q is error correc-
tion capability. For example a (15,7,2) BCH code transmits 7 information bits by using a
15-bit code word, and it can correct any random errors up to two errors in the code word.
IS-95 systems use convolutional code based on the Viterbi algorithm.

3.3.1 Convolutional Code

Convolutional encoders can be thought of as finite-state machines that change states as the
function of the input sequence. A convolutional code [6,7] is generated by passing an informa-
tion sequence through a finite-state shift register. The shift register contains K stages and m lin-
ear algebraic function generators based on generator polynomials. The number of output bits for
each k-bit input data sequence is n bits. The code rate is r = k/n. The parameter K is called the
constraint length and indicates the number of input data bits upon which the current output is
dependent (see Fig. 3-11).
The error-correcting capability of a convolutional coding scheme increases as rate r
decreases. However, the channel bandwidth and decoder complexity both increase with K. The
advantage of lower code rates when using convolutional code with coherent phase-shift keying
(PSK) is that the required Eb/N0 is decreased, permitting the transmission of higher data rates for
a given amount of power or permitting reduced power for a given data rate. Simulation studies

Rate = 1/2 g0 c0
Constraint length K = 9

X1 X2 X3 X4 X5 X6 X7 X8

Shift Register

Information Bits (input)

g1 c1
Code Symbols (output)

Figure 3-11 Convolutional Encoder

Channel Coding 53

have indicated that, for a fixed constraint length, a decrease in the code rate from 1/2 to 1/3
results in a reduction of the required Eb/N0 of about 0.4 dB. However, the corresponding increase
in decoder complexity is about 17%. For a smaller value of code rate, the improvement in per-
formance relative to the increased decoding complexity diminishes rapidly. Eventually, a point is
reached where a further decrease in code rate is characterized by a reduction in coding gain.
The Viterbi algorithm performs maximum-likelihood decoding. It reduces the computa-
tional load by taking advantage of the special structure in coder trellis. The complexity of a Viterbi
decoder is not a function of the number of symbols in the code word sequence. The Viterbi algo-
rithm removes from consideration those trellis paths that could not possibly be candidates for the
maximum-likelihood choice. When two paths enter the same state, the one having the best metric
is selected; this path is called the surviving path. This selection of surviving paths is performed for
all the states. The decoder continues in this way to advance deeper into the trellis, making deci-
sions by eliminating the least likely paths. The major drawback of the Viterbi algorithm in that,
while error probability decreases exponentially with constraint length, the number of code
states—and consequently decoder complexity—grows exponentially with constraint length.
In IS-95, traffic data frames on uplink and downlink are fed to convolutional encoders.
Both uplink and downlink encoders use an 8-bit shift register with a constraint length of 9. The
rate of the uplink coder is 1/3—it outputs 3 bits for every input bit. At a rate below 9.6 kbps, out-
put bits are repeated to bring the number of bits in a 20-ms block to 576, for a gross rate of 28.8
kbps. The rate of the downlink encoder is 1/2—it outputs 2 bits for every input bit. At a rate
below 9.6 kbps, output bits are repeated to bring the number of bits in a 20-ms block to 384, for
a gross rate of 19.2 kbps.

EXAMPLE 3.1
Compare the error probability for a CDMA link with and without the use of error correction coding.
Assume that the uncoded transmission characteristics are: QPSK modulation, Gaussian noise, Eb/N0
= 10 dB, data rate = 9.6 kbps. For the coded case, assume the use of a (15,11) error-correcting code,
which is capable of correcting any single error pattern within a block of 15 bits. Also assume the
demodulator makes hard decisions and feeds the demodulated code bit directly to the decoder,
which in turn outputs an estimate of the original message.
Uncoded Condition

2E –6
p bu = Q --------b- = Q [ 20 ] = 4.05 × 10
N0
The probability that the uncoded message block will be received in error is
u k – 6 11 –5
P M = 1 – ( 1 – p bu ) = 1 – ( 1 – 4.05 × 10 ) = 4.455 × 10
This is the probability that at least 1 bit out of 11 is in error.
Coded Condition
9600 × 15
R c = ------------------------ = 13,091 bps
11
54 Ch. 3 • Speech and Channel Coding

E 10
-----b- = -------------------------------- = 7.333 ( 8.653 dB )
N0 ( 13.091 ⁄ 9.6 )

2E –5
pc = Q --------b- = Q [ 14.666 ] = 6.83 × 10
N0
15
 15 
∑ 
c j 15 – j
P M =  ( pc ) ( 1 – pc )
2 j 

The summation starts with j = 2 since the code corrects all single errors within a block of 15
bits.
 
P M ≈  15  ( p c ) ( 1 – p c ) = 1.96 × 10
c 2 13 –6

 2 
–5
4.455 × 10
The probability of message error has improved by -----------------------------
–6
- = 22.73 due to error-correct-
ing code. 1.96 × 10

3.4 Summary
This chapter discussed the digital voice encoding systems used for both CDMA and wideband
CDMA. Conventional wireline systems transmit voice by digitizing the voice signal using PCM
at a rate of 64 kbps. While it is possible to use PCM in wireless systems, the capacity of the
wireless system is lower compared to using other digitizing methods for voice. The CDMA sys-
tem uses different approaches to digitizing the voice signal—it uses a CELP at 8 or 13.3 kbps to
digitize voice. CELP systems model the operation of the human vocal tract to efficiently code
speech. We described the operation of the CELP encoder and decoder in detail and provided a
high-level description of EVRC. Throughout the descriptions on voice coding, the goal has been
to explain the coding systems at a high level so that you can understand the operation of the sys-
tem and read the standards with some understanding of the motivation for them. If you need to
design systems or want additional information, we encourage you to read the standards.

3.5 Problems
1. Repeat Example 3.1 for Eb/N0 = 6.8 dB and a data rate of 14.4 kbps.
2. Determine the resulting output from a one-half convolutional encoder for the input 1 0
1 1 0 0 … 0 0. The left-most bit is transmitted first.

g0 = 1 + D + D2 + D3 + D5 + D7 + D8, and

g1 = 1 + D2 + D3 + D4 + D8

Assume that the encoder is initially reset (all registers are 0).
References 55

3.6 References
1. ITU Recommendation G.711.
2. Recommendation G162, CCITT Plenary Assembly, Geneva, May–June 1964, Blue Book,
Vol. 111, p. 52.
3. Skalar, B., Digital Communication—Fundamentals and Applications, Prentice Hall, Engle-
wood Cliffs, NJ, 1988.
4. TIA-IS-96A, “Speech Service Option Standard for Wideband Spread Spectrum Digital Cel-
lular System.”
5. TIA-IS-127, “Enhanced Variable Rate Codec (EVRC) 8.5 kbps Speech Coder.”
6. TIA-IS-665, “W-CDMA (Wideband Code Division Multiple Access) Air Interface Compat-
ibility Standard for 1.85–1.99 GHz PCS Applications.”
7. Ziemer, R. E., and Peterson, R. L., Introduction to Digital Communication, Macmillan Pub-
lishing Co., New York, 1992.
This page intentionally left blank
 C H A P T E R 4

Diversity, Combining,
and Antennas

4.1 Introduction
In this chapter, we discuss the concepts of diversity reception where multiple signals are com-
bined to improve the SNR of the system. Time diversity is used to improve system performance
for IS-95 CDMA systems; therefore we explore that system in more detail. We then describe
various combining schemes that are used to combine the signals. Finally, we consider some
practical antennas used in the cellular telephone industry today.

4.2 Diversity Reception

Buildings and other obstacles in built-up areas scatter the signal. Furthermore, because of the
interaction between the several incoming waves, the resultant signal at the antenna is subject to
rapid and deep fading. Average signal strength can be 40 to 50 dB below the free-space path
loss. Fading is most severe in heavily built-up areas in an urban environment. In these areas, the
signal envelope follows a Rayleigh distribution over short distances and a log-normal distribu-
tion over large distances [10].
Diversity reception techniques are used to reduce the effects of fading and improve the
reliability of communication without increasing either the transmitter’s power or the channel
bandwidth.
The basic idea of diversity reception is that, if two or more independent samples of a sig-
nal are taken, these samples will fade in an uncorrelated manner. This means that the probability
of all the samples being simultaneously below a given level is much lower than the probability
of any individual sample being below that level. The probability of M samples all being simulta-
neously below a certain level is p M, where p is the probability that a single sample is below that
level. Thus, we can see that a signal composed of a suitable combination of the various samples
will have much less severe fading properties than any individual sample.

57
58 Ch. 4 • Diversity, Combining, and Antennas

In principle, diversity reception techniques can be applied either at the base station or at
the mobile station, although each type of application has different problems that must be
addressed. Typically, the diversity receiver is used in the base station instead of the mobile sta-
tion. The cost of the diversity combiner can be high, especially if multiple receivers are required.
Also the power output of the mobile station is limited by its battery life. The base station, how-
ever, can increase its power output or antenna height to improve coverage to a mobile station.
Most diversity systems are implemented in the receiver instead of the transmitter since no extra
transmitter power is needed to install the receiver diversity system. Since the path between the
mobile station and the base station is assumed to be reciprocal, diversity systems implemented
in a mobile station work similarly to those in a base station.

4.3 Types of Diversity

There are different ways by which diversity can be achieved: time, frequency, space, angle, mul-
tipath, and polarization [1]. In order to gain complete advantage of diversity, combining must be
performed at the receiving end. Combiners are designed so that input signal levels, after phase
and time delay corrections for the multipath effects, add vectorially while noise outputs are
added randomly. Thus, on the average, the combined output SNR will be greater than that
present at the input of a single receiver.

4.3.1 Macroscopic Diversity

Macroscopic diversity is used to reduce large-scale fading caused by shadowing. The local
mean signal strength varies because of variations in the terrain between the mobile station trans-
mitter and the base station receiver. If only one antenna site is used, the traveling mobile unit
may not be able to transmit a signal to the base station at certain geographical locations because
of terrain variations such as hills or mountains. Therefore, two separated antenna sites can be
used to receive two signals and to combine them to reduce long-term fading. The selective com-
bining technique discussed later in this chapter works best in the macroscopic diversity scheme
since other methods require coherent combining which is difficult to achieve when the receivers
are some distance apart. Macroscopic diversity is often used in shortwave radio systems to
reduce the effects of fading from the ionosphere. Cellular and PCS systems achieve the same
effect by handoffs to nearby cell sites when the signal strength becomes weak.
With CDMA systems, macroscopic diversity (i.e., soft handoff) is essential in order to
achieve reasonable system performance because of frequency reuse of unity and fast power con-
trol. If the mobile station is not connected to the base station that has the lowest attenuation,
unnecessary interference is generated in adjacent cells. In the reverse direction (MS to BS), the
macroscopic diversity is beneficial because the more base stations try to detect the signals, the
higher the probability is for at least one to succeed. In the reverse direction, the detection process
itself does not utilize the information from the other base stations receiving the same signal, but
the diversity is selection diversity in which the best frame is utilized in the network based on the
frame error rate (FER) indication from a CRC.
Types of Diversity 59

In the forward direction, macroscopic diversity is different because the transmission origi-
nates from several sources and diversity reception is handled by one receiver in the mobile sta-
tion. All extra transmissions contribute to interference. Capacity improvement is based on a
principle similar to a RAKE receiver in a multipath channel, in which the received power-level
fluctuations tend to decrease as separable paths increase. With forward-link macroscopic diver-
sity, the RAKE receiver capability to gain from extra diversity depends also on the number of
available RAKE fingers. If the RAKE receiver is not able to collect enough energy from trans-
missions from two or, in some cases, three base stations due to a limited number of RAKE fin-
gers, the extra transmissions to the mobile station can have a negative effect on total system
capacity due to increased interference. This is most likely in the macroscopic cellular environ-
ment because the typical number of RAKE fingers considered adequate to capture the channel
energy in most cases is four. If all connections offered that amount of diversity, then the receiver
would have only one or two branches to allocate for each connection.

4.3.2 Microscopic Diversity

Microscopic diversity uses two or more antennas that are at the same site (co-located) but
that are designed to exploit differences in arriving signals from the receiver. Microscopic diver-
sity techniques are used to prevent deep fades from occurring. Once the diversity branches are
created, any of the combining schemes (e.g., selective, maximal-ratio, or equal-gain) can be
used. The following methods are used to obtain uncorrelated signals for combining:
Space diversity. Several transmission paths are used. Two antennas separated physically
by a short distance d can provide two signals with low correlation between their fades. The
separation d in general varies with antenna height h and with frequency. The higher the
frequency, the closer the two antennas can be to each other. Typically a separation of a few
wavelengths is enough to obtain uncorrelated signals.
Frequency diversity. Signals received on two frequencies, separated by the coherence
bandwidth, Bc , are uncorrelated. To use frequency diversity in an urban or suburban envi-
ronment for cellular and PCS frequencies, the frequency separation must be 300 kilohertz
(kHz) or more. The use of frequency hopping typical of TDMA system (such as GSM)
provides frequency diversity.
Polarization diversity. Horizontally or vertically polarized carrier waves are used. The
horizontal and vertical polarization components, Ex and Ey , respectively, transmitted by
two polarized antennas at the base station and received by two polarized antennas at the
mobile unit, can provide two uncorrelated fading signals. Polarization diversity results in a
3-dB power reduction at the transmitting site since the power must be split between two
polarized antennas.
Angle diversity. When the operating frequency is ≥10 gigahertz (GHz), the scattering of
the signals from transmitter to receiver generates received signals from different directions
that are uncorrelated with each other. Thus, two or more directional antennas can be
60 Ch. 4 • Diversity, Combining, and Antennas

pointed in different directions at the receiving site and can provide signals for a combiner.
This scheme is more effective at the mobile unit than at the base station since the scatter-
ing is from local buildings and vegetation and is more pronounced at street level than at
the height of base station antennas.
Time diversity. The transmission of a symbol is spread out over time. If the identical sig-
nal is transmitted in different time slots, the received signals will be uncorrelated. This
system will work for an environment where the fading occurs independent of the move-
ment of the receiver. In a mobile radio environment, the mobile unit may be at a standstill
at a location having a weak local mean or caught in deep fade. Although fading still occurs
even when the mobile is still, the time-delayed signals are correlated and time diversity
will not reduce the fades.
Time diversity is achieved by coding, interleaving, and retransmitting. Channel coding is
applied to achieve lower power levels and required signal quality in terms of bit error rate
(BER)/FER. Interleaving and channel coding processes are used to correct errors due to channel
fades and interference peaks.

4.3.3 RAKE Receiver

In 1958, Price and Green [9] proposed a method of resolving multipath problems using
wideband pseudorandom sequences modulated onto a transmitter using other modulation meth-
ods (AM or FM). The pseudorandom sequence has the property that time-shifted versions of
itself are almost uncorrelated. Thus, a signal that propagates from transmitter to receiver over
multiple paths (hence multiple different time delays) can be resolved into separately fading sig-
nals by cross-correlating the received signal with multiple time-shifted versions of the pseudo-
random sequence. Fig. 4-1 shows a block diagram of a typical system. In the receiver, the
outputs are time shifted and, therefore, must be sent through a delay line before entering the

Taped Delay Line

r (t ) T T T T T T T T T ...

... ...

a (t ) a (t ) a (t )
X X X

... ...

Diversity Combiner
Combiner Output

Figure 4-1 RAKE Receiver

Types of Diversity 61

diversity combiner. The receiver is called a RAKE receiver since the block diagram looks like a
garden rake.
When the CDMA systems were designed for cellular systems, the inherent wide-bandwidth
signals with their orthogonal Walsh functions were natural for implementing a RAKE receiver. In
addition, the RAKE receiver mitigates the effects of fading and is in part responsible for the
claimed 10:1 spectral efficiency improvement of CDMA over analog cellular.
In the CDMA system, the bandwidth (1.25 to 15 MHz) is wider than the coherence band-
width of the cellular or PCS channel. Thus, when the multipath components are resolved in the
receiver, the signals from each tap on the delay line are uncorrelated with each other. The
receiver can then combine them using any of the combining schemes. The CDMA system then
uses the multipath characteristics of the channel to its advantage to improve the operation of the
system (see Fig. 4-2).
The combining scheme used governs the performance of the RAKE receiver. An important
factor in the receiver design is synchronizing the signals in the receiver to match that of the
transmitted signal. Since adjacent cells are also on the same frequency with different time delays
on the Walsh codes, the entire CDMA system must be tightly synchronized.
A RAKE receiver uses multiple correlators to separately detect the M strongest multipath
components. The relative amplitudes and phases of the multipath components are found by cor-
relating the received waveform with delayed versions of the signal or vice versa. The energy in
the multipath components can be recovered effectively by combining the (delay-compensated)
multipath components in proportion to their strengths. This combining is a form of diversity and
can help to reduce fading. Multipath components with relative delays of less than ∆t = 1/Bw can-

Z1
Correlator 1
a1

Z'
Baseband Z2
∑
CDMA r (t )
Correlator 2
Signal with
Multipath a2
Components

ZM
Correlator M
aM

Figure 4-2 RAKE Receiver Correlator

62 Ch. 4 • Diversity, Combining, and Antennas

not be resolved and, if present, contribute to fading; in such cases forward error-correction cod-
ing and power control schemes play the dominant role in mitigating the effects of fading.
The outputs of the M correlators are denoted as Z1, Z2, …, and ZM. The weights of the out-
puts are a1, a2, …, and aΜ, respectively [10] (see Fig. 4-2). The weighting coefficients are based
on the power or the SNR from each correlator output. If the power or SNR is small from a par-
ticular correlator, it is assigned a small weighting factor. The composite signal, Z , is given by
M
Z = ∑ ak ⋅ Z k (4.1)
k=1

The weighting coefficients, ak , are normalized to the output signal power of the correlator
in such a way that the coefficients sum to unity, as shown in Eq. (4.2).
2
Zk
a k = --------------
M
- (4.2)

∑
2
Zk
k=1

In CDMA cellular/PCS systems, the forward link (BS to MS) uses a three-finger RAKE
receiver, and the reverse link (MS to BS) uses a four-finger RAKE receiver [5]. In the IS-95
CDMA system, the detection and measurement of multipath parameters are performed by a
searcher receiver, which is programmed to compare incoming signals with portions of I- and Q-
channel PN codes. Multipath arrivals at the receiver unit manifest themselves as correlation
peaks that occur at different times. A peak’s magnitude is proportional to the envelope of the
path signal. The time of each peak, relative to the first arrival, provides a measurement of the
path’s delay.
The PN chip rate of 1.2288 Mcps allows for resolution of multipath components at time
intervals of 0.814 µs. Because all of the base stations use the same I and Q PN codes, differing
only in code phase offset, not only multipath components but also other base stations are
detected by correlation (in a different search window of arrival times) with the portion of the
codes corresponding to the selected base stations. The searcher receiver maintains a table of the
stronger multipath components and/or base station signals for possible diversity combining or
for handoff purposes. The table includes time of arrival, signal strength, and the corresponding
PN code offset.
On the reverse link, the base station’s receiver assigned to track a particular mobile trans-
mitter uses the I- and Q-code times of arrival to identify mobile signals from users affiliated with
that base station. Of the mobile signals using the same I- and Q-code offsets, the searcher
receiver at the base station can distinguish the desired mobile signal by means of its unique
scrambling, long-PN-code offset, acquired before voice transmission begins on the link using a
special preamble for that purpose. As the call proceeds, the searcher receiver is able to monitor
the strengths of the multipath components from the mobile unit to the base station and to use
more than one path through diversity combining.
Basic Combining Methods 63

4.4 Basic Combining Methods

After obtaining the necessary signal samples, we need to consider the question of processing
these samples to obtain the best results. For most communication systems, the process can be
broadly classified as the linear combination of the samples. In the combining process, the vari-
ous signal inputs are individually weighted and added together as [8]

r ( t ) = a1 r1 ( t ) + a2 r2 ( t ) + … + a M r M ( t ) (4.3a)
M
r(t ) = ∑ ai ri ( t ) (4.3b)
i=1

where ri (t) = the envelope of the ith signal, and

ai = the weight factor applied to the ith signal.
We make the following assumptions in the analysis of a combiner:
1. The noise in each branch is independent of the signal and is additive.
2. The signal amplitudes change because of fading, but the fading rate is much smaller
than the lowest modulation frequency present in the signal.
3. The noise components are locally incoherent and have 0 mean, with a constant local
mean-square value (i.e., constant noise power).
4. The local mean-square values (powers) of the signals are statistically independent.
Since the goal of the combiner is to improve the noise performance of the system, the
analysis of combiners is generally performed in terms of SNR. We will examine several different
types of combiners and compare their SNR improvements over no diversity.

4.4.1 Selection Combiner

The selection combiner is the simplest of all the schemes. An ideal selection combiner
chooses the signal with the highest instantaneous SNR, so the output SNR is equal to that of the
best incoming signal. In practice, the system cannot function on an instantaneous basis; to be
successful, it is essential that the internal time constants of a selection system are substantially
shorter than the reciprocal of the signal fading rate.
We assume that the signal received by each branch is statistically independent of the sig-
nals in other branches and is Rayleigh distributed with equal mean signal power P0. The proba-
bility density function of the signal envelope, on branch i, is given by
r i –r2i ⁄ 2P0
p ( r i ) = ------ e (4.4)
P0
where 2 P0 = mean-square signal power per branch = <ri>, and
ri2 = instantaneous power in the ith branch.

Let ξi = ri2 /2Ni and ξ0 = 2P0 /2Ni , where Ni is the noise power in the ith branch.
64 Ch. 4 • Diversity, Combining, and Antennas

ξi 2
∴----- = r i ⁄ 2P 0 (4.5)
ξ0

The probability density function for ξi is given by

1 ( –ξi ⁄ ξ0 )
p ( ξ i ) = ----- e (4.6)
ξ0

We assume that the signal in each branch has a constant mean; thus, the probability that
the SNR on any one branch is less than or equal to any given value ξg is
ξg ( –ξg ⁄ ξ0 )
P [ ξi ≤ ξg ] = ∫0 p ( ξ i ) dξ i = 1 – e (4.7)

Therefore, the probability that the SNRs in all branches are simultaneously less than or equal to
ξg is given by
( –ξg ⁄ ξ0 ) M
P M ( ξ g ) = P [ ξ 1, ξ 2, …, ξ M ≤ ξ g ] = [ 1 – e ] (4.8)

The probability that at least one branch will exceed the threshold SNR value of ξg is given by

P (at least one branch ≥ ξ g ) = 1 – P M ( ξ g ) (4.9)

The percentage of time the instantaneous output SNR ξM is below or equal to the threshold
value, ξg, is equal to P(ξM ≤ ξg). We plot results for M = 1, 2, and 4 in Fig. 4-3. Note that the larg-
est gain occurs for the two-branch combiner. By differentiating Eq. (4.8) we get the probability
density function
( –ξg ⁄ ξ0 ) M – 1 ( –ξg ⁄ ξ0 )
p M ( ξg ) = ( M ⁄ ξ0 ) [ 1 – e ] e (4.10)

The mean value of the SNR can be given as

ξg ( –ξg ⁄ ξ0 ) M – 1 ( –ξg ⁄ ξ0 )
∫0 M  ----
- [ 1 – e
∞
ξM = ] e dξ g (4.11)
ξ 0

Let x = ξg/ξ0 and dx = dξg/ξ0

ξM ∞ –x M – 1 –x
∴------ = M ∫0 x [ 1 – e ] e dx (4.12)
ξ0
–x –x
Substituting y = 1 – e or x = – ln ( 1 – y ) ; then dy = e dx
∞
ξM 1
∴------ = M ∫0 [ – ln ( 1 – y ) ]y ∑ ∫0 ---K- y
1 M–1 1 M+K–1
dy = M dy (4.13)
ξ0
K=1
Basic Combining Methods 65

1E+00

1E-01

M=1

1E-02 M=2
Probability

M=4

1E-03

1E-04

1E-05
-40 -35 -30 -25 -20 -15 -10 -5 0 5 10

10log (ξg /ξ0 )

Figure 4-3 Probability for Different Values of M-Selection Combiner

M
ξM 1
∴------ =
ξ0 ∑ ---K- (4.14)
K=1

Table 4-1 shows that the mean SNR increases slowly with M.

4.4.2 Maximal-Ratio Combiner

Maximal-ratio combining was first proposed by Kahn [3]. The M signals are weighted
proportional to their signal voltage-to-noise power ratios and then summed.
M
rM = ∑ ai ri ( t ) (4.15)
i=1
Since noise in each branch is weighted according to noise power
M M

∑ ∑ ai a j ni ( t )n j ( t )
2
ni ( t ) = (4.16)
j = 1i = 1
66 Ch. 4 • Diversity, Combining, and Antennas

Table 4-1 Number of Branches vs. Mean SNR (dB)

ξM ξ
------ 10 log -----M-
M ξ0 ξ0

1 1.000 0.000
2 1.500 1.761
3 1.833 2.632
4 2.083 3.187
5 2.283 3.585
6 2.450 3.892

The average noise power

M M

∑ ai ni ( t ) = 2 ∑ ai N i
2 2 2
NT = (4.17)
i=1 i=1

where
2
n i ( t ) = 2N i (4.18)
The SNR at the output is given as
M 2

∑ ai ri ( t )
1 i=1
ξ M = --- ----------------------------
- (4.19)
2 M
∑ ai N i
2

i=1

We want to maximize ξM. This can be done by using the Schwartz inequality.
M 2 M M

∑ ai ri ≤ ∑ ∑ ai
2 2
ri (4.20)
i=1 i=1 i=1

If a i = r i ⁄ N i , then
M M 2
ri
∑ ri ∑ -----
2

1 =1 i=1 i N
ξM = --- i----------------------------
M
- (4.21)
2
∑ ri
2

i=1
M 2 M
1 r
∴ξ M = --- ∑ -----i = ∑ ξi (4.22)
2 Ni
i=1 i=1
Basic Combining Methods 67

Thus, the SNR at the combiner output equals the sum of the SNR of the branches.
M M
ξM = ∑ ξi = ∑ ξ0 = Mξ 0 (4.23)
i=1 i=1

ξM
∴------ = M (4.24)
ξ0
The probability density function of the combiner output SNR is given by
ξM
– -------
M – 1 ξ0
ξM e
p(ξM ) = ---------------------------
M
- , ξM ≥ 0 (4.25)
ξ 0 ( M – 1 )!
The probability that ξM ≤ ξg is given by

 ξ----g-
K–1
ξg
– ----- M
ξ0  ξ 0
P ( ξ M ≤ ξg ) = 1 – e ∑ --------------------
( K – 1 )!
(4.26)
K=1

 ξ----g-
K–1
ξg
– ----- M
ξ0  ξ 0
P ( ξ M > ξg ) = e ∑ --------------------
( K – 1 )!
(4.27)
K=1

The plot of P for M = 1, 2, and 4 is shown in Fig. 4-4.

4.4.3 Equal-Gain Combining

Equal-gain combining is similar to maximal-ratio combining, but there is no attempt to
weight the signal before addition; thus ai = 1 [2]. The envelope of the output signal is given by
Eq. (4.3b) with all ai = 1
M
r = ∑ ri (4.28)
i=1

and the mean output SNR is given as

M 2

∑ ri
1 i=1
ξM = --- -------------------
- (4.29)
2 M
∑ Ni
i=1

If we assume that the mean noise power in each branch is the same (i.e., N), then Eq.
(4.29) becomes
68 Ch. 4 • Diversity, Combining, and Antennas

1E+00

1E-01

M=1

1E-02

M=2
Probability

M=4
1E-03

1E-04

1E-05
-40 -35 -30 -25 -20 -15 -10 -5 0 5 10

10log (ξg /ξ )
0

Figure 4-4 Probability for Different Values of Maximal-Ratio Combiner

M 2 M
1 1
ξ M = -------------
2NM ∑ ri = -------------
2NM ∑ r j ri (4.30)
i=1 j, i = 1

2 πP
but r i = 2P 0 ; and r i = --------0- .
2
Since the various branch signals are uncorrelated, r j r i = r i r j = r i r j , for i not equal to j.
Therefore Eq. (4.30) will be

1 πP 0 π
ξ M = ------------- 2MP 0 + M ( M – 1 ) --------- = ξ 0 1 + ( M – 1 ) --- (4.31)
2NM 2 4

ξM π
------ = 1 + ( M – 1 ) --- (4.32)
ξ0 4
Basic Combining Methods 69

For M = 2, the probability P can be written in closed form as

 2ξ g ξg
–  ---------
 ξ0  ξ g – ----
-
ξ0 ξ
P ( ξ M ≤ ξg ) = 1 – e – π  ----- e ⋅ erf ----g- (4.33)
 ξ 0 ξ0
For M > 2, the probability can be obtained by numerical integration techniques. The plot
of probability P(ξM ≤ ξg) is shown in Fig. 4-5.
Table 4-2 shows M versus SNR at 1% probability for the selection, maximal-ratio, and
equal-gain combiners. Table 4-3 shows SNR improvement for M = 2, 4, and 6 at 1% probabil-
ity for the selection, maximal-ratio, and equal-gain combiners. Note that the selection diver-
sity scheme has the poorest performance and the maximal-ratio combiner has the best. The
performance of equal-gain combining is only marginally inferior to maximum-ratio. The
implementation complexity for equal-gain combining is significantly less than maximal-ratio
combining because of the requirement of correct weighting factors. The data is compared in
Fig. 4-6.

1E+00

1E-01

1E-02

M=2
Probability

1E-03

1E-04 Equal Gain

1E-05
-40 -35 -30 -25 -20 -15 -10 -5 0 5 10

10log (ξg /ξ )
0

Figure 4-5 Probability for M = 2 Equal Gain Combiner

70 Ch. 4 • Diversity, Combining, and Antennas

Table 4-2 SNR (dB)

M Selection Maximal-Ratio Equal-Gain
1 –20.0 –20.0 –20.0
2 –10.0 –8.5 –9.2
4 –4.0 –1.0 –2.0
6 –2.0 2.0 1.5

Table 4-3 SNR Improvement (dB)

M Selection Maximal-Ratio Equal-Gain
2 10.0 11.5 10.8
4 16.0 19.0 18.0
6 18.0 22.0 21.5

1E+00

1E-01
M=1
M=2 M=4

1E-02
All
Probability

1E-03
Selection

Selection Maximal
Maximal Ratio
Ratio

1E-04
Equal Gain

1E-05
-50 -40 -30 -20 -10 0 10 20

10log (ξg /ξ0 )

Figure 4-6 Performance Improvement Comparison of Various Combining Schemes

BPSK Modulation and Diversity 71

4.5 BPSK Modulation and Diversity

As we discussed earlier, to prevent deep fades from occurring, microscopic diversity techniques
can exploit the rapidly changing signal. Macroscopic diversity can be used to reduce large-scale
fading caused by shadowing due to variations in the terrain between the transmitter and receiver.
Macroscopic diversity is useful at the base station receiver. By using base station antennas suffi-
ciently separated in space, the base station improves the reverse link performance choosing the
antenna with strongest signal from the mobile.
For a BPSK system, the bit error probability is given by

 2E 
p BPSK = Q  --------b- (4.34)
 N0 
The bit error probabilities over M-branch diversity with selection, equal-gain, and maximal-
ratio combining are given as [5]
• Selection combining
M–1
M ( –1 )
k
 M – 1 ⋅ ------------ ρc
p BPSK, M = -----
2 ∑  K  k+1
- ⋅ 1 – -----------------------
k + 1 + ρc
(4.35)
k=0

• Equal-gain combining
M–1

∑ 
M M + k – 1 k
p BPSK, M = ( p BPSK ) ⋅ ⋅ ( 1 – p BPSK ) (4.36)
k 
k=0

• Maximal-ratio combining

1 M
2M – 1  -------
p BPSK, M ≈  ⋅ - (4.37)
 M   4ρ c

Eb
where ρ c = ---------------
- = average SNR per diversity branch
N0 ⋅ M
ρc 
p BPSK = 1--- ⋅  1 – ---------------
2  1 + ρ c

EXAMPLE 4.1
Compare the bit error performance of BPSK modulation having SNR =10 dB with 2-branch diver-
sity using selection, equal-gain, and maximal-ratio combining.
• Selection combining
1 k
1 ( – 1 )-
∑  k ⋅ ------------
10 1
p BPSK, 2 = ⋅ 1 – ------------------------ = [ 1 – 0.95346 ] – --- [ 1 – 0.91287 ] = 0.002976
k+1 k + 1 + 10 2
k=0
72 Ch. 4 • Diversity, Combining, and Antennas

Table 4-4 Bit Error Performance Comparison

Performance with Respect to

Combining Type PBPSK,2 Maximal-Ratio Combining
Maximal-ratio 0.001875 1.0
Equal-gain 0.002153 1.148
Selection 0.002976 1.587

• Equal-gain combining

p BPSK = --- ⋅  1 – ------ = 0.02327

1 10
2  11
1

∑  1 +k k ⋅ ( 1 – 0.02327 )
2 k
p BPSK, 2 = ( 0.02327 ) ⋅ = 0.002153
k=0

• Maximal-ratio combining
2
p BPSK, 2 ≈   ⋅  --------------- = 0.001875
3 1
 2  4 × 10

The bit error performance of equal-gain and selection combining is about 15% and 59%
worse than the maximal-ratio combining, respectively (see Table 4-4).

4.6 Examples of Base Station and Mobile Antennas

While the simple dipole antenna (Fig. 4-7) is the reference example for antenna specifications,
most practical antenna designs aim to improve on the gain of the dipole antenna. Practical
antenna design must consider the following issues:

λ/4 λ/4

Feed Line

Figure 4-7 Dipole Antenna

Examples of Base Station and Mobile Antennas 73

• Antenna pattern. Closely related to the gain of the antenna is the antenna pattern. As
gain is increased, the beam width is decreased. This can be an advantage or a disadvan-
tage depending on the antenna orientation and the needs of the system design.
• Bandwidth. The antenna must operate over the full range of frequencies in use for the
cellular or PCS system. If the antenna bandwidth is small, channels at the edge of the
band may not receive signals as well as those near the band center.
• Gain. The higher the gain of the antenna, the lower the power that is necessary at the
transmitter. Since the antenna is purchased once and the transmitter power is purchased
continuously, high-gain antennas save money by using less electricity, thus conserving
the natural resources used to create the electricity.
• Ground plane. Some antennas must be mounted above a reflecting surface to function
correctly. For example, a quarter-wave antenna is one-half of a dipole and requires that
the other half of the dipole be developed by a mirror image below a ground plane. This
can be used to advantage in designing antennas for vehicles, but it is a disadvantage
when base station antennas, which are high above the surface of the earth, are
designed.
• Height. The higher the antenna, the better the coverage of the system. However, if the
coverage of the system is too good, interference from other cells may become trouble-
some. In an interference-limited system, all levels scale equally so that, at the first
order, there will not be a problem. However, since radio wave propagation is statistical,
there may be locations where good propagation exists from a point far removed from a
base station. The higher the base station antenna, the more likely that these anomalous
events will occur.
• Input impedance. Most cables used as feed line from the transmitter/receiver to the
antenna are either 50 ohms or 72/75 ohms. If the input impedance of the antenna is far
removed from either of these values, it will be difficult to get the antenna to accept the
power delivered to it and its efficiency η will be low.
• Mechanical rigidity. If the antenna flexes in the wind, it will introduce an additional
fading component to the received signal. Ultimately, the continuous flexing will cause
metal fatigue and mechanical failure of the antenna.
• Polarization. For wireless cellular and PCS communications, a vertical antenna is the
easiest to mount on a vehicle; therefore vertical polarization has been standardized. In
general, horizontal or vertical polarization works equally well.
With this background, we will examine some simple antennas that are used for base and
mobile operation.

4.6.1 Quarter-Wave Vertical Antenna

The simplest antenna for a vehicle is the quarter-wave vertical (see Fig. 4-8). A length of
wire a quarter of a wavelength long is mounted on the roof of the vehicle. With metal vehicles,
the other half of the dipole is developed in the image in the ground plane. Since a vertical dipole
74 Ch. 4 • Diversity, Combining, and Antennas

antenna has an omnidirectional pattern [6], the quarter-wave vertical has an omnidirectional
pattern. The gain of the antenna is the same as that of a dipole (0 dB dipole [dBd] or 2.1 isotro-
pic dipole [dBi]). The impedance of a quarter-wave vertical is 36.5 ohms and requires a match-
ing transformer for proper feeding of the antenna.

λ/4

Antenna

Image of Antenna Ground Plane

Figure 4-8 Quarter-Wave Vertical Antennas

λ/2

λ/2

λ/2

Each antenna element

is 1/2 wavelength long.
λ/2
Spacing between
elements is 1/2 wavelength.

λ/2

λ/2

λ/2

Figure 4-9 Stacked Vertical Dipoles

Examples of Base Station and Mobile Antennas 75

4.6.2 Stacked Dipoles

Since a vertical dipole has an omnidirectional pattern, three or more dipoles can be
stacked vertically to produce increased gain and maintain the omni pattern. A typical base sta-
tion antenna will have four half-wave dipoles spaced apart by one wavelength vertically (see
Fig. 4-9). All the antenna elements are fed in phase with a signal from the transmitter. The result-
ant pattern is omnidirectional in the horizontal plane and has an 8.6-dBi gain on the horizon (0-
degree elevation) with a vertical beamwidth of ±6.5 degrees. Stacked dipoles have an impedance
of 63 ohms; thus a matching transformer is necessary, but it is easier to build than the one for the
quarter-wave vertical.
A variation of the base station antenna for vehicles uses a half-wave dipole above a quarter-
wave vertical (see Fig. 4-10). The other half of the antenna is in the ground plane image. The two
elements are decoupled from each other by a quarter-wavelength-long decoupling coil. This is the
most common cellular antenna for vehicles. It has a gain of 7–10 dBi.

λ/2

λ/4
Antenna

λ/4

Ground Plane

Image of Antenna

Figure 4-10 Half-Wave over Quarter-Wave Antenna Mounted on Vehicle

76 Ch. 4 • Diversity, Combining, and Antennas

4.6.3 Corner Reflectors

The antennas we have discussed so far have omnidirectional patterns. A directional
antenna at a base station can improve the SNR of the system and thus improve its spectral effi-
ciency. In 1939 Kraus [4] designed the corner reflector antenna consisting of a vertical dipole
and two sheets of metal at a 45, 60, or 90 degree angle (see Fig. 4-11). The impedance and gain
of the antenna depend on the angle of the corner and the spacing from the corner to the dipole
antenna. The gain of the antenna varies from 7 to 13 dBd, and the impedance varies from 0 to
150 ohms. A colinear antenna can be used in place of the dipole for additional gain.

4.6.4 Smart Antenna

Smart antennas include either a Switched Beam System (SBS) or an adaptive antenna sys-
tem [7] (see Figs. 4-12 and 4-13). The SBS uses multiple fixed beams in a sector and a switch to
select the best beam for receiving a particular signal. In an adaptive antenna system, the received
signals by multiple antennas are weighted and combined to maximize the SNR using either the
minimum mean-square error (MMSE) or least squares (LS) criterion. The advantage of the
adaptive antenna system over the SBS is that, in addition to the M-fold antenna gain, it provides
an M-fold diversity gain.
Smart antennas offer a broad range of methods to improve system performance. They pro-
vide enhanced coverage through range expansion, hole filling, and better building penetration.
Given the same transmitter power output at the base station and the mobile unit, smart
antennas can increase range by increasing the gain of the base station antenna. The M-fold
antenna gain increases the range by a factor of M1/γ where γ is the path loss exponent; it reduces
the number of base stations to cover a given area by M2/γ [7]. An SBS with M beams can increase
the system capacity by a factor of M by reducing the number of interferers. An adaptive antenna

Reflector

Angle of reflector equals

45, 60, or 90 degrees.

Antenna

Figure 4-11 Corner Reflector

Examples of Base Station and Mobile Antennas 77

Switched Beam System (SBS) Adaptive Beam

Mobile 1
Interferer
Mobile Direct Ray
Mobile
1
Direct Ray

Reflected
Mobile Rays
2

Reflected Ray Mobile 2 Interferer

Select from—or use—multiple fixed Adaptively weight and combine multiple
antenna beams to optimize performance. antenna elements to optimize performance.

Figure 4-12 Smart Antennas

Spot Beam on
a Handset

Antenna Elements

Gain and Phase Shift

To Transmitter
or Receiver
Figure 4-13 Adaptive Array Smart Antenna
78 Ch. 4 • Diversity, Combining, and Antennas

system can provide some additional gain by suppressing interferers further. However, since there
are so many interferers, the additional gain may not be worth the complexity.

4.7 Summary
In this chapter we considered the role of antennas in the wireless system. We also presented the
concepts of diversity reception where multiple signals are combined to improve the SNR of the
system.

4.8 References
1. Garg, V. K., and Wilkes, J. E., Wireless and Personal Communications Systems, Prentice
Hall, Upper Saddle River, NJ, 1996.
2. Halpern, S. W., “The Theory of Operation of an Equal-Gain Predication Regenerative Diver-
sity Combiner with Rayleigh Fading Channel,” IEEE Transactions on Communication Tech-
nology, COM-22 (8), August 1974, pp. 1099–106.
3. Kahn, L. R., “Radio Squarer,” Proceedings of the IRE 42, November 1954, p. 1704.
4. Kraus, J. D., Antenna, McGraw-Hill, New York, 1988.
5. Lee, J. S., and Miller, L. E., CDMA Systems Engineering Handbook, Artech House, Boston,
1998.
6. Lee, W. C. Y., “Antenna Spacing Requirements for a Mobile Radio Base Station Diversity,”
Bell System Technical Journal 50(6), July–August 1971.
7. Liberti, J. C. Jr., and Rappaport, T. S., Smart Antennas for Wireless Communications, Pren-
tice Hall, Upper Saddle River, NJ, 1999.
8. Mahrotra, A., Cellular Radio Performance Engineering, Artech House, Boston, 1994.
9. Price, R., and Green, P. E. Jr., “A Communication Technique for Multipath Channels,” Pro-
ceedings of the IRE 46, March 1958, pp. 555–70.
10. Rappaport, T. S., Wireless Communications: Principle and Practice, Prentice Hall, Upper
Saddle River, NJ, 1996.
11. Special Issue on Smart Antennas, IEEE Personal Communications 5(1), February 1998.
 C H A P T E R 5

IS-95 System
Architecture

5.1 Introduction
A wireless system—whether for cellular operation (Band Class 0), i.e., 850-MHz band, or for
PCS operation (Band Class 1), i.e., 1.8-GHz band—must support communication with the
mobile station and interact with the Public-Switched Telephone Network (PSTN). As the mobile
station changes its location during a call, the wireless system must insure that the connection
between the mobile station and the PSTN is maintained.
A wireless system consists of discrete logical components. These components may be dis-
crete physical entities or they may be physically located with another logical entity. These func-
tional entities must interact in order to coordinate operation. Such interaction is achieved by
messaging over interfaces between two entities. If two functional entities are physically separate
and if the interface is standardized, the service provider can purchase products from different
manufacturers. However, successful operation is not guaranteed since the associated standard
often does not cover all facets of operation. Manufacturers need to cooperate to eliminate differ-
ences that jeopardize proper interaction.
This chapter first will cover the functional entities of the wireless network and the inter-
faces between the entities that have been standardized by the wireless communication industry.
Then it examines the activities of the International Telecommunication Union (ITU) to add Intel-
ligent Network (IN) to wireless systems.

5.2 TR-45/TR-46 Reference Model

Key to the North American systems is the use of a common reference model from the cellular
standards group TR-45. When work started on PCS, the TR-46 standards group adopted the TR-
45 reference model for PCS, but with some minor changes in the names of the elements (see Fig.
5-1). A second reference model was proposed by T1P1, but it is similar to the TR45/46 model.

79
 80

OS AUX IWF

PSPDN Rx
TE2
WPT0 O X L
Pi
Base Station

Sm WPT1 Um BTS A bis BSC A MSC Ai PSTN W DCE Rv

TE1 TE2

Figure 5-1 TR-45/46 Reference Model

Di
TE2
Rm TAP F B
C
ISDN S
TE1
MSC Mi
Rm WPT2
TE2

I HLR D VLR PLMN TA R TE2

MS
H External Networks
I
DMH
EIR AC
G
Mobility Managers
(Terminal & Personal) Other
WPT0 : Wireless Personal Terminal VLRs
Um: Interface Between BTS and Wireless Personal Terminal
Sm, Rm: Interface Between WPT and Terminal Equipment
DCE: Data Communication Equipment
Rv: Interface Between DCE and Terminal Equipment Type 2
Rx: Interface Between PSPDN and Terminal Equipment Type 2
The names of each of the network elements are similar, and some of the functionality is parti-
Ch. 5 • IS-95 System Architecture

how mobility is handled. Mobility is the capability for users to place and receive calls in systems
tioned differently between the models. The main difference between the two reference models is
TR-45/TR-46 Reference Model 81

other than their home system. In the T1P1 reference model, the user data and the terminal data
are separate; thus, users can communicate with the network via different mobile stations. In the
TR-45/46 reference model, only terminal mobility is supported. A user can place or receive calls
at only one terminal (the one the network has identified as owned by the user). The T1P1 func-
tionality is migrating toward independent terminals and user mobility, but all aspects of it are not
currently supported. We also discuss a Wireless Intelligent Network (WIN) architecture and ref-
erence model that overcomes some of the mobility problems in the current architecture.
The main elements of the reference model are
• Mobile Station (MS). The MS terminates the radio path on the user side and enables the
user to gain access to services from the network. The MS can be a stand-alone device or
can have other devices (e.g., personal computers, fax machines) connected to it.
• Base Station (BS). The BS terminates the radio path and connects to the Mobile
Switching Center (MSC). The BS is often segmented into the BTS and the BSC.
◆ Base Transceiver System (BTS). The BTS consists of one or more transceivers
placed at a single location and terminates the radio path on the network side. The
BTS may be co-located with a BSC or may be independently located.
◆ Base Station Controller (BSC). The BSC is the control and management system
for one or more BTSs. The BSC exchanges messages with both the BTS and the
MSC. Some signaling messages may pass through the BSC transparently.
• Mobile Switching Center (MSC). The MSC is an automatic system that interfaces the
user traffic from the wireless network with the wireline network or other wireless net-
works. The MSC functions as one or more of the following:
◆ Anchor MSC: first MSC providing radio contact in a call
◆ Border MSC: an MSC controlling BTSs adjacent to the location of a mobile
station
◆ Candidate MSC: an MSC that could possibly accept a call or a handoff
◆ Originating MSC: the MSC directing an incoming call toward a mobile station
◆ Remote MSC: the MSC at the other end of an intersystem handoff trunk
◆ Serving MSC: the MSC currently providing service to a call
◆ Tandem MSC: an MSC providing only trunk connections for a call in which a
handoff has occurred
◆ Target MSC: the MSC selected for a handoff
◆ Visited MSC: an MSC providing service to the mobile station
• Home Location Register (HLR). The HLR is the functional unit that manages mobile
subscribers by maintaining all subscriber information (e.g., electronic serial number,
directory number, international mobile station identification, user profiles, current loca-
tion). The HLR may be co-located with an MSC as an integral part of the MSC or may
be independent of the MSC. One HLR can serve multiple MSCs, or an HLR may be
distributed over multiple locations.
82 Ch. 5 • IS-95 System Architecture

• Data Message Handler (DMH). The DMH is used to collect billing data.
• Visited Location Register (VLR). The VLR is linked to one or more MSCs and is the
functional unit that dynamically stores subscriber information (e.g., the user’s elec-
tronic serial number [ESN], directory number, user profile information) obtained from
the user’s HLR when the subscriber is located in the area covered by the VLR. When a
roaming MS enters a new service area covered by an MSC, the MSC informs the asso-
ciated VLR about the MS by querying the HLR after the MS goes through a registration
procedure.
• Authentication Center (AC). The AC manages the authentication or encryption infor-
mation associated with an individual subscriber. The AC may be located within an HLR
or MSC or may be located independently of both.
• Equipment Identity Register (EIR). The EIR provides information about the mobile
station for record purposes. The EIR may be located within an MSC or may be located
independently of it.
• Operations System (OS). The OS is responsible for overall management of the wire-
less network.
• Interworking Function (IWF). The IWF enables the MSC to communicate with other
networks.
• External Networks. These are other communications networks—the Public-Switched
Telephone Network (PSTN), the Integrated Services Digital Network (ISDN), the Pub-
lic Land Mobile Network (PLMN), and the Public-Switched Packet Data Network
(PSPDN).
The following interfaces are defined between the various elements of the system:
• BS to MSC (A-Interface). This interface between the base station and the MSC sup-
ports signaling and traffic (both voice and data). A-Interface protocols have been
defined using SS7, ISDN BRI/PRI, and frame relay transport (TIA IS-634).
• BTS-to-BSC Interface (Abis). If the base station is segmented into a BTS and BSC,
this internal interface is defined.
• MSC-to-PSTN Interface (Ai). This interface is defined as an analog interface using
either Dual-Tone Multifrequency (DTMF) signaling or Multifrequency (MF) signaling.
• MSC to VLR (B-Interface). This interface is defined in the TIA IS-41 protocol speci-
fication [8].
• MSC to HLR (C-Interface). This interface is defined in the TIA IS-41 protocol speci-
fication [8].
• HLR to VLR (D-Interface). This interface is the signaling interface between an HLR
and a VLR and is based on SS7. It is currently defined in the TIA IS-41 protocol speci-
fication [8].
• MSC to ISDN (Di-Interface). This is the digital interface to the PSTN and is a T1
interface (24 channels of 64 kbps) and uses Q.931 signaling.
Functional Model Based on Reference Model 83

• MSC-MSC (E-Interface). This interface is the traffic and signaling interface between
wireless networks. It is currently defined in the TIA IS-41 protocol specification.
• MSC to EIR (F-Interface). Since the EIR is not yet defined, the protocol for this inter-
face is not defined.
• VLR to VLR (G-Interface). When communication is needed between VLRs, this
interface is used. It is defined by TIA IS-41.
• HLR to AC (H-Interface). The protocol for this interface is not defined.
• DMH to MSC (I-Interface). The protocol for this interface is defined in the IS-124
[9].
• MSC to IWF (L-Interface). This interface is defined by the interworking function.
• MSC to PLMN (Mi-Interface). This interface is to another wireless network.
• MSC to OS (O-Interface). This is the interface to the OS. It is currently being defined
in ATIS standard body T1M1.
• MSC to PSPDN (Pi-Interface). This interface is defined by the packet network that is
connected to the MSC.
• Terminal Adapter (TA) to Terminal Equipment (TE) (R-Interface). These inter-
faces are specific for each type of terminal connected to an MS.
• ISDN to TE (S-Interface). This interface is outside the scope of PCS and is defined
within the ISDN system.
• BS to MS (Um-Interface). This is the air interface.
• PSTN to DCE (W-Interface). This interface is outside the scope of PCS and is defined
within the PSTN system.
• MSC to AUX (X-Interface). This interface depends on the auxiliary equipment con-
nected to the MSC.
• BS Management Application Part (BSMAP). Message sent between BS and MSC.
• Direct Transfer Application Part (DTAP). Message sent between MS and MSC.

5.3 Functional Model Based on Reference Model

Fig. 5-2 shows the functional model derived from the reference model. Several physical scenarios
can be developed using the functional entities shown in Fig. 5-2. Fig. 5-3 shows the Functional
Entity (FE) grouping in which the physical interface between the radio system (RS) and the
switching system platform (SSP) carries both the call control and mobility management messages.
• Radio Terminal Function (RTF) FE. This is the subscriber unit (SU). The only phys-
ical interface is to the Radio System (RS) through the air interface.
• Radio Control Function (RCF) FE and Radio Access Control Function (RACF)
FE. These are included in the RS. Combining these FEs onto the same platform allows
air-interface-specific functions (such as those that would impact handoff) to be isolated
from the other interfaces. OS information, including performance data and accounting
records, is generated, collected, and formatted on this platform. There is only one phys-
ical interface to SSP to carry both the call control and mobility management signaling.
84 Ch. 5 • IS-95 System Architecture

• Service Switching Function (SSF)/ Call Control Function (CCF) FE. The SSF/CCF
FE is contained in SSP and provides interfaces to operator services, E911, international
calls, and network repair/maintenance centers. Physical interfaces for this collection
include: to the RS, to the mobility management platform, to the internal peripheral (IP),
and to other SSPs and external networks.

TMD
PMD FE6
To other FE5
RACFs SCF SDF

TMD
PMD
FE3
RACF SRF

FE7
RASC
FE4
FE1 FE2 SSF FE4 SSF

RTF RCF CCF CCF

RT RT PSC PSC
RPT RPI
RPC

CCAF

FE9
CCAF: Call Control Agent Function RPT: Radio Port Terminal Transport & Signaling
CCF: Call Control Function RT: Radio Terminal Signaling Relationship
PMD: Personal Mobility Data SCF: Service Control Function Air Interface
PSC: Personal Switching Center SDF: Service Data Function
RACF: Radio Access Control Function SRF: Specialized Resource Function
RASC: Radio Access System Control SSF: Service Switching Function
RCF: Radio Control Function TMD: Terminal Mobility Data
RPI: Radio Port Interface

Figure 5-2 Functional Model Derived from the Reference Model

Other Platform
Other SDF
RACFs

SCF
MMAP
Interface
SCP/HLR/VLR SRF

RACF
IP
RTF SSF CCF

RCF
Other
SSP SSPs
SU
RS

MMAP: Mobility Management Application Port Transport & Signaling

Signaling Relationship

Figure 5-3 Functional Entity Groupings

Wireless Intelligent Network 85

• Specialized Resource Function (SRF) FE and Data Interworking Function. These

are contained in the IP. Physical interfaces for this collection include one to the SSP
and another to the mobility management platform.
Individually the SSF/CCF FE and CCF FE represent interswitch and internetwork func-
tional entity collections and physical interfaces.
As shown in Fig. 5-3, the only interface to the RS is from the SSP. There is no direct phys-
ical path between the RS and the switch control point (SCP)/VLR. All operations to or from the
RS pass through the SSP, whether or not the SSP consumes or produces the operation.

5.4 Wireless Intelligent Network

For the wireline network, BellCore has defined a set of protocols called Intelligent Network (IN)
[4] that enable a rich set of new telephony capabilities to be generated without additional soft-
ware development in the central switch. As IN has grown in popularity for the wireline network,
the wireless network has also embraced the concept. IN improves the ability to locate and effi-
ciently direct calls to roaming mobile stations and provides other advanced features. In this sec-
tion we examine the work of the ITU to add IN to the wireless systems.
An examination of the needs of intelligent networking and the mobility aspects of wireless
communications shows several similarities. The wireline user should have one phone number
anywhere in the world. Anyone dialing that number would have their calls routed to the owner’s
destination independent of number’s owner’s location. The services that a user has available in
the home or office should also be available in a home or office where the user is visiting as well
as at public phones.
The functions of the HLR/VLR in a wireless network are similar to the functions of an
SCP in a wireline network.
The ITU has adapted IN for use in wireless networks. Question 8 of Study Group (SG) 11
[7] has defined a communications control plane (Fig. 5-4) and a radio resource control plane
(Fig. 5-5) using IN concepts.
On the network side of the communications control plane, the following functional ele-
ments are defined:

• The Bearer Control Function (BCF) provides those bearer functions needed to process
handoffs. A common example of this is a conference bridge to support soft handoffs.
• The Bearer Control Function for the radio bearer (BCFr) provides the functions
necessary to select bearer functions and radio resources. It also detects and responds to
pages from the network and performs handoff processing. Some example bearer func-
tions are PCM voice, ADPCM voice, packet data, and circuit-switched data.
• The Call Control Function-Enhanced (CCF') provides the call and connection con-
trol in the network. For example, it establishes, maintains, and releases call instances
requested by the CCAF'; provides IN triggers to the SSF; controls bearer connection
elements in the network.
86 Ch. 5 • IS-95 System Architecture

• The Service Access Control Function (SACF) provides the network side of mobility
management functions. Some examples are registration and location of the mobile station.
• The Service Control Function (SCF) contains the service and mobility control logic
and call processing to support the functions of a mobile terminal.

SDF
Mobile Side Network Side

MCF SACF SCF

SSF

UIMF CCAF' CCF' SRF

TIMF TACAF TACF RCF

BCAF BCFr

Figure 5-4 Communications Control Plane

Mobile Side Network Side

Mobile
MRRC RRC Radio Resource Control
Radio Resource Control

Mobile
Radio Frequency
Radio
MRTR RFTR Transmission
Transmission
and Reception
and Reception

Figure 5-5 Radio Resource Control Plane

Wireless Intelligent Network 87

• The Service Data Function (SDF) provides data storage and data access in support of
mobility management and security data for the network.
• The Specialized Resource Function (SRF) provides the specialized functions needed
to support execution of IN services. Some examples are dialed digit receivers, confer-
ence bridges, and announcement generators.
• The Service Switching Function (SSF) provides the functions required for interaction
between the CCF' and the SCF. It supports extensions of the CCF' logic to recognize
IN triggers and interact with the SCF. It manages the signaling between the CCF' and
the SCF and modifies functions in the CCF' to process IN services under control of the
SCF.
• The Terminal Access Control Function (TACF) provides control of the connection
between the mobile station and the network. It provides paging of mobile stations, page
response handling, handoff decision, and completion. It also provides trigger access to
IN functionality.
On the mobile side, the following functional elements are defined:
• The Bearer Control Agent Function (BCAF) establishes, maintains, modifies, and
releases bearer connections between the mobile station and the network.
• The Call Control Agent Function-Enhanced (CCAF') supports the call processing
functions of the mobile station.
• The Mobile Control Function (MCF) supports the mobility management functions of
the mobile station.
• The Terminal Access Control Agent Function (TACAF) provides the functions nec-
essary to select bearer functions and radio resources. It also detects and responds to
pages from the network and performs handoff processing.
• The Terminal Identification Management Function (TIMF) stores the terminal-
related security information. It provides terminal identification to other functional ele-
ments and provides the terminal authentication and cryptographic calculations.
• The User Identification Management Function (UIMF) provides user-related secu-
rity information similar to the TIMF.
Both the TIMF and the UIMF can be stored in either the mobile station or a separate secu-
rity module often implemented in a smart card.
The radio resource control plane (Fig. 5-5) is responsible for assigning and supervising
radio resources. Four function entities (two on the mobile side and two on the network side) per-
form the functions of the radio access subsystem.
• The Radio Resource Control (RRC) provides functionality in the network to select
radio resources (channels, spreading codes, etc.), make handoff decisions, control the
RF power of the mobile station and provide system information broadcasting.
• The Radio Frequency Transmission and Reception (RFTR) provides the network
side of the radio channel. It provides the radio channel encryption and decryption (if
88 Ch. 5 • IS-95 System Architecture

used) and channel quality estimation (data error rates for digital channels), sets the RF
power of the mobile station, and detects when the mobile station accesses the system.
• The Mobile Radio Resource Control (MRRC) processes the mobile side of the radio
resource selection. It provides base station selection during start-up, mobile-assisted
handoff control, and system access control.
• The Mobile Radio Transmission and Reception (MRTR) provides the mobile side of
the radio channel and performs functions similar to RFTR.
The two control planes interact to provide services to the mobile station and the network.
The TR-46 PCS network reference model working group has generated a simplified ver-
sion of the ITU model and has explicitly shown the operations functions on the model. The IN
Function Reference Model for PCS (Fig. 5-6) has many of the functional elements with the same
name as those of the ITU model. The differences are that
• The Radio Terminal Function (RTF) contains all functionality of the mobile side of
the reference model.
• The Radio Access Control Function (RACF) is similar to the SACF and provides
mobility management functions.

Mobile Side Network Side

SDF

RACF SCF

SSF

RTF RCF CCF SRF

to other FEs
CCAF

SMF

Figure 5-6 IN Functional Reference Model for PCS

Summary 89

• The Radio Control Function (RCF) provides the capabilities of the TACF, the BCFr,
and the BCF in the ITU model. It provides the radio ports and the radio port controller
capabilities in the PCS network.
• The Call Control Agent Function (CCAF) provides access to the wireless network by
wireline users.
• The Service Management Function (SMF) provides the network management func-
tions for each functional element.

With the ITU reference model, the functionality to support all of the features and capabili-
ties for a wireless network can be partitioned into different functional elements and still meet the
variety of national and worldwide standards. Therefore no exact partitioning of the functions of
the network reference models can be made. We encourage you to examine the standards and
implementations of the various manufacturers for partitioning examples.
Around the world, two common mobile application parts (MAP) are used. In GSM sys-
tems deployed worldwide, the GSM MAP is used. In the systems that were originally deployed
in the United States, the IS-41 MAP is used. The use of two different MAPs makes roaming
between the two systems difficult. The use of the IN application part with enhancements will
offer the opportunity for a common MAP for all mobile and wireline services.

5.5 Summary
This chapter presented the TR-45/46 reference model, which is used as the basis by standards
committees. The main elements of this model are the Mobile Station (MS), Base Station (BS),
Mobile Switching Center (MSC), Home Location Register (HLR), and Visited Location Regis-
ter (VLR). Next we discussed the MSC-BS interface (TIA IS-634), which standardizes the mes-
saging between the base station and the MSC. Messages between the BS and the MSC are
categorized into two types: Base Station Application Part (BSAP) and Direct Transfer Applica-
tion Part (DTAP). Messages can be associated with the functions of call processing and supple-
mentary services, radio resource management, mobility management, and terrestrial facility
management. The effects of a CDMA system upon the architecture are emphasized. Finally, we
focused on the concepts of wireless intelligent networking. Current systems have difficulty with
roaming because of the support of two MAPs in the world (GSM and IS-41). With wireless
intelligent networking and extensions to the IN MAP, true worldwide roaming may become pos-
sible once a common frequency band is available.

5.6 References
1. American National Standards Institute, Inc., “Signaling System No. 7 (SS7)—Message
Transfer Part (MTP),” ANSI T1.111-1992, June 1992.
2. American National Standards Institute, Inc., “Signaling System No. 7 (SS7)—Signaling
Connection Control Part (SCCP),” ANSI T1.112-1992, June 1992.
90 Ch. 5 • IS-95 System Architecture

3. American National Standards Institute, Inc., “Synchronization Interface Standards for Digi-
tal Networks,” 1087, ANSI T1.101-1987, March 1987.
4. Bellcore, “Advanced Intelligent Network (AIN) Release 1 Switching System Generic
Requirements,” TA-NWT-001123, Issue 1, May 1991.
5. Committee T1, “Stage 2 Service Description for Circuit Mode Switched Bearer Services,”
Draft T1.704.
6. Committee T1—Telecommunications, “A Technical Report on Network Capabilities, Archi-
tectures, and Interfaces for Personal Communications,” T1 Technical Report #34, May
1994.
7. ITU Study Group 11, “Version 1.1.0 of Draft New Recommendation Q.FNA, Network
Functional Model for FPLMTS,” Document Q8/TYO-50, September 15, 1995.
8. TIA Interim Standard, IS-41 C, “Cellular Radiotelecommunications Intersystem Opera-
tions,” January 1996.
9. TIA IS-124, “Cellular Radio Telecommunications Intersystem Non-Signaling Data Message
Handlers (DMH),” 1994.
10. TIA IS-634, “MSC-BS Interface for Public 800 MHz,” Revision 0, 1995.
11. TIA SP-2977, “Cellular Features Description,” Prepublication Version, March 14, 1995.
12. TIA TR-45 Reference Model, 1990.
13. TIA TR-46 Reference Model, 1991.
 C H A P T E R 6

IS-95 CDMA Air

Interface

6.1 Introduction
This chapter provides a high-level description of the IS-95 CDMA air interface. The intent of
this chapter is to lay a foundation for the system details given in chapters 7–11. We describe
important aspects of the forward link (base station to mobile) and reverse link (mobile to base
station) and include modulation parameters for the channels.

6.2 TIA IS-95 CDMA System

The TIA IS-95 CDMA system operates on the same frequency band as the Advanced Mobile
Phone System (AMPS) using Frequency Division Duplex (FDD) with 25 MHz in each direc-
tion.* For Band Class 0 the reverse link (mobile to base station) uses frequencies from 869 to
894 MHz, and forward link (base station to mobile) uses frequencies from 824 to 849 MHz. The
mobile station supports CDMA operations on AMPS channel numbers 1013 through 1023, 1
through 311, 356 through 644, 689 through 694, and 739 through 777 inclusive. The PCS ver-
sion of the specification J-STD-008 for use at 1800-MHz (Band Class 1) has been also standard-
ized. The CDMA channels are defined in terms of an RF frequency and code sequence. Sixty-
four Walsh functions are used to identify the forward (down) link channels, whereas 64 long PN
codes are used for the identification of the reverse (up) link channels. Figs. 6-1 and 6-2 show the
CDMA forward (down) link and reverse (up) link channels structure. The modulation and cod-
ing features of the IS-95 CDMA system are listed in Table 6-1. We will discuss the system in
detail in chapters 7–11.

* The frequency spectrum for the A-System cellular service provider is split such that the spectrum is not divisible by
1.25 MHz. Thus, the A-System cellular provider cannot partition the spectrum into 10 1.25-MHz CDMA channels.
This restriction is not imposed for the B-System, however.

91
92 Ch. 6 • IS-95 CDMA Air Interface

Modulation and coding details for the reverse link and forward link channels differ. Pilot
signals are transmitted by each cell to assist the mobile radio in acquiring and tracking the cell
site forward link signals. The strong coding helps these radios to operate effectively at Eb /N0
ratio of 5- to 7-dB range.

Forward CDMA Link (1.23-

MHz channel transmitted
by base station)

PCH : Paging Channel

FTCH : Forward Traffic Channel

Pilot Sync PCH PCH Code Code Code Code Code

#1 #7 #1 #N #P #S #55

W0 W32 W1 W7 W8 W 63

Fundamental Mobile Power Fundamental Mobile Power Supplementary

Code Channel Control Code Channel Control Code Channel
Data Subchannel Data Subchannel Data

FTCH with One Code Channel FTCH with Multiple Code Channels

Figure 6-1 Forward Link Channels Structure

Reverse CDMA Link (1.23-MHz

Channel Received by Base Station)

Access Access Access Access Traffic Traffic

Channel Channel Channel Channel Channel Channel
(PCH1) (PCH1) (PCHP) (PCHP) #1 #T

Fundamental Supplementary Supplementary Supplementary

Code Code Code Code
Channel Channel Channel Channel

Addressed by Long Code PN

Figure 6-2 Reverse Link Channels Structure

TIA IS-95 CDMA System 93

Table 6-1 Modulation and Coding Features of the IS-95 CDMA System
Modulation Quadrature Phase-Shift Keying (QPSK)
Chip rate 1.2288 Mcps
Nominal data rate (RS1) 9600 bps
Filtered bandwidth 1.23 MHz
Coding convolution with Viterbi decoding
Interleaving with 20-ms span

The CDMA system uses power control and voice activation to minimize mutual interfer-
ence. Voice activation is provided by using a variable-rate vocoder that operates at a maximum
rate of 8 kbps to a minimum rate of 1 kbps for Rate Set 1 (RS1), depending on the level of voice
activity. With the decreased data rate, the power control circuits reduce the transmitter power to
achieve the same bit error rate. A precise power control, along with voice activation circuits, is
critical to avoid the excessive transmitter signal power responsible for contributing the overall
interference in the system. A coding algorithm at 13.3 kbps for RS2 is also supported.
A time interleaver with a 20-ms span is used with error-control coding to overcome rapid
multipath fading and shadowing. The time span used is the same as the time frame of the voice
compression algorithm.
The CDMA radio uses a RAKE receiver to take advantage of a multipath delay greater
than 1 µs, which occurs commonly in cellular/PCS networks in urban and suburban areas.

6.2.1 Forward Link

In this section, we summarize the operation of the forward link. For more details refer to
chapter 7. The forward link channels (see Fig. 6-1) include one pilot channel, one synchroniza-
tion (sync) channel, up to seven paging channels, and a number of forward traffic channels (if
multiple carriers are implemented, the pilot channel and sync channels do not need to be dupli-
cated). Each forward traffic channel contains one forward fundamental code channel and may
contain one to seven forward supplemental code channels. The information on each channel is
modulated by the appropriate Walsh function and then modulated by a quadrature pair of PN
sequences at a fixed chip rate of 1.2288 Mcps. The pilot channel is always assigned to code
channel number 0. If the sync channel is present, it is given the code channel number 32. When-
ever paging channels are present, they are assigned the code channels number 1 through 7
(inclusive), in sequence. The remaining code channels are used by forward traffic channels.
The sync channel always operates at a fixed data rate of 1200 bps and is convolutionally
encoded to 2400 bps, repeated to 4800 bps, and interleaved over the period of the pilot pseudo-
random binary sequence. Each of the interleaved symbols uses four Walsh symbols.
The forward traffic channels are grouped into sets. RS1 has four elements—9600, 4800,
2400, and 1200 bps. RS2 contains four elements—14,400, 7200, 3600, and 1800 bps. All radio
94 Ch. 6 • IS-95 CDMA Air Interface

systems support RS1 on the forward traffic channels. RS2 is optionally supported on the forward
traffic channels. When a radio system supports a rate set, it supports all four elements of the set.
Speech is encoded using a variable-rate vocoder to generate forward-traffic-channel data
depending on voice activity. Since the frame duration is fixed at 20 ms, the number of bits per
frame varies according to the traffic rate. Half-rate convolutional encoding is used, which dou-
bles the traffic rate to give rates from 2400 to 19,200 symbols per second. Interleaving is per-
formed over 20 ms. A long code of 242 – 1 (= 4.4 × 1012) is generated containing the user’s ESN
embedded in the MS long code mask (with voice privacy, the MS long-code mask does not use
the ESN). The scrambled data is multiplexed with power control information that steals bits
from the scrambled data. The multiplexed signal remains at 19,200 bps and is changed to 1.2288
Mcps by the Walsh code Wi assigned to the ith user traffic channel (TCH). The signal is spread at
1.2288 Mcps by pilot quadrature pseudorandom binary sequence signals, and the resulting
quadrature signals are then weighted. The power level of the traffic channel depends on its data
transmission rate.
The paging channels (PCH) provide the mobile stations with system information and
instructions, in addition to acknowledging messages following access requests on the mobile
stations’ access channels. The paging channel data is processed in a manner similar to the traffic
channel data. However, there is no variation in the power level on a per-frame basis. The 42-bit
mask is used to generate the long code. The paging channel operates at a data rate of 9600 or
4800 bps.
All 64 channels are combined to give single I and Q channels. The signals are applied to
quadrature modulators and the resulting signals are summed to form a QPSK signal, which is
linearly amplified.
The pilot CDMA signal transmitted by a base station provides a reference for all mobile
stations. It is used in the demodulation process. The pilot signal level for all base stations is about
4 to 6 dB higher than the traffic channel with a constant value. The pilot signals are quadrature
pseudorandom binary sequence signals with a period of 32,768 chips. Since the chip rate is
1.2288 Mcps, the pilot pseudorandom binary sequence corresponds to a period of 26.66 ms,
which is equivalent to 75 pilot channel code repetitions every 2 seconds. The pilot signals from
all base stations use the same pseudorandom binary sequence, but each base station is identified
by a unique time offset of its pseudorandom binary sequence. These offsets are in increments of
64 chips, providing 511 unique offsets relative to 0 offset code. These large numbers of offsets
ensure that unique base station identification can be obtained, even in dense microcellular envi-
ronments.Tables 6-2 through 6-5 provide modulation parameters for the forward link channels.
A mobile station processes the pilot channel to find the strongest signal components. The
processed pilot signal provides an accurate estimation of time delay and the phase and magni-
tude of the three multipath components. These components are tracked in the presence of fast
fading, and coherent reception with combining is used. The chip rate on the pilot channel and on
all channels is locked to precise system time, e.g., by using the Global Positioning System
(GPS). Once the mobile station identifies the strongest pilot offset by processing the multipath
components from the pilot channel correlator, it examines the signal on its sync channel which is
TIA IS-95 CDMA System 95

Table 6-2 Sync Channel Modulation Parameters

Parameter Data Rate 1200 bps Units
PN chip rate 1.2288 Mcps
Code rate 1/2 bits per code symbol
Code symbol repetition 2 modulation symbols per code symbol
Modulation symbol rate 4800 symbols per second (sps)
PN chips per modulation symbol 256 PN chips per modulation symbol
PN chips per bit 1024 PN chips per bit

Table 6-3 Paging Channel Modulation Parameters

Data Rate (bps)
Parameter 9600 4800 Units
PN chip rate 1.2288 1.2288 Mcps
Code rate 1/2 1/2 bits per code symbol
Code symbol repetition 1 2 modulation symbols per code symbol
Modulation symbol rate 19,200 19,200 sps
PN chips per modulation symbol 64 64 PN chips per modulation symbol
PN chips per bit 128 256 PN chips per bit

Table 6-4 Forward Traffic Channel Modulation Parameters for RS1

Data Rate (bps)
Parameters 9600 4800 2400 1200 Units
PN chip rate 1.2288 1.2288 1.2288 1.2288 Mcps
Code rate 1/2 1/2 1/2 1/2 bits per code symbol
Code symbol 1 2 4 8 repeated symbols per
repetition code symbol
Modulation symbols 19,200 19,200 19,200 19,200 sps
rate
PN chips per 64 64 64 64 PN chips per
modulation symbol modulation symbol
PN chips per bit 128 256 512 1024 PN chips per bit
96 Ch. 6 • IS-95 CDMA Air Interface

Table 6-5 Forward Traffic Channel Modulation Parameters for RS2

Data Rate (bps)
Parameters 14,400 7200 3600 1800 Units
PN chip rate 1.2288 1.2288 1.2288 1.2288 Mcps
Code rate 1/2 1/2 1/2 1/2 bits per code symbol
Code symbol 1 2 4 8 repeated symbols per
repetition code symbol
Puncturing rate 4/6 4/6 4/6 4/6 modulation symbols
per repeated symbol
Effective code rate 3/4 3/4 3/4 3/4
Modulation symbol 19,200 19,200 19,200 19,200 sps
rate
PN chips per 64 64 64 64 PN chips per
modulation symbol modulation symbol
PN chips per bit 85.33 170.67 341.33 682.67 PN chips per bit

locked to the pseudorandom binary sequence signal on the pilot channel. Since the sync channel
is time aligned with its base station’s pilot channel, the mobile station finds the information per-
tinent to this particular base station on the sync channel. The sync channel message contains
time of day and long-code synchronization to ensure that long-code generators at the base sta-
tion and mobile station are aligned and identical. The mobile station now attempts to access the
paging channel and listens for system information. The mobile station enters the idle state when
it has completed acquisition and synchronization. It listens to the assigned paging channel and is
able to receive and initiate the calls. When informed by the paging channel that voice traffic is
available on a particular channel, the mobile station recovers the speech data by applying the
inverse of the spreading procedures.

6.2.2 Reverse Link

In this section, we summarize the operation of the reverse link. For a more detailed discus-
sion see chapter 7. The reverse link is separated from the forward link by 45 MHz at cellular fre-
quencies and 80 MHz at PCS frequencies. The reverse link uses the same 32,768-chip code as
that used on the forward link. The reverse link channels are either access channels or reverse
traffic channels. A reverse traffic channel is further subdivided into a single fundamental code
channel and 0 through 7 supplemental code channels (see Fig. 6-2). There are 62 traffic channels
and up to 32 access channels. The access channel is used by the mobile station to communicate
nontraffic information, such as originating calls and responding to paging. The access rate is
fixed at 4800 bps. All mobile stations accessing a radio system share the same frequency assign-
ment. Each access channel is identified by a distinct access-channel long-code sequence having
TIA IS-95 CDMA System 97

an access number, a paging channel number associated with the access channel, and other sys-
tem data. Each mobile station uses a different PN code; therefore the radio system can correctly
decode the information from an individual mobile station. Data transmitted on the reverse chan-
nel is grouped into 20-ms frames. All data on the reverse channel is convolutionally encoded,
block interleaved, and modulated by modulation symbols transmitted for each 6 code symbols.
The modulation symbol is one of the 64 mutually orthogonal waveforms that are generated
using Walsh functions.
The reverse traffic channel may use either 9600-, 4800-, 2400-, or 1200-bps data rates for
transmission. The duty cycle for transmission varies proportionally with data rate, being 100%
at 9600 bps and dropping to 12.5% at 1200 bps. An optional second rate set is also supported in
the PCS version of CDMA and new versions of cellular CDMA (see chapter 7 for details). The
actual burst transmission rate is fixed at 28,800 code symbols per second. Since 6 code symbols
are modulated as one of 64 modulation symbols for transmission, the modulation symbol trans-
mission rate is fixed at 4800 modulation symbols per second. This results in a fixed Walsh chip
rate of 307.2 kilo-chips per second (kcps). The rate of spreading PN sequence is fixed at 1.2288
Mcps, so that each Walsh chip is spread by 4 PN chips. Table 6-6 provides the signal rates and

Table 6-6 CDMA Reverse Traffic Channel Modulation Parameters for RS1
Parameter 9600 bps 4800 bps 2400 bps 1200 bps units
PN chip rate 1.2288 1.2288 1.2288 1.2288 Mcps
Code rate 1/3 1/3 1/3 1/3 bits per code sym
Transmitting 100 50 25 12.5 %
duty cycle
Code symbol 1 2 4 8 repeated code symbol
repetition per code symbol
Code symbol rate 3 × 9600 = 28,800 28,800 28,800 28,800 sps
Modulation 6 6 6 6 code symbol per mod.
symbol
Modulation 28,800/6 = 4800 4800 4800 4800 sps
symbol rate
Walsh chip rate 64 × 4800 = 307.2 307.2 307.2 307.2 kcps
Mod. symbol 1/4800 = 208.33 208.33 208.33 208.33 µs
duration
PN chips per 12,288/288 = 42.67 42.67 42.67 42.67 PN chips per code
code symbol symbol
PN chips per 1,228,800/4800 = 256 256 256 256 PN chips per mod.
mod. symbol symbol
PN chips per 4 4 4 4 PN chips per Walsh
Walsh chip chip
98 Ch. 6 • IS-95 CDMA Air Interface

their relationship for the various transmission rates on the reverse traffic channel for RS1.
Reverse traffic channel modulation parameters for RS2 are given in Table 6-7.
Following the orthogonal spreading, the reverse traffic channel and access channel are
spread in quadrature. Zero-offset I and Q pilot PN sequences are used for spreading. These
sequences are periodic with 215 chips (32,768 PN chips in length) and are based on characteristic
polynomials gI (x) and gQ (x).
15 13 9 8 7 5
gI ( x ) = x + x + x + x + x + x + 1 (6.1)

15 12 11 10 6 5 4 3
gQ ( x ) = x + x + x + x + x + x + x + x + 1 (6.2)

The maximum-length linear feedback register sequences I(n) and Q(n), based on these
polynomials, have a period 215 – 1 and are generated by using the following recursions:

I ( n ) = I ( n – 15 ) ⊕ I ( n – 10 ) ⊕ I ( n – 8 ) ⊕ I ( n – 7 ) ⊕ I ( n – 6 ) ⊕ I ( n – 2 ) (6.3)

Table 6-7 CDMA Reverse Traffic Channel Modulation Parameters for RS2
Parameter 14,400 bps 7200 bps 3600 bps 1800 bps units
PN chip rate 1.2288 1.2288 1.2288 1.2288 Mcps
Code rate 1/2 1/2 1/2 1/2 bits per code symbol
Transmitting duty 100 50 25 12.5 %
cycle
Code symbol 1 2 4 8 repeated code symbols
repetition per code symbol
Code symbol rate 28,800 28,800 28,800 28,800 sps
Modulation 6 6 6 6 repeated code symbol
per modulation symbol
Modulation symbol 4800 4800 4800 4800 sps
rate
Walsh chip rate 307.2 307.2 307.2 307.2 kcps
Mod. symbol duration 208.33 208.33 208.33 208.33 µs
PN chips per code 42.67 42.67 42.67 42.67 PN chips per repeated
symbol code symbol
PN chips per mod. 256 256 256 256 PN chips per mod.
symbol symbol
PN chips per 4 4 4 4 PN chips per Walsh chip
Walsh chip
TIA IS-95 CDMA System 99

based on gI (x) as the characteristic polynomial, and

Q ( n ) = Q ( n – 15 ) ⊕ Q ( n – 12 ) ⊕ Q ( n – 11 ) ⊕
Q ( n – 10 ) ⊕ Q ( n – 9 ) ⊕ Q ( n – 5 ) ⊕ (6.4)
Q(n – 4) ⊕ Q(n – 3)
based on gQ (x) as the characteristic polynomial, where I(n) and Q(n) are binary numbers (0 and
1) and the additions are modulo-2. To obtain the I and Q pilot PN sequences (of period 215), a 0
is inserted in I(n) and Q(n) after 14 consecutive 0 outputs (this occurs only once in each period).
Therefore, the pilot PN sequences have one run of 15 consecutive 0 outputs instead of 14.
The chip rate for the pilot PN sequence is 1.2288 Mcps and its period is 26.666 ms. There
are exactly 75 pilot PN repetitions every 2 seconds. The spreading modulation is Offset Quadra-
ture Phase-Shift Keying (OQPSK). The data spread by Q pilot PN sequence is delayed by half a
chip time (406.901 nanoseconds [ns]) with respect to the data spread by I pilot PN sequence (see
chapter 7). Fig. 6-3 and Table 6-8 describe the characteristics of OQPSK.

Table 6-8 CDMA Reverse Traffic Channel Modulation Parameters for RS2
I Q Phase
0 0 π/4
1 0 (3π)/4
1 1 –(3π)/4
0 1 –π/4

Q-Channel

(1,0) (0,0)

I-Channel

(1,1) (0,1)

Figure 6-3 Signal Constellation and Phase Transition of OQPSK Used on Reverse CDMA
Channel
100 Ch. 6 • IS-95 CDMA Air Interface

Table 6-9 CDMA Access Channel Modulation Parameters

Parameter Data Rate 4800 bps Units
PN chip rate 1.2288 Mcps
Code rate 1/3 bits per code symbol
Code symbol repetition 2 symbols per code symbol
Transmit duty cycle 100 %
Code symbol rate 28,800 sps
Modulation 6 code symbol per modulation symbol
Modulation symbol rate 4800 sps
Walsh chip rate 307.2 kcps
Modulation symbol duration 208.33 µs
PN chips per code symbol 42.67 PN chips per code symbol
PN chips per modulation symbol 256 PN chips per modulation symbol
PN chips per Walsh chip 4 PN chips per Walsh chip

Each base station transmits a pilot signal of constant power on the same frequency. The
received power level of the received pilot signal enables the mobile station to estimate the path
loss between the base station and the mobile station. Knowing the path loss, the mobile station
adjusts its transmitted power such that the base station will receive the signal at the requisite
power level. The base station measures the mobile station’s received power and informs the
mobile station to make the necessary adjustment to its transmitter power. One command every
1.25 ms adjusts the transmitted power from the mobile station in ±0.5 dB steps. The base station
uses frame errors reported by the mobile station to increase or decrease the transmitted power.
CDMA access channel modulation parameters are listed in Table 6-9.
CDMA provides a soft handoff. As the mobile station moves to the edge of its single cell,
the adjacent base station assigns a modem to the call; meanwhile, the current base station contin-
ues to handle the call. The call is handled by both base stations on a make-before-break basis.
Handoff diversity occurs with both base stations handling the call until the mobile station moves
sufficiently close to one of the base stations for it to then exclusively handle the call. This hand-
off procedure is different from conventional break-before-make or hard-handoff procedures. The
soft-handoff procedure will be discussed in more detail in chapter 10.

6.3 Summary
A CDMA system operates with a low Eb /N0 ratio, exploits voice activity, and uses sectorization
of cells. Each sector has 64 CDMA channels. It is a synchronized system with three receivers to
provide path diversity at the mobile station and four receivers at the cell site.
References 101

This chapter provided a high-level description of the IS-95 CDMA air interface. The for-
ward link channels are pilot, sync, paging, and forward traffic (fundamental and supplementary)
channels. The reverse link channels include access and reverse traffic (fundamental and supple-
mentary) channels.

6.4 References
1. Garg, V. K., Smolik, K. F., and Wilkes, J. E., Applications of CDMA in Wireless Communi-
cations, Prentice Hall PTR, Upper Saddle River, NJ, 1997.
2. TIA/EIA IS-95, “Mobile Station–Base Station Compatibility Standard for Dual-Mode
Wideband Spread Spectrum Cellular Systems,” 1998.
This page intentionally left blank
 C H A P T E R 7

Physical and Logical

Channels of IS-95
CDMA

7.1 Introduction
This chapter first covers how to introduce a CDMA carrier in an existing AMPS or TDMA sys-
tem, establishing the number of AMPS or TDMA channels that should be removed in order to
introduce the first and second CDMA carrier of 1.23 MHz without interfering with the remain-
ing AMPS or TDMA carriers. After this, we briefly describe modulation schemes, bit repetition,
block interleaving, and channel coding that are used in processing logical channels on the IS-95
CDMA forward and reverse links. Details about information processing, message types, and
message framing are then presented for the pilot, sync, paging, and traffic channels on the for-
ward link. This chapter also provides similar details for the access and traffic channels on the
reverse link.

7.2 Physical Channels

In IS-95, physical channels are defined in terms of an RF frequency and a code sequence. There
are 64 Walsh codes available for the forward link (BS to MS), providing 64 logical channels. On
the reverse link, channels are identified by long PN code (see chapter 6) sequences.
A CDMA system is implemented using N wideband frequency carriers, each capable of
supporting M circuits that can be accessed by any user. A unique circuit is defined by a different
code sequence for each user. Frequency assignment remains under control of the system in both
forward link (BS to MS) and reverse link (MS to BS) directions. Fig. 7-1 shows N RF carriers
for each direction. This is a Frequency Division Duplex (FDD) arrangement and is called the
CDMA/FDD system.
In IS-95, CDMA carrier band center frequencies are denoted in terms of AMPS channel
numbers. We refer to Fig. 7-2 where the AMPS channel number 283 is the center of a CDMA
carrier band. To introduce one CDMA carrier, we need 41 30-kHz AMPS channels to provide a

103
104 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

Code Code Code

RF Carrier #1
(1.2288 MHz) Sequence Sequence Sequence
#1 #2 #M

Code Code Code

RF Carrier #2 Sequence Sequence Sequence
(1.2288 MHz) #1 #2 #M

45 MHz

Code Code Code

RF Carrier #N Sequence Sequence Sequence
(1.2288 MHz) #1 #2 #M

Forward Link (BS to MS)

Code Code Code

RF Carrier #1 Sequence Sequence Sequence
(1.2288 MHz) #1 #2 #M

Code Code Code

RF Carrier #2 Sequence Sequence Sequence
(1.2288 MHz) #1 #2 #M

RF Carrier #N Code Code Code

(1.2288 MHz) Sequence Sequence Sequence
#1 #2 #M

Reverse Link (MS to BS)

Figure 7-1 CDMA/FDD System

Physical Channels 105

283

22 22 33 33
55 66 00 11
34 23 34 23

Guard CDMA Carrier Band Guard

9 AMPS 41 AMPS Channels 9 AMPS
Channels Channels

59 AMPS Channels (1.77 MHz)

Figure 7-2 Introducing One CDMA Carrier

CDMA carrier bandwidth of 1.23 MHz. The recommended guard band between the CDMA
carrier band edge and an AMPS or a TDMA carrier is 0.27 MHz—that is equal to 9 AMPS or
TDMA channels. Thus to introduce the first CDMA carrier without interfering with the remain-
ing AMPS or TDMA channels, it is necessary that 59 AMPS channels be removed. In order to
introduce the second CDMA carrier, we should remove only 41 additional AMPS channels. We
see from Fig. 7-3 that, to introduce two CDMA carriers, we should remove 100 AMPS chan-
nels, a total of 3 MHz.
The primary and secondary CDMA carriers are the preassigned frequencies (AMPS chan-
nel numbers) that allow the mobile to acquire the CDMA system. A base station can support pri-
mary, secondary, or both types of channels.
The 1.23-MHz bandwidth for a CDMA carrier suggests that the minimum center fre-
quency separation between two carrier frequencies is 1.23 MHz. The MS and BS frequencies

Guard CDMA Carrier # 2 CDMA Carrier # 1 Guard

22 22 22 33 33
11 22 66 00 11
23 12 23 34 23

9 AMPS 41 AMPS 41 AMPS 9 AMPS

Channels Channels Channels Channels

100 AMPS Channels (3 MHz)

Figure 7-3 Introducing Two CDMA Carriers

106 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

for cellular band (Band Class 0) are specified in Table 7-1. The channel spacings, CDMA
channel designations, and transmit center frequencies of Band Class 0 are given in Table 7-2.
The valid CDMA carrier frequencies are on AMPS channel numbers 1013–1023, 1–311, 356–
644, 689–694, and 739–777 (see Table 7-3). Only the primary and secondary CDMA carrier

Table 7-1 Band Class 0 System Frequencies

Transmit Frequency Band (MHz)
System Mobile Station Base Station
A 824.025–835.005 869.025–880.005
844.995–846.495 889.995–891.495
B 835.005–844.995 880.005–889.995
846.495–848.985 891.495–893.985

Table 7-2 CDMA Channel Number to CDMA Frequency Assignment for Band Class 0
Transmitter CDMA Channel Number CDMA Frequency Assignment (MHz)
Mobile station 1 ≤ N ≤ 777 0.030 N + 825.000
1013 ≤ N ≤ 1023 0.030 [N – 1023] + 825.000
Base station 1 ≤ N ≤ 777 0.030 N + 870.000
1013 ≤ N ≤ 1023 0.030 [N – 1023] + 870.000

Table 7-3 Channel Numbers for Band Class 0

Frequency Block Valid CDMA Frequency Assignment CDMA Channel Number
A" (1 MHz) not valid 991–1012
valid 1013–1023
A (10 MHz) valid 1–311
not valid 312–333
B (10 MHz) not valid 334–355
valid 356–644
not valid 645–666
A' (1.5 MHz) not valid 667–688
valid 689–694
not valid 695–716
B' (2.5 MHz) not valid 717–738
valid 739–777
not valid 778–799
Physical Channels 107

center frequencies are specified in IS-95 standard. Other center frequencies are selected by
each system operator.
Table 7-4 shows the CDMA center frequency (AMPS channel number) assignments for
systems A and B with 41 AMPS channel separation.
The mobile station and base station frequencies for the PCS band (Band Class 1) are spec-
ified in Table 7-5. The channel spacings, CDMA channel designations, and transmit center fre-
quencies of Band Class 1 are given in Table 7-6. Mobile stations in Band Class 1 support
operations on channel numbers 25 through 1175 as shown in Table 7-7. Note that certain chan-
nel assignments are not valid and that others are conditionally valid. Transmission on condition-
ally valid channels is allowed if the adjacent block is allocated to the licensee or if other valid
authorization has been obtained.
A preferred set of CDMA frequency assignments for Band Class 1 is given in Table 7-8.

Table 7-4 CDMA Center Frequency Assignment for Systems A and B for Cellular Band Class 0
CDMA Frequency AMPS Channels
CDMA Channel Type System A System B
Primary 283 384
242 425
201 466
160 507
119 548
78 589
37 630
1019
Secondary 691 777
Total number of CDMA channels 9 8

Table 7-5 Band Class 1 System Frequencies

Transmit Frequency Band (MHz)
Block Mobile Station Base Station
A 1850–1865 1930–1945
D 1865–1870 1945–1950
B 1870–1885 1950–1965
E 1885–1890 1965–1970
F 1890–1895 1970–1975
C 1895–1910 1975–1990
108 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

Table 7-6 CDMA Channel Number to CDMA Frequency Assignment for Band Class 1
Transmitter CDMA Channel Number Center Frequency of CDMA Channel (MHz)
Mobile station 0 ≤ N ≤ 1199 1850.000 + 0.050 N
Base station 0 ≤ N ≤ 1199 1930.000 + 0.050 N

Table 7-7 CDMA Channel Numbers for Band Class 1

Block Valid CDMA Frequency Assignment CDMA Channel Number
A (15 MHz) not valid 0–24
valid 25–275
conditionally valid 276–299
D (5 MHz) conditionally valid 300–324
valid 325–375
conditionally valid 376–399
B (15 MHz) conditionally valid 400–424
valid 425–675
conditionally valid 676–699
E (5 MHz) conditionally valid 700–724
valid 725–775
conditionally valid 776–799
F (5 MHz) conditionally valid 800–824
valid 825–875
conditionally valid 876–899
C (15 MHz) conditionally valid 900–924
valid 925–1175
not valid 1176–1199

Table 7-8 CDMA Preferred Set of Frequency Assignments for Band Class 1
Block Preferred Set of Channel Numbers
A 25, 50, 75, 100, 125, 150, 175, 200, 225, 250, 275
D 325, 350, 375
B 425, 450, 475, 500, 525, 550, 575, 600, 625, 650, 675
E 725, 750, 775
F 825, 850, 875
C 925, 950, 975, 1000, 1025, 1050, 1075, 1100, 1125, 1150, 1175
Modulation 109

7.3 Modulation
The signals from each channel (pilot, sync, paging, and traffic) are modulo-2 added to I and Q
PN short-code sequences. The I and Q spread signals are baseband filtered and sent to a linear
adder with gain control (see Fig. 7-4). The gain control allows individual channels to have differ-
ent power levels assigned to them. The CDMA system assigns power levels to different channels
depending on the quality of the received signal at a mobile station. The algorithms for determin-
ing the power levels are proprietary to each equipment manufacturer. The I and Q baseband sig-
nals are modulated by I and Q carrier signals, combined together, amplified, and sent to the base
station antenna. The net signal from the CDMA modulator is a complex quadrature signal that
looks like noise.
The same PN short-code sequences are used on all channels (i.e., pilot, sync, paging, and
traffic) of the forward link. All base stations in a system are synchronized using the GPS satel-
lite. Different base stations use time-shifted versions of these PN sequences to allow mobile sta-
tions to select the appropriate base station.
Unlike the forward direction, the CDMA system uses a different modulation scheme to
generate the signal in the reverse direction. The net signal from modulator is a 4-phase quadra-
ture signal. The output from either the access channel or the traffic channel is sent to two mod-
ulo-2 adders—one for the in-phase channel and the other for the quadrature channel. Two
different short-code PN sequences are modulo-2 added to the data and filtered by a baseband fil-
ter. For a quadrature channel, a delay of 1/2 PN symbol (406.9 ns) is added before the filter.
Thus, the reverse channel uses OQPSK (see Fig. 7-5). No pilot signal is used on the reverse link.

7.4 Bit Repetition

The nominal RS1 data rate on the forward and reverse traffic channels is 9600 bps. If data is
transmitted at a lower rate (4800, 2400, or 1200 bps), then the data bits are repeated n times to
increase the rate to 9600 bps. Likewise, the nominal RS2 data rate on the forward and reverse

cos (ωt + φ) sin (ωt + φ)

I/2

I LPF

Q LPF
Q/2
sin (ωt + φ) cos (ωt + φ)
Transmitter LPF = Linear Pass Filter Receiver

Figure 7-4 QPSK Modulator

110 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

I-Channel Sequence
1.2288 Mcps sin (ωt + φ)

I
Baseband
Filter

s(t )
1/2 PN Chip
Delay = 406.9 ns
Q Baseband
D
Filter

Q-Channel Sequence cos (ωt + φ)

1.2288 Mcps

Figure 7-5 OQPSK Modulation

traffic channels is 14,400 bps. If the data is transmitted at the lower rates (7200, 3600, or 1800
bps), then the data bits are repeated either 2, 4, or 8 times to increase the rate to 14,400 bps.

7.5 Block Interleaving

Communications over a radio channel are characterized by deep fades that can cause large num-
bers of consecutive errors. Most coding schemes perform better on random data errors than on
blocks of errors. By interleaving the data, no two adjacent bits are transmitted near to each other,
and the data errors are randomized.
The interleaver spans a 20-ms frame. In the reverse direction, the output of the interleaver
is 28.8 kbps if the data rate is 9.6 kbps (RS1). The resultant signal transmits with 100% duty
cycle. If the data rate is lower (4800, 2400, or 1200 bps), the interleaver plus randomizer deletes
redundant bits and transmits with a lower duty cycle (50%, 25%, 12.5%). Thus, bits are not
repeated on the reverse CDMA traffic channel. On the access channel, the data bits are repeated.
In the forward direction, the nominal data rate for RS1 is 19.2 kbps; lower data rates use a lower
duty cycle.
On the reverse traffic channel, the output of the interleaver is processed by a data random-
izer. The randomizer removes redundant data blocks generated by the code repetition. It uses a
masking pattern determined by the data rate and the last 14 bits of the long code. For a 20-ms
block (192 bits at 9600 bps), the data randomizer segments the block into 16 blocks of 1.25 ms
each. At a data rate of 9600 bps (RS1), all blocks are filled with data. At a data rate of 4800 bps,
8 out of 16 blocks are filled with data in a random manner. Similarly, for 2400 and 1200 bps, 4 of
16 and 2 of 16 blocks, respectively, are randomly filled with data. Thus, no redundant data are
transmitted over the reverse channel.
Channel Coding 111

7.6 Channel Coding

In IS-95, traffic data frames on uplink and downlink are fed to convolutional encoders. Both
uplink and downlink encoders use an eight-stage shift register with constraint length of 9. The
rate of the uplink coder is 1/3—it outputs 3 bits for every input bit. At rates below 9.6 kbps, out-
put bits are repeated to bring the number of bits in a 20-ms block to 576, for a gross rate of 28.8
kbps (RS1). The rate of the downlink encoder is 1/2—it outputs 2 bits for every input bit. At
rates below 9.6 kbps (RS1), output bits are repeated to bring the number of bits in a 20-ms block
to 384, for a gross rate of 19.2 kbps.

7.7 Logical Channels

Logical channels in CDMA are the control and traffic channels. The control channels are the
pilot channel (downlink), the paging channels (downlink), the sync channels (downlink), and
the access channels (uplink) (see Fig. 7-6)
The traffic channels are used to carry user information (speech or data), along with signal-
ing traffic, between the base station and the mobile station. Four different rates are used—1, 1/2,

Control Channels

Downlink Uplink

Pilot Sync Paging Access

Traffic Channels

Speech or Data Associated Signaling

1 1/2 1/4 1/8 Blank & Power Control

Burst (downlink)

Dim &
Burst

Figure 7-6 IS-95 Logical Channels

112 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

1/4, and 1/8. The downlink traffic channel is called the forward link, whereas the uplink traffic
channel is referred to as the reverse link.
When user speech or data is replaced by associated signaling, it is called blank-and-burst
signaling. When part of the speech is replaced by signaling information, it is called dim-and-
burst signaling. All associated signaling is sent in Rate 1. In addition, on the downlink there is a
power control subchannel that allows the mobile to adjust its transmitted power by ±1 dB every
1.25 ms (i.e., 16 times during a 20-ms speech frame).
In the forward direction there are the pilot channel, the sync channel, up to seven paging
channels, and a number of forward traffic channels that all share the same center frequency. Out
of the 64 Walsh-coded channels available for use (W0, W1, W2, …, W63), the pilot channel on W0
is always required. There can be one sync channel (W32) and seven paging channels (W1 to W7—
the maximum allowed); the remaining channels are traffic channels. A traffic channel is further
subdivided into a fundamental code channel and a supplementary code channel (see Figs. 6-1
and 6-2). The primary paging channel is always assigned the Walsh code W1. The mobile exam-
ines the number of paging channel parameters in the system parameter message. If this value is
not 1, a hashing algorithm is invoked to determine the correct paging channel number.

7.7.1 Pilot Channel

The pilot channel is used by a base station to provide a reference for all mobile stations. It
provides a phase reference for coherent demodulation at the mobile receiver to enable coherent
detection. Note that the pilot channel does not carry any information and is assigned the Walsh
code W0 (see Fig. 7-7). The pilot signal level for all base stations is kept about 4 to 6 dB higher
than a traffic channel with a constant signal power. The pilot signal is used for comparisons of
signal strength between different base stations to decide when to perform handoff.
The pilot channel is needed to lock onto other logical channels on the same RF carrier. It
carries an unmodulated DSSS signal that is transmitted continuously by each base station. The

I Pilot PN
Walsh W0 Sequence at 1.2288 Mcps

BB
1.2288
All 0s Mcps
To QPSK
Modulator

BB
BB = baseband filter
Q Pilot PN
Sequence at 1.2288 Mcps

Figure 7-7 Pilot Channel Processing

Logical Channels 113

pilot signals are quadrature pseudorandom (PN) binary sequence signals with a period of 32,768
(215) chips. Since the chip rate is 1.2288 Mcps, the pilot PN sequence corresponds to a period of
26.667 ms. This is equivalent to 75 pilot channel code repetitions every 2 seconds. The pilot sig-
nals from all base stations use the same PN sequences, but each base station is identified by a
unique time offset. These offsets are in increments of 64 chips to provide 512 unique offsets (see
Fig. 7-8). The high number of offsets ensures that unique base station identification can be
obtained even in dense microcellular environments.

7.7.2 Sync Channel

The sync channel is an encoded, interleaved, and modulated spread spectrum signal that
may be used with the pilot channel to acquire initial time synchronization. It is assigned the
Walsh code W32. The sync channel always operates at a fixed rate of 1200 bps and is convolu-
tionally encoded to 2400 bps, repeated to 4800, and interleaved. (see Fig. 7-9). The sync channel
is used with the pilot channel to acquire initial time synchronization. Only the Sync Channel
message is transmitted over this channel.

PN Offsets
0 i–1 i k 511

64
chips 32,768 Chips

Figure 7-8 Short PN Offsets

I Pilot PN
Walsh Code W32 1.2288 Mcps
BB
Sync Code Modulated
Channel Message Symbol Symbol
1.2288
Convolutional Mcps
Encoder
Symbol Block To QPSK
Rate = 1/2, K = 9 Repetition Interleaver Modulator
1.2 2.4 4.8 4.8
kbps ksps ksps ksps

BB
BB = baseband filter
ksps = kilosymbols per second
Q Pilot PN
1.2288 Mcps
Figure 7-9 Sync Channel Processing
114 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

The Sync Channel message parameters are

• System Identification (SID): identifier number for the system
• Network Identification (NID): subidentifier for the system
• Pilot short PN sequence offset (PILOT_PN) index: offset index, in units of 64 chips,
for the base station or sector
• Long-code state (LC_STATE): long code at the time specified in system time
parameter
• System time (SYS_TIME)
• Leap seconds (LP_SEC): number of leap seconds that have occurred since the start of
system
• Offset of local time: offset from the system time
• Daylight saving time indicator
• Paging channel data rate (PRAT): 4.8 or 9.6 kbps
The Sync Channel message itself is long and may occupy more than one sync channel
frame.The Sync Channel message is organized in a Sync Channel message capsule. A Sync
Channel message capsule contains the Sync Channel message and padding. When the Sync
Channel message occupies more than one sync channel frame, padding is used to fill the bit
positions up to the beginning of the next sync channel superframe, where the next Sync Channel
message starts.
The Sync Channel message has an 8-bit message-length header, a message body of a min-
imum of 2 bits and a maximum of 1146 bits, and a cyclic redundancy check (CRC) code of 30
bits (see Fig. 7-10). The message length includes the header, body, and CRC, but not the pad-
ding. The CRC is computed on the message-length header and the message body using the fol-
lowing polynomial:
30 29 21 20 15 13 12 11 8 7 6 2
g( x ) = x + x + x + x + x + x + x + x + x + x + x + x + x + 1 (7.1)
After a message is formed, it is segmented into 31-bit groups and sent in a sync frame (see
Fig. 7-11) consisting of a 1-bit start of message (SOM) field and 31 of the sync channel frame
bodies. The value of 1 for SOM indicates that the frame is the start of the Sync Channel mes-
sage, whereas a value of 0 for SOM indicates that the frame is a continuation of a Sync Channel
message or padding. The sync channel frames are transmitted in groups of sync channel super-
frames. Three sync frames are combined to form a superframe. Each superframe carries 96 bits
and lasts for 80 ms (see Fig. 7-12). The entire Sync Channel message is then sent in N super-

8 bits NMSG = 2 – 1146 bits 30 bits

Message Padding =
Data CRC 0…000
Length (bytes)

Figure 7-10 Message Framing on Sync Channel and Paging Channel

Logical Channels 115

1 bit 31 bits
SOM Data

Note: SOM = 1 for first body of Sync Channel message

SOM = 0 for all other bodies in Sync Channel message

Figure 7-11 Sync Channel Frame

96 bits

Sync Frame #1 Sync Frame #2 Sync Frame #3

80 ms
Figure 7-12 Sync Channel Superframe

frames. The padding bits are used so that the Start message always starts at 1 bit after the begin-
ning bit of the superframe.

7.7.3 Paging Channel

The paging channel is used to transmit control information to the mobile station. When a
mobile station is to receive a call, it will receive a page from the base station on an assigned pag-
ing channel. The paging channel provides the mobile station with system information and
instructions, in addition to acknowledging messages following access requests on the mobile
station’s access channels. Paging channel data are processed in a manner similar to the process-
ing of traffic channel data. However, there is no power control on a per-frame basis. The 42-bit
mask is used to generate the long code (see Fig. 7-13). The paging channel operates at a data rate
of 4800 or 9600 bps.
The Paging Channel message is similar in form to the Sync Channel message. It has
an 8-bit message-length header, a message body of a minimum of 2 bits and a maximum of
1146 bits, and a CRC code of 30 bits. The message length includes the header, body, and
CRC, but not the padding. The CRC is computed on the message-length header and the
message body using the same code as the sync channel.
A Paging Channel message can use synchronized capsules that end on a half-frame
boundary or unsynchronized capsules that can end anywhere within a half-frame. If synchro-
nized Paging Channel messages are less than an integer multiple of 47 bits for 4800-bps trans-
mission (or 95 bits for 9600-bps transmission), they are padded with 0 bits at the end of the
message. Unsynchronized messages are not padded.
After a message is formed, it is segmented into 47- or 95-bit chunks and sent in a paging
channel half-frame (see Fig. 7-14) that consists of a 1-bit synchronized capsule indicator (SCI)
field and 47 or 95 bits of the paging channel frame body. A value of 1 for SCI indicates that the
116 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

I Pilot PN
Walsh Code 1–7 1.2288 Mcps
Paging Channel
BB
Message Code Modulation
(4.8 or 9.6 kbps) Symbol Symbol
Convolutional
Encoder
Symbol Block To QPSK
Rate = 1/2, K = 9 Repetition Interleaver Modulation
9.6 19.2 19.2
or ksps ksps
19.2 ksps
BB
19.2
64:1 ksps

Long-Code Mask Q Pilot PN

Long-Code Long-Code 1.2288 Mcps
for Paging
Generator Decimator
Channel
1.2288
Mcps BB = baseband filter

Figure 7-13 Paging Channel Processing

One Paging Channel Half-Frame = 10 ms = 96 bits

at 9600 bps or 48 bits at 4800 bps

SCI Paging Channel Half-Frame Body

1 bit 47 bits at 4800 bps

95 bits at 9600 bps
Note: SCI =1 for first new capsule of synchronized Paging Channel message
SCI = 0 for all other capsules in Paging Channel message

Figure 7-14 Paging Channel Half-Frame

frame is the start of a Paging Channel message (either synchronized or unsynchronized). Mes-
sages can also start in the middle of a frame and immediately after the end of an unsynchronized
message (with no padding bits). A value of 0 for SCI indicates that the frame is not the start of a
message and that it can include a message (with or without padding), padding only, or the end of
one message and start of another.
Eight paging channel half-frames are combined to form a paging channel slot (see Fig. 7-
15) of length 80 ms (384 bits at 4800 bps and 768 bits at 9600 bps). The entire Paging Channel
message is then sent in N slots. The maximum number of slots that a message can use is 2048.
The base station always starts a slot with an asynchronized message capsule that starts at 1 bit
after the beginning of a slot. The first bit in a slot is SCI = 1.
The paging channel carries information to allow the network to
Logical Channels 117

One Paging Channel Slot = 80 ms (384 bits at 4800 bps or 768 bits at 9600 bps)

Paging Channel Paging Channel

Half-Frame Half-Frame
48 bits at 4800 bps 48 bits at 4800 bps
96 bits at 9600 bps 96 bits at 9600 bps

Figure 7-15 Paging Channel Slot Structure

• Supply display information to be displayed by the mobile station

• Identify the called party’s number
• Identify the calling party’s number
• Convey information to the mobile station by means of tones or other alerting signals
• Indicate the number of messages waiting
The messages carried by the paging channel include
1. System Parameter message. Provides overhead information such as the pilot PN
sequence offset index, i, in PN-I-i (t) and PN-Q-i (t), base station identifier, the number
of paging channels, PAGE_CHAN, and other system information. PAGE_CHAN and a
hashing algorithm (IS-95) are used by the mobile station to determine the correct pag-
ing channel number. Initially, the mobile station expects a PAGE_CHAN equal to 1—
the primary paging channel. A different value invokes the hashing algorithm.
2. Access Parameters message. Defines parameters required by a mobile station to trans-
mit on an access channel.
3. Neighbor List message. Provides information about neighbor base station parameters,
e.g., the neighbor pilot PN sequence offset index, i. If the neighbor does not have a pag-
ing channel on the current CDMA carrier frequency, the CDMA Channel List message
contains this information.
4. CDMA Channel List message. Provides the list of CDMA carriers.
5. Slotted Page or Page message. Provides data used to inform the mobile that it can
receive a call. The mobile station monitors for its identification number in the mobile
identification number (MIN) field. With the slotted page message, the mobile needs to
monitor only specific time slots of the Paging Channel message.
6. Page message. Provides a page to the mobile station. The mobile monitors every time
slot of the Paging Channel message.
7. Typical Order message. Several order messages can be carried on the paging channel,
such as abbreviated alert, base station challenge confirmation, reorder, audit, intercept,
base station acknowledgment, lock until power cycled, maintenance required, unlock,
release (with or without reason), registration accepted, registration request, registration
rejected, and local control.
118 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

8. Channel Assignment message. Message to inform the mobile station to tune to a new
frequency.
9. Data Burst message. Data message sent by base station to the mobile station.
10. Authentication Challenge message. Allows the base station to validate the mobile
identity. The unique mobile authentication keys and/or Shared Secret Data (SSD) for
each mobile registered in the system will be used to perform the authentication calcula-
tions. These are then sent back to the base station in an Authentication Challenge
Response message.
11. SSD Update message. Request by the base station for the mobile station to update the
SSD.
12. Feature Notification message. Contains information records to allow the network to
supply information to be displayed by the mobile, to identify the called party’s number,
to identify the calling party’s number, to convey information to the mobile by means of
tones or other alerting signals, and to indicate the number of messages waiting.
The paging channel is divided into 80-ms slots called the paging channel slots. IS-95
allows two modes of paging—slotted and nonslotted. In the slotted mode, a mobile listens for
pages only at certain times (i.e., during its page slot). This feature allows the mobile to turn off
its receiver for most of the time, saving the battery power and increasing time between battery
charging. In the nonslotted mode of operation, the mobile is required to monitor all paging slots.
In the slotted mode, a mobile generally monitors the paging channel for one or two slots
per slot cycle. The mobile can specify its preferred slot using the SLOT_CYCLE_INDEX field
in the Registration message, in the Origination message, in the Page Response message, in traf-
fic signaling to the base station. The length of the slot cycle, T, in units of 1.28 seconds is given
by T = 2i, where i is the selected slot cycle index.
There are 16T slots in a cycle for a particular mobile using some value of i, and four 20-ms
full frames in an 80-ms slot. A value of i = 0 means that the mobile listens to every 16th paging
slot; i = 1 implies that the mobile monitors every 32nd slot; for i = 2, the mobile monitors every
64th slot. The i = 0 value ensures that the pages are not missed by the mobile, but it is a drain on
mobile’s battery power. The value of i = 1 is suggested. PGSLOT is a randomly calculated num-
ber that specifies the slot out of the 16T slots to be monitored by the mobile. This number is
fixed for each mobile.

7.7.4 Access Channel

The access channel is used by the mobile station to transmit control information to the
base station. The access channel allows the mobile station to communicate nontraffic informa-
tion (e.g., call origination and response to page). The access rate is fixed at 4800 bps. All mobile
stations accessing a system share the same frequency assignment. Each access channel is identi-
fied by a distinct access channel long-code sequence having an access number, a paging channel
number associated with the access channel, and other system data. There are many messages
that can be carried on the access channel. When a mobile places a call it uses the access channel
to inform the base station. This channel is also used to respond to a page.
Logical Channels 119

The messages carried by access channel include

1. Registration message. The mobile station sends this message to inform the base sta-
tion about its location, status, identification, and other parameters required to register
with the system. This is necessary so that the base station can page the mobile when-
ever a call is to be delivered to the mobile.
2. Order message. Typical order messages include base station challenge, SSD update
confirmation, SSD update registration, mobile station acknowledgment, local control
response, mobile station reject (with or without reason).
3. Data Burst message. This is a user-generated data message sent by the mobile station
to base station.
4. Origination message. This message allows the mobile station to place a call—sending
dialed digits.
5. Page Response message. This message is used by the mobile to respond to a page or
slotted page in continuation of the process of receiving a call.
6. Authentication Challenge Response message. This message contains necessary
information to validate the mobile station’s identity.
Fig. 7-16 shows the processing of the access channel. The baseband information is error
protected using a convolutional encoder of rate r = 1/3. The lower encoding rate on the reverse
link is used to make error protection more robust. The symbol repetition repeats the symbol,
yielding a code symbol rate of 28.8 kilosymbols per second (ksps). The data is then interleaved
to combat fading, after which 64-ary orthogonal modulation is used. In this case, for each exited

I-Channel PN
Sequence
1.2288 Mcps
1.2288 Mcps
Access BB
Message
Convolutional 64-ary
Encoder
Symbol Block
Orthogonal To OQPSK
Rate = 1/3, K = 9 14.4 Repetition Interleaver Modulation
4.8 28.8 28.8 Modulator
kbps ksps ksps ksps

BB

Q-Channel PN
Sequence
Long-Code
1.2288 Mcps
Generator 1.2288 Mcps
BB = baseband filter

Long-Code Mask

Figure 7-16 Access Channel Processing

120 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

6 input symbols to the modulator, one output Walsh function is generated. The reason for using
orthogonal modulation of the symbol is the noncoherent nature of the reverse link. Since each
group of 6 symbols is represented by a unique Walsh function, it is much easier for the base sta-
tion to detect 6 symbols at a time by deciding which 64-bit Walsh function was transmitted dur-
ing that period. An exited of 6 coded symbols is used as 6 binary symbols corresponding to a
decimal value between 1 and 64. The pattern of the 6-symbol group determines the choice of the
particular Walsh function transmitted.
The Walsh function is defined by

W i = c 0 + 2c 1 + 4c 2 + 8c 3 + 16c 4 + 32c 5 (7.2)

where c5 is the most recent and c0 the oldest of the 6 symbols to be transmitted, and
Wi is selected from 1 of 64 orthogonal Walsh functions.
As an example suppose an exited of 6 symbols (–1, 1, 1, –1, –1, –1) is an input. The correspond-
ing bit values are (1 0 0 1 1 1). The output of the modulator in terms of the Walsh function will be

W 39 = 1 + 2 × 1 + 4 × 1 + 8 × 0 + 16 × 0 + 32 × 1 = 39

The 64-ary modulated data at 4.8 ksps (modulation symbols) or at 307.2 ksps (code sym-
bols) is spread using the long PN sequence at 1.2288 Mcps. The long code has a length of 242 – 1
bits and is generated by the following polynomial:
42 35 31 27 26 25 22 21 19 18
x +x +x +x +x +x +x +x +x +x +
L( x ) = (7.3)
17 16 10 7 6 5 3 2
x +x +x +x +x +x +x +x +x+1
The output of the long-code generator is modulo-2 added with a shared 42-bit long-code
mask to generate a long code. The long PN sequence is used to distinguish the access channel
from all other channels occupying the reverse link. The data is scrambled in the I and Q paths by
the short PN sequence. Since the reverse link uses OQPSK modulation, the data in the Q path is
delayed by 1/2 a PN chip. The primary reason for this delay is to prevent the collapse of the
QPSK signal envelope to 0. This is essential because the power amplifier of the mobile is typi-
cally small and limited in performance.
The message on the access channel consists of an access preamble of multiple frames of
96 0 bits with a length of 1_PAM_SZ frames (Fig. 7-17), followed by an Access Channel

Access Channel Preamble

= 00000…0000
96 × (1 + PAM_SZ) bits
(1 + PAM_SZ) frames

Figure 7-17 Access Channel Preamble

Logical Channels 121

message capsule of length 3 + MAX_CAP_SZ frames. The message capsule also consists of
frames 96 bits in length. Since the data rate on the access channel is 4800 bps, each frame has
a duration of 20 ms.
The entire access channel transmission, therefore, occurs in an access channel slot that has
a length of

4 + MAX_CAP_SZ + PAM_SZ frames (7.4)

where the values of MAX_CAP_SZ and PAM_SZ are received on the paging channel.
An access channel slot nominally begins at a frame where

t mod (4 + MAX_CP_SZ + PAM_SZ) = 0 (7.5)

where t is the system time in frames.

The actual start of transmission on the access channel is randomized to minimize colli-
sions between multiple mobiles accessing the channel at the same time.
All access channels corresponding to a paging channel have the same slot length. Different
base stations may have different slot lengths.
The Access Channel message (Fig. 7-18) is similar in form to the Sync Channel message
and has an 8-bit message-length header, a message body of a minimum of 2 bits and a maximum
of 842 bits, and a CRC code of 30 bits. Following the message are padding bits to make the mes-
sage end on a frame boundary. The message length includes the header, body, and CRC, but not
the padding bits. The CRC is computed on the message-length header and message body using
the same code as the sync channel (Eq. 7.1).
Each access channel frame contains either preamble bits (all 0s) or message bits. Frames
containing message bits (Fig. 7-19) have 88 message bits and 8 encoder tail bits (set to all 0s).
Multiple frames are combined with an access channel preamble to form an access channel slot
(Fig. 7-20).

8 bits N MSG = 2–842 bits 30 bits

Message
Length Data CRC padding
(in bytes) = … 000

NMSG = Message length in bits (including length field and CRC)

Figure 7-18 Message Framing on Access Channel

88 bits 8 bits

Access Channel Frame Body Encoder Tail Bits

Figure 7-19 Access Channel Framing

122 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

96 (1 + PAM_SZ) bits or (1 + PAM_SZ) frames 96 bits 96 bits

Access Access
Access Channel Pramble Channel Channel
Frame Frame

96 × (4 + PAM_SZ + MAX_CAP_SZ) bits

Figure 7-20 Access Channel Slot

7.7.5 Forward Traffic Channels

The forward traffic channels are grouped into rate sets. RS1 has four elements: 9600,
4800, 2400, and 1200 bps. RS2 uses four elements: 14,400, 7200, 3600, and 1800 bps. All sys-
tems support RS1 on the forward traffic channels. RS2 is optionally supported on the forward
traffic channels. When a system supports a rate set, it supports all four elements. Walsh codes
that can be assigned to forward traffic channels are available at a cell or sector (W2 through W31
and W33 through W63). After the first paging channel, each additional paging channel consumes
one traffic channel Walsh code. So, if all seven paging channels are used, only 55 Walsh codes
will be available for the forward traffic channels.
Each forward traffic channel contains one fundamental code channel and may contain one
to seven forward supplemental code channels. Each of these code channels is orthogonally
spread by the appropriate Walsh function and is then spread by a quadrature pair of PN
sequences at a fixed chip rate of 1.2288 Mcps. Multiple forward CDMA channels may be used
within a base station in a frequency-division-multiplexed manner.
The speech is encoded using a variable-rate vocoder to generate the forward traffic data
depending on voice activity. Since frame duration is fixed at 20 ms, the number of bits per frame
varies according to the traffic rate. Since half-rate convolutional encoding is used, it doubles the
traffic rate to provide rates from 2400 to 19,200 symbols per second. Interleaving is performed
over 20 ms. A long code of 242 – 1 = 4.4 × 1012 is generated containing the user’s ESN embedded
in the mobile station long-code mask (with voice privacy, the mobile station long-code mask does
not use the ESN). The scrambled data are multiplexed with power control information that steals
bits from the scrambled data. The multiplexed signal remains at 19,200 bps and is changed to
1.2288 Mcps by the Walsh code Wi assigned to the ith user traffic channel. The signal is spread at
1.2288 Mcps by quadrature PN binary sequence signals, and the resulting quadrature signals are
then weighted. The power level of the traffic channel depends on its data transmission rate.
A power control subchannel is continuously transmitted on the forward traffic channel. A
0 specifies that the mobile increases its mean output power level by 1 dB (nominal), and a 1 indi-
cates a decrease in mean output power level by 1 dB (nominal).
The power control bits puncture the modulated data symbols at a rate of 800 bps; a single
power control bit replaces 2 data symbols. The location of a power control bit in a power control
group is determined by bits 20–23 of the 1/64 long code in the previous 1.25-ms control group.
Logical Channels 123

The 19.2-ksps long code is decimated to 800 bps (see Fig. 7-21) to establish the location of the
bits in the power control subchannel. This is done to randomize the location of the power control
bits to avoid any spikes due to periodic repetition. Fig. 7-22 shows forward traffic channel pro-
cessing for RS2.
Each power control group contains 24 scrambled bits. The 24-bits position in a 1.25-ms
period is numbered 0, 1, 2, …, 23 (see Fig. 7-23). The power control bits are always trans-
mitted at full power.
Channels not used for paging or sync can be used for traffic. Thus, the total number of traf-
fic channels at a base station is 63 minus the number of paging plus sync channels in operation at
that base station. Information on the forward traffic channels includes the primary traffic
(voice or data), secondary traffic (data), and signaling in frames 20 ms in length.
When the data rate on the forward traffic channel is 9600 bps (RS1), each frame of 192
bits carries 172 information bits, 12 frame-quality bits, and 8 encoder tail bits (set of all 0s). At
4800 bps, there are 80 information bits, 8 frame-quality bits, and 8 tail bits for a total of 96 bits.
At 2400 and 1200 bps, there are 40 and 16 information bits and 8 tail bits, for a total of 48 and
24 bits, respectively. The base station can select the data transmission rate on a frame-by-frame
basis. A data rate of 9600 bps can support multiplexed traffic and signaling. Data rates of 1200,
2400, and 4800 bps can support only primary traffic information.
The frame-quality indicator is a CRC on the information bits in the frame. At 9600 bps the
generator polynomial is

I-Channel PN
Convolutional Sequence
Symbol
Encoder Walsh i (1.2288 Mcps)
Rate = 1/2, K = 9 Repetition

Power Control Bits BB

800 bps
Modulation
Symbol
Block To QPSK
Interleaver MUX*
Modulation
19.2 19.2
ksps ksps

19.2 BB
ksps 800
64:1 24:1 bps
I-Channel PN
Long-Code Sequence
Decimator Decimator (1.2288 Mcps)
Generator

1.2288
Mcps
* Power control bits are not multiplexed in for supplemental code channel of the forward traffic channels.
BB = baseband filter
MUX = multiplexer

Figure 7-21 Forward Traffic Channel Processing for RS1

124 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

I-Channel PN
Convolutional Sequence
Symbol
Encoder 28.8 Walsh i (1.2288 Mcps)
Rate = 1/2, K = 9 Repetition
ksps
Power Control Bits BB
800 bps
Modulation
Symbol
Puncture 2 of Block To QPSK
Every 6 Inputs Interleaver MUX*
Modulation
19.2 19.2
ksps ksps

19.2 BB
ksps 800
64:1 24:1 bps
I-Channel PN
Long-Code Sequence
Decimator Decimator (1.2288 Mcps)
Generator

1.2288
Mcps
* Power control bits are not multiplexed in for supplemental code channel of the forward traffic channels.
BB = baseband filter
MUX = multiplexer

Figure 7-22 Forward Traffic Channel Processing for RS2

1 Frame (20 ms): 16 power control groups

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1.25
ms

1.25 ms = 24 scrambled traffic data bits

1 power control bit = 2 data bits

20 21 22 23 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 0 1 2 3

1/64 Long Code Used for Scrambling

20 21 22 23
1 1 0 1

These 4 bits of previous long code specified the starting point of the power
control bits value: 1 1 0 1 = 11; the power control bit starts at position 11.

Figure 7-23 Position of the Power Control Bits

Logical Channels 125

12 11 10 9 8 4
g( x ) = x + x + x + x + x + x + x + 1 (7.6)
At 4800 bps, the generator polynomial is
8 7 4 3
g( x ) = x + x + x + x + x + 1 (7.7)
The generator polynomials for the 10-bit and 6-bit frame-quality indicators are as follows:
• 10-bit frame-quality indicator
8 7 4 3
g( x ) = x + x + x + x + x + 1 (7.8)
• 6-bit frame-quality indicator
6 2
g( x ) = x + x + x + 1 (7.9)
When the data rate on the forward traffic channel is 14,400 bps (RS2), each frame of 288
bits carries 267 information bits, 12 frame-quality bits, and 8 encoder tail bits. Table 7-9 pro-
vides a summary of the forward traffic channel frame structure for RS1 and RS2. Figs. 7-24 and
7.25 show the forward traffic channel frame structure for RS1 and RS2, respectively.
At 9600 bps, the 172 information bits consist of 1 or 4 format bits and 171 or 168 traffic
bits. A variety of different multiplexing options are supported. The entire 171 information bits can
be used for primary traffic, or the 168 bits can be used for 80 primary traffic bits and 88 signaling
traffic bits or 88 secondary traffic bits. Other options use 40 and 128 or 16 and 152 bits for pri-
mary and signaling/secondary traffic. Alternatively, the entire 168 bits can be used for signaling
or secondary traffic.

Table 7-9 Forward Traffic Channel Frame Structure Summary

Number of Bits per Frame

Rate Transmission Reserved/ Frame-Quality Encoder

Set Rate (bps) Total Flag Information Indicator Tail Bits
1 9600 192 0 172 12 8
a
4800 96 0 80 8 8
a
2400 48 0 40 0 8
a
1200 24 0 16 0 8
2 14,400 288 1 267 12 8
a
7200 144 1 125 10 8
a
3600 72 1 55 8 8
a
1800 36 1 21 6 8
a
Applicable to forward fundamental code channel only; not allowed on forward supplemental code channels.
126 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

192 bits (20 ms)

9600-bps
Frame 172 12 8
F T

96 bits (20 ms)

4800-bps
Frame 80 8 8

F T

48 bits (20 ms)

2400-bps
Frame 40 8

24 bits (20 ms)

1200-bps F = frame-quality indicator (CRC)

Frame 16 8
T = encoder tail bits
T

Figure 7-24 Forward Traffic Channel Frame Structure for RS1

288 bits (20 ms)

14,400-bps
Frame 1 267 12 8

R/F F T

144 bits (20 ms)

7200-bps
Frame 1 125 10 8
R/F F T

72 bits (20 ms)

3600-bps
Frame 1 55 8 8

R/F F T

36 bits (20 ms)

F = frame-quality indicator (CRC)
1800-bps T = encoder tail bits
Frame 1 21 6 8
R/F = reserved/flag bit
R/F F T

Figure 7-25 Forward Traffic Channel Frame Structure for RS2

Logical Channels 127

When the forward traffic channel is used for signaling, the message is similar in form to
the paging channel (see Fig. 7-10) and has an 8-bit message-length header, a message body of a
minimum of 16 bits and a maximum of 1160 bits, and a CRC code of 16 bits. Following the
message are padding bits to make the message end on a frame boundary. The message length
includes the header, body, and CRC, but not the padding. The CRC is computed on the message-
length header and the message body using
16 12 5
g( x ) = x + x + x + 1 (7.10)

When the forward traffic channel is used for signaling, the following are typical messages
that can be sent:
1. Order message. Similar to the Order message on the paging channel.
2. Authentication Challenge message. When the base station suspects the validity of the
mobile, it can challenge the mobile to prove its identity.
3. Alert with Information message. Allows the base station to validate the mobile identity.
4. Data Burst message. A data message sent by the base station to the mobile.
5. Handoff Direction message. Provides the mobile with information needed to begin the
handoff process.
6. Analog Handoff Direction message. Tells the mobile to switch to the analog mode
and begin the handoff process.
7. In-Traffic System Parameters message. Updates some of the parameters set by the
System Parameters message in the paging channel.
8. Neighbor List Update message. Updates the neighbor base station parameters set by
the Neighbor List message on the paging channel.
9. Power Control message. Tells the mobile how long the period is or what threshold is
to be used in measuring frame-error statistics that will be sent in the mobile’s power
measurement report message.
10. Send Burst Dual-Tone Multifrequency (DTMF) message. When the base station
needs dialed digits, it can request them in this message. For example, this message
would be used for digits for a three-way call.
11. Retrieve Parameters message. Requests the mobile to report on any of the retrievable
and settleable parameters (refer to IS-95 appendix E).
12. Set Parameter message. Informs the mobile to adjust any of the retrievable and settle-
able parameters (refer to IS-95 appendix E).
13. SSD Update message. A request from the base station for the mobile to update the
shared secret data.
14. Flash with Information message. Contains information records to allow the network
to supply display information to be displayed by the mobile, to identify the responding
party’s number (the connected number), to convey information to the mobile by means
of tones or other alerting signals, and to indicate the number of messages waiting.
128 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

15. Mobile Registration message. Informs the mobile that it is registered and supplies the
necessary system parameters.
16. Extended Handoff Direction message. One of several handoff messages sent by the
base station.

7.7.6 Reverse Traffic Channels

For the RS1, the reverse traffic channel may use either 9600-, 4800-, 2400-, or 1200-bps
data rates for transmission. The duty cycle for transmission varies proportionally with the data
rate, from 100% at 9600 to 12.5% at 1200 bps. An optional RS2 is also supported. The actual
burst transmission rate is fixed at 28,800 code symbols per second. Reverse traffic channel pro-
cessing is similar to that of the access channel. The major difference is that the reverse traffic
channel (fundamental code) uses a data burst randomizer (see Fig. 7-26). The reverse CDMA
channel structure for the supplementary code channel is shown in Fig. 7-27.
Since hextets (6 code symbols) are modulated as one of 64-ary modulation symbols for
transmission, the modulation symbol transmission rate is fixed at 4800 modulation symbols per
second. This results in a fixed Walsh chip rate of 307.2 kcps. The data from the 64-ary modulator
is fed into the data burst randomizer. The data burst randomizer takes advantage of the voice
activity on the reverse link. This is used to reduce reverse link power during a quiet period of
speech by pseudorandom masking out redundant symbols produced by symbol repetition. This
is achieved by the data burst randomizer. The data burst randomizer generates a masking pattern
of 0s and 1s to randomly mask out redundant data. The masking pattern depends on the vocoder

1.2 kbps (1.8) 3.6 ksps (3.6)

2.4 kbps (3.6) 7.2 ksps (7.2)
4.8 kbps (7.2) 14.4 ksps (14.4)
9.6 kbps (14.4) 28.8 ksps (28.8)
I-Channel PN
Convolutional Sequence
Symbol (1.2288 Mcps)
Encoder
Rate = 1/3, K = 9 Repetition 28.8
ksps
BB

64-ary Data
Block
Othogonal Burst
To QPSK
Interleaver Modulation
28.8 Modulator 4.8 Randomizer
ksps ksps

BB

Q-Channel PN
Long-Code Sequence
Note: Numbers in ( ) represent RS2. Generator 1.2288 (1.2288 Mcps)
BB = baseband filter Mcps

Figure 7-26 Reverse Traffic Channel Processing for Fundamental Code Channel
Logical Channels 129

I-Channel PN
Sequence
(1.2288 Mcps)

BB

Convolutional 64-ary
Encoder
Block To QPSK
Orthogonal
Rate = 1/3, K = 9 Interleaver Modulation
9.6 28.8 Modulator 4.8
kbps ksps ksps

BB

Q-Channel PN
Long-Code Sequence
Generator 1.2288 (1.2288 Mcps)
BB = baseband filter Mcps

Figure 7-27 Reverse CDMA Channel Structure for Supplementary Code Channel

rate. For a vocoder operating at 9.6 kbps, no data is masked out, whereas, if the vocoder is oper-
ating at 1.2 kbps, then the symbols are repeated seven times and the data burst randomizer masks
out seven out of eight groups of symbols.
Each 20-ms traffic channel frame is divided into 16 power control groups of 1.25 ms each
(as discussed earlier). The data burst randomizer pseudorandomly masks out individual power
control groups (PCG). With 9.6 kbps, no PCG is masked out; with 4.8 kbps, 8 PCGs are masked
out in a frame; with 2.4 kbps, 12 PCGs are masked; with 1.2 kbps, 14 PCGs are masked. An
example of this operation with a vocoder operating at 2.4 kbps is shown in Fig. 7-28.

20 ms (16 PCGs)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1.25
ms Data Burst
(1 PCG) Randomizer (operating at 2.4-kbps frame)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Transmitted PCG

Figure 7-28 Data Burst Randomizer Operation at 2.4-kbps Frame

130 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

The reverse channel structure for RS2 is similar. The RS2 vocoder supports 14.4, 7.2, 3.6,
and 1.8 kbps data rates. For RS2, the convolutional encoder is 1/2 instead of 1/3 as in RS1.
The system can multiplex primary (voice) and secondary (data) or signaling traffic on the
same traffic channel. Multiplex option 1 is used to transmit primary and secondary traffic. This
option is also used to transmit primary (voice) and signaling (messaging) traffic. Multiplex
option 1 uses the following methods to simultaneously transmit primary and secondary traffic.
• Blank and burst. The entire traffic channel frame is used to send only secondary data
or only signaling data. The secondary or signaling data blank out the primary data (see
Fig. 7-29).
• Dim and burst. The traffic frame is used to send both primary and secondary data or
both primary and signaling data (see Fig. 7-30). Fig. 7-31 shows the traffic channel
frame structure for the reverse link.

172 bits

1 1 2 Signaling Traffic 168 bits

MM TT TM MM = mixed mode = 0; full speech
= 1 = 0 = 11 MM = 1; signaling only or lower-rate speech + signaling
TT = traffic type
TM = traffic mode

Figure 7-29 Blank and Burst for Speech RS1

Rate 1/2

1 1 2 Primary Traffic (80) Signaling Traffic (88)

MM TT TM
= 1 = 0 = 00
Rate 1/4
Primary
1 1 2 Traffic (40) Signaling Traffic (128)

MM TT TM
= 1 = 0 = 01
Rate 1/8
Primary
1 1 2 Traffic Signaling Traffic (152)
(16)
MM TT TM MM = mixed mode = 0; full speech
= 1 = 0 = 10 MM = 1; signaling only or lower-rate speech + signaling
TT = traffic type
TM = traffic mode

Figure 7-30 Dim and Burst for Speech RS1

Logical Channels 131

192 bits (20 ms)

Traffic Rate—
1 171 bits F = 12 T = 8 9600 bps

96 bits (20 ms)

Traffic Rate—
80 bits F = 8 T = 8 4800 bps

48 bits (20 ms)

Traffic Rate—
40 bits T = 8 2400 bps

24 bits (20 ms)

Traffic Rate—
16 bits T = 8 1200 bps

Figure 7-31 Traffic Channel Frame Structure for Reverse Link

The reverse traffic channel carries the following typical messages:

1. Order messages on the reverse traffic channel typically include base station challenge,
SSD update confirmation, SSD update rejection, parameter update confirmation,
mobile station acknowledgment, service option request, service option response,
release (normal with power-down indication), long-code transition request (public and
private), connect, continuous DTMF tone (start and stop), service option control,
mobile station reject (with and without a reason), and local control.
2. Authentication Challenge Response message contains the information to validate the
mobile’s identity.
3. Flash with Information message contains information records from the mobile station
concerning mobile features, mobile key pay facility, called party number, calling party
number, and the connected number (i.e., the responding party).
4. Data Burst message is a user-generated data message sent by the mobile to the base
station.
5. Pilot Strength Measurement message sends information about the strength of other
pilot signals that are not associated with the serving base station.
6. Power Measurement Report message sends FER statistics to the base station. The
report is generated at specified intervals or when a threshold is reached.
132 Ch. 7 • Physical and Logical Channels of IS-95 CDMA

7. Send Burst DTMF message uses two tones DTMF—one low- and one high-frequency
tone to represent a dialed digit—and transmits dialed digits to the base station.
8. Status message contains information records from the mobile about mobile identifica-
tion, mobile call mode, mobile terminal information, and security status.
9. Origination Continuation message is a continuation of the origination message that
was sent on the access channel if additional dialed digits need to be sent.
10. Handoff Completion message is the mobile response to a Handoff Direction message.
11. Parameter Response message is the mobile response to the base station for a Retrieve
Parameters message.

7.8 Summary
This chapter discussed the implementation of CDMA carriers in an existing AMPS or TDMA
system, establishing that 59 AMPS carrier channels must be removed to introduce the first
CDMA carrier without interfering with the remaining AMPS or TDMA channels. The chapter
provided details of information processing and message framing for the pilot, sync, paging, and
traffic channels on the forward link. The pilot channel (Walsh code 0) is unmodulated; it consists
of only short-code-spreading sequences.
The pilot channel is used by all mobiles attached to a cell as a coherent phase reference
and also provides a unique identifier for different base stations. The sync channel (Walsh code
32) transmits system timing information to allow mobiles to synchronize themselves with base
stations. The paging channel(s) (Walsh codes 1–7) are the digital control channel(s) for the
CDMA forward link. One base station can have up to 7 paging channels. The first paging chan-
nel is always assigned Walsh code 1. The traffic channels (Walsh 8–31 and 33–63) carry digi-
tized voice data.
On the reverse link, the access channel is used by the mobile to register with the system, to
access the system before assignment of a traffic channel, to originate a call, or to respond to a
page. Several access channels can be used per paging channel. The reverse traffic channels are
used to deliver encoded voice and reverse link signaling from mobile to base station.
We also presented typical messages that are carried over the logical channels of the for-
ward and reverse links.

7.9 References
1. Garg, V. K., Smolik, K., and Wilkes, J. E., Applications of CDMA to Wireless/Personal
Communications, Prentice Hall PTR, Upper Saddle River, NJ, 1997.
2. Garg, V. K., and Wilkes, J. E., Wireless and Personal Communications Systems, Prentice
Hall PTR, Upper Saddle River, NJ, 1996.
3. Rappaport, T. S., Wireless Communications, Prentice Hall PTR, Upper Saddle River, NJ,
1996.
4. TIA/EIA/SP-3693, “Mobile Station–Base Station Compatibility Standard for Dual-Mode
Wideband Spread Spectrum Cellular Systems,” November 1997.
 C H A P T E R 8

IS-95 CDMA Call

Processing

8.1 Introduction
This chapter discusses IS-95 CDMA call processing states that a mobile station goes through in
getting to a traffic channel. These include the system initialization state, the system idle state, the
system access state, and the traffic channel state. Each of the call processing states has several
substates that are also discussed, in addition to idle handoff, slotted paging operation, CDMA
registration, and authentication procedures. The chapter covers messages used to exchange data
in different call processing states and concludes by providing call flows for CDMA call origina-
tion, call termination, call release, and authentication.

8.2 CDMA Call Processing State

In getting to a traffic channel, a mobile station in CDMA goes through several states: system ini-
tialization state, system idle state, system access state, and traffic channel state (see Fig. 8-1).
In the system initialization state the mobile acquires a pilot channel by searching all the
PN-I and PN-Q possibilities and selecting the strongest pilot signal. Once the pilot is acquired, the
synchronization (sync) channel is acquired using the W32 Walsh function and the detected time off-
set of the pilot channel. Then the mobile obtains the system configuration and timing information.
Next the mobile enters the system idle state where it monitors the paging channel. The
mobile can receive messages from the base station containing necessary parameters to initiate or
to receive a call.
If a call is being placed or received, the mobile enters the system access state where the
necessary parameters are exchanged. The mobile transmits its response on the access channel
and the base station transmits its response on the paging channel.
When the access attempt is successful, the mobile enters the traffic channel state. In this
state, speech communication takes place, with associated control messages replacing digital
speech by either of two methods:

133
134 Ch. 8 • IS-95 CDMA Call Processing

• Blank-and-burst signaling. The case where the complete speech packet is replaced
with signaling.
• Dim-and-burst signaling. The case where part of the speech packet is replaced with
signaling.

CDMA RF Carrier
Selection

Tune to primary/secondary
CDMA carrier.

System Initialization State

Mobile acquires pilot channel

and
sync channel.

System Idle State

Mobile acquires paging channel

and
monitors for messages.

System Access State

Mobile sends messages on

access channel; BS sends
messages on paging
channel.

Traffic Channel State

Speech communication
on forward (downlink)
and reverse (uplink)
traffic channels;
Associated control message exchange using:
- blank and burst
- dim and burst
- power control (downlink)

Figure 8-1 Getting to a Traffic Channel in CDMA

CDMA Call Processing State 135

In addition, power control messages are sent by a method called bit puncturing on the for-
ward link channel. In bit puncturing two gross-data bits are replaced by a single power control
bit.

8.2.1 System Initialization State

In the system initialization state, the mobile selects a system to use and then proceeds to
acquire and then synchronize to a CDMA carrier. The system initialization state consists of the
following substates (Fig. 8-2):
• System determination substate. In this substate, the mobile selects the system to use
(analog or CDMA) if it is a dual-mode cellular unit. The mobile choices include service
provider preference. If a CDMA system is selected, the mobile sets the CDMA channel
parameters (CDMA_CH) to Ni , where Ni is either a primary or secondary CDMA chan-
nel number.
• Pilot channel acquisition substate. The mobile acquires the pilot channel of the
CDMA system. The mobile tunes to the CDMA_CH, sets its code channel for the pilot
channel, and searches for pilot channel. If the mobile acquires the pilot channel within
T20m seconds, the mobile enters the sync channel acquisition substate. If the mobile
does not acquire the pilot channel within T20m seconds, the mobile enters the system
determination substate with an acquisition failure indication.
• Synchronization (sync) channel acquisition substate. The mobile acquires the syn-
chronization channel and obtains system configuration and timing information for the

Power-up and
Mobile Enters Analog
Initialization Initialization
State Enter
Idle State
Begin Analog End Analog Mode—
Mode No CDMA Pilot
Found

System Pilot Sync Channel

Determination Acquisition Acquisition Timing Change
Substate Substate Substate Substate

CDMA System Pilot Channel Sync Channel System Timing

Is Selected Is Acquired Is Acquired Is Acquired

Figure 8-2 System Initialization State

136 Ch. 8 • IS-95 CDMA Call Processing

CDMA system. Upon entering the sync channel acquisition substate, the mobile sets
its code channel for the sync channel. If the mobile does not receive a valid sync
channel message within T21m seconds, the mobile enters the system determination
substate with a protocol mismatch indication. If the mobile receives a valid sync
channel message within T21m seconds, the mobile stores system configuration and tim-
ing information.
• Timing change substate. The mobile synchronizes its long-code timing and system
timing to those of the CDMA system after receiving and processing the synchroniza-
tion message.
Fig. 8-3 shows the system states and the activities associated with them.
Synchronization occurs when the phase (t – T) of the locally generated PN code is equal to
the phase (t – Ti – Tp) of the incoming code.

(t – T ) = (t – Ti – T p) (8.1)

∴T = T i + T p (8.2)

where T = phase of local PN code,

TP = propagation delay, and
Ti = pilot offset for ith pilot.
At this stage of the system, we observe two points:

Begin Analog Power-up

Mode

1
Analog CDMA
Initialization Initialization
Idle
End Analog 2 Handoff
Mode
Idle
State
Call Origination, System
Page Response, 3 Access Ack
or Registration Other Than
System
Origination
Access or Page
State Response
Traffic Channel
Acquisition
4
Control on End of Traffic
Traffic Channel Channel Use

Figure 8-3 CDMA Call Processing States for Dual Mode Mobile
CDMA Call Processing State 137

1. When the locally generated code phase matches with the incoming pilot, pilot acquisi-
tion occurs. Thus, PN (t – Ti – Tp) is known.
2. Although total phase (Ti + Tp) is known, the pilot offset Ti is not known. We get the pilot
offset Ti from the Sync message of the sync channel.
Once the sync channel is acquired and the Sync message is received and processed, the
mobile stores the following information from the Sync message:
• Protocol revision level (MIN_P_REVstored = MIN_P_REVreceived)
• System identification (SID)
• Network identification (NID)
• Pilot PN offset (Ti )
• System time (Ts )
• Long-code state at system time (LC_STATE)
• Paging channel data rate (PRAT)
• Number of leap seconds that have occurred since the start of system time (LP_SEC)
• Offset of local time from system time (LTM_OFF)
• Daylight saving time indicator (DAY_LT)
In the timing change substate, the mobile uses the pilot offset, the system time, and the
long-code state information, obtained from the Sync message, to synchronize its timing to the
system time and to synchronize its long-code phase to that of the system.
In IS-95 the long code is generated with a 42-stage shift register. The mobile knows the
generation polynomial. The problem is to get the correct code phase. The mobile obtains the
long-code state from the Sync message. The long code is a 42-bit sequence that corresponds to
the contents of a shift register at the system time, Ts. The mobile loads its shift registers with the
42-bit long-code state. The mobile waits and, at system time Ts, it starts shifting the contents of
the shift registers at 1.2288 Mcps. At this point, long-code synchronization is achieved. The
mobile may now tune to a paging channel in order to enter the idle state.

8.2.2 Idle State

In the idle state, the mobile monitors the paging channel. In this state, the mobile can
1. Receive messages and orders from the base station
2. Receive an incoming call
3. Initiate a registration process
4. Initiate a call
5. Initiate a message transmission
Fig. 8-4 summarizes the activities in the idle state. Upon entering the idle state, the mobile
sets its Walsh code to the primary paging channel (W1) and sets its paging channel rate to the
rate obtained from the Sync message.
138 Ch. 8 • IS-95 CDMA Call Processing

CDMA Initialization

Acquire
Primary Paging
Channel (W1)

Idle Activities
- Monitor Paging Channel
- Authentication
- Idle Handoff
- Update Overhead Information
- Page Response
- Mobile Origination
- Registration

Figure 8-4 Idle State Activities in CDMA

The paging channel is subdivided into 80-ms slots called paging channel slots. In the non-
slotted mode, paging and control data for the mobile can be received in any of the paging chan-
nel slots. Therefore, the mobile monitors all slots on a continuous basis. In IS-95, the paging
channel protocol also allows scheduling the transmission of messages for a given mobile in cer-
tain assigned slots. A mobile station that monitors the paging channel only during certain
assigned slots is referred to as operating in the slotted mode. During the slots in which the pag-
ing channel is not monitored, the mobile can stop or reduce its processing activities to save bat-
tery power. This is sometimes referred to as a discontinuous reception mode of operation.
In the slotted mode operation, the mobile monitors the paging channel for 1 or 2 slots per
cycle. Slotted page messages contain a field called more-pages. When this field is set to 0, it
indicates that the remainder of the slots will contain no more messages addressed to the mobile.
This allows the mobile to stop monitoring the paging channel as soon as possible. If a Slotted
Page message with the more-pages field set to 0 is not received in the assigned slot, the mobile
continues to monitor the paging channel for one additional slot. For each of its assigned slots,
the mobile begins monitoring the paging channel in time to receive the first bit of the assigned
slot. The mobile then continues to monitor the paging channel until one of the following condi-
tions is satisfied:
1. The mobile receives a Slotted Page message with the more-pages field set to 0.
2. The mobile monitors the assigned slot and the slot following the assigned slot and
receives at least one valid message.
The mobile can specify its preferred slot cycle using the SLOT_CYCLE_INDEX field in
CDMA Call Processing State 139

the Registration message, Origination message, or Page Response message. When the mobile
station is in the control on traffic channel state, it can also specify the preferred slot cycle using
the SLOT_CYCLE_INDEX field of the terminal information record of the Status message. The
length of the slot cycle, T, in units of 1.28 seconds, is given as
i
T = 2 (8.3)
where i = selected slot cycle index.
There are 16 × T slots in a slot cycle for a particular mobile using some value of i, and four
20-ms full frames in 80-ms slot. SLOT_NUM is the paging slot number. To determine the
assigned slots, the mobile uses a Hash algorithm to select a slot number in the range 0 to 2048.
The minimum and maximum cycles are 16 slots (1.28 seconds) and 2048 slots (163.84 seconds),
respectively. The value of SLOT_NUM is given as

SLOT_NUM =  --- mod2048

t
(8.4)
 4

where t = system time in frames.

For each mobile station, the starting time of its slot is offset from the beginning of each of
its slot cycles by a fixed, randomly selected number of slots, called PGSLOT. As an example, let
the selected slot cycle index i = 0, T = 20 = 1, so 16 × T = 1.28 seconds for a slot cycle, and the
computed value of PGSOLT equal to 6. The mobile begins monitoring the paging channel at the
start of the slot in which SLOT_NUM equals 6 (see Fig. 8-5). The next slot in which the mobile
must begin monitoring the paging channel is 16 slots later, i.e., the slot in which SLOT_NUM is
22, since the slot cycle length is 16 for T = 1.

1 Slot Cycle 1.28 Seconds

2047 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Paging Channel Slot

Mobile in Mobile in
Reacquisition 8 Paging Channel
Nonactive Nonactive
Time Half Frames
State State

80 ms
0
i = 0, T = 2 = 1; PGSLOT = slot number 6 out of every 16 slots

Figure 8-5 Slot Cycle with i = 0

140 Ch. 8 • IS-95 CDMA Call Processing

8.2.2.1 Acknowledgment Procedure

We consider a message sent to the mobile by the base station on the paging channel. The
message typically contains an address field, acknowledgment required sequence number, mes-
sage sequence number, and other fields including the data field (see Fig. 8-6). Whether or not the
mobile sends an acknowledgment to the base station depends on the value of the ACK_REQ
field. Acknowledgment of messages received on the paging channel are sent to the base station
on an access channel.
If the ACK_REQ field of the message from the base station is 1, the mobile transmits an
acknowledgment with the following parameters (see Fig. 8-7):
• VALID_ACK field: 1
• ADDR_TYPE field: I (address type of message being acknowledged)
• ACK_SEQ field: I (message sequence number of the message being acknowledged)
When the Page messages or Slotted Page messages addressed to a mobile do not have an
ACK_REQ field, the mobile transmits a Page Response message including an acknowledgment
in response to each record of a page message or slotted message.
When a message does not include an acknowledgment, the mobile sets the VALID_ACK
field to 0. The ADDR_TYPE and ACK_SEQ fields are then set to the ADDR_TYPE and
MSG_SEQ values of the last message that requires an acknowledgment.
• VALID_ACK field: 0
• ADDR_TYPE field: I ( address type of last message requiring acknowledgment)
• ACK_SEQ field: I (message sequence number of last message requiring acknowledg-
ment)
When a message does not include an acknowledgment, the mobile sets the VALID_ACK
field to 0. The ADDR_TYPE and ACK_SEQ fields then set to 000 and 111, respectively if there
is no previous message that requires an acknowledgment.

8.2.2.2 Idle Handoff

An idle handoff, or change of paging channel, occurs when a mobile has moved from the
coverage area of one base station to the coverage area of another base station during the idle

Other Field ADDR_TYPE ACK_REQ MSG_SEQ

Figure 8-6 A Typical Message from Base Station to Mobile

Other Field ADDR_TYPE VALID_ACK ACK_SEQ

Figure 8-7 A Typical Message with Acknowledgment from Mobile

CDMA Call Processing State 141

state. The mobile determines that a handoff should occur when it detects a new pilot that is suffi-
ciently stronger than the current pilot.
Pilot channels are identified by short PN offsets. They are grouped into sets describing
their status with respect to pilot searching procedures. In the idle state, three pilot sets are main-
tained: active, neighbor, and remaining.
Using a strategy similar to a sliding correlator, it is possible to acquire a pilot if its short
PN code’s correct phase is known. For each pilot set, a search window is specified. This allows
the mobile to search for the direct path as well as multipath components of the pilot signal. The
search window is centered either on the earliest arriving multipath or the short PN offset.
If the mobile determines that a neighbor set or remaining set pilot is sufficiently stronger
than the active set pilot, idle handoff is performed. While performing an idle handoff, the mobile
operates in nonslotted mode until at least one valid message is received from the new paging
channel. On receiving a valid message from the new paging channel, the mobile may resume
slotted mode operation. After performing an idle handoff, the mobile discards all unprocessed
messages received on the old paging channel.
The paging channel is used to transmit control information to the mobiles that have not
been assigned to traffic channels.Two types of control messages are sent:
• Overhead messages that are broadcast messages for all mobile stations.
• Directed messages addressed to a particular mobile or a specific group of mobiles.
There are four overhead messages that are continuously broadcast on the paging channel.
1. System Parameter message
2. Neighbor List message
3. CDMA Channel List message
4. Access Parameters message
The first three messages are called configuration messages. A configuration message sequence
number (CONFIG_MSG_SEQ) is associated with a set of configuration messages sent on the
paging channel. When the contents of one or more configuration messages change, the configu-
ration message sequence number is incremented. The mobile stores the sequence number con-
tained in each configuration message received.
Access Parameters messages are sequenced by the access parameter sequence number
(ACC_MSG_SEQ). The mobile stores the most recently received Access Parameters message
sequence number in the ACC_MSG_SEQ field.
Configuration and access parameters from one paging channel are not used while
monitoring a different paging channel. If the stored parameters are current, the mobile
processes the messages on the paging channel. When a System Parameter message
(SYS_PAR_MSG_SEQr) is received, its associated configuration message sequence number
(CONFIG_MSG_SEQr) is compared to the stored value of the System Parameter message
(SYS_PAR_MSG_SEQs). As long as there is a match, the received System Parameter message
142 Ch. 8 • IS-95 CDMA Call Processing

is ignored. If the result is a mismatch, the mobile stores the configuration message sequence
number, the SID, NID, and the base station identification (BASE_ID).
When an Access Parameters message is received, the ACC_MSG_SEQr is compared to
the stored value of the Access Parameters message sequence value (ACC_MSG_SEQs). If there
is a match, the received Access Parameters message is ignored. However, if the result is a mis-
match, the mobile stores the following parameters:
• Access Parameters message sequence number (ACC_MSG_SEQ)
• Number of access channel (ACC_CHAN)
• Nominal transmit power (NOM_PWR)
• Initial power offset for access (INIT_PWR)
• Power increment or power step (PWR_STEP)

8.2.2.3 Neighbor List Message

When a Neighbor List message (NGHBR_LST_MSG_SEQr) is received, its associated
configuration message sequence number (CONFIG_MSG_SEQ) is compared with the stored
value of the Neighbor List message (NGHBR_LST_MSG_SEQs). If there is a match, the
received Neighbor List message is ignored. If the result is a mismatch, the mobile stores the fol-
lowing parameters:
• Configuration message sequence number (CONFIG_MSG_SEQ)
• Short PN offsets of neighbor list members
• Short PN sequence offset increment (PILOT_INC)
The mobile updates the idle handoff neighbor set so that it contains only the pilot offsets
listed in the Neighbor List message.

8.2.2.4 CDMA Channel List Message

When a CDMA Channel List message (CHAN_LST_MSG_SEQr) is received, its associ-
ated configuration message sequence number (CONFIG_MSG_SEQ) is compared with the
stored value of the CDMA Channel List message (CHAN_LST_MSG_SEQs). If there is a
match, the received CDMA Channel List message is ignored. If the result is a mismatch, the
mobile stores the following parameters:
• Configuration message sequence number (CONFIG_MSG_SEQ)
• CDMA channel list message sequence (CHAN_LST_MSG_SEQ)

8.2.3 System Access State

The system access state (see Fig. 8-8) includes the following substates:
• Update overhead information substate. In this substate, the mobile monitors the pag-
ing channel until it has received a current set of configuration messages
• Mobile station origination attempt substate. The mobile station sends an Origination
message to the base station.
CDMA Call Processing State 143

Update
Overhead Information Received General Page
Substate Message, Page Message,
or Slotted Page Message
Registration
Access Page
User Initiated a Response
Received Message or Order
User Cancelled Call, or PACA Call Substate
Requiring Acknowledgment
a PACA Call Reoriginated
or Response

User-Generated
Mobile Data Burst Message
Order/Message
Response Mobile
Substate Origination
Registration Attempt
Access Substate
Substate

Mobile
Message PACA
Transmission Cancel
Substate Substate

(Enter Mobile (Enter Mobile Station Control

Station Idle State) PACA: Priority Access Channel Assignment on Traffic Channel State)

Figure 8-8 CDMA Mobile System Access State

• Page response substate. The mobile sends a Page Response message to the base station.
• Registration access substate. The mobile station sends a Registration message to the
base station.
• Mobile station order/message response substate. The mobile sends a response to a
message received from the base station.
• Mobile station message transmission substate. The mobile sends a Data Burst mes-
sage to the base station.
• PACA cancel substate. The mobile sends a Priority Access Channel Assignment
(PACA) Cancel message. If the base station responds with an authentication request,
the mobile responds in this substate.

8.2.4 Mobile Control on Traffic Channel State

Call Origination: The mobile station control on the traffic channel state for call origina-
tion consists of the following substates (see Figs. 8-9 and 8-10):
144 Ch. 8 • IS-95 CDMA Call Processing

Origination
Attempt

Traffic Channel
Initialization
Mobile receives a
base station ACK order.

Conversation

User initiates disconnect

or mobile receives release order.

Release

To Initialization State
Figure 8-9 Mobile Control on Traffic Channel (Call Origination)

• Traffic channel initialization substate. In this substate, the mobile verifies that it can
receive the forward traffic channel and begins to transmit on the reverse traffic channel.
• Conversation substate. The mobile station exchanges primary traffic packets with the
base station.
• Release substate. The mobile station disconnects the call.
Call Termination: The mobile station control on traffic channel for call termination con-
sists of the following substates (see Figs. 8-11 and 8-12):

• Traffic channel initialization substate (same as in “Call Origination”)

• Waiting for order substate. In this substate, the mobile waits for an alert with an infor-
mation message. The information may be data such as calling party number and/or
voice.
• Waiting mobile station answer substate. The mobile waits for the user to answer the
call.
• Conversation substate. The mobile station’s primary service option application
exchanges primary traffic packets with the base station.
• Release substate. The mobile disconnects the call.
Flow diagrams for CDMA far-end-initiated and mobile-initiated call release are shown in
Figs. 8-13 and 8-14, respectively.
CDMA Registration 145

MS BTS BSC MSC

Origination
A
(Access Channel) Origination

Subscriber Validation

Origination ACK
Null Data B
Channel Assignment
(Paging Channel)
Traffic Channel Preamble Frame Selector (FS) Request
FS Assignment C
FS Assignment

FS Connect and Clock Adjust

ACK

Channel Confirm

Origination Continuation
(Optional)
Origination Continuation
(Optional)
Setup

TO PSTN
ACK
Service Option Response
Service Option Response

Speech Packets (if vocoder is located in MSC)

Speech Frames
PSTN
Answer

Answer

A Channel element assigned to call

B Walsh code assigned to call
C Speech handler assigned to call

Figure 8-10 Flow Diagram for CDMA Call Origination

8.3 CDMA Registration

The registration process is used by the mobile to notify the base station of its location, status,
identification, slot cycle, and other characteristics. The base station can efficiently page the
mobile station when establishing a mobile-terminated call. For operation in the slotted mode, the
mobile supplies the SLOT_CYCLE_INDEX parameter so that the base station can determine
146 Ch. 8 • IS-95 CDMA Call Processing

Page Response

Traffic
Channel
Initialization
Mobile terminates call and receives
a base station ACK order on
forward traffic channel.
Waiting for
Order
Mobile receives an alert
with information message.
Mobile receives
Waiting for
release order.
Mobile
Mobile Answer
receives
release Mobile user answers call.
order.
Conversation

Mobile user initiates disconnect,

or mobile receives release order.

Release

Figure 8-11 Mobile Control on Traffic Channel (Call Termination)

which slots the mobile is monitoring. The mobile supplies the station class mark and protocol
revision number so that the base station knows the capabilities of the mobile station.
IS-95 supports nine different forms of registration:
1. Power-up registration. The mobile registers when it powers up, switches from using
an alternative serving system, or switches from using the analog system with dual-
mode cellular mobile.
CDMA Registration 147

MS BTS BSC MSC

PSTN
Incoming Call

Tone Request
Page Request Tone
Page
(Broadcast Channel)
(Paging Channel) Tone ACK
Page Response
(Access Channel)
ACK Page Response
(Paging Channel)
(Pref. Mode CDMA)
Call Treatment
Channel Assignment
(Paging Channel)
Traffic Preamble
FS (Frame Selector)Request
ACK
FS Assignment
FS Assignment
Transport Setup
ACK
ACK
Service Option Response

Alert

Connect Order

Answer
Answer
PSTN
Answer
ACK

Figure 8-12 Flow Diagram for CDMA Call Termination

2. Power-down registration. The mobile registers when it powers down, informing the
system that it is no longer active.
3. Timer-based registration. The mobile registers at regular intervals. Its use also allows
the system to automatically deregister mobile stations that did not perform a successful
power-down registration.
4. Distance-based registration. A mobile station registers when the distance between the
current base station and the base station in which it last registered exceeds a threshold.
The mobile determines that it has moved a certain distance by computing a distance
based on the difference in latitude and longitude between the current base station and
the base station where the mobile last registered. If this distance exceeds the threshold
value, the mobile station registers.
5. Zone-based registration. A mobile registers when it enters a new zone. Zones are
groups of base stations within a given system and network. A base station’s zone
148 Ch. 8 • IS-95 CDMA Call Processing

MS BTS BSC MSC

Disconnect
Release
Release
Release
ACK
FS (Frame Selector) Remove

ACK
ACK
Clear
Disconnect
ACK

Figure 8-13 Flow Diagram for CDMA Call Release (Far-End Initiated)

MS BTS BSC MSC

Release

ACK
FS (Frame Selector) Remove

ACK

Release

ACK

Clear
Disconnect

ACK

Disconnect

Figure 8-14 Flow Diagram for CDMA Call Release (Mobile Initiated)

assignment is identified by the REG_ZONE field of the system parameters message.

Zone-based registration causes a mobile to register whenever it moves into a new zone
that is not on its internally stored list of visited registration zones. A zone is added to
the list whenever a registration (including implicit registration) occurs and is deleted
upon expiration of a timer. After a system access, timers are enabled for every zone
except one.
6. Parameter-change registration. This occurs when a mobile station modifies any of
the following stored parameters:
Authentication 149

◆ the preferred slot cycle index

◆ the station class mark
◆ the call termination enabled indicator
7. Ordered registration. The mobile registers when the base station requests that it do so.
8. Implicit registration. When a mobile successfully sends an Origination message or
Page Response message, the base station can infer the mobile station’s location. This is
considered an implicit registration.
9. Traffic channel registration. Whenever the base station has registration information
for a mobile that has been assigned to a traffic channel, the base station can notify the
mobile that it is registered.

8.4 Authentication
To authenticate a mobile station, secret data known as the A-key is used. The A-key is known
only to the Authentication Center (AC) of the mobile’s home system and to the mobile station. It
is the most secure piece of secret data. The A-key and a special random number (RANDSSD)
can be used by the AC and mobile to generate Shared Secret Data (SSD). The AC may send SSD
to the serving system, but it never sends it over the air link.
RAND is a 32-bit random number issued periodically by the base station in the system
overhead data in two 16-bit segments: RAND_A and RAND_B. The mobile stores and uses the
most recent version of RAND in the authentication process. The last RAND received by the
mobile station is confirmed from the mobile with an 8-bit number RANDC, a part of RAND,
since the current system RAND and the one used by the mobile station could differ when the
base station receives mobile station results.
The Electronic Serial Number (ESN) is a 32-bit binary number that uniquely identifies the
mobile to any system. It is factory set and not readily alterable in the field. Modification of the
ESN requires a special facility not normally available to subscribers.
The Mobile Identification Number (MIN) is derived from mobile station’s 10-digit direc-
tory telephone number. The first 3 digits map into the 10 most significant bits, the second 3 dig-
its map into the next 10 bits, while the last 4 digits map into the remaining 14 bits.
The SSD is a 128-bit pattern stored in the semipermanent memory of the mobile and is
known by the base station. SSD is a concatenation of two 64-bit subsets: SSD_A and SSD_B.
SSD_A is used to support the authentication procedure, and SSD_B is used to support voice pri-
vacy and message confidentiality.
SSD is maintained during power off. It is generated using a 56-bit random number
(RANDSSD created by the home AC), the mobile’s A-key, and the ESN. The A-key is a 64-bit
secret pattern assigned and stored in mobile station’s permanent security and identification
memory. The need to pass the A-key itself from system to system as the subscriber roams is
eliminated. SSD updates are carried out only in the mobile station and its associated home sys-
tem’s HLR/AC, not in the serving system. The AC manages the encrypting keys associated with
an individual subscriber when such functions are provided within the network.
150 Ch. 8 • IS-95 CDMA Call Processing

All mobiles are assigned an ESN when they are manufactured. They are also assigned a
15-digit International Mobile Subscriber Identity (IMSI). When the mobile is turned on, it must
register with the system. When it registers, it sends its IMSI and other data to the network. The
VLR in the visited system then queries the home system’s HLR for the security data and service
profile information. The VLR then assigns a Temporary Mobile Subscriber Identity (TMSI) to
the mobile station. The mobile station uses the TMSI for further accesses to that system. The
TMSI provides anonymity of communications since only the mobile and the network know the
identity of the mobile with a given TMSI. When a mobile roams into the new system, some air
interfaces use TMSI to query the old VLR and then assign a new TMSI; other air interfaces
request that the mobile station send its IMSI, and then they assign a new TMSI.
The network transmits a random number RAND that is received by all mobile stations.
When a mobile station accesses that system, it calculates AUTHR, an encrypted version of
RAND using SSD_A. It then transmits to the network the desired message with its authentication.
For additional details on authentication refer to chapter 11.
The network performs the same calculation and confirms the identity of the mobile. All
communications between the mobile and network are encrypted to prevent someone’s decoding
of the data and using the data to clone other mobile stations. Furthermore, each time a mobile
places or receives a call, a call history count is incremented (CHCNT). The counter is also used
for clone detection since clones will not have a call history identical to the legitimate mobile.
Procedures have been designed to allow a system to challenge an individual mobile with a
unique challenge and to update the SSD.

MS BSS MSC New VLR HLR Old VLR

Global Challenge

Service Request

Validate
RAND

AUTH Request

AUTH Request

Calculate
AUTHR
AUTH Response

Call
Proceeds
AUTH Response (+)
Regist. Accept

Figure 8-15 Call Flows for a Global Challenge

Summary 151

MS BSS MSC New VLR HLR Old VLR

Decision to
Challenge
Unique Challenge
Unique Challenge

Challenge Response

Challenge Response

AUTH Request

Calculate
AUTHC
AUTH Response

Compare

Call Proceeds or
Terminates

Figure 8-16 Call Flows for Unique Challenge

All mobile stations accessing the network must respond to the global challenge as part of
their access. The global challenge response is an integral part of the network access (call origi-
nation, page response, registration, and so on). The call flow for global challenge is given in
Fig. 8-15.
The unique challenge can be sent to a mobile station at any time. It is typically initiated by
the MSC in response to the some event (registration failure and after a successful handoff are the
most typical cases). This is used to challenge the mobile as to its identity. Fig. 8-16 shows call
flows for the unique challenge.

8.5 Summary
This chapter discussed the CDMA call processing states and provided details of different mes-
sages that are used to exchange information in these states. We presented the registration and
authentication procedures of CDMA and concluded the chapter by discussing mobile substates
and flow diagrams for CDMA call origination, call termination, and call release.

8.6 References
1. Garg, V. K., Smolik, K. F., and Wilkes, J. E., Applications of CDMA in Wireless Com-
munications, Prentice Hall PTR, Upper Saddle River, NJ, 1997.
2. Garg, V. K., and Wilkes, J. E., Wireless and Personal Communications Systems, Pren-
tice Hall PTR, Upper Saddle River, NJ, 1996.
3. TIA IS-95 B, “Mobile Station and Radio Interface Specifications.”
This page intentionally left blank
 C H A P T E R 9

Signaling Applications in
IS-95 CDMA

9.1 Introduction
In chapter 5, we discussed the CDMA system architecture used for cellular and PCS in North
America as embodied in the design standardized by TIA and ATIS in IS-95A for the cellular sys-
tem and J-STD-008 for PCS. Chapter 8 covered IS-95 CDMA call processing states that a
mobile station goes through in getting to a traffic channel. These states include the system ini-
tialization state, system idle state, system access state, and traffic channel state. It also covered
CDMA registration and authentication procedures and included the messages that are used to
exchange information in different call processing states. We then provided call flows for CDMA
call origination, call termination, call release, and authentication.
In this chapter, we first describe the layering concept that has been used to develop the
protocols for IS-95 CDMA and then focus on three functional entities—Mobile Station (MS),
Base Station System (BS), and Mobile Switching Center (MSC)—discussing the standardized
interfaces between these entities. The focus is mainly on the A-interface and TIA IS-634-defined
MSC-BS messages, message sequencing, and mandatory timers at the BS and the MSC. The
chapter concludes with call flow diagrams for typical supplementary services, handoff scenarios,
and Over-the-Air Service Provisioning (OTASP).

9.2 Layered Structure

TIA IS-95 CDMA has a layered structure that is designed to provide voice, packet data (up to 64
kbps), simple circuit data (example async fax), and simultaneous voice and packet data services
(see Fig. 9-1).
At the basic level, IS-95 provides protocols and services that correspond to the bottom two
layers of the International Organization for Standardization (ISO)/open system interconnection
(OSI) reference model, i.e., physical and link layer. The physical layer performs coding, inter-
leaving, modulation, and spreading functions for the physical channels.

153
154 Ch. 9 • Signaling Applications in IS-95 CDMA

Upper Layers Upper Layers Layer 3

OSI Layers 3–7 (Primary Traffic) (Secondary Traffic) (Call Processing and Control)

Layer 2 Layer 2 Layer 2

(Primary Traffic) (Secondary Traffic) (Signaling) Layer 2
OSI Layer 2 (Secondary Traffic)
Multiplex Sublayer

OSI Layer 1 Physical Layer (OSI Layer 1)

OSI = open system interconnections

Figure 9-1 TIA IS-95 Layered Structure

The link layer provides protocol support and control mechanisms for data transport ser-
vices and maps the data transport needs of the upper layers into the specific capabilities and
characteristics of the physical layer. It maps logical and signaling channels into code channels
specifically supported by the coding and modulation functions of the physical layer.
The link layer (layer 2) is subdivided into the Link Access Control (LAC) and Medium
Access Control (MAC) sublayers (see Fig. 9-2). Applications and upper-layer protocols corre-
sponding to OSI layers 3 through 7 utilize the services provided by LAC. The LAC sublayer per-
forms the functions essential to set up, maintain, and release a logical link connection.
The MAC sublayer provides a control function that manages resources supplied by the
physical layer (e.g., physical code channels for communication of information over the air inter-
face) and coordinates the usage of those resources desired by various LAC service entities. This
coordination function (which operates under direct control of the BS MAC function) resolves
contention issues between LAC service entities within a single mobile station, as well as
between competing mobile stations. The MAC sublayer is also responsible for delivering the
Quality of Service (QoS) level requested by a LAC service (e.g., by reserving air interface
resources or by resolving priorities between competing LAC service entities).
In IS-95 CDMA the MAC uses Radio Link Protocol (RLP), which provides a highly effi-
cient streaming service that makes a best effort to deliver data between peer entities. The RLP
provides both a transparent and nontransparent mode of operation. In the nontransparent mode
RLP uses Automatic Repeat Request (ARQ) protocols to retransmit data segments that were not
delivered properly by the physical layer. In this mode RLP introduces some delay. In the trans-
parent mode, RLP does maintain byte synchronization between the sender and receiver and noti-
fies the receiver of the missing parts of the data stream. Transparent RLP does not introduce any
transmission delay and is useful in implementing voice services over RLP.
Upper-layer entities provide support for multiple concurrent active sessions with any com-
bination of service type, such as

• Voice service including voice telephony, PSTN access, mobile-to-mobile voice ser-
vices, and Internet telephony
Layered Structure 155

• End-user data-bearing services including packet data, circuit data services, and short
message service (SMS)
• Signaling services that control all aspects of operation of the mobile station
Packet data services conform to industry standard connection-oriented and connectionless
packet data including IP-based protocol (e.g., Transmission Control Protocol [TCP] and User
Data Protocol [UDP]) and ISO/OSI Connectionless Interworking Protocol (CLIP).
Circuit data services emulate international-standard-defined connection-oriented services
such as asynchronous dial-up access, fax, etc.
Fig. 9-2 shows IS-95 (2G CDMA system) and cdma2000 (3G CDMA) layered structure
with protocols.

Packet Data Voice Circuit Data

Application Services Application

Upper Signaling
Layers Services TCP UDP
(OSI 3–7) High-Speed
Circuit Network
Layer Services
IP
PPP

LAC LAC Protocol Null LAC

Link
Layer MAC
(OSI 2) Control Best Effort Delivery RLP
State
MAC

Multiplexing QoS Control

Physical
Layer Physical Layer
(OSI 1)

IP—Internet Protocol QoS—Quality of Service

LAC—Link Access Control RLP—Radio Link Protocol
MAC—Medium Access Control TCP—Transmission Control Protocol
OSI—Open System Interconnect UDP—User Datagram Protocol
PPP—Point-to-Point Protocol —Unique to cdma2000

Figure 9-2 TIA IS-95 and cdma2000 Layered Structure with Protocols
156 Ch. 9 • Signaling Applications in IS-95 CDMA

9.3 A-Interface
North American standards, until recently, have not addressed the standardization of the BS-MSC
interface (the A-Interface in the network reference model). However, wireless service providers
are experiencing explosive growth in North America and are consequently finding it necessary to
purchase equipment from multiple equipment manufacturers. Thus, the wireless industry has
pressed for standards specifying the A-Interface. At this time, however, the BTS-BSC interface,
i.e., the Abis interface, is not being addressed. TIA IS-634 defines MSC-BS messages, message
sequencing, and mandatory timers at the base station and the mobile switching center. Call pro-
cessing, radio resource management, mobility management, and transmission facilities manage-
ment are separate functions that are supported by the applications layer (see Fig. 9-3).
The MSC-BS interface (A-Interface) utilizes point-to-point signaling between the BS and
the MSC. The normal routing situation is that there are one or more signaling links available
between the BS and the MSC, and these constitute a link set. They run in a load-sharing mode,
and the change-over and change-back procedures are supported between these signaling links.
Load sharing is performed on the BS with more than one signaling link by means of the Signal-
ing Link Selection (SLS) field.
The underlying transport mechanism for the applications layer is ISDN with the physical
layer specified by American National Standard Institute (ANSI) T1.101, the Message Transfer
Part (MTP) specified by ANSI T1.111, and the Signaling Connection Control Part (SCCP) spec-
ified by ANSI T1.112. The physical interface supports one or more 1.544-Mbps digital transmis-
sion facilities, each providing twenty-four 56-kbps or 64-kbps channels. Each channel can be
used for traffic or for signaling. The MTP and the SCCP support only signaling messages, while

Base Station MSC

Supplementary Services

Supplementary Services
Transmission Facilities

Transmission Facilities
Radio Resource

Radio Resource
Call Processing

Call Processing
Management

Management

Management

Management

Management

Management
Mobility

Mobility

SCCP SCCP

MTP MTP

Physical Facility

MTP: message transfer part

SCCP: signaling connection control part
MSC: mobile switching center

Figure 9-3 TIA IS-634 Functions

A-Interface 157

the physical layer supports both signaling messages and traffic messages. Traffic messages carry
voice transmission. TIA IS-634 allows the transcoder (vocoder) to reside either at the base sta-
tion or “very near” to the mobile switching center. In the first case, an entire DS0* (64-kbps)
connection is required for each call, while the second case does not necessitate an entire DS0
connection.
At the applications layer, the call processing and mobility management functions are con-
nected between the mobile station and the mobile switching center, while the radio resource
management and the transmission facilities management functions are connected between the
base station and the mobile switching center. Accordingly, the Base Station Application Part
(BSAP), which is the applications-layer signaling protocol, is divided into two subapplications
parts. The first is called the BS Management Application Part (BSMAP). BSMAP messages are
sent between the base station and the mobile switching center. The second is the Direct Transfer
Application Part (DTAP) in which messages are sent between the mobile station and the mobile
switching center. The base station acts as a transparent conduit for DTAP messages. The base
station merely maps the messages going to/from the mobile switching center into the appropri-
ate air interface signaling protocol, e.g., TIA IS-95A. This approach simplifies the role of the
base station for call processing and mobility management.
The DTAP is used to transfer call control and mobility management signaling messages to
and from the MS. The BSMAP supports other procedures between the MSC and the BS related
to the MS, or to a cell within the BS, or to the whole BS.
The base station associates the DTAP messages with a particular mobile station and call
using a transaction identification. BSAP messages are transferred over an SCCP connection. The
DTAP and BSMAP layer 3 messages between the base station and the mobile switching center
are contained in the user data field of the SCCP frames. The data field is supported in Connec-
tion Request (CR), Connection Confirm (CC), Released (RLSD), and Data (DT) SCCP frames
for mobile stations having one or more active transactions. The layer 3 user data field is parti-
tioned into three components (see Fig. 9-4):
1. BSAP message header
2. Distribution data unit (this component includes the length indicator and the Data Link
Connection Identifier [DLCI]—applies only to DTAP messages)
3. Layer 3 message
The BSAP message header consists of the message discrimination and the data link con-
nection identifier, which is applicable only for DTAP messages. The D-bit (bit 0) of the message
discrimination octet is set to 1 for a DTAP message and is set to 0 for a BSMAP message. The
distribution data unit consists of the length indicator octet, which gives the number of octets fol-
lowing the length indicator.
DTAP messages are applicable only to mobility management and call processing (includ-
ing supplementary services) functions, while BSAP messages are associated with radio

* A DS0 is a 64-kbps Pulse Code Modulation (PCM) transport facility and is a single 64-kbps time slot on a T1 carrier.
158 Ch. 9 • Signaling Applications in IS-95 CDMA

DTAP Message Header BSMAP Message Header

Message Message
Discriminator Discriminator

DLCI

Length Length
Indicator Indicator

Layer 3 Layer 3
Message Message

DLCI: data link connection identifier

BSMAP: base station management application part
DTAP: direct transfer application part

Figure 9-4 Layer 3 Data Field

resource management and to call processing (to a lesser degree). Each DTAP message contains
the protocol discriminator octet, which identifies the associated procedure, i.e., call control,
mobility management, radio resource management, and facilities management. All DTAP and
BSMAP messages are identified by the message-type octet.
The remaining part of this section provides greater detail for each function supported by
the BSAP.

9.3.1 Supported Architectural Configurations

TIA IS-634 makes a number of assumptions regarding the underlying CDMA architec-
ture. The basic architecture is shown in Fig. 9-5.
The main entities are the mobile switching center, the transcoder (XC), the base station,
the base transceiver system, and the mobile station. The MS is not shown in Fig. 9-5, but the MS
communicates with the BTS over the air interface. TIA IS-634 assumes that the base station is
really the BSC. One or more BTSs are connected to the BSC.
The transcoder supports both voice coding (vocoder) and diversity reception. Diversity
reception allows the transcoder to pick the best frame when multiple connections are established
during a soft handoff. Diversity reception distinguishes CDMA technology from other digital
technologies. To be more specific, the XC is responsible for the following:
A-Interface 159

MSC

Transcoder

BSC

BTS BTS

Figure 9-5 Basic CDMA Architecture

• Distribute speech/data on the forward traffic channel to all BTSs associated with a call.
During a soft handoff, multiple BTSs are simultaneously assigned to the call. The XC
selects the best speech/data frame from all the BTSs associated with the call on the
reverse traffic channel. This implies that signal quality characteristics of the speech/
data frame are provided to the transcoder.·
• Decode QCELP* format and change to PCM format for voice frames sent on the
reverse traffic channel. If the call is a data call, this task is bypassed.
• Decode PCM format and change to QCELP format for voice frames sent on the for-
ward traffic channel. If the call is a data call, this task is bypassed.
• Rate adapt voice frames to fully utilize the transmission bandwidth of the assigned ter-
restrial circuits. This task is bypassed for data calls.
• Rate adapt compressed voice PCM format into a circuit-switched subrate channel on a
DS0 facility. One common compression approach is Adaptive Differential Pulse Code
Modulation (ADPCM). Compression utilizes the fact that voice activity is less than
100% of the total duration. Typically, the actual voice activity is approximately 50%.
• Provide a control capability of inserting blank-and-burst or dim-and-burst signaling
into the voice transmission on the forward traffic channel.
The transcoder is considered as a logical part of the BS, although the transcoder can be
physically located at the BS, or at the MSC, or somewhere between the BS and the MSC. The

* Qualcomm Code-Excited Linear Prediction (QCELP) is the CDMA speech processing algorithm that is specified in
TIA IS-96A. The algorithm is based upon code-excited linear prediction. See chapter 3 for more information.
160 Ch. 9 • Signaling Applications in IS-95 CDMA

terrestrial facility connects the transcoder to the MSC. The terrestrial facility may be full-rate
(56 kbps or 64 kbps), subrate, PSTN, bypass, or PSTN/bypass. If only the BS is associated with
a call, the terrestrial facility is connected to the PSTN. If a call is configured for a soft handoff
between two base stations, another terrestrial facility is required to connect the transcoder at the
target BS with the transcoder at the source BS. This terrestrial facility may be a full-rate, subrate,
bypass, or bypass/PSTN facility. The connection between the transcoder and the BS is not
addressed in TIA IS-634.
The transcoder and the BS may be physically co-located or may be externally connected
by a full-rate or subrate facility if the transcoder is located near or at the MSC. The BSC may
support multiple BTSs. Thus, if a call is in a soft handoff using only BTSs connected to a given
BSC, no messaging between the MSC and the BS is necessary.
BTSs are uniquely identified by the Cell Global Identification (CGI). CGI is composed of
four components:·
• Mobile Country Code (MCC)
• Mobile Network Code (MNC)
• Location Area Code (LAC)
• Cell Identity (CI)
TIA IS-634 supports addressing modes so that a BTS can be identified by the CGI, by the
CI, by a combination of the LAC, MCC, and MNC, or by the associated BS.

9.3.2 Call Processing and Supplementary Services

TIA IS-634 supports call setup (mobile origination and mobile termination) as well as
supplementary services (e.g., call waiting) and call release. However, the support of handoffs
during a call is considered to be part of the radio resource management function.
Most of the messages that are associated with call processing and supplementary services
are DTAP messages. For these messages, the role of the base station is minimized since the base
station passes the messages to the mobile station.
The initial BS-MSC message in the call setup procedure includes the mobile identity. The
mobile identity can be the Mobile Identification Number (MIN), the mobile station Electronic
Serial Number (ESN), or the International Mobile Subscriber Identifier (IMSI). The identity
type is selected by either the mobile station or the wireless network. For a mobile origination,
the initial BS-MSC message is the Connection Management (CM) service request; and for a
mobile termination, the initial BS-MSC message is the paging request. However, the initial mes-
sage from the base station to the mobile switching center at call setup is an encapsulated DTAP
message within a BSMAP message. The mobile switching center sends an assignment request
message, which contains the terrestrial channel. Also, the mobile switching center may select the
radio channel or provide channel parameters and permit the base station to choose the radio
channel at the appropriate BTS.
Tables 9-1 and 9-2 list the messages defined by TIA IS-634 for call processing and supple-
mentary services, respectively.
A-Interface 161

Table 9-1 Call Processing Messages

Message Name Direction Message Type
CM Service Request BS → MSC DTAP
Paging Request MSC → BS BSMAP
Paging Response BS → MSC DTAP
Setup BS ↔ MSC DTAP
Emergency Setup BS → MSC DTAP
Alerting BS ↔ MSC DTAP
Call Confirmed BS → MSC DTAP
Call Proceeding MSC → BS DTAP
Connect MSC → BS DTAP
Connect Acknowledge BS ↔ MSC DTAP
Progress MSC → BS DTAP
Release MSC ↔ BS DTAP
Release Complete MSC ↔ BS DTAP
Assignment Request MSC ↔ BS BSMAP
Assignment Complete BS → MSC BSMAP
Assignment Failure BS → MSC BSMAP
Privacy Mode Command MSC → BS BSMAP
Privacy Mode Complete BS → MSC BSMAP
Clear Request BS → MSC BSMAP
Clear Command MSC → BS BSMAP
Clear Complete BS → MSC BSMAP

Table 9-2 Supplementary Service Messages

Message Name Direction Message Type
Send Burst DTMF a
BS ↔ MSC DTAP
Send Burst Acknowledge BS ↔ MSC DTAP
Start DTMF BS → MSC DTAP
Start DTMF Acknowledge MSC → BS DTAP
Stop DTMF BS → MSC DTAP
Stop DTMP Acknowledge MSC → BS DTAP
Flash with Information BS ↔ MSC DTAP
a
Dual-tone multifrequency
162 Ch. 9 • Signaling Applications in IS-95 CDMA

9.3.3 Radio Resource Management

After call setup, the base station is responsible for maintaining a reliable radio link
between the mobile station and the base station. This responsibility requires that the base station
perform the following tasks:
• Radio channel supervision
• Radio channel management
• Initiation and execution of handoffs
The objective of each of these tasks is common for all radio technologies, although the actual
implementation is dependent on the associated technology.
The support of soft handoffs is one capability that distinguishes CDMA from other multi-
ple access technologies. Thus TIA IS-634 supports the procedures associated with soft handoffs.
These procedures are:
• IS-95 add target procedure
• IS-95 drop target procedure
• IS-95 drop source procedure
The source BS is the BSC that controls the transcoder. If either a hard or soft handoff is to be
configured with a target BTS that is connected to another BSC (target BS), then a handoff
required message is sent to the mobile switching center. The mobile switching center then sends
a handoff request to the target BS. At the same time, only one target BS can be addressed.
Table 9-3 summarizes messages associated with radio resource management.

9.3.4 Mobility Management

Mobility management is implemented using DTAP messages. The purpose of the mobility
management function is to support registration and deregistration of a mobile. In addition, this
function encompasses authentication and voice privacy. Authentication includes the authentica-
tion challenge and the SSD update. Supporting this function has little differential impact upon
the BS-MSC architecture.
Messages associated with mobility management are listed in Table 9-4.

9.3.5 Transmission Facilities Management

The transmission facilities management function is responsible for the management of ter-
restrial circuits. Terrestrial circuits are transmission facilities that carry traffic (voice or data) and
signaling information between the MSC and the BS. Furthermore, different facilities may carry
traffic information from facilities carrying signaling information. Each facility may be blocked/
unblocked and allocated/deallocated by the transmission facilities management function. For
digital technologies, e.g., CDMA, this function can disable the transcoders at both the originat-
ing end and the terminating end for mobile-to-mobile calls. This action eliminates the need for
vocoder tandeming, which degrades the voice quality of a call. However, TIA IS-634 does not
explicitly address this capability for calls spanning multiple BSs.
A-Interface 163

Table 9-3 Handoff Messages

Message Name Direction Message Type

Strength Measurement Request BS ↔ MSC BSMAP
Strength Measurement Response BS ↔ MSC BSMAP
Strength Measurement Report BS ↔ MSC BSMAP
Handoff Required BS → MSC BSMAP
Handoff Request MSC → BS BSMAP
Handoff Request Acknowledge BS → MSC BSMAP
Handoff Failure BS → MSC BSMAP
Handoff Command MSC → BS BSMAP
Handoff Required Reject MSC → BS BSMAP
Handoff Commenced BS → MSC BSMAP
Handoff Complete BS → MSC BSMAP
Handoff Performed BS → MSC BSMAP
Soft Handoff Drop Target BS → MSC → BS BSMAP
Soft Handoff Drop Source BS → MSC → BS BSMAP

Table 9-4 Mobility Management Messages

Message Name Direction Message Type

Authentication Request MSC → BS DTAP
Authentication Reject MSC → BS DTAP
SSD Update Request MSC → BS DTAP
Base Station Challenge BS → MSC DTAP
Base Station Challenge Response MSC → BS DTAP
SSD Update Response BS → MSC DTAP
Location Updating Request BS → MSC DTAP
Location Updating Accept MSC → BS DTAP
Location Updating Reject MSC → BS DTAP
Parameter Update Request MSC → BS DTAP
Parameter Update Confirm BS → MSC DTAP
164 Ch. 9 • Signaling Applications in IS-95 CDMA

Table 9-5 summarizes the message types that are associated with transmission facilities
management.
The MTP provides a mechanism that makes the transfer of signaling messages reliable.
The SCCP is used to provide a referencing mechanism to identify a particular transaction relat-
ing to, for instance, a particular call. The SCCP can also enhance the message routing for opera-
tion and maintenance information.
At the BS, only messages with a correct Destination Point Code (DPC) are accepted.
Other messages are discarded. At an MSC (with the capability of acting as a signal transfer point
[STP]) each message received from a BS signaling link is passed through a screening function
which checks that the DPC of the message is the same as the signaling point (SP) code of the
exchange. If it is the same, the message is sent to the normal MTP message-handling functions;
otherwise, the message is discarded. The SP code for signaling may be included in the national
SP code scheme or in a separate signaling network.
The BS exchanges signaling messages only with its MSC, where a protocol conversion
may be needed in some cases. Therefore, no SCCP translation function is required in the MS
between the national and the local SCCP and the MTP within the MSC area.
Several functions of the SCCP (such as error detection, receipt confirmation, and flow
control) are not used on the MSC-BS interface. The segmenting/reassembling function is used if
the total message length exceeds the MTP’s maximum allowed message length.

9.3.6 Use of the SCCP

The MTP and the SCCP are used to support signaling messages between the MSC and the
BS. Fig. 9-6 shows the A-Interface signaling protocol stack.

Table 9-5 Transmission Management Messages

Message Name Direction Message Type
Overload MSC ↔ BS BSMAP
Block BS → MSC BSMAP
Block Acknowledge MSC → BS BSMAP
Unblock BS → MSC BSMAP
Unblock Acknowledge MSC → BS BSMAP
Reset BS ↔ MSC BSMAP
Reset Acknowledge BS ↔ MSC BSMAP
Reset Circuit BS ↔ MSC BSMAP
Reset Circuit Acknowledge BS ↔ MSC BSMAP
Transcoder Control Request MSC ↔ BS BSMAP
Transcoder Control Acknowledge MSC ↔ BS BSMAP
A-Interface 165

BS MSC
BSAP BSAP
DTAP BSMAP DTAP BSMAP
IS-634 IS-634
SCCP SCCP
MTP MTP
Physical Physical

A-Interface
Figure 9-6 A-Interface Signaling Protocol Stack

The initial messages exchanged in call setup are used to establish an SCCP connection for
subsequent signaling communications relating to the call. A new connection is established when
individual information related to an MS transaction has to be exchanged between a BS and an
MSC and no such transaction exists between the MSC and that BS. We need to distinguish the
reasons for connection establishment:
1. A new transaction (e.g., location update, incoming or outgoing call) is initiated on the
radio path.
2. Following an access request made by the MS on the random access channel, the con-
nection establishment is then initiated by the MSC.
The BS initiates a connection establishment when it receives the first layer 3 mes-
sage from the MS. The message contains the mobile identity parameter (MIN, ESN, or
IMSI). The BS then constructs the first MSC-BS Interface BSMAP message (Complete
Layer 3 Information) which includes one of the appropriate DTAP messages (Location
Update Request, CM Service Request, or Paging Response) depending on whether the
MS is accessing the network for the purpose of registration, call origination, or call ter-
mination. The Complete Layer 3 Information message is sent to the MSC in the user
data field of the SCCP Connection Request message. The Complete Layer 3 Informa-
tion message includes cell identity and the layer 3 message that was received from the
mobile.
At the reception of the SCCP Connection Request message, the MSC may check,
based on the received identity, whether another association already exists for the same
MS. If that is the case, the connection establishment is refused. Otherwise, an SCCP
Connection Confirm message is sent back to the BS. This message may optionally con-
tain a BSMAP or DTAP message in the user data field (Fig. 9-7).
3. The MSC decides to perform an inter-BS handoff. The connection establishment is
then initiated by the MSC.
The connection establishment is undertaken by the MSC as soon as the MSC
decides to perform an inter-BS handoff. An SCCP Connection Request message is sent
166 Ch. 9 • Signaling Applications in IS-95 CDMA

BS MSC

SCCP Connection Request (Complete Layer 3 Information)

Timer T3230

SCCP Connection Confirm

Figure 9-7 SCCP Connection Establishment

to the BS. The user data field of this message may contain the BSMAP Handoff Request
Message. However, it is preferable to transfer the layer 3 messages in the user data field
of the SCCP Connection Request in order to complete the establishment of the relation-
ship between the radio channel requested and SCCP connection as soon as possible.
When it receives the SCCP Connection Request message, the BS performs the
necessary checking and, in the successful case, reserves a radio channel for the
requested handoff. An SCCP Connection Confirm message is also returned to the MSC
and may contain the BSMAP Handoff Request Acknowledge message in the user field
(Fig. 9-8).
This procedure is initiated by the MSC in normal conditions for all calls. A con-
nection is released when a given signaling connection is no longer required. This may
happen in normal cases:
◆ At the end of a transaction (call, location update)
◆ After completion of a successful external handoff—the connection with the old
BS is released.
The MSC/BS sends an SCCP released (RLSD) message. The user data field of
the message is optional and may contain a transparent layer 3 message (e.g., DTAP) or
be empty.

BS MSC

SCCP Connection Request: Handoff Request

SCCP Connection Confirm

Figure 9-8 SCCP Connection Establishment during Handoff

Roaming 167

When receiving this message, the BS/MSC releases all the radio resources allo-
cated to the relevant MS, if there are still any left, and sends an SCCP release complete
(RLC) back to the MSC/BS.
The normal release of SCCP connections is initiated by MSC. Under abnormal
conditions, the SCCP connection may be released by the BS in order to clear resources.
Whenever an SCCP connection is abnormally released, all resources associated with
that connection are cleared. Abnormal release could result from, for example, resource
failure, protocol error, or unexpected receipt of an SCCP released (RLSD) or SCCP
RLC command.
The SCCP local reference number (source/destination) is a 3-byte element inter-
nally chosen by the MSC or BS to uniquely identify a signaling connection. In the
direction MSC to BS, the source local reference is selected by the MSC and the desti-
nation local reference is chosen by the BS. In the direction BS to MSC, the source local
reference is chosen by the BS and the destination local reference is chosen by the MSC.
Note that it is the responsibility of the BS and MSC to insure that no two calls have
identical SCCP local reference numbers.

9.4 Roaming
Roaming enables mobile stations to receive services outside of their home areas. When an MS is
roaming, registration, call origination, and call delivery will take extra steps. Whenever the VLR
tries to retrieve MS data, and data is not available, then the VLR sends a message to the appro-
priate HLR to retrieve the necessary data. The data consists of IMSI-to-MIN conversion, service
profiles, SSD for authentication, and other data needed to process calls. The most logical time to
retrieve this data is when the MS registers with the system.
Once the data on a roaming MS is stored in the VLR, then call processing for any originat-
ing services (basic or supplementary) is identical to that of home MS. However, there may be
times when the MS originates a call before registration has been accomplished or when the VLR
data is not available. At those times, an extra step will be added for the VLR to retrieve the data
from the HLR. Thus any originating service has two optional steps where the VLR sends a mes-
sage (using IS-41 signaling over SS7) to the HLR requesting data on the roaming MS, and the
HLR returns a message with the proper call information.
Call delivery is not possible to an unregistered MS since the network does not know where
the MS is located. Once the MS is registered with a system, then call delivery to the roaming MS
is possible. This section will discuss call delivery to roaming MS in detail.
There are two types of call delivery to roaming MS—when the MS has a geographic-
based directory number (indistinguishable from a wireline number), and when the MS has a
nongeographic number.
We will describe the call flows for both operations.
When the MS has a geographic number, then the MSC is assigned a block of numbers that
are within the local numbering plan for the area of the world where the MSC is located. Call
168 Ch. 9 • Signaling Applications in IS-95 CDMA

routing to the MS is then done according to the procedures for that of a wireline telephone.* If an
MS associated with an MSC is not in its home area, the MSC will query the HLR for the loca-
tion of the MS. The MSC then invokes call forwarding to the MSC at the MS’s location, and the
connection is made to the second MSC where call-terminating services are delivered (refer to
the procedures in section 8.2.4). This procedure is inefficient because it results in two sets of net-
work connections—originating switch to home MSC, and home MSC to visited MSC.
Call delivery to a roaming MS is a cooperative effort among the home and visited MSCs,
the VLR and HLR, and the radio system. The detailed call flow steps, for call delivery to a roam-
ing MS with a geographic directory number (see Fig. 9-9 for the call flow diagram) are:
1. A user in the worldwide phone network (wired or wireless) dials the directory number
of the MS.
2. The originating switch sends an SS7 initial address message (IAM) to the home MSC.
3. The home MSC queries the HLR for the location of the MS.
4. The HLR returns the location of the visited system.
5. The MSC invokes call forwarding to the MSC in the visited system, and the forwarding
(home) MSC switch sends an SS7 IAM to the visited MSC.
6. Call processing continues to the terminating call flow (see Fig 8.12).

Mobile Station Base Visited Home Originating

VLR HLR
Station MSC MSC Switch

User
1.
Dials Call

2. SS7 IAM

3. HLR Query

HLR
4.
Response

5. SS7 IAM

6. Normal Call Processing Continues

Figure 9-9 Call Termination to a Roaming Mobile Station with a Geographic Number

*For example, in Chicago, 312-944-XXXX is used by the local cellular provider for cellular phones in the
downtown area. The wireline network routes calls to those numbers in a normal fashion, and calls terminate
on the cellular switch.
Roaming 169

When the MS has a nongeographic number, then calls can be directed from an originating
switch directly to the visited switch. Call delivery to a nongeographic number requires the origi-
nating switch to recognize the number as a nongeographic number and do special call process-
ing for routing. This special processing is known as Intelligent Network (IN) processing. If the
originating switch does not support IN, then it will route the call to a switch that supports IN.
With IN support, the originating switch will recognize the nongeographic number and send an
SS7 message to the HLR with a request for the location of the MS. The HLR will return a tem-
porary directory number (on the visited MSC) that can be used to route to the MS in the visited
system. Calls then proceed according to normal termination. Call delivery to a roaming MS with
a nongeographic number is, therefore, a cooperative effort between the visited MSC, the VLR
and HLR, and the radio system. The detailed call flow steps for call delivery to a roaming MS
with a nongeographic directory number (see Fig. 9-10 for call flow) are:
1. A user in the worldwide phone network (wired or wireless) dials the directory number
of the MS.
2. The originating switch recognizes the number as a nongeographic number and sends an
SS7 query message to the HLR at the home MSC.
3. The HLR returns the location of the visited system with a directory number to use for
further call processing.
4. The originating switch sends an SS7 IAM to the visited MSC.
5. Call processing continues to the terminating call flow (see Fig 8.12).

9.4.1 Call Waiting

Call waiting notifies a wireless subscriber of an incoming call while the user’s mobile sta-
tion is in the busy state. The user can either answer or ignore the incoming call. Once the call is

Mobile Station Base Visited Home Originating

VLR HLR
Station MSC MSC Switch

User
1.
Dials Call

2. SS7 HLR
Query
3. SS7 HLR
Response
4. SS7 IAM

5. Normal Call Processing Continues

Figure 9-10 Call Termination to a Roaming Mobile Station with a Nongeographic Number
170 Ch. 9 • Signaling Applications in IS-95 CDMA

answered, the user can switch between the calls until one or more parties hang up. When either
party hangs up, then the call reverts to a normal (non-call-waiting) call. If the MS user hangs up,
then both calls are cleared according to normal call-clearing functions.
The detailed call flow steps for the delivery of call waiting to a mobile station (see Fig. 9-
11) are:
1. User dials a call.
2. The originating switch sends an SS7 IAM to the MSC.
3. The MSC queries the VLR.
4. The VLR returns with a location of the MS that is within the serving system. If the MS
is not inside the serving system, then the call is forwarded to the serving MSC.

Mobile Station Base Originating

VLR HLR
Station MSC Switch

User
1.
Dials Call
2. SS7 IAM

3. VLR Query

4. VLR Response

MSC Applies
5. Call Waiting Tone

6. Flash

7. ISDN Hold

MSC Puts
8. Call 1 on Hold
Connects Call 2
9. ISDN Hold Ack

10. Flash

11. ISDN Hold

MSC Puts
12. Call 2 on Hold
Connects Call 1
13. ISDN Hold Ack

14. Drop

15. ISDN Drop

MSC Drops
16. Current Call and
Connects Other Call
17. ISDN Drop Ack

Figure 9-11 Call Waiting

Roaming 171

5. The MSC determines that the MS is busy, subscribes to call waiting, and thus applies
the call waiting tone.
6. The user presses the FLASH button (may be SEND on some MSs) to answer the call
waiting indication, and the MS sends a Flash message to the base station.
7. The base station sends an ISDN Hold message to the MSC.
8. The MSC puts the first call on hold and connects the second call.
9. The MSC sends a Hold Acknowledge message to the base station.
10. The user presses the FLASH button (may be SEND on some MSs) to talk to caller 1,
and the MS sends a Flash message to the base station.
11. The base station sends an ISDN Hold message to the MSC.
12. The MSC puts the second call on hold and connects the first call.
13. The MSC sends a Hold Acknowledge message to the base station.
14. The user wants to drop the current call (either 1 or 2) and pushes the DROP (or END)
key, and the MS sends a Drop message to the base station.
15. The base station sends an ISDN Drop message to the MSC.
16. The MSC drops the current call and connects the other call (the one currently on hold).
17. The MSC sends an ISDN Drop Acknowledge message to the base station.

9.4.2 Handoffs
A wireless telephone (mobile station) moves around a geographic area. When the station is
idle, it periodically reregisters with the system. When a call is active, then the combined mobile
station, the base station, and the MSC manage the communications between the base station and
mobile station so that good radio link performance is maintained. The process whereby a mobile
station moves to a new traffic channel is called handoff. The original analog cellular system pro-
cessed handoffs by commanding the mobile station to tune to a new frequency. For analog cellu-
lar, the handoff process caused a small break in time on the voice path and a noticeable “click”
was heard by both parties in the telephone call. For data modems, the click often caused data
errors or loss of data synchronization.
For CDMA systems, the characteristics of spread spectrum communications permit the
system to simultaneously receive mobile transmissions on two or more base stations. In addi-
tion, the mobile station can simultaneously receive the transmissions of two or more base sta-
tions. With these capabilities, it is possible to process a handoff from one base station to another,
or from one antenna face to another on the same base station, without any perceptible distur-
bance in voice or data communications.
During handoff, the signaling and voice information from multiple base stations must be
combined (or bridged) in a common point with decisions made on the quality of the data. Simi-
larly, voice and signaling information must be sent to multiple base stations and the mobile sta-
tion must combine the results. The common point could be anywhere in the network, but it is
typically at the mobile switching center. The call flows described here for handoff assume the
MSC contains the bridging circuitry.
172 Ch. 9 • Signaling Applications in IS-95 CDMA

In CDMA, both the base station and the mobile station monitor the performance of the
radio link and can request handoffs. Handoffs requested by a mobile station are called mobile-
assisted handoffs; those requested by the base station are called base-station assisted handoffs.
Either side can initiate the handoff process whenever the following triggers occur:
• Base station traffic load not balanced. The network can monitor loads at all base
stations and trigger handoffs to balance loads between them to achieve higher traffic
efficiency.
• Distance limits exceeded. Since all base stations and mobile stations are synchronized,
both sides can determine base-to-mobile range. When the distance limit is exceeded,
either side can request a handoff.
• Pilot signal strength below threshold. When the received signal strength of the pilot
signal falls below a threshold, either side can initiate a handoff.
• Power level exceeded. When the base station commands a mobile station to increase
its power and the maximum power level of the mobile station is exceeded, then either
side can request a handoff.
The mobile station determines the parameters for the handoff request from the System
Parameters message in the CDMA system. The message is transmitted on the system’s paging
channel.
As we have described, the handoff process is a cooperative effort between the old and new
base stations, the mobile station, and the MSC. The following call flows are based on a frame
relay A-Interface [9] between the base station and the MSC and are included as representative
call flows. Actual call flows may be either standard or proprietary to an equipment vendor.
The detailed call flow steps for a CDMA soft handoff (beginning) (see Fig. 9-12) are:
1. The mobile station determines that another base station has sufficient pilot signal to be
a target for handoff.
2. The mobile station sends a Pilot Strength Measurement message to the serving base
station.
3. The serving base station sends an inter-BS Handoff Request message to the MSC.
4. The MSC accepts the handoff request and sends an inter-BS Handoff Request message
to the target base station.
5. The target base station establishes communication with the mobile station by sending it
a Null Traffic message.
6. The target base station sends a Join Request message to the MSC.
7. The MSC conferences the connections from the two base stations so the handoff can be
processed without a break in the connection (i.e., soft handoff) and sends a Join
Acknowledge message to the target base station.
8. The target base station sends an inter-BS Handoff Acknowledgment message to the
MSC.
9. The MSC sends an inter-BS Handoff Acknowledgment message to the serving base
station.
Roaming 173

Mobile Station Primary (Serving) Target

Base Station MSC Base Station

Decision to
1.
Hand Off

2. Pilot Strength
Measurement
3. Inter_BS HO Request

4. Inter_BS HO Request

5. Null Traffic

6. Join_REQ

7. Join_ACK

8. Inter_BS_HO_ACK

9. Inter_BS_HO_ACK

10. Handoff Direction

11. Handoff Comp

12. HO_INFO

13. HO_INFO_ACK

14. Pilot Measurement Request Order

15. Pilot Strength Measurement

Figure 9-12 CDMA Soft Handoff—Beginning

10. The serving base station sends a Handoff Direction message to the mobile station.
11. The mobile station sends a Handoff Complete message to the serving base station.
12. The serving base station sends a Handoff Information message to the MSC.
13. The MSC confirms the message with a Handoff Information Acknowledgment message.
14. The target base station sends a Pilot Measurement Request Order message to the
mobile station.
15. The mobile station sends a Pilot Strength Measurement message to the target base
station.
The mobile unit is now communicating with two base stations (i.e., it is in soft handoff).
Both base stations must communicate with the MSC, which then uses the highest quality signals
from the two base stations and sends transmitted signals to both base stations.
After the mobile station is in soft handoff, one of the signals may fall below a predeter-
mined threshold (based on information sent in overhead messages on the control channel) and
the mobile stations will request that one base station be removed from the connection. The
detailed call flow steps for a CDMA soft handoff with the serving base station dropping off (see
Fig. 9-13) are:
174 Ch. 9 • Signaling Applications in IS-95 CDMA

Mobile Station Primary (Serving) Target

Base Station MSC Base Station

Decision to
1.
Hand Off

2. Pilot Strength
Measurement

3. Handoff Direction

4. Handoff Comp

5. Inter_Prim_XFER

6. Inter_Prim_XFER_ACK

7. HO_INFO

8. HO_INFO_ACK

9. Pilot Measurement Request Order

10. Pilot Measurement

11. Remove_REQ

12. Remove_ACK

Figure 9-13 CDMA Soft Handoff—Serving Base Station Dropping Off

1. The mobile station determines that the serving base station has insufficient pilot signal
to continue to be a base station in the soft handoff.
2. The mobile station sends a Pilot Strength message to the serving base station. The mes-
sage requests that the base station drop off from the handoff.
3. The serving base station sends a Handoff Direction message to the mobile station that
indicates which base station is to be dropped from the soft handoff (in this case, the
serving base station).
4. The mobile station sends a Handoff Complete message to the serving base station.
5. The serving base station sends an Interface Primary Transfer message to the target base
station with relevant call record information.
6. The target base station confirms the message with an Interface Primary Transfer
Acknowledge message.
7. The target base station then sends a Handoff Information message to the MSC.
8. The MSC sends a Handoff Information Acknowledge message to the target base
station.
9. The target base station sends a Pilot Measurement Request Order message to the
mobile station.
Roaming 175

10. The mobile station sends a Pilot Strength Measurement message to the target base sta-
tion.
11. The old serving base station sends a Remove Request message to the MSC that
requests that the base station be dropped from the connection.
12. The MSC confirms the message by sending a Remove Acknowledge message to the old
serving base station.
The mobile station is now communicating with the target base station (new serving base
station). If additional soft handoffs are needed, the handoff beginning procedure is repeated.
The procedures to drop a target base station from a soft handoff are similar to those that
drop the serving base station. The detailed call flow steps for a CDMA soft handoff with the tar-
get base station dropping off (see Fig. 9-14) are:
1. The mobile station determines that the target base station has an insufficient pilot signal
to continue to be a base station in the soft handoff.

Mobile Station Primary (Serving) Target

Base Station MSC Base Station

Decision to
1.
Hand Off

2. Pilot Strength
Measurement
3. Handoff Direction

4. Handoff Comp

5. Inter_BS_Remove

6. Inter_BS_Remove

7. HO_INFO

8. HO_INFO_ACK

9. Remove_REQ

10. Remove_ACK

11. Inter_BS_Remove_ACK

12. Inter_BS_Remove_ACK

13. Pilot Measurement

Request Order
14. Pilot Strength
Measurement

Figure 9-14 CDMA Soft Handoff—Target Base Station Dropping Off

176 Ch. 9 • Signaling Applications in IS-95 CDMA

2. The mobile station sends a Pilot Strength message to the serving base station. The mes-
sage requests that the target base station drop off from the handoff.
3. The serving base station sends a Handoff Direction message to the mobile station that
indicates which base station is to be dropped from the soft handoff (in this case, the tar-
get base station).
4. The mobile station sends a Handoff Complete message to the serving base station.
5. The serving base station sends an Inter-BS Remove message to the MSC.
6. The MSC sends an Inter-BS Remove message to the appropriate base station (in this
case, the target base station).
7. The serving base station then sends a Handoff Information message to the MSC.
8. The MSC sends a Handoff Information Acknowledge message to the serving base sta-
tion.
9. The target base station sends a Remove Request message to the MSC.
10. The MSC sends a Remove Acknowledge to the target base station.
11. After the target base station removes its resource from the call, it sends an Inter-BS
Remove Acknowledge message to the MSC.
12. The MSC sends a Remove Acknowledge message to the serving base station.
13. The serving base station sends a Pilot Measurement Request Order message to the
mobile station.
14. The mobile station sends a Pilot Strength Measurement message to the serving base
station.
The mobile station is now communicating only with the serving base station. If additional
soft handoffs are needed, the handoff beginning procedure is repeated.

9.4.3 Over-the-Air Service Provisioning (OTASP) [8]

Successful OTASP processing involves the following procedures and the exchange of
some A-Interface messages (see Fig. 9-15):
• OTASP Call Setup
• OTASP Data Exchange, which includes exchange of the following messages:
◆ ADDS Deliver
◆ ADDS Deliver ACK
Some existing procedures may also be applied as subprocedures:
◆ SSD Update procedure (see chapter 11)
◆ Privacy Mode Request procedure (see chapter 11)
• OTASP Call Clearing
1. The MS sends an Origination message over the access channel to the BS to initiate the
OTASP process (see chapter 8).
Roaming 177

MS BS MSC
Origination Message
1

Normal Call Setup

2

ADD Deliver
3
Data Burst (OTA)
4
Layer 2 ACK/Reject Order
5
with Layer 2 ACK ADD Deliver ACK
6
Rejection Order
7
Rejection Order
8
Data Burst (OTA)
9
Layer 2 ACK
10
ADD Deliver
11
SSD Update
12 Optional

Terminal Authentication
13 Optional

Privacy Mode Request

14 Optional
Repeat Steps 3 to 11
15 Optional

Normal Call Clearing

16

Figure 9-15 CDMA OTSAP Message Flow

2. The MSC and BS use a normal call setup procedure to establish the OTASP call.
3. Upon request from Over-the-Air Function (OTAF), the MSC encapsulates an OTASP
data message within an ADD Deliver message and sends it to the BS.
4. The BS extracts the OTASP data message, places it in the CDMA Data Burst message,
and transmits it over the traffic channel to the MS.
5. The MS may respond with a Layer 2 ACK or a Reject Order containing a Layer 2 ACK
acknowledging the Data Burst message.
6. When the BS receives a Layer 2 ACK, or when a BS receives a Reject Order containing
a Layer 2 ACK acknowledging the Data Burst message from the MS in response to an
ADD Deliver message containing a Tag information element, it sends an ADD Deliver
ACK message to the MSC with the corresponding Tag value.
7. The MS may return a Reject Order message.
178 Ch. 9 • Signaling Applications in IS-95 CDMA

8. The BS will send a Rejection message to the MSC to convey the information contained
in the Reject Order message.
9. The OTASP application in the MS responds by sending an OTASP data message. The
MS places the OTASP data message in the CDMA Data Burst message and transmits it
over the traffic channel to the BS.
10. Upon reception of the CDMA Data Burst message, the BS responds with a Layer 2
ACK.
11. The BS extracts the OTASP data message and places it in the ADDS Deliver message
to the MSC. Steps 12 through 15 are optional.
—————————————
12. After the A-key has been derived from information transferred via the ADDS Deliver
message, an SSD Update procedure over the traffic channel may also be used to
exchange authentication information (RANDSSD, RANDBS, AUTHBS) (see chapter
11).
13. After an SSD Update procedure, terminal authentication needs to be performed to gen-
erate the cipher key that will be used for privacy (see chapter 11).
14. After terminal authentication, Privacy Mode procedures over the traffic channel may
also be applied to specify the use of either Signaling Message Encryption (SME) or
Privacy for the call.
15. Multiple forward and reverse OTASP messages can be sent between the OTASP end
point in the network and the MS. The MSC and the BS will transfer the messages
whenever they are received.
—————————————
16. Once the OTASP service programming has been successful, the call can be cleared
using a regular Call Clear procedure (see chapter 8).

9.5 Summary
First this chapter described the layering concept used to develop the protocols for the IS-95
CDMA and followed with a discussion of the signaling applications for a CDMA wireless tele-
phony system. Since end-to-end call flows are not presented in any of the standards but are dis-
tributed across several standards, we described several basic and supplementary call flows.
An important component of wireless services is the ability to find and place calls to a
roaming mobile station. Most mobile stations have a geographic number, but many will have
nongeographic numbers in the future, so we described call flows for both. Geographic numbers
are phone numbers that are located to a specific point on the worldwide phone system. Nongeo-
graphic numbers do not have a location associated with them, and the network maintains a data-
base of the location of the phone. Additional routing steps are necessary to place a call to a
mobile station with a nongeographic number.
While cellular and PCS systems (and CDMA, in particular) have a rich set of supplemen-
tary features, the most common feature is call waiting. The various standards describe additional
References 179

procedures for all of the basic and supplementary services. We encourage you to consult the
standards [1–7] for additional information.
Finally, since the CDMA system processes handoffs in a different way than analog cellular
or TDMA cellular/PCS systems, we described the soft handoff process for CDMA and present
call flows for soft handoff, beginning and ending. Also discussed was the Over-the-Air Service
Provisioning (OTASP) procedure.

9.6 References
1. Committee T1, “Stage 2 Service Description for Circuit Mode Switched Bearer Services,”
Draft T1.704.
2. Committee T1—Telecommunications, “A Technical Report on Network Capabilities, Archi-
tectures, and Interfaces for Personal Communications,” T1 Technical Report #34, May
1994.
3. EIA/TIA-553, “Cellular System Mobile Station–Land Station Compatibility Specification.”
4. Garg, V. K., and Wilkes, J. E., Wireless and Personal Communications Systems, Prentice
Hall PTR, Upper Saddle River, NJ, 1996.
5. Report of the Joint Experts Meeting on Privacy and Authentication for PCS, Phoenix, Ari-
zona, November 8–12, 1993.
6. TIA Interim Standard, IS-41 C, “Cellular Radio Telecommunications Intersystem Opera-
tions.”
7. TIA IS-95A, “Mobile Station–Base Station Compatibility Standard for Dual Mode Spread
Spectrum Cellular System.”
8. TIA IS-634, “MSC-BS Interface for 800 MHz,” 1995.
9. TR-45 Contribution, “Frame Relay A-Interface.”
This page intentionally left blank
 C H A P T E R 1 0

Soft Handoff and Power

Control in IS-95 CDMA

10.1 Introduction
Soft handoff is different from the traditional hard-handoff process. With hard handoff, a definite
decision is made on whether to hand off or not. The handoff is initiated and executed without the
user attempting to have simultaneous traffic channel communications with the two base stations.
With soft handoff, a conditional decision is made on whether to hand off. Depending on the
changes in pilot signal strength from the two or more base stations involved, a hard decision will
eventually be made to communicate with only one. This normally happens after it is evident that
the signal from one base station is considerably stronger than those from the others. In the interim
period, the user has simultaneous traffic channel communication with all candidate base stations.
It is desirable to implement soft handoff in power-controlled CDMA systems because
implementing hard handoff is potentially difficult in such systems. A system with power control
attempts to dynamically adjust transmitter power while in operation. Power control is closely
related to soft handoff. IS-95 uses both power control and soft handoff as an interference-
reduction mechanism. Power control is the main tool used in IS-95 to combat the near-far prob-
lem. It is theoretically unnecessary to have power control if one can successfully implement a
more intelligent receiver than that used in IS-95, which is the subject of the field of multiuser
detection (MUD), a feature being proposed for the 3G CDMA systems. Power control is neces-
sary in order for a CDMA system to achieve a reasonable level of performance in practice. The
use of power control in the CDMA system necessitates the use of soft handoff when the origi-
nal and new channels occupy the same frequency band. For power control to work properly, the
mobile must attempt to be linked at all times to the base station from which it receives the
strongest signal. If this does not happen, a positive power control feedback loop could inadvert-
ently occur, causing system problems. Soft handoff can guarantee that the mobile is indeed
linked at all times to the base station from which it receives the strongest signal, whereas hard
handoff cannot guarantee this.

181
182 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

The performance of CDMA systems is very sensitive to differences in received signal

powers from various users on the reverse link. Due to the nonorthogonality of the spreading PN
codes used by different users, a strong interfering signal may mask out a weak desired signal,
causing unreliable detection of the latter. This is called the near-far problem.
This chapter first covers handoff strategy used in IS-95 CDMA and then focuses on power
control schemes for the reverse and forward link.

10.2 Types of Handoff

There are four types of handoff:

1. Intersector or softer handoff. The mobile communicates with two sectors of the same
cell (see Fig. 10-1). A RAKE receiver at the base station combines the best versions of
the voice frame from the diversity antennas of the two sectors into a single traffic frame.
2. Intercell or soft handoff. The mobile communicates with two or three sectors of differ-
ent cells (see Fig. 10-2). The base station that has the direct control of call processing

Figure 10-1 Softer Handoff

Cell A Cell B

α α α
α
γ
γ β β
β β
γ γ

α
Cell A Cell B
γ
β
Cell C

Two-Way Soft Handoff Three-Way Soft Handoff

Figure 10-2 Soft Handoff

Types of Handoff 183

during handoff is referred to as the primary base station. The primary base station can
initiate the forward control message. Other base stations that do not have control over
call processing are called the secondary base stations. Soft handoff ends when either
the primary or secondary base station is dropped. If the primary base station is dropped,
the secondary base station becomes the new primary for this call. A three-way soft
handoff may end by first dropping one of the base stations and becoming a two-way
soft handoff.
The base stations involved coordinate handoff by exchanging information via
SS7 links. A soft handoff uses considerably more network resources than the softer
handoff.
3. Soft-softer handoff. The mobile communicates with two sectors of one cell and one
sector of another cell (see Fig. 10-3). Network resources required for this type of hand-
off include the resources for a two-way soft handoff between cell A and B plus the
resources for a softer handoff at cell B.
4. Hard handoff. Hard handoffs are characterized by the break-before-make strategy. The
connection with the old traffic channel is broken before the connection with the new
traffic channel is established. Scenarios for hard handoff include
◆ Handoff between base stations or sectors with different CDMA carriers
◆ Change from one pilot to another pilot without first being in soft handoff with the
new pilot (disjoint active sets)
◆ Handoff from CDMA to analog, and analog to CDMA
◆ Change of frame offset assignment—CDMA traffic frames are 20 ms long. The
start of frames in a particular traffic channel can be at 0 time in reference to a sys-
tem or it can be offset by up to 20 ms (allowed in IS-95). This is known as the
frame offset. CDMA traffic channels are assigned different frame offset to avoid
congestion. The frame offset for a particular traffic channel is communicated to
the mobile. Both forward and reverse links use this offset. A change in offset

γ
γ α

β
β

Cell A Cell B
Figure 10-3 Soft-Softer Handoff
184 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

assignment will disrupt the link. During soft handoff the new base station must
allocate the same frame offset to the mobile as assigned by the primary base sta-
tion. If that particular frame offset is not available, a hard handoff may be
required. Frame offset is a network resource and can be used up.

10.2.1 Soft Handoff (Forward Link)

In this case all traffic channels assigned to the mobile are associated with pilots in the
active set and carry the same traffic information with the exception of power control subchannel.
When the active set contains more than one pilot, the mobile provides diversity by combining its
associated forward traffic channels.

10.2.2 Soft Handoff (Reverse Link)

During intercell handoff, the mobile sends the same information to both base stations.
Each base station receives the signal from the mobile with appropriate propagation delay. Each
base station then transmits the received signal to the vocoder/selector. In other words, two copies
of the same frame are sent to the vocoder/selector. The vocoder/selector selects the better frame
and discards the other.

10.2.3 Softer Handoff (Reverse Link)

During intersector handoff, the mobile sends the same information to both sectors. The
channel card/element at the cell site receives the signals from both sectors. The channel card
combines both inputs, and only one frame is sent to the vocoder/selector. It should be noted that
extra channel cards are not required to support softer handoff as is the case for soft handoffs. The
diversity gain from soft handoffs is more than the diversity gain from softer handoffs because
signals from distinct cells are less correlated than signals from sectors of the same cell.

10.2.4 Benefit of Soft Handoff

A key benefit of soft handoff is the path diversity on the forward and reverse traffic chan-
nels. Diversity gain is obtained because less power is required on the forward and reverse links.
This implies that total system interference is reduced. As a result, the average system capacity
is improved. Also less transmit power from the mobile results in longer battery life and longer
talk time.
In a soft handoff, if a mobile receives an up power control bit from one base station and a
down control bit from the second base station, the mobile decreases its transmit power. The
mobile obeys the power down command since a good communications link must have existed to
warrant the command from the second base station.

10.3 Pilot Sets

The term pilot refers to a pilot channel identified by a pilot sequence offset and a frequency assign-
ment. A pilot is associated with the forward traffic channels in the same forward CDMA link.
Search Windows 185

Each pilot is assigned a different offset of the same short PN code. The mobile search for
pilots is facilitated by the fact that the offsets are the integer multiples of a known time delay (64
chips offset between adjacent pilots). All pilots in a pilot set have the same CDMA frequency
assignment. The pilots identified by the mobile, as well as other pilots specified by the serving
sectors (neighbors of the serving base stations/sectors), are continuously categorized by the
mobile into four groups.

• Active set. It contains the pilots associated with the forward traffic channels (Walsh
codes) assigned to the mobile. Because there are three fingers of the RAKE receiver in
the mobile, the active set size is a maximum of three pilots. IS-95 allows up to six pilots
in the active set, with two pilots sharing one RAKE finger.The base station informs the
mobile about the contents of the active set by using the Channel Assignment message
and/or the Handoff Direction message (HDM). An active pilot is a pilot whose paging
or traffic channels are actually being monitored or used.
• Candidate set. This set contains the pilots that are not currently in the active set. How-
ever, these pilots have been received with sufficient signal strength to indicate that the
associated forward traffic channels could be successfully demodulated. Maximum size
of the candidate set is six pilots.
• Neighbor set. This set contains neighbor pilots that are not currently in the active or the
candidate set and are likely candidates for handoff. Neighbors of a pilot are all the sec-
tors/cells that are in its close vicinity. The initial neighbor list is sent to the mobile in
the System Parameter message on the paging channel. The maximum size of the neigh-
bor set is 20.
• Remaining set. This set contains all possible pilots in the current system, excluding
pilots in the active, candidate, or neighbor sets.
While searching for a pilot, the mobile is not limited to the exact offset of the short PN
code. The short PN offsets associated with various multipath components are located a few chips
away from the direct path offset. In other words, the multipath components arrive a few chips
later relative to the direct path component. The mobile uses the search window for each pilot of
the active and candidate set, around the earliest arriving multipath component of the pilot.
Search window sizes are defined in number of short PN chips. The mobile should center the
search window for each pilot of the neighbor set and the remaining set around the pilot’s PN off-
set using the mobile time reference.

10.4 Search Windows

The mobile uses the following three search windows to track the received pilot signals:

• SRCH_WIN_A: search window size for the active and candidate sets
• SRCH_WIN_N: search window size for the neighbor set
• SRCH_WIN_R: search window size for the remaining set
186 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

10.4.1 SRCH_WIN_A
SRCH_WIN_A is the search window that the mobile uses to track the active and candidate
set pilots. This window is set according to the anticipated propagation environment—it should
be large enough to capture all usable multipath signal components of a base station, and at the
same time it should be as small as possible in order to maximize searcher performance.

EXAMPLE 10.1
Consider the propagation environment of a CDMA network, where the signal with a direct path trav-
els 1 kilometer (km) to the mobile, whereas the multipath travels 5 km before reaching the mobile.
What should be the size of SRCH_WIN_A?
1000
Direct path travels a distance of ------------ = 4.1 chips
244
5000
Multipath travels a distance of ------------ = 20.5 chips
244
The difference in distance traveled between the two paths = 20.5 – 4.1 = 16.4 chips
The window size ≥ 2 × 16.4 = 32.8 chips
Use window size = 33 chips

EXAMPLE 10.2
Consider cells A and B separated by a distance of 12 km. The mobile travels from cell A to cell B.
The RF engineer wishes to contain the soft handoff area between points X and Y located at distance
6 and 10 km from cell A (see Fig. 10-4). What should be the search window size?
At point X the mobile is 6000/244 = 24.6 chips from cell A
At point X the mobile is 10,000/244 = 41.0 chips from cell B
Path difference = 41.0 – 24.6 = 16.4 chips
At point Y the mobile is 10,000/244 = 41.0 chips from cell A
At point Y the mobile is 6000/244 = 24.6 chips from cell B
Path difference = 41.0 – 26.4 = 16.4 chips
The SRCH_WIN_A > 2 × 16.4 > 32.8 chips
This way, as the mobile travels from cell A to cell B, the mobile can ensure that, beyond Y, the
pilot from cell A drops out of the search window.

10.4.2 SRCH_WIN_N
SRCH_WIN_N is the search window that the mobile uses to monitor the neighbor set
pilots. The size of this window is typically larger than that of SRCH_WIN_A. The window
needs to be large enough not only to capture all usable multipath of the serving base station’s
signal, but also to capture the potential multipath of neighbors’ signals. In this case, we need to
take into account multipath and path differences between the serving base station and neighbor-
Handoff Parameters 187

Cell A Cell B
Soft Handoff Region

X Y

6 km
10 km
16 km

Figure 10-4 SRCH_WIN_A for Soft Handoff between X and Y

ing base stations. The maximum size of this search window is limited by the distance between
two neighboring base stations. Let’s consider two neighboring base stations located at a distance
of 6 km. The mobile is located right next to base station 1, and, therefore, the propagation delay
from base station 1 to the mobile is negligible. The distance between base station 2 and mobile is
6 km. The distance in chips is 6000/244 = 24.6 chips. The search window shows that the pilot
from cell 2 arrives 24.6 chips later at the mobile. Thus, in order for a mobile (located within cells
1 and 2) to search pilots of potential neighbors, SRCH_WIN_N needs to be set according to the
physical distances between the current base station and its neighboring base station. The actual
size may not be this large, since this is an upper bound for SRCH_WIN_N.

10.4.3 SRCH_WIN_R
SRCH_WIN_R is the search window that the mobile uses to track the remaining set pilots.
A typical requirement for the size of this window is that it is at least as large as SRCH_WIN_N.

10.5 Handoff Parameters

There are four handoff parameters. T_ADD, T_COMP, and T_DROP relate to the measurement
of pilot Ec /It and T_TDROP is a timer. Whenever the strength of a pilot in the active set falls
below a value of T_DROP, a timer is started by the mobile. If the pilot strength goes back above
T_DROP, the timer is reset; otherwise the timer expires when a time T_TDROP has elapsed
since the pilot strength has fallen below T_DROP. Mobile maintains a handoff drop timer for
each pilot in the active set and in the candidate set.

10.5.1 Pilot Detection Threshold (T_ADD)

Any pilot that is strong but is not in the HDM is a source of interference. This pilot must
be immediately moved to the active set for handoff to avoid voice degradation or a possible
dropped call. T_ADD affects the percentage of mobiles in handoff. It should be low enough to
quickly add useful pilots and high enough to avoid false alarms due to noise.

10.5.2 Comparison Threshold (T_COMP)

It has effect on handoff percentage similar to T_ADD. It should be low for faster handoff
and should be high to avoid false alarms.
188 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

10.5.3 Pilot Drop Threshold (T_DROP)

It affects the percentage of mobiles in handoff. It should be low enough to avoid dropping
a good pilot that goes into a short fade. It should be high enough not to quickly remove useful
pilots in the active or candidate set. The value of T_DROP should be carefully selected by con-
sidering the values of T_ADD and T_TDROP.

10.5.4 Drop Timer Threshold (T_TDROP)

It should be greater than the time required to establish handoff. T_TDROP should be small
enough not to quickly remove useful pilots. A large value of T_TDROP may be used to force a
mobile to continue in soft handoff in a weak coverage area.
Table 10-1 provides typical values of the handoff parameters.

10.6 Handoff Messages

Handoff messages in IS-95 are Pilot Strength Measurement message (PSMM), Handoff Direc-
tion message (HDM), Handoff Completion message (HCM), and Neighbor List Update message
(NLUM).
The mobile detects pilot strength (Ec /It ) and sends the PSMM to the base station. The
base station allocates the forward traffic channel and sends the HDM to the mobile. On receiv-
ing the HDM, the mobile starts demodulation of the new traffic channel and sends HCM to the
base station.
The PSMM contains the following information for each of the pilot signals received by the
mobile:
• Estimated Ec /It
• Arrival time
• Handoff drop timer
The HDM contains the following information:
• HDM sequence number
• CDMA channel frequency assignment
• Active set (now has old and new pilots [PN offsets])
• Walsh code associated with each pilot in the active set

Table 10-1 Handoff Parameter Values

Parameter Range Suggested Value
T_ADD –31.5 to 0 dB –13 dB
T_COMP 0 to 7.5 dB 2.5 dB
T_DROP –31.5 to 0 dB –15 dB
T_TDROP 0 to 15 seconds 2 seconds
Handoff Messages 189

• Window size for the active and candidate sets

• Handoff parameters (T_ADD, T_DROP, T_COMP, T_TDROP)
The HCM contains the following information:
• A positive acknowledgment
• PN offset of each pilot in the active set
The NLUM is sent by the base station. It contains the latest composite neighbor list for the
pilots in the active set.
The mobile continuously tracks the signal strength for all pilots in the system. The signal
strength of each pilot is compared with the various thresholds such as the pilot detection thresh-
old, the pilot drop threshold, the comparison threshold, and the drop timer threshold.
A pilot is moved from one set to another depending on its signal strength relative to the
thresholds. Fig. 10-5 shows a sequence on the threshold.
1. Pilot strength exceeds T_ADD. Mobile sends a PSMM and transfers pilot to the candi-
date set.
2. Base station sends an HDM to the mobile with the pilot to be added in active set.
3. Mobile receives HDM and acquires the new traffic channel. Pilot goes into the active
set and mobile sends HCM to the base station.
4. Pilot strength drops below T_DROP; mobile starts the handoff drop timer.

Base station
Pilot sends HDM Mobile sends Mobile sends
Strength HCM Base station HCM
sends HDM
Mobile sends Mobile
PSMM Mobile sends receives
PSMM NLUM

T_ADD

T_DROP

Active
Candidate
Neighbor

Neighbor

(1) (2) (3) (4) (5) (6) (7) (8)

PSMM: Pilot Strength Measurement Message Handoff

HDM: Handoff Direction Message Drop Timer
HCM: Handoff Completion Message

Figure 10-5 Handoff Threshold Example: Pilot Thresholds

190 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

5. Handoff drop timer expires. Mobile sends a PSMM to the base station.
6. Base station sends an HDM without related pilot to the mobile.
7. Mobile receives HDM. Pilot goes into the neighbor set and mobile sends HCM to the
base station.
8. The mobile receives an NLUM which does not include the pilot. Pilot goes into the
remaining set.
The mobile maintains a T_TDROP for each pilot in the active set and candidate set. The
mobile starts the timer whenever the strength of the corresponding pilot becomes less than a pre-
set threshold. The mobile resets and disables the timer if the strength of the corresponding pilot
exceeds the threshold.
When a member of the neighbor or remaining set exceeds T_ADD, the mobile moves the
pilot to candidate set (Fig. 10-6) and sends a PSMM to the base station. As the signal strength of
candidate pilot Pc gradually increases, it rises above the active set pilot, Pa. A PSMM is sent to
the base station only if
Pc – Pa > T_COMP x 0.5 dB
where Pa and Pc are the strength of pilots in active and candidate sets.

10.7 Handoff Procedures

10.7.1 Mobile-Assisted Soft-Handoff (MASHO) Procedures
The mobile monitors the Forward Pilot Channel (FPICH) level received from neighboring
base stations and reports to the network those FPICHs that cross a given set of thresholds. Two
types of thresholds are used: the first to report FPICHs with sufficient power to be used for coher-

Pilot Strength

T_COMP × 0.5 dB
Pa

T_ADD

1 2

Neighbor or Remaining Candidate Active

Set Set

Figure 10-6 Pilot Movement from Neighbor or Remaining Set to Active Set
Handoff Procedures 191

ent demodulation, and second to report those FPICHs whose power has dropped to a level where
it is not beneficial to use them for coherent demodulation. The margin between the two thresholds
provides a hysteresis to avoid a ping-pong effect due to variations in FPICH power. Based on this
information, the network instructs the mobile to add or remove FPICHs from its active set.
The same user information, modulated by the appropriate base station code, is sent from
multiple base stations. Coherent combining of different signals from different sectorized anten-
nas, from different base stations, or from the same antennas but on different multiple path compo-
nents is performed in the mobile using RAKE receivers. A mobile will typically place at least one
RAKE receiver finger on the signal from each base station in the active set. If the signal from the
base station is temporarily weak, then the mobile can assign the finger to a stronger base station.
The signal transmitted by a mobile is processed by base stations with which the mobile is
in soft handoff. The received signal from different sectors of a base station is combined in the
base station on a symbol-by-symbol basis. The received signal from different base stations can
be selected in the infrastructure (on a frame-by-frame basis). Soft handoff results in increased
coverage range and capacity on the reverse link.

10.7.2 Dynamic Soft-Handoff Thresholds

While soft handoff improves overall system performance, it may in some situations nega-
tively impact system capacity and network resources. On the forward link, excessive handoff
reduces system capacity whereas, on the reverse link, it costs more network resources (backhaul
connections).
Adjusting the handoff parameters at the base stations will not necessarily solve the prob-
lem. Some locations in the cell receive only weak FPICHs (requiring lower handoff thresholds),
and other locations receive a few strong and dominant FPICHs (requiring higher handoff thresh-
olds). The principle of dynamic threshold for adding FPICHs is as follows:
• The mobile detects FPICHs that cross a given static threshold, T1. The metric for the
FPICH in this case is the ratio of FPICH energy per chip to total received power (Ec /It ).
• On crossing the static threshold, the FPICH is moved to a candidate set. It is then
searched more often and tested against a second dynamic threshold, T2.
• Comparison with T2 determines if the FPICH is worth adding to the active set. T2 is a
function of the total energy of FPICHs demodulated coherently (in the active set).
• The condition of an FPICH for crossing T2 is expressed as
NA
   
10 log ( P cj ) ≥ Max  SOFT-SLOPE ⋅ 10 log  ∑ P ai + ADD-INTERCEPT, T 1  (10.1)
 i = 1  

where Pcj = strength of the jth FPICH in the coordinate set,

Pai = strength of the ith FPICH in the active set,
NA = number of FPICHs in the active set, and
SOFT-SLOPE and ADD-INTERCEPT = adjustable system parameters.
192 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

When FPICHs in the active set are weak, adding an additional FPICH (even weak) will
improve performance. However, when there is one or more dominant FPICHs, adding an addi-
tional weaker FPICH above T 1 will not improve performance, but will use more network
resources. The dynamic soft-handoff thresholds reduce and optimize the network resource utili-
zation.
• After detecting an FPICH above T2, the mobile reports it back to the network. The net-
work then sets up the handoff resources and orders the mobile to coherently demodu-
late this additional FPICH. Pilot 2 is added to active set.
• When the FPICH (pilot 1) strength decreases below a dynamic threshold T3, the hand-
off connection is removed. The FPICH is moved back to the candidate set. The thresh-
old T 3 is a function of the total energy of FPICHs in the active set. FPICHs not
contributing sufficiently to total FPICH energy are dropped. If it decreases below a
static threshold T4, an FPICH is removed from the candidate set.
• An FPICH dropping below a threshold (e.g., T3 and T4) is reported back to the network
only after being below the threshold for a specific period of time. This timer allows for
a fluctuating FPICH not to be prematurely reported.
Fig. 10-7 shows a time representation of soft handoff and associated events when the
mobile station moves away from a serving base station (FPICH 1) toward a new base station
(FPICH 2). The combination of static and dynamic thresholds (vs. static thresholds only) results
in reduced soft-handoff regions (see Fig. 10-7). The major benefit of this is to limit soft handoff
to areas and times when it is most beneficial.
1. When pilot 2 exceeds T1, mobile moves it to the candidate set.
2. When pilot 2 exceeds T2 (dynamic), mobile reports it back to the network.
3. Mobile receives an order to add pilot 2 to the active set.
4. Pilot 1 drops below T3 (relative pilot 2).
5. Handoff timer expires on pilot 1. Mobile reports pilot strength to the network.
6. Mobile receives an order to remove pilot 1.
7. Handoff timer expires after pilot 1 drops below T4.

10.8 Setup and End of Soft Handoff

10.8.1 Setup
One of the major benefits of a CDMA system is the ability of a mobile to communicate
with more than one base station at one time during a call. This functionality allows the CDMA
network to perform soft handoff. In soft handoff a controlling primary base station coordinates
with other base stations as they are added or deleted for the call. This allows the base stations (up
to three, total) to receive/transmit voice packets with a single mobile for a single call.
Each base station transmits the received mobile voice packets to the BSC/MSC. The BSC/
MSC selects the best voice frame from one of the three base stations. This provides the PSTN
party with the best-quality voice.
Setup and End of Soft Handoff 193

Active Set Total Ec /It

Pilot 1 Pilot 2

Ec /It

T1

T4

Dynamic threshold—mobile in Time

soft handoff between 3 and 6
1 2 3 4 5 6 7
Static threshold only—mobile in
soft handoff between 1 and 7

Figure 10-7 Dynamic Thresholds Handoff Procedure

Fig. 10-8 shows a mobile communicating with two base stations for one call. This is called
a two-way soft handoff. Steps of soft handoff are
• The mobile detects a pilot signal from a new cell and informs primary base station A.
• A communications path from base station B to the original frame selector is established.
• The frame selector selects frames from both streams.
• The mobile detects that base station A’s pilot is failing and requests that this path be
dropped.
• The path from original base station A to the frame selector is dropped.
Base station B gives base station A its assigned Walsh code. Base station A gives the
mobile the Walsh code of B as part of the HDM. Now the mobile can listen to base station B.
Base station A gives the user’s long-code mask to base station B. Now B can listen to the
mobile. Both base stations A and B receive forward link power control information back from
the mobile and act accordingly. The mobile receives independent puncture bits from both A and
B. If directions conflict, the mobile decreases power; otherwise the mobile obeys directions.

10.8.2 End of Soft Handoff

Fig. 10-9 shows the process used by a mobile communicating with two base stations A and
B to end handoff when the signal from base station A is not strong enough. When the mobile
194 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

MS Primary BS A Secondary BS B BSC/MSC

Pilot Strength Measurement

Handoff Request
Frame Selector Join

ACK

ACK
Handoff Direction

ACK
Handoff Information

ACK
Pilot Measurement Request

Pilot Strength Measurement

Figure 10-8 Soft Handoff Setup

MS Primary BS A Secondary BS B BSC/MSC

PSMM

HDM (Drop A)

HCM
Master Transfer

Master Transfer ACK

ACK

ACK
New Primary BS
Handoff Information

ACK

Pilot Measurement Request

PSMM

Frame Selector Remove

ACK

Figure 10-9 End of Soft Handoff

Maintenance of Pilot Sets 195

entered into soft handoff with base stations A and B, the primary base station was A. However,
when the mobile drops A and starts communicating with base station B alone, B becomes the
new primary base station.

10.9 Maintenance of Pilot Sets

10.9.1 Active Set Maintenance
The active set is initialized to contain only one pilot (e.g., the pilot associated with the
assigned forward traffic channel). This occurs when the mobile is first assigned a forward traffic
channel. As the mobile processes HDMs, it updates the active set with the pilots listed in the
HDMs.
A pilot Pc from the candidate is added to the active set when Pc exceeds a member of the
active set by T_COMP. A pilot Pa from the active set is removed when Pa has dropped below
T_DROP and the drop timer (T_TDROP) has expired (see Fig. 10-10).

10.9.2 Candidate Set Maintenance

The candidate set is initialized to contain no pilot. This happens when the mobile is first
assigned a forward traffic channel. A pilot Pn from the neighbor set is added to the candidate set
when its strength exceeds T_ADD. Also, a pilot Pr from the remaining set is moved to the candi-
date set when its strength exceeds T_ADD. A pilot Pc is deleted from the candidate set when the
handoff drop timer corresponding to Pc has expired. Also, when the candidate set size has been
exceeded, the pilot Pc , whose handoff drop timer is close to expiring, is deleted from the candi-
date set (see Fig. 10-11).

10.9.3 Neighbor Set Maintenance

The neighbor set is initialized to contain the pilots specified in the most recently received
Neighbor List message. This happens when the mobile is first assigned a forward traffic channel.
The mobile maintains a counter—AGE—for each pilot in the neighbor set. If a pilot moves from
the active set or candidate set to neighbor set, its counter is initialized to 0. However, if a pilot

Pilot exceeds active by T_COMP

Remaining Set Neighbor Set Candidate Set Active Set

Pilot below T_DROP and T_TDROP timer has expired

Figure 10-10 Active Set Maintenance

196 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

Pilot exceeds T_ADD Pilot exceeds active by T_COMP

Remaining Set Neighbor Set Candidate Set Active Set

Pilot below T_DROP and T_TDROP timer has expired

Pilot exceeds T_ADD

Figure 10-11 Candidate Set Maintenance

moves from the remaining set to the neighbor set, its counter is set to the maximum age value
(see Fig. 10-12). The mobile adds a pilot in the neighbor set under the following conditions:
• A pilot in the active set is not contained in the HDM, and the corresponding handoff
drop timer has expired.
• The handoff drop timer of a pilot in the candidate set has expired.
• A new pilot to the candidate set causes the candidate set size limit to be exceeded.
• The pilot is contained in the Neighbor List message and is not already a pilot of the
candidate set or neighbor set.
The mobile deletes a pilot in the neighbor set under the following conditions:
• The HDM contains a pilot from the current neighbor set.
• The strength of a pilot in the neighbor set exceeds T_ADD.
• A new pilot to the neighbor set causes the size limit of the neighbor set to be exceeded.
• A neighbor set pilot’s AGE exceeds the maximum value of the AGE counter.

10.10 The Need for Power Control

CDMA is an interference-limited system—since all mobiles transmit at the same frequency,
internal interference generated within the system plays a critical role in determining system
capacity and voice quality. The transmit power from each mobile must be controlled to limit
interference. However, the power level should be adequate for satisfactory voice quality.
As the mobile moves around, the RF environment changes continuously due to fast and
slow fading, shadowing, external interference, and other factors. The objective of power control
is to limit transmitted power on the forward and reverse links while maintaining link quality
under all conditions. Due to noncoherent detection at the base station, interference on the reverse
link is more critical than it would be on the forward link. Reverse link power control is therefore
essential for a CDMA system and is enforced by the IS-95 standard.
The Need for Power Control 197

AGE = MAX_AGE AGE = 0

Remaining Set Neighbor Set Candidate Set Active Set

AGE = 0

Active Set

Candidate Set Pn > T_ADD

Pn – Pa > T_COMP

Neighbor Set
Set Size Exceeded
or
AGE > AGE_MAX
Remaining Set

Figure 10-12 Neighborhood Set Maintenance

Power control is also needed in CDMA systems to resolve the near-far problem. To mini-
mize the near-far problem, the goal in a CDMA system is to assure that all mobiles achieve the
same received power levels at the base station. The target value for the received power level must
be the minimum level possible that allows the link to meet user-defined performance objectives
(BER, FER, capacity, dropped-call rate, and coverage). In order to implement such a strategy,
the mobiles closer to the base station must transmit less power than those far away.
Voice quality is related to frame-error rate (FER) on both the forward and reverse link. The
FERs are largely correlated to Eb /It . The FER also depends on vehicle speed, local propagation
conditions, and distribution of other cochannel mobiles. Since the FER is a direct measure of
signal quality, the voice quality performance in a CDMA system is measured in terms of FERs
rather than Eb /It . Thus, to assure good signal quality, it is not sufficient to maintain a target Eb /It ;
198 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

it is also necessary to respond to specific FERs as they occur. The recommended performance
bounds are
• A typical recommended range for FER—0.2% to 3% (optimum power level is achieved
when FER ≤ 1%)
• A maximum length of burst error—3 to 4 frames (optimum value of burst error ≈ 2
frames)

10.11 Reverse Link Power Control

The reverse link power control affects the access and reverse traffic channels. It is used for estab-
lishing the link while originating a call and reacting to large path-loss fluctuations. The reverse
link power control includes the open-loop power control (also known as autonomous power con-
trol) and the closed-loop power control. The closed-loop power control involves the inner-loop
power control and the outer-loop power control.

10.11.1 Reverse Link Open-Loop Power Control

The open-loop power control is based on the principle that a mobile closer to the base sta-
tion needs to transmit less power as compared to a mobile that is farther away from the base sta-
tion or is in fade. The mobile adjusts its transmit power based on total power received in the
1.23-MHz band (i.e., power in pilot, paging, sync, and traffic channels). This includes power
received from all base stations on the forward link channels. If the received power is high, the
mobile reduces its transmit power. On the other hand, if the power received is low, the mobile
increases its transmit power.
In open-loop power control the base station is not involved. The mobile determines the ini-
tial power transmitted on the access channel and traffic channel through open-loop power con-
trol. A large dynamic range of 80 dB is allowed to provide an ability to guard against deep fades.
The mobile acquires the CDMA system by receiving and processing the pilot, sync, and
paging channels. The paging channel provides the Access Parameters message which contains
the parameters to be used by the mobile when transmitting to the base station on an access chan-
nel. The access parameters are
• The access channel number
• The nominal power offset (NOM_PWR)
• The initial power offset step size
• The incremental power step size
• The number of access probes per access probe sequence
• The time-out window between access probes
• The randomization time between access probe sequences
Based on the information received on the pilot, sync, and paging channels, the mobile
attempts to access the system via one of several available access channels. During the access
state, the mobile has not yet been assigned a forward link traffic channel (which contains the
Reverse Link Power Control 199

power control bits). Since the reverse link closed-loop power control is not active, the mobile
initiates, on its own, any power adjustment required for a suitable operation.
The prime goal in CDMA systems is to transmit just enough power to meet the required
performance objectives. If more power is transmitted than necessary, the mobile becomes a jam-
mer to other mobiles. Therefore, the mobile tries to get the base station attention first by trans-
mitting at very low power. The key rule is that the mobile transmits in inverse proportion to what
it receives.
When receiving a strong pilot from the base station, the mobile transmits a weak signal
back to the base station. A strong signal at the mobile implies a small propagation loss on the
forward link. Assuming the same path loss on the reverse link, only a low transmit power is
required from the mobile in order to compensate for the path loss.
When receiving a weak pilot from the base station, the mobile transmits back a strong sig-
nal. A weak received signal at the mobile indicates a high propagation loss on the forward link.
Conversely, a high transmit power level is required from the mobile.
The mobile transmits the first access probe at a mean power level defined by

T x = – R x – K + ( NOM-PWR – 16 × NOM-PWR-EXT ) + INIT-PWR (dBm) (10.2)

where Tx = mean output transmit power (dBm),

Rx = mean input receive power (dBm),
NOM-PWR = nominal power (dB),
NOM-PWR-EXT = nominal power for extended handoff (dB),
INIT-PWR = initial adjustment (dB),
K = 73 for cellular (Band Class 0), and
K = 76 for PCS (Band Class 1).
If INIT-PWR were 0, then NOM-PWR – 16 × NOM-PWR-EXT would be the correction
that should provide the correct received power at the base station. NOM-PWR – 16 × NOM-
PWR-EXT allows the open-loop estimation process to be adjusted for different operating envi-
ronment.
The values for NOM-PWR, NOM-PWR-EXT, INIT-PWR, and the step size of a single
access probe correction PWR-STEP are system parameters specified in the Access Parameters
message. These are obtained by the mobile station prior to transmitting. If, as the result of an
Extended Handoff Direction message or a General Handoff Direction message, the NOM-PWR
and NOM-PWR-EXT values change, the mobile uses the NOM-PWR and NOM-PWR-EXT val-
ues from the Extended Handoff Direction message or a General Handoff Direction message.
The total range of the NOM-PWR – 16 × NOM-PWR-EXT correction is –24 to 7 dB. While
operating in Band Class 0, NOM-PWR-EXT is set to 0, making the total range of correction from
–8 to 7 dB. The range of the INIT-PWR parameter is –16 to 15 dB, with a nominal value of 0 dB.
The range of the PWR-STEP parameter is 0 to 7 dB. The accuracy of the adjustment to the mean
output power due to NOM-PWR, NOM-PWR-EXT, INIT-PWR, or a single access probe correc-
tion of PWR-STEP should be ±0.5dB or ±20%, whichever is greater.
200 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

The major flaw with this criterion is that reverse link propagation statistics are estimated
based on forward link propagation statistics. But, since the two links are not correlated, a signif-
icant error may result from this procedure. However, these errors will be corrected once the
closed-loop power control mechanism becomes active as the mobile seizes a forward traffic
channel and begins to process power control bits.
After the Acknowledgment time window (Ta ) has expired, the mobile waits for an addi-
tional random time (RT) and increases its transmit power by a step size. The mobile tries again.
The process is repeated until the mobile gets a response from the base station. However, there is
a maximum number of probes per probe sequence and a maximum number of probe sequences
per access attempt.
The entire process to send one message and receive an acknowledgment for the message is
called an access attempt. Each transmission in the access attempt is referred to as an access
probe. The mobile transmits the same message in each access probe in an access attempt. Each
access probe contains an access channel preamble and an access channel capsule (see Fig. 10-
13). Within an access attempt, access probes are grouped into access probe sequences. Each
access probe sequence consists of up to 16 access probes, all transmitted on the same access
channel.
There are two reasons that could prevent the mobile from getting an acknowledgment after
the transmission of a probe.

1 2 15 (Max.)
Back-off
Delay

Probe Sequence Probe # 16

Probe # 1 Probe # 3
Probe # 2

Random Time (RT) Access Preamble Message Capsule

Acknowledgment
Window (Ta) 1–16 Frames 3–16 Frames
Figure 10-13 Access Attempt, Probe Sequence, and Probe in Open-Loop Power Control
Reverse Link Power Control 201

1. The transmit power level might be insufficient. In this case, the incremental step power
strategy helps to resolve the problem.
2. There might be a collision due to the random contention of the access channel by sev-
eral mobiles. In this case, the random waiting time minimizes the probability of future
collisions.
The process is shown by the access probe ladder in Fig. 10-14.
The transmit power is defined by

T x = – R x – K + ( NOM-PWR – 16 × NOM-PWR-EXT )
+ Sum of Access Probe Corrections (10.3)
where the access probe correction is the sum of all the appropriate incremental power steps prior
to receiving an acknowledgment at the mobile.
For every access probe sequence, a back-off delay is generated pseudorandomly. Timing
between access probes of an access probe sequence is also generated pseudorandomly. After
transmitting each access probe, the mobile waits for Ta . If an acknowledgment is received, the
access attempt ends. If no acknowledgment is received, the next access probe is transmitted after
an additional random time (see Fig. 10-13).
If the mobile does not receive an acknowledgment within an access attempt, the attempt is
considered as a failure and the mobile tries to access the system at another time. If the mobile
receives an acknowledgment from the base station, it proceeds with the registration and traffic
channel assignment procedures. The initial transmission on the reverse traffic channel shall be at
a mean output power defined by Eq. (10.3).
The mobile station supports a total combined range of initial offset parameters, closed
NOM-PWR, and access probe corrections of at least ±32dB for mobile stations operating in
Band Class 0 and ±40dB for mobile stations operating in Band Class 1.

Tx
Initial Power
+ Open-Loop
Correction

Initial Power Open-Loop Correction

Access Probe
Figure 10-14 Access Probe Ladder
202 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

The sources of error in the open-loop power control are

• Assumption of reciprocity on the forward and reverse links
• Use of total received power including power from other base stations
• Slow response time ~ 30 ms to counter fast fading due to multipath

10.11.2 Reverse Link Closed-Loop Power Control

Fading sources in multipath require a much faster power control than the open-loop power
control. The additional power adjustments required to compensate for fading losses are handled
by the reverse link closed-loop power control mechanism, which has a response time of 1.25 ms
for 1-dB steps and a dynamic range of 48 dB (covered in 3 frames). The quicker response time
gives the closed-loop power control mechanism the ability to override the open-loop power con-
trol mechanism in practical applications. Together, two independent power control mechanisms
cover a dynamic range of at least 80 dB. The closed-loop power control provides correction to
the open-loop power control. Once on the traffic channel, the mobile and base stations engage in
closed-loop power control.
The reverse link closed-loop power control mechanism consists of two parts—inner-loop
power control and outer-loop power control. The inner-loop power control keeps the mobile as
close to its target (Eb /It )setpoint as possible, whereas the outer-loop power control adjusts the base
station target (Eb /It )setpoint for a given mobile.
To understand the operation of the closed-loop power control mechanism, let’s review the
structure of the forward traffic channel and its operation. The areas of focus are the output of the
interleaver and the input to the MUX. A power control subchannel continuously transmits on the
forward traffic channel. This subchannel runs at 800 power control bits per second. Therefore, a
power control bit (0 or 1) is transmitted every 1.25 ms. A 0 bit indicates to the mobile that it
should increase its mean output power level, whereas 1 indicates to the mobile to decrease its
mean output power level.
A 20-ms frame is organized into 16 time intervals of equal duration (see Fig. 10-15).
These time intervals, each of 1.25 ms, are called Power Control Groups (PCGs). Thus, a frame
has 16 PCGs. Prior to transmission, the reverse traffic channel interleaver output data stream is
gated with a time filter. The time filter allows transmission of some symbols and deletion of

One Frame (20 ms) = 16 PCG

Gated-on PCGs
Gated-off PCGs 1 PCG = 1.25 ms PCG = power control group

Figure 10-15 Power Control Groups

Reverse Link Power Control 203

others. The duty cycle of the transmission gate varies with the transmit data rate, i.e., variable
rate vocoder output, which, in turn, depends on the voice activity. Table 10-2 indicates the num-
ber of PCGs that are sent at different frame rates.
The assignment of the gated-on and gated-off groups is determined by the Data Burst Ran-
domizer (DBR). At the base station, the reverse link receiver estimates the received signal
strength by measuring Eb /It during each power group (1.25 ms).
• If the signal strength exceeds a target value, a power-down power control bit 1 is sent.
• Otherwise a power-up control bit 0 is transmitted to the mobile via the power control
subchannel on the forward link.
Similar to the reverse link transmission, the forward link transmissions are organized in
20-ms frames. Each frame is subdivided into 16 PCGs. The transmission of a power control bit
occurs on the forward traffic channel in the second PCG following the corresponding reverse
link PCG in which the signal strength was estimated. For example, if the signal strength is esti-
mated on PCG #2 of a reverse link frame, then the corresponding power control bit must be sent
on PCG #4 of the forward link frame (see Fig. 10-16). Once the mobile receives and processes
the forward link channel, it extracts the power control bits from the forward traffic channel. The
power control bits then allow the mobile to fine-tune its transmit power on the reverse link.
Based on the power control bit received from the base station, the mobile either increases
or decreases transmit power on the reverse traffic channel as needed to approach the target value
of (Eb /It )nom or set point that controls the long-term FER. Each power bit produces a 1-dB change
in mobile power, i.e., it attempts to bring the measured Eb /It value 1 dB closer to its target value.
Note that it might not succeed because It is also always changing. Therefore, further adjustments

Table 10-2 Power Control Groups vs. Frame Rate

Frame Rate Rate (kbps) No. of PCGs Sent
Full 9.6 16
1/2 4.8 8
1/4 2.4 4
1/8 1.2 2

Reverse Link Frame 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Forward Link Frame 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Figure 10-16 PCG Location in Reverse and Forward Link Frames

204 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

may be required to achieve the desired Eb /It . The base station, through the mobile, can directly
change only Eb, not It , and the objective is the ratio of Eb to It , not any particular value for Eb or It.
The base station measures Eb /It 16 times in each 20-ms frame. If the measured Eb /It is
greater than the current target value of Eb /It , the base station informs the mobile to decrease its
power by 1 dB. Otherwise, the base station orders the mobile to increase its power by 1 dB (see
Fig. 10-17).
The relationship between Eb /It and the corresponding FER is nonlinear and varies with
vehicle speed and RF environment. Performance deteriorates with increasing vehicle speed. The
best performance corresponds to a stationary vehicle where additive white Gaussian noise domi-
nates. Thus, a single value of Eb /It is not satisfactory for all conditions. The use of a single, fixed
value for Eb /I t could reduce channel capacity by 30% or more by transmitting excessive,
unneeded power.
The value of the variable a is kept very small (see Fig. 10-18), so it may take 35 frames to
reduce the Eb /It set point by 1 dB. Typically, the value of 100a is set at about 3 dB. The set point
value is reduced by a for each consecutive frame until a frame error occurs. The set point is then
increased by a relatively large amount and the process is repeated. The set point can range from
3 dB to 10 dB. A value of Eb /It ≥ 5 dB corresponds to good voice quality.
Since FER is a direct measure of link quality, the system is controlled using the measured
FERs rather than Eb /It . FER is the key parameter in controlling and assuring a satisfactory voice
quality. It is not sufficient to maintain a target Eb /It , but it is necessary to control FERs as they
occur. The objective of the Reverse Outer-Loop Power Control (ROLPC) is to balance the
desired FER on the reverse link and system capacity. System capacity can be controlled with the

Eb Eb
Mobile
Transmit
Power
(dBm) 1.25 ms

Next
Eb /It
Target Set Point
Eb /It

Eb /It Set Point

It

Frame i Frame i + 1 Time

Figure 10-17 Target Eb /It

Reverse Link Power Control 205

Set Point Value

Maximum
10 dB

100a

100a
Frame Error
20 ms Occurs
Minimum
3 dB
Time
Figure 10-18 Set Point Value vs. Time

ROLPC parameters by increasing the acceptable FER. Changing FER can be accomplished by
setting the ratio of down_ frr to up_ frr. The down_ frr is calculated by the system by using the
desired reverse FER (rfer) and up_ frr as
down_ frr = (rfer × up_ frr)/2 (10.4)
Based on simulations, the following values for up_ frr are suggested:
If (0.2% ≤ rfer ≤ 0.4%), up_ frr = 6000
If (0.6% ≤ rfer ≤ 1.0%), up_ frr = 5000
If (1.2% ≤ rfer ≤ 2.0%), up_ frr = 3000
If (2.2% ≤ rfer ≤ 3.0%), up_ frr = 1000
Tables 10-3 and 10-4 lists the range and default values of different parameters for RS1 and
RS2.
The inner-loop power control is also responsible for detecting the mobile that fails to
respond to power control and that may be causing interference to other mobiles. The base station
counts the number of consecutive power decrease commands, and, if the count exceeds the spec-
ified threshold value, the base station will send a Lock until Power Cycle message to the mobile.
This message disables the mobile until the user turns the power off and on. Fig. 10-19 gives the
flow chart for the reverse link closed-loop power control.
206 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

Table 10-3 ROLPC Parameters for RS1

Parameter Range Suggested Value Description of Parameter
rfer 1 0.2–3.0% 1% target reverse link FER (rfer)
(Eb /It)nom 1 (dB) 3.5–8.0% 6.5% initial (Eb /It )set point
(Eb /It)max 1 (dB) 5.5—9.5% 8.5% maximum (Eb /It)set point
(Eb /It)min 1 (dB) 3.0–5.8% 3.5% minimum (Eb /It)set point

Table 10-4 ROLPC Parameters for RS2

Parameter Range Suggested Value Description of Parameter
rfer 2 0.2–6.0% 1% target reverse link FER
(Eb /It )nom 2 (dB) 3.8–8.3% 6.8% initial (Eb /It )set point
(Eb /It )max 2 (dB) 5.8–9.8% 8.8% maximum (Eb /It )set point
(Eb /It )min 2 (dB) 3.0–5.8% 3.8% minimum (Eb /It )set point

The mobile power output with both open-loop and closed-loop power control is given as

T x = – R x – K + ( NOM-PWR – 16 × NOM-PWR-EXT ) + INIT-PWR

+ Sum of Access Probe Corrections
+ Sum of all Closed-Loop Power Control Corrections (10.5)

10.12 Forward Link Power Control

Forward link power control (FLPC) aims at reducing interference on the forward link. The
FLPC not only limits the in-cell interference, but it is especially effective in reducing other cell/
sector interference.
The forward link power control attempts to set each traffic channel transmit power to the
minimum required to maintain the desired FER at the mobile. The mobile continuously mea-
sures forward traffic channel FER. It reports this measurement to the base station on a periodic
basis. After receiving the measurement report, the base station takes the appropriate action to
increase or decrease power on the measured logical channel. The base station also restricts the
power dynamic range so that the transmitter power never exceeds a maximum value that would
cause excessive interference or so that it never falls below the minimum value required for ade-
quate voice quality.
Since FERs are measured (not Eb /It as in the closed inner-loop strategy), this process is a
direct reflection of voice quality. However, it is a much slower process. Because orthogonal
Walsh codes are employed for the forward link instead of long PN codes, cochannel interference
is not an urgent issue. Therefore, slow measurements do not add much degradation to system
performance. Fig. 10-20 is a flow chart for the FLPC process.
Forward Link Power Control 207

Initialize Eb /It
Set Point to (Eb /It)nom

Measure Eb /It on
PCG (1.25 ms)

Yes No
(Eb /It)m > set point

Order mobile to
power up 1 dB.

Order mobile to
power down 1 dB.

No No One Frame
No. of Commands Received
> Threshold

Yes
Yes

Send a Lock until Calculate new (Eb /It)

Power Cycle set point.
message to mobile.

Inner Loop (Every 1.25 ms)

Outer Loop (20 ms)

Figure 10-19 Flow Chart for Reverse Link Closed-Loop Power Control

Set (ffer)T and Timer

Measure ffer
(ffer)m

No No
(ffer)m< (ffer)T Timer Has
Expired

Yes
Yes
Report Measurement

Repeat Process

Figure 10-20 Flow Chart for Forward Link Power Control

208 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

Forward link power control is expressed in terms of parameters N, D, U, and V (see Fig.
10-21), which may be adjusted to various values for the operation of an actual system.
For RS1, the Power Measurement Report message (PMRM) contains the number of error
frames received and the total number of frames received during the interval covered by the
report (frame counters then are reset for the next report interval). The FER is equal to the num-
ber of error frames divided by the total number of frames received in the reporting interval. The
following are the steps for forward link power control for RS1 (see Fig. 10-21).
Action by Mobile
• Mobile keeps track of the number of error frames in a period of length pwr_rep_ frame.
• If error frames > a specified number, the mobile sends a PMRM containing:
◆ Total number of frames in pwr_rep_ frame
◆ Number of error frames in pwr_rep_ frame
◆ FER
• If error frames < a specified number, a PMRM is not sent.
• After sending a PMRM, the mobile waits for a period—pwr_rep_delay—before start-
ing a new period.
Action by Base Station
• On receiving the PMRM, the base station compares the reported FER as follows and
adjusts traffic channel power.
◆ FER < fer_small → reduce power by D
◆ fer_small < FER < fer_big → increase power by U
◆ FER > fer_big → increase power by V

Traffic PMRM Not Received

Channel
Digital Key Parameters
Gain FER < fer-small N ~ 80 frames
V (dB) D ~ 0.25 dB
D (dB)
U ~ 1.0 dB
U (dB)
V ~ 2.0 dB
FER Too High
FER > fer_big

FER Too High

fer_small < FER < fer_big

N Frames

Time
Figure 10-21 Forward Link Power Control for RS1
Forward Link Power Control 209

• If no PMRM is received
◆ Base station starts a timer fpc_step.
◆ When timer expires, power level is reduced by D.
◆ The timer resets after it expires or after receipt of a PMRM.
• Digital gain is never set below min_gain or above max_gain.
• If flpc_enable = 0, digital gain is set to nom_gain.
For RS2, 1 bit per reverse link frame (the E or erasure bit) is dedicated to inform the base
station whether or not the last forward link frame was received without error at the mobile. This
allows more rapid and precise control of forward link power than the scheme used for RS1. The
following are the steps for forward link power control for RS2 (see Fig. 10-22).
Forward Link Power Control with RS2
• Uses erasure indicator bit instead of PMRM
• Much faster than RS1 implementation
◆ Forward link power control could change every 2 frames; thus, its response is
very fast.
• Process
◆ In each frame, the mobile sends an erasure indicator bit showing whether the pre-
vious forward frame had an erasure bit or not.
◆ If an erasure is indicated by the mobile, the base station increases traffic channel
digital gain by dn_adj.
Tables 10-5 and 10-6 list the values of the parameters for forward link power control for
RS1 and RS2, respectively.

Bad Traffic Frame Received with

Erasure Bit Unknown
Traffic
Channel
Digital Gain Good Traffic Frame Received
with Erasure Bit False

down_adj
up_adj

Good Traffic Frame Received

with Erasure Bit True
20 ms

Time
Figure 10-22 Forward Link Power Control for RS2
210 Ch. 10 • Soft Handoff and Power Control in IS-95 CDMA

Table 10-5 Forward Link Power Control Parameters for RS1

Parameters Range Suggested Value Description
FER 0.2–3% 1% target forward FER
fer_small 0.2–5% 2% lower forward link FER threshold minimum PMRM
FER required to increase gain by U
fer_big 2–10% 6% upper forward link FER threshold minimum PMRM
FER required to increase gain by V
min_gain 34–50 40 minimum traffic channel digital gain
max_gain 50–108 80 maximum traffic channel digital gain
nom_gain 34–108 57 nominal traffic channel digital gain
fpc_step 20–5000 ms 1600 ms forward power control timer value which determines
when gain is decreased by D

Table 10-6 Forward Link Power Control Parameters for RS2

Parameter Range Suggested Value Description
FER 0.2–6% 1% target forward FER
up_adj 1–50 15 gain increase when forward erasure is observed
dn_adj — N/A gain decrease when no forward erasure is observed
min_gain 30–50 30 minimum traffic channel digital gain
max_gain 50–127 127 maximum traffic channel digital gain
nom_gain 40–108 80 nominal traffic channel digital gain
where dn_adj = (up_adj × FER)/100

10.13 Summary
This chapter covered soft handoff and power control in IS-95 CDMA. Soft handoff provides
path diversity on the forward and reverse links. Diversity gains are achieved because less power
is required on the forward and reverse links. This results in the reduction of total system interfer-
ence and an increase in system capacity.
Since the RF environment changes continuously due to fast and slow fading, shadowing,
external interference, and other factors, the aim of power control is to adjust the transmitted
power on the forward and reverse link while maintaining link quality under all operating condi-
tions. Power control in the CDMA system is required to resolve the near-far problem. To mini-
mize the near-far problem, the goal in a CDMA system is to assure that all mobile stations
achieve the same received power levels at the base station.
References 211

The reverse link power control includes the open-loop power control and the closed-loop
power control. The open-loop power control is too slow to counter fast fading due to multipath.
The closed-loop power control provides correction to the open-loop power control. It begins
after acquiring the traffic channel and is directed by the base station. The closed-loop power con-
trol occurs every 1.25 ms and is much faster and more effective than the open-loop power con-
trol. With the closed-loop power control, power can change ±16 dB per 20-ms frame.

10.14 References
1. Garg, V. K., Smolik, K. F., and Wilkes, J. E., Application of CDMA in Wireless/Personal
Communications, Prentice Hall PTR, Upper Saddle River, NJ, 1997.
2. TIA/EIA IS-95A, “Mobile Station–Base Station Compatibility Standard for Dual-Mode
Wideband Spread Spectrum Cellular Systems,” May 1995.
3. TIA/EIASP-3693, “Mobile Station–Base Station Compatibility Standard for Dual-Mode
Wideband Spread Spectrum Cellular Systems,” November 18, 1997.
4. Wang, S. W., and Wang, I., “Effects of Soft Handoff, Frequency Reuse and Non-Ideal
Antenna Sectorization on CDMA System Capacity,” Proc. IEEE VTC, Secaucus, NJ, May
1993, pp. 850–54.
5. Wong, Daniel, and Lim, T. J., “Soft Handoff in CDMA Mobile Systems,” IEEE Personal
Communications, 4(6), December 1997.
This page intentionally left blank
 C H A P T E R 1 1

Security and
Identification in IS-95
CDMA

11.1 Introduction
This chapter looks at various parameters that are used to identify a mobile station, including
International Mobile Station Identity (IMSI), Mobile Directory Number (MDN), Electronic
Serial Number (ESN), and station class mark. Then the focus shifts to authentication proce-
dures—the authentication of MS registration, of MS originations, of MS terminations, of MS
data bursts, and of Temporary Mobile Station Identity (TMSI) assignment—and unique chal-
lenge response procedures. This chapter also covers the procedure to update Shared Secret Data
(SSD) as well as the parameter update and voice privacy procedures.

11.2 Mobile Identification Parameters

11.2.1 Mobile Station Identification Number
Mobile stations operating in the CDMA mode are identified by International Mobile Sta-
tion Identity (IMSI). The IMSI contains up to 15 numerical characters (0–9). The first three dig-
its of the IMSI are the Mobile Country Code (MCC), and the remaining digits are the National
Mobile Station Identity (NMSI). The NMSI consists of the Mobile Network Code (MNC) and
Mobile Station Identification Number (MSIN). The IMSI structure is shown in Fig. 11-1.
An IMSI that is 15 digits in length is called a Class 0 IMSI (the NMSI is 12 digits in
length); an IMSI that is less than 15 digits in length is called a Class 1 IMSI (the NMSI is less
than 12 digits in length).
IMSI_M is an IMSI that contains a Mobile Identification Number (MIN) in the lower 10
digits of the NMSI. An IMSI_M can be a Class 0 or Class 1 IMSI.
IMSI_T is an IMSI that is associated with the MIN assigned to the mobile. An IMSI_T
can be a Class 0 or Class1 IMSI. When operating in the CDMA mode, MS sets its operational
IMSI value, IMSI_O, to either IMSI_M or IMSI_T depending on the capabilities of the BS.

213
214 Ch. 11 • Security and Identification in IS-95 CDMA

MCC MNC MSIN

MCC = Mobile Country Code
3 digits MNC = Mobile Network Code
NMSI MSIN = Mobile Station Identification Number
NMSI = National Mobile Station Identity
IMSI (less than or equal to 15 digits) IMSI = International Mobile Station Identity

Figure 11-1 IMSI Structure

An IMSI_S is a 10-digit (34-bit) number derived from an IMSI. When an IMSI has 10 or
more digits, IMSI_S is equal to the last 10 digits. When an IMSI has fewer than 10 digits, the
least significant digits of IMSI_S are equal to the IMSI and 0s are added to the most significant
side to obtain a total of 10 digits. A 10-digit IMSI_S consists of 3- and 7-digit parts, called
IMSI_S2 and IMSI_S1, respectively, as shown in Fig. 11-2.
The IMSI_S derived from IMSI_M is designated IMSI_M S. The IMSI_S derived from
IMSI_T is designated IMSI_T S. The IMSI_S derived from IMSI_O is designated IMSI_O S.
When an IMSI has 12 or more digits, IMSI_11 12 is equal to the 11th and 12th digits of the
IMSI. When an IMSI has fewer than 12 digits, digits with value equal to 0 are added to the
most significant side to obtain a total of 12 digits and the IMSI_11 12 is equal to the 11th and
12th digits of the resulting number. For encoding various types of IMSI, refer to the IS-95B
standards [5].

11.2.2 Mobile Directory Number

A Mobile Directory Number (MDN) is a dialable number associated with the mobile sta-
tion through a service subscription. An MDN is not necessarily the same as the mobile station’s
identification on the air interface, i.e., MIN, IMSI_M, or IMSI_T. An MDN consists of up to 15
digits. The MS should have memory to store at least one MDN.

11.2.3 Electronic Serial Number (ESN)

The ESN is a 32-bit binary number that uniquely identifies the mobile station to any wire-
less system.

IMSI_S2 IMSI_S1
IMSI_S Digits First 3 Digits Second 3 digits Thousand digits Last 3 digits

xxx xxx x xxx

Bits 10 10 4 10
Figure 11-2 IMSI_S Binary Mapping
Mobile Identification Parameters 215

11.2.4 Station Class Mark

Class-of-station information referred to as the station class mark is stored in an MS. The
digital representation of this class mark is given for Band Class 0 and Band Class 1 in Tables 11-1
and 11-2.

11.2.5 Registration Memory

The MS should have memory to store one element in the zone-based registration list. This
stored element includes both REG_ZONE and the corresponding (SID, NID) pair. The data
should be retained for at least 48 hours under power-off conditions. If, after 48 hours, the data
integrity cannot be guaranteed, then the entry should be deleted upon power-on.
The MS should also have memory to store one element in the system/network registration
list SID_NID_LIST. The data should be retained for at least 48 hours under power-off condi-
tions. If, after 48 hours, the data integrity cannot be guaranteed, then the entry in
SID_NID_LIST should be deleted upon power-on.

Table 11-1 Station Class Mark for Band Class 0

Function Bits Setting
Extended SCM indicator 7 always 0 0xxxxxxx
Dual mode 6 CDMA only x0xxxxxx
dual mode x1xxxxxx
Slotted class 5 nonslotted xx0xxxxx
slotted xx1xxxxx
IS-54 power class 4 always 0 xxx0xxxx
25-MHz bandwidth 3 always 1 xxxx1xxx
Transmission 2 continuous xxxxx0xx
discontinuous xxxxx1xx
Power class 1–0 Class I xxxxxx00
Class II xxxxxx01
Class III xxxxxx10
Class IV xxxxxx11

Table 11-2 Station Class Mark for Band Class 1

Function Bits Setting
Extended SCM indicator 7 always 1 1xxxxxxx
Reserved 6 always 0 x0xxxxxx
Slotted class 5 nonslotted xx0xxxxx
slotted xx1xxxxx
Reserved 4 – 0s all 0s xxx00000
216 Ch. 11 • Security and Identification in IS-95 CDMA

The MS should have memory to store the distance-based registration variables

BASE_LAT_REG, BASE_LONG_REG, and REG_DIST_REG. The data should be retained for
at least 48 hours under power-off conditions. If, after 48 hours, the data integrity cannot be guar-
anteed, then REG_DIST_REG should be set to 0 upon power-on.

11.2.6 Access Overload Class

The 4-bit access overload class (ACCOLC) indicator is used to identify which overload
class controls access attempts by the MS; it is also used to identify redirected overload classes in
global service redirection. For mobile stations that are classified as overload classes ACCOLC 0
through ACCOLC 9, the MS’s 4-bit ACCOLC indicator is derived from the last digit of the asso-
ciated decimal representation of the IMSI_M by a decimal-to-binary conversion (see Table 11-
3). When a mobile station’s IMSI_M is updated, the mobile station must recalculate the
ACCOLC as indicated above. Mobile stations designated for test use are assigned to ACCOLC
10; mobile stations designated for emergency use are assigned to ACCOLC 11. ACCOLC 12
through ACCOLC 15 are reserved. The 4-bit ACCOLC indicators for overload Classes 10
through 15 are specified in Table 11-4.

11.2.7 Home System and Network Identification

In addition to the HOME_SID parameter that the mobile stores for 800-MHz analog oper-
ation, the mobile station stores at least one home (SID, NID) pair. The mobile station stores the
1-bit parameters MOB_TERM, MOB_TERM_FOR_SID, and MOB_TERM_FOR_NID.

11.2.8 Local Control Option

When a mobile station supports the local control option, a means is provided within the
mobile station to enable or disable the local control option.

Table 11-3 ACCOLC Indicator Mapping for Overload Classes 0 through 9

Last Digit of the Decimal Representation of the IMSI ACCOLC
0 0000
1 0001
2 0010
3 0011
4 0100
5 0101
6 0110
7 0111
8 1000
9 1001
Authentication Procedures 217

Table 11-4 ACCOLC Indicator Mapping for Overload Classes 10 through 15

Overload Class ACCOLC
10 1010
11 1011
12 1100
13 1101
14 1110
15 1111

11.2.9 Preferred Operation Selection

When the mobile supports operation in Band Class 0, a means is provided within the
mobile to identify the preferred system as either system A or system B. In addition, the mobile
station may provide a means to allow operation with only system A or system B.
When the mobile station supports operation in Band Class 0, means may be provided
within the mobile station to identify the preferred operation type—either CDMA or analog
mode. In addition, the mobile station may provide a means to allow operation with only the ana-
log or CDMA mode.

11.2.10 Discontinuous Reception

The mobile station provides memory to store the preferred slot cycle index:
SLOT_CYCLE_INDEX.

11.3 Authentication Procedures

In the authentication process, information is exchanged between an MS and a BS to confirm the
identity of the MS. A successful authentication process occurs only when it is demonstrated that
the MS and BS possess identical sets of Shared Secret Data (SSD).
The MS uses the operational IMSI (IMSI_O) for authentication purposes, and the BS uses
the IMSI associated with the last MS registration. Table 11-5 summarizes the input parameters
of the AUTH_SIGNATURE procedure for each of its uses in IS-95 standards.
SSD is a 128-bit quantity that is stored in semipermanent memory in the MS and is readily
available to the BS. The SSD is not passed across the air interface between the MS and the net-
work nor across the MSC-BS interface. As shown in Fig. 11-3, SSD is divided into two sub-
sets—SSD_A and SSD_B. SSD_A is used to support the authentication procedures and SSD_B
is used to support voice privacy and signaling message encryption. SSD is generated using the
procedure discussed in section 11.4.
• Rand Challenge Memory (RAND). When operating in 800-MHz analog mode,
RAND is a 32-bit value held in the MS. When operating in CDMA mode, it is equal to
218 Ch. 11 • Security and Identification in IS-95 CDMA

Table 11-5 AUTH_SIGNATURE Input Parameters

RAND_ SAVE_
Procedure CHALLENGE ESN AUTH_DATA SSD_AUTH REGISTERS
Registration RAND ESN IMSI_S1 SSD_A False
Unique RANDU & 8 ESN IMSI_S1 SSD_A False
challenge LSBs of IMSI_S2
Originations RAND ESN Digits SSD_A True
Terminations RAND ESN IMSI_S1 SSD_A True
MS data bursts RAND ESN Digits SSD_A False
BS challenge RANDBS ESN IMSI_S1 SSD_A_NEW False
TMSI RAND ESN IMSI_S1 SSD_A False
assignment

SSD_A SSD_B

64 bits 64 bits
Figure 11-3 Partitioning of SSD

the RAND value received in the last Access Parameters message of the CDMA paging
channel. RAND is used along with SSD_A and other parameters, as appropriate, to
authenticate MS originations, terminations, and registrations.
• Call History Parameter (COUNT). COUNT is a modulo-64 count held in the MS and
is updated by the MS when a Parameter Update order is received on the CDMA for-
ward traffic channel.
• A-Key. A-key is 64 bits long and is assigned to the mobile station (Fig. 11-4). It is
stored in the mobile station’s permanent security identification memory and is known
only to the mobile station and its associated HLR/AC.
• Temporary Mobile Station Identity (TMSI). TMSI is a temporary, locally assigned
number used to address the MS. The MS obtains a TMSI when assigned by the BS. As
a number, the TMSI does not have any association with the MS, IMSI, ESN, or direc-
tory numbers, all of which are permanent identifications.

A-Key
64 bits

Figure 11-4 A-key

Authentication Procedures 219

11.3.1 Authentication of MS Registrations

The following authentication procedures are performed when AUTH is set to 01 (standard
authentication mode), and the MS attempts to register by sending a Registration message on the
access channel.
The MS sets the input parameters of the AUTH_SIGNATURE procedure (see Fig. 11-5)
and sets the SAVE_REGISTERS input parameter to FALSE. The MS then executes the
AUTH_SIGNATURE procedure. The 18-bit output AUTH_SIGNATURE is used to fill the
AUTHR field of the Registration message. The RANDC (the 8 most significant bits of the
RAND) and COUNT fields of the message are filled with the current values stored in the MS.
The BS compares the received value of RANDC to the most significant 8 bits of its inter-
nally stored value of RAND. The BS may also compare the received value of COUNT with its
internally stored value associated with the received IMSI/ESN. The BS computes the value of
AUTHR in the same manner as the MS, but it uses its internally stored value of SSD_A. The BS
compares its computed value of AUTHR to the value received from the MS.
If any comparisons fail, the BS may deem the registration attempt unsuccessful, initiate
the unique challenge response procedure, or commence the process of updating the SSD.

11.3.2 Unique Challenge Response Procedure

The unique challenge response procedure is initiated by the BS and can be carried out
either on the paging and access channels or on the forward and reverse traffic channels. The BS
generates the 24-bit RANDU and sends it to the MS in the Authentication Challenge message on
either the paging channel or the forward traffic channel. Upon receipt of the Authentication
Challenge message, the MS sets the input parameters of the AUTH_SIGNATURE procedure
(Fig. 11-6). The 24 most significant bits of the RAND_CHALLENGE input parameter are used
with RANDU, and the 8 least significant bits of RAND_CHALLENGE are used with the 8 least
significant bits of IMSI_S2. The MS sets the SAVE_REGISTERS input parameter to FALSE.
The MS then executes the AUTH_SIGNATURE procedure. The 18-bit output
AUTH_SIGNATURE is used to fill the AUTHU field of the Authentication Challenge Response
message, which is sent to the BS. The BS computes the value of AUTHU in the same way as the
MS, but uses its internally stored value of SSD_A. The BS compares its computed value of

RAND_CHALLENGE ESN AUTH_DATA SSD_AUTH

RAND (32) ESN (32) IMSI_S1 (24) SSD_A (64)

AUTH_SIGNATURE
Procedure
AUTH_SIGNATURE
AUTHR (18)

Figure 11-5 Calculation of AUTHR for Authentication of MS Registration

220 Ch. 11 • Security and Identification in IS-95 CDMA

RAND_CHALLENGE ESN AUTH_DATA SSD_AUTH

IMSI_S2
RAND (24) (8*) ESN (32) IMSI_S1 (24) SSD_A (64)

AUTH_SIGNATURE
Procedure
AUTH_SIGNATURE
AUTHU (18)
*8 least significant bits of IMSI_S2

Figure 11-6 Calculation of AUTHU for Unique Challenge Response Procedure

AUTHU to the value received from the MS. If the comparison fails, the BS may deny further
access attempts by the MS, drop the call in process, or initiate the process of updating SSD.

11.3.3 Authentication of MS Originations

When AUTH is set to 01 and the MS attempts to originate a call by sending an Origination
message on the access channel, the following authentication procedure is performed. The MS
sets the input parameters of the AUTH_SIGNATURE procedure, as shown in Fig. 11-7. The
AUTH_DATA input parameter consists of the last 6 digits contained in the CHAR field of the
Origination message. If fewer than 6 digits are included in the Origination message, the most
significant bits of IMSI_S1 are used to replace the missing digits. IMSI_S1 is used initially to fill
the AUTH_DATA input parameter, and then the last dialed digits entered by the subscriber are
used to replace all or part of this initial value. If a full 6 digits are dialed, the first digit of the six
that were dialed is used as the most significant 4 bits of AUTH_DATA, the second digit is the
next less-significant 4 bits of AUTH_DATA, and so forth. If fewer than 6 digits are dialed, then
the least significant 4 bits of AUTH_DATA are the last dialed digit, the second-to-the-last dialed
digit becomes the next more-significant 4 bits of AUTH_DATA, and so on up to the first of the
dialed digits. The MS sets the SAVE_REGISTERS input parameter to TRUE.
The MS then executes the AUTH_SIGNATURE procedure. The 18-bit output
AUTH_SIGNATURE is used to fill the AUTHR field of the Origination message. The RANDC

RAND_CHALLENGE ESN AUTH_DATA SSD_AUTH

RAND (32) ESN (32) Digits* (24) SSD_A (64)

AUTH_SIGNATURE
Procedure
AUTH_SIGNATURE
AUTHR (18)
*Last 6 digits transmitted by the MS

Figure 11-7 Calculation of AUTHR for Authentication of MS Originations

Authentication Procedures 221

(8 most significant bits of RAND) and COUNT fields of the message are filled with the current
value stored in the MS. The BS compares the received value of the RANDC to the most signifi-
cant 8 bits of the internally stored value of the RAND. The BS may also compare the received
value of COUNT with its internally stored value associated with the received IMSI/ESN. The
BS computes the value of AUTHR in the same manner as the MS, but uses its internally stored
value of SSD_A. The BS compares its computed value of AUTHR to the value received from the
MS. If the comparisons executed at the BS are successful, the BS may initiate the appropriate
channel assignment procedures. After channel assignment, the BS may issue a Parameter Update
order on the forward traffic channel, updating the value of COUNT in the MS. If any of the com-
parisons fail, the BS may deny service, initiate the unique challenge procedure, or commence
the process of updating the SSD.

11.3.4 Authentication of MS Terminations

When AUTH is set to 01 and the MS responds to a page by sending a Page Response mes-
sage on the access channel, the following authentication procedures are performed.
The MS sets the input parameters of the AUTH_SIGNATURE procedure (refer to Fig.
11-8) and the SAVE_REGISTERS input parameter to TRUE. The MS then executes the
AUTH_SIGNATURE procedure. The 18-bit output AUTH_SIGNATURE is used to fill the
AUTHR field of the Page Response message. The RANDC (8 most significant bits of RAND)
and COUNT fields of the message are filled with the current values stored in the MS. The BS
compares the received value of RANDC to the 8 most significant bits of its internally stored
value of RAND. The BS may also compare the received value of COUNT with its internally
stored value associated with the received IMSI/ESN.
The BS computes the value of AUTHR in the same manner as the MS, but uses its inter-
nally stored value of SSD_A. The BS compares its computed value of AUTHR to the value
received from the MS. If the comparisons executed at the BS are successful, the BS initiates the
appropriate channel assignment procedures. After channel assignment, the BS may issue a
Parameter Update order on the forward traffic channel updating the value of COUNT in the MS.
If any of the comparisons fail, the BS may deny service, initiate the unique challenge response
procedure, or commence the process of updating the SSD.

RAND_CHALLENGE ESN AUTH_DATA SSD_AUTH

RAND (32) ESN (32) IMSI_S1 (24) SSD_A (64)

AUTH_SIGNATURE
Procedure
AUTH_SIGNATURE
AUTHR (18)

Figure 11-8 Calculation of AUTHR for Authentication of MS Terminations

222 Ch. 11 • Security and Identification in IS-95 CDMA

11.3.5 Authentication of MS Data Bursts

When AUTH is set to 01 and the MS attempts to send a Data Burst message on the access
channel, the following authentication procedures are performed.
The MS sets the input parameters of AUTH_SIGNATURE procedure (refer to Fig. 11-9).
The AUTH_DATA input is generated by first filling the AUTH_DATA parameter with the 24 bits
of IMSI_S1 and then replacing part or all of the prefilled value with up to six 4-bit digits that are
provided by the procedure (according to BURST_TYPE) requesting the Data Burst message.
The MS sets the SAVE_REGISTERS input parameter to FALSE. The MS then executes
the AUTH_SIGNATURE procedure. The 18-bit output AUTH_SIGNATURE is used to fill the
AUTHR field of the Data Burst message. The RANDC (8 most significant bits of RAND) and
COUNT fields of the message are filled with the current values stored in the MS. The BS com-
pares the received value of RANDC to the 8 most significant bits of its internally stored value of
RAND. The BS may also compare the received value of COUNT with its internally stored value
associated with the received IMSI/ESN.
The BS computes the value of AUTHR in the same way as the MS, but using its internally
stored value of SSD_A and by generating the AUTH_DATA input in the same way as discussed
for the MS. The BS compares its computed value of AUTHR to the value received from the MS.
If comparisons executed at the BS are successful, the BS may process the message. If any of the
comparisons fail, the BS may ignore the message, initiate the unique challenge response proce-
dure or commence the process of updating the SSD.

11.3.6 Authentication of TMSI Assignment

The MS sets the input parameters of AUTH_SIGNATURE procedure (see Fig. 11-5). The
MS sets the SAVE_REGISTERS input parameter to FALSE and executes the AUTH_SIGNATURE
procedure. The 18-bit output AUTH_SIGNATURE is used to fill the AUTHR field of the TMSI
Assignment Completion message. The RANDC (8 most significant bits of RAND) and COUNT
fields of the message are filled with the current values stored in the MS. The BS compares the
received value of RANDC to the 8 most significant bits of its internally stored value of RAND.
The BS may also compare the received value of COUNT with its internally stored value
associated with the received IMSI/ESN. The BS computes the value of AUTHR in the same way

RAND_CHALLENGE ESN AUTH_DATA SSD_AUTH

RAND (32) ESN (32) Digits* (24) SSD_A (64)

AUTH_SIGNATURE
Procedure
AUTH_SIGNATURE
AUTHR (18)
*Digits from the requesting procedure

Figure 11-9 Calculation of AUTHR for Authentication of MS Data Burst

Shared Secret Data 223

as the MS, but uses its internally stored value of SSD_A. The BS compares its computed value
of AUTHR to the value received from the MS.
If any of the comparisons fail, the BS may deem the TMSI assignment unsuccessful, ini-
tiate the unique challenge response procedure, or commence the process of SSD update.

11.4 Shared Secret Data

Air interface procedures have been defined to update the SSD at the MS. A new SSD is gener-
ated at the HLR/AC, which initiates the SSD update procedure. The SSD update procedure
involves exchange of the following MSC-BS messages:
• SSD Update Request message
• Base Station Challenge Order message
• Base Station Challenge Response message
• SSD Update Response message
The call flow for the SSD update procedure, shown in Fig. 11-10, is as follows:
1. The MSC sends an SSD Update Request message to the BS to indicate that the SSD at
the MS needs updating. The update information is in the form of a random number
(RANDSSD).
2. The BS sends an SSD Update Order message to the MS on either the paging channel or
the forward traffic channel. The RANDSSD field of the message contains the same
value used for the HLR/AC computation of SSD.
3. Upon receipt of the SSD Update Order message from the BS, the MS uses the
RANDSSD as input to the SSD_Generation procedure (algorithm) to generate the new
SSD.

MS BS MSC
SSD Update Request
SSD Update Order

Base Station Challenge Order

Base Station Challenge Order

Base Station
Challenge Response
Base Station Challenge
Confirmation Order

SSD Update Confirmation Order

SSD Update Response

Figure 11-10 Call Flow for SSD Update—Successful Case

224 Ch. 11 • Security and Identification in IS-95 CDMA

4. The MS sets SSD_A_NEW and SSD_B_NEW to the outputs of the SSD_Generation

procedure.
5. The MS selects a 32-bit random number (RANDBS) and sends it to the BS in a Base
Station Challenge Order message on the access channel or reverse traffic channel.
6. The BS forwards the Base Station Challenge Order message to the MSC to verify that
the new SSD calculated at the MS is the same as the number in the MSC.
7. On reception of the Base Station Challenge Order message, the MSC uses the new SSD
as input to the algorithm to generate the authentication response signature (AUTHBS)
(Fig. 11-11). The MSC then sends the authentication response signature (AUTHBS) to
the BS in the Base Station Challenge Response message.
8. Upon receipt of the Base Station Challenge Response message from the MSC, the BS
transmits this information in a Base Station Challenge Confirmation Order message to
the MS on the paging channel or the forward traffic channel.

Computation of Shared Secret Data (SSD)

RANDSSD ESN A-Key AAV

56 32 64 8

SSD_Generation
Procedure

SSD_A_NEW SSD_B_NEW
64 64

Computation of AUTHBS

Boolean RANDBS ESN IMSI_S1 SSD_A_NEW

Value 32 32 24 64

AUTH_SIGNATURE
Process

AUTHBS
18

Figure 11-11 Computation of SSD and AUTHBS

Parameter Update 225

9. Upon receipt of the Base Station Challenge Confirmation Order message, the MS com-
pares the received value of AUTHBS to its internally computed value.
10. If the MS receives a Base Station Challenge Confirmation Order message when an SSD
update is not in progress, the MS responds with an SSD Update Rejection order.
11. If the comparison is successful, the MS executes the SSD_update procedure to set
SSD_A and SSD_B to SSD_A_NEW and SSD_B_NEW, respectively. The MS then
sends an SSD Update Confirmation Order message to the BS, indicating a successful
completion of the SSD update.
12. If the comparison is not successful, the MS discards SSD_A_NEW and SSD_B_NEW
and then sends an SSD Update Rejection Order message to the BS indicating unsuc-
cessful completion of the SSD update.
13. Upon receipt of the SSD Update Confirmation Order message, the BS sets SSD_A and
SSD_B to the value received from the HLR/AC.
14. If the MS fails to receive the Base Station Challenge Confirmation Order message
within T64m seconds of the reception of the acknowledgment to the Base Station Chal-
lenge Order message, the MS discards SSD_A_NEW and SSD_B_NEW. The mobile
then terminates the SSD update process.
Message flow for an SSD update is shown in Fig. 11-12.

11.5 Parameter Update

This procedure is performed when the MSC needs to instruct the MS to update the call history
count (COUNT). This is done at the earliest convenient time after a traffic channel is allocated
for either call origination or termination.
The MSC sends a Parameter Update Request message to the BS and starts timer T3220.
Upon receipt of this message, the BS instructs the MS to update its count by sending a Parameter
Update Order. When the MS receives this order, it increments its call history count and immedi-
ately sends the Parameter Update Confirm message to the MSC. Upon receipt of this message,
the MSC increments its count and stops timer T3220.

11.6 Voice Privacy

If provided by the air interface, there can be voice privacy between the BS and MS. In order to
provide this protection, the relevant device used for the voice privacy needs to be loaded with the
appropriate key. This is supplied by the MSC over the MSC-BS interface. Voice privacy cannot
be invoked unless broadcast authentication is activated in the system.
The privacy mode procedure should be completed either before handoff is initiated or after
a handoff operation is complete.
The following call flow indicates voice privacy invoked during an established call (see Fig.
11-13).
226 Ch. 11 • Security and Identification in IS-95 CDMA

SSD Update Order (RANDSSD)

RANDSSD A-Key A-Key RANDSSD

SSD_GENERATION_PROCEDURE SSD_GENERATION_PROCEDURE

SSD_B_NEW SSD_A_NEW
SSD_A_NEW SSD_A_NEW

Base Station Challenge

Order (RANDBS)
RANDBS

AUTH_SIGNATURE AUTH_SIGNATURE

AUTHBS

Base Station Challenge Order

Confirmation (AUTHBS)
AUTHBS = AUTHBS?

SSD Update Order Confirmation

(Success/Failure)

Figure 11-12 SSD Update Message Flow

MS BS MSC

Privacy Mode
Command
Mobile Order
Timer
Mobile Response T3280
Privacy Mode
Complete

Figure 11-13 Privacy Mode Procedure

Summary 227

1. At any point during the call following the receipt of the Assignment Complete message
the MSC may send the Privacy Mode Command message to the BS to specify that pri-
vacy is to be provided for traffic information. The MSC then starts timer T3280.
2. After the radio traffic channel has been acquired, voice privacy can be established when
the BS transmits a voice privacy request order to the MS.
3. The MS performs the required privacy mode procedures and acknowledges the BS with
a voice privacy response order.
4. The BS returns the Privacy Mode Complete message to the MSC to indicate successful
receipt of the Privacy Mode Command message. The MSC stops timer T3280 upon
receipt of the Privacy Mode Complete message.
The Privacy Mode Command Optional BSMAP message may be sent by the MSC to the
BS after receipt of the Assignment Complete message. Its typical application is to specify the
use of encryption/privacy parameters for the call. It may be sent in the following cases:
• To preload the BS with encryption/privacy parameters during call setup
• To enable or disable the use of encryption/privacy during conversation
The preloading of the BS with parameters allows the BS to immediately initiate privacy upon
request by the mobile user or immediately following assignment to a traffic channel.
Where the MSC needs to provide encryption/privacy information to the BS during call
setup, it may place the information in the Privacy Mode Command message or place it in the
Assignment Request message. Which message is chosen for including encryption information is
a BS and MSC manufacturers’ choice.
The privacy mode procedures may be invoked by the MSC during the conversation state to
enable or disable the use of encryption/privacy. This may be initiated by the MSC or sent in
response to a request for privacy by mobile user.

11.7 Summary
This chapter introduced various parameters used to identify a mobile station. Because fraud is a
serious problem in the analog AMPS system, IS-95 CDMA uses cryptographic methods for
combating fraud. At any time during call processing, the MSC can present a unique challenge to
a mobile station to confirm its identity. We discussed the unique challenge response procedures
used by a mobile station.
The chapter described various types of authentication procedures in which information is
exchanged between an MS and a BS to confirm the identity of the MS. A successful authentica-
tion process occurs only when it is demonstrated that the MS and BS possess identical sets of
Shared Secret Data (SSD).
Also discussed was the procedure to update the SSD, including a call flow diagram. The
procedure involves exchange of the SSD update request, BS challenge, BS challenge response,
and SSD update response messages between the MSC and the BS. The chapter concluded with a
228 Ch. 11 • Security and Identification in IS-95 CDMA

presentation of procedures used when the MSC instructs the mobile station to update the call
history count (COUNT) and when voice privacy is requested between the BS and MS.

11.8 References
1. Garg, V. K., and Wilkes, J. E., Wireless and Personal Communications Systems, Prentice
Hall PTR, Upper Saddle River, NJ, 1996.
2. Report of the Joint Experts Meeting on Privacy and Authentication for PCS, Phoenix, Ari-
zona, November 8–12, 1993.
3. TIA Interim Standard, IS-41C, “Cellular Radio Telecommunications Intersystem Opera-
tions,” 1997.
4. TIA IS-95A, “Mobile Station–Base Station Compatibility Standard for Dual Mode Spread
Spectrum Cellular System,” 1992.
5. TIA IS-95B, “Mobile Station–Base Station Compatibility Standard for 800 MHz Cellular
Mobile Telecommunications System and 1.8 to 2.0 GHz CDMA PCS Systems,” 1998.
 C H A P T E R 1 2

RF Engineering and
Network Planning

12.1 Introduction
This chapter presents basic guidelines for engineering a CDMA system. The topic is extremely
complex and cannot be covered extensively in a single chapter. This chapter discusses several
topics that are germaine to the engineering of a CDMA system: propagation models, link bud-
gets, the transition from analog operation to CDMA operation, radio link capacity, facility engi-
neering, border cells on a boundary between two service providers, and interfrequency handoff.

12.2 Radio Design for a Cellular/PCS Network

A designer needs to consider many factors early in the design of a cellular/PCS network for an
urban area. For example, all prospective service providers must carefully evaluate the extent of
radio coverage for indoor locations, the quality of service for different environments, efficient
use of the spectrum, and the evolution of the network. Often, these factors are further compli-
cated by the constraints imposed by the operating environments and regulatory issues. A system
designer must carefully balance all the trade-offs to ensure that the network is robust, future
proof, and of high service quality.

12.3 Radio Network Planning

The grade-of-service (GOS) performance measures include area coverage probability and
blocking. The area coverage probability is related to the quality of network planning and the
network capacity. Blocking is based on the available resources. Area coverage probability can
also be defined by outage. Outage occurs when the network is not able to provide the specified
quality of service. If the system is coverage limited, outage can be defined as the probability
when path loss and shadowing exceed the difference between the maximum transmitted power
and the required received signal level. Coverage and capacity objectives require a trade-off

229
230 Ch. 12 • RF Engineering and Network Planning

between the desired quality and overall network cost. A smaller signal outage probability means
smaller cells and therefore higher overall network cost; smaller interference outage probability
implies lower capacity and thus higher network cost. An outage probability of 5 to 10% corre-
sponding to 90 to 95% coverage probability is often used. The coverage probability could be
different for different services.
Many factors must be included in radio network planning. The network planning should
address issues such as traffic distribution, macro- and micro-cell deployment, provision for
indoor and high-bit-rate coverage, cell locations and cost of sites, and environmental concerns
such as cell tower appearance.

12.4 Radio Link Design

For any wireless communications system, the first important step is to design the radio link. This
is required to determine base station density in different environments as well as the correspond-
ing radio coverage. For a wireless network to provide good-quality indoor and outdoor service in
an urban environment, flexibility and resilience should be incorporated into the design. The
transmit power of handsets will be the determining factor for a CDMA system with balanced up/
downlink power.
Although the mobile antenna gain does not affect the balancing of the link budget, it is an
important factor in the design of the power budget for handset coverage. From a user point of
view, a cellular/PCS network should imply that there is little restriction on making or receiving
calls within a building or in a moving vehicle using handsets. A system should be designed to
allow the antenna of a handset to be placed in nonoptimal positions. In addition, the antenna may
not even be extended when calls are being made or received. In normal system designs, it is
assumed that the gain of a mobile antenna is 0 dBi.* However, allowing for the handset antennas
to be placed in suboptimal positions, a more conservative gain of –3 dBi should be used. In real-
ity, because of the positioning of an antenna in an arbitrary position or with the antenna retracted
into the handset housing antenna gain could be as low as –6 to –8 dBi depending on specific
handsets and their corresponding housing designs.

12.5 Estimation of Cell Count

The number of users and offered traffic load per user are used to determine the overall traffic load.
Knowing the cell capacity and cell coverage, an estimate of the number of cells can be made.
The cell capacity is determined from simulations or analytical formulas. User information
rate, traffic characteristics, QoS requirements (delay, BER/FER), and outage probability are the
important factors in determining the cell capacity.
The link budget is used to determine maximum cell coverage. In addition to Eb /It ,
equipment-specific factors such as cable losses, antenna gains, and the receiver noise figure
are needed for link budget calculations.

* dBi refers to the gain relative to an isotropic antenna.

Radio Coverage Planning 231

Soft-handoff gain has a large impact on link budget. The soft-handoff gain depends on
shadowing correlation and coverage probability. Soft handoff provides macroscopic diversity
gain through increased diversity.The actual gain depends on the radio environment and number
of RAKE receiver fingers. Since each radio environment has its own characteristics, for the
detailed coverage prediction, some correction factors for the path-loss models are needed.
For the reverse link, the impact of load factor ρ in the link budget for interference margin
Im (dB) can be determined from

I m = 10 log  ------------
1
(12.1)
 1 – ρ

Since the interference margin increases with ρ, cell range would decrease with the increas-
ing load factor. Asymmetric traffic load should be taken into consideration in link budget calcu-
lations. CDMA can trade the reverse link capacity for coverage. This is useful since usually the
mobile transmission power limits the maximum cell range.
After the cell count has been obtained, detailed radio network planning can be initiated by
taking into account the exact radio environment where each cell will be located. Due to cost of
sites, zoning requirements, building restrictions, or other reasons, it may not be possible to
achieve optimum cell sites in a real network. This may impact the initial coverage plan. For
detailed network planning, a network planning tool should be used. A network planning tool has
a digital map of the area to be planned. Building heights and antenna pattern are also modeled.
The optimization process of radio network coverage generally includes
• Detailed description of radio environment
• Control channel power planning
• Soft-handoff parameters planning
• Interfrequency handoff planning
• Iterative network coverage analysis
• Network testing

12.6 Radio Coverage Planning

The most important design objective of a cellular/PCS network is to provide near-ubiquitous
radio coverage. One of the most crucial considerations in the radio coverage planning process is
the propagation model. The accuracy of the prediction by a particular model depends on its abil-
ity to account for the detailed terrain, vegetation, and buildings. This accuracy is of vital impor-
tance in determining the path loss and, hence, the cell sizes and the infrastructure requirement of
a cellular/PCS network. An overestimation will lead to an inefficient use of the network
resources, whereas an underestimation will result in poor radio coverage. Propagation models
generally tend to oversimplify real-life propagation conditions and may be grossly inaccurate in
complex metropolitan urban environments. Empirical propagation models only provide general
guidelines and are too simplistic for accurate network design. Accurate field measurements must
be made to provide information on radio coverage in an urban environment. Measured data can
232 Ch. 12 • RF Engineering and Network Planning

be used either directly in the planning process to access the feasibility of individual cell sites or
to calibrate the coefficients of the empirical propagation model to achieve better characterization
of a specific environment.
Radio propagation in an urban environment is subject to shadowing. To ensure that the sig-
nal level in 90% of the cell area is equal to or above the specified threshold, a shadow fading
margin, which is dependent on the standard deviation of the signal level, must be included in the
link budget. For a typical urban environment, a shadow fading margin of 8 to 9 dB should be
used based on the assumption that the path loss follows an inverse 2–5 exponent law—the path
loss is inversely proportional to the distance of separation raised to a power between 2 and 5.
The value of the power is dependent upon propagation characteristics.
Another critical factor that affects radio coverage is the penetration loss for both buildings
and vehicles. If radio coverage for the outer portion of a building is sufficient, then an assumed
penetration loss of 10 to 15 dB should be adequate. However, if calls will be received and origi-
nated within the inner core of the building, a penetration loss of about 30 dB should be used.
Similarly, for in-vehicle coverage, the penetration loss is equally important. A car could experi-
ence a penetration loss of 3 to 6 dB, whereas vans and buses have even larger variations. The
penetration loss at the front of a van should be no more than that experienced in a car, but the
loss at the back of a van could be as high as 10 to 12 dB, depending on the amount of window
space. Thus, the designer should assume a high penetration loss to ensure good service quality.
For an urban environment, because building penetration loss is the dominant factor in designing
the system, in-vehicle penetration will generally be sufficient as a consequence.

12.7 Propagation Models

Designers use propagation models to determine how many cell sites they need to satisfy the cov-
erage requirements for the network. Initial network design typically engineers for coverage.
Later, growth engineers for capacity. Some systems may need to start with wide area coverage
and high capacity and therefore may be at a later stage of the growth curve right from start-up.
The coverage requirement is coupled with traffic loading requirements, which rely on the
propagation model chosen to determine the traffic distribution and the off-loading from an exist-
ing cell site to new cell sites as part of the capacity relief program. The propagation model helps
to determine where the cell sites should be located to achieve an optimal position in the network.
If the propagation model used is not effective in helping to place cell sites correctly, the proba-
bility of incorrectly deploying a cell site into the network is high.
The performance of the network is affected by the propagation model chosen because that
model is used for interference predictions. As an example, if the propagation model is inaccurate
by 4 dB, then Eb /It could be 11 dB or 3 dB (assuming that Eb /It = 17 dB is the design requirement).
Based on traffic loading condition, designing for a high Eb /It level could negatively affect financial
feasibility. On the other hand, designing for a low Eb /It would degrade the quality of service.
The propagation model is also used in other system performance aspects including hand-
off optimization, power-level adjustments, and antenna placements. Although no propagation
Propagation Models 233

model can account for all perturbations experienced in the real world, it is essential to use one or
several models for determining the path losses in the network. Each of the propagation models
being used in the industry has pros and cons. It is through a better understanding of the limita-
tions of each of the models that a good RF engineering design can be achieved in a network.

12.7.1 Modeling for the Outside Environment

12.7.1.1 Analytical Model
The propagation loss between the base station and the mobile station in the outside envi-
ronment has been studied extensively. Propagation loss is generally expressed by the following
expression [2,4,7]:

r
P ( r ) = N ( r 0, σ ) + 10γ log ---- dB (12.2)
r0

where P(r) = loss at distance r relative to the loss at a reference distance r0,
γ = path-loss exponent, and
σ = standard deviation, typically 8 dB.
The second term on the right side of Eq. (12.2) represents a constant attenuation in the outside
environment between the base station and the mobile station. Typically, γ approximately
equals 4, although it may range between 2 (which equals the loss in free space) and 5. If γ is
equal to 4, then the signal will be attenuated 40 dB if the distance increases ten times with
respect to the reference distance. The first term in Eq. (12.2) represents the variation in the loss
about the average path loss. This function is an approximate log-normal distribution with an
average equal to the second term and a standard deviation of approximately 8 dB. It has been
found that this value is applicable for a wide range of radio environments, including urban and
rural areas.

12.7.1.2 Empirical Models

Several empirical models have been suggested and used to predict propagation path losses.
The two most widely used models are the Hata-Okumura model and the Walfisch-Ikegami
Model.

The Hata-Okumura Model [3]

Most of the propagation tools use a variation of the Hata model. Hata’s model is an empir-
ical relationship derived from the technical report made by Okumura [6] so that the results could
be used in computational tools. Okumura’s report consists of a series of charts that have been
used in radio communication modeling. The following are the expressions used in the Hata
model in order to determine the mean loss L50:
Urban area:

L 50 = 69.55 + 26.16 log f c – 13.82 log h b – a ( h m ) + ( 44.9 – 6.55 log h b ) log r dB (12.3)
234 Ch. 12 • RF Engineering and Network Planning

where fc = frequency (MHz),

L50 = mean path loss (dB),
hb = base station antenna height (m),
a(hm) = correction factor for mobile antenna height (dB), and
r = distance from base station (km).
The range of the parameters for which the Hata model is valid is
150 ≤ fc ≤ 1,500 MHz
30 ≤ hb ≤ 200 m
1 ≤ hm ≤ 10 m
1 ≤ r ≤ 20 km
a(hm) is computed as
For a small or medium-sized city:

a ( h m ) = ( 1.1 log f c – 0.7 )h m – ( 1.56 log f c – 0.8 ) dB (12.4)

For a large city:

2
a ( h m ) = 8.29 ( log 1.54h m ) – 1.1 dB, fc ≤ 200 MHz (12.5)

or
2
a ( h m ) = 3.2 ( log 11.75h m ) – 4.97 dB, fc ≥ 400 MHz (12.6)

Suburban area:

fc 2
L 50 = L 50 ( urban ) – 2 log  ------ – 5.4 dB (12.7)
 28

Open area:
2
L 50 = L 50 ( urban ) – 4.78 ( log f c ) + 18.33 log f c – 40.94 dB (12.8)

Hata’s model does not account for any of the path-specific correction used in Okumura’s
model.
Okumura’s model [6] tends to average some of the extreme situations and does not
respond sufficiently quickly to rapid changes in the radio path profile. The distance-dependent
behavior of Okumura’s model is in agreement with the measured values. Okumura’s measure-
ments are valid only for the building types found in Tokyo. Experience with comparable mea-
surements in the United States has shown that the typical U.S. suburban situation is often
somewhere between Okumura’s suburban and open areas. Okumura’s suburban definition is
more representative of a U.S. residential metropolitan area with large groups of row houses.
Propagation Models 235

Okumura’s model requires that considerable engineering judgment be used, particularly in

the selection of the appropriate environmental factors. Data is needed in order to be able to pre-
dict the environmental factors from the physical properties of the buildings surrounding a mobile
receiver. In addition to the appropriate environmental factors, path-specific corrections are
required to convert Okumura’s mean path-loss predictions to the predictions applicable to the
specific path under study. Okumura’s techniques for correction of irregular terrain and other
path-specific features require engineering interpretations and are thus not readily adaptable for
computer use.

The Walfisch-Ikegami (or COST 231) Model [14]

This model is used to estimate the path loss in an urban environment for cellular commu-
nication (see Fig. 12-1). It is a combination of the empirical and deterministic model for estimat-
ing the path loss in an urban environment over the frequency range of 800 MHz to 2000 MHz.
This model is used primarily in Europe for the GSM system and in some propagation models in
the United States. The model contains three elements: free-space loss, rooftop-to-street diffrac-
tion and scatter loss, and multiscreen loss. The expressions used in this model are

L 50 = L f + L rts + L ms (12.9)

or

L 50 = L f when L rts + L ms ≤ 0 (12.10)

where Lf = free-space loss,

Lrst = rooftop-to-street diffraction and scatter loss, and
Lms = multiscreen loss.
Free-space loss is given as

L f = 32.4 + 20 log r + 20 log f c dB (12.11)

The rooftop-to-street diffraction and scatter loss is given as

L rts = – 16.9 – 10 log W + 10 log f c + 20 log ∆h m + L 0 dB (12.12)

where W = street width (m), and

∆hm = hr – hm (m).

L 0 = – 9.646 dB 0 ≤ φ ≤ 35 degrees

L 0 = 2.5 + 0.075 ( φ – 35 ) dB 35 ≤ φ ≤ 55 degrees

L 0 = 4 – 0.114 ( φ – 55 ) dB 55 ≤ φ ≤ 90 degrees

where φ = incident angle relative to the street.

236 Ch. 12 • RF Engineering and Network Planning

R
∆hb θ

hb ∆hm
hR hm

Cell Site b w Mobile

Building Building
Mobile
φ

Building Building

Figure 12-1 The Walfisch-Ikegami Propagation (COST 231) Model

The multiscreen loss is given as

L ms = L bsh + k a + k d log r + k f log f c – 9 log b (12.13)

where b = distance between buildings along the radio path (m),

Lbsh = –18log11 + ∆hb, hb > hr ,
Lbsh = 0, hb < hr ,
ka = 54, hb > hr ,
ka = 54 – 0.8hb, r ≥ 500 m, hb ≤ hr ,
ka = 54 – 1.6∆hbr, r < 500 m, hb ≤ hr ,
kd = 18, hb < hr ,
15∆h
kd = 18 – --------------b- , hb ≥ hr ,
∆h m
fc
kf = 4 + 0.7  -------- - – 1 for midsize city and suburban area with moderate tree density, and
 925 
fc
kf = 4 + 1.5  -------- - – 1 for metropolitan center.
 925 
Note: Both Lbsh and ka increase the path loss with lower base station antenna heights.
The range of parameters for which the Walfisch-Ikegami model is valid is
800 ≤ fc ≤ 2000 (MHz)
4 ≤ hb ≤ 50 (m)
Propagation Models 237

1 ≤ hm ≤ 3 (m)
0.02 ≤ r ≤ 5 (km)
The following default values can be used for the model:
b = 20–50 (m)
W = b/2
φ = 90 degrees
Roof = 3 m for pitched roof and 0 m for flat roof
hr = 3 (number of floors) + roof
Table 12-1 uses the following data to show a comparison of the path loss from the Hata
and Walfisch-Ikegami models.
fc = 880 MHz; hm = 1.5 m; hb = 30 m; roof = 0 m; hr = 30 m; φ = 90 degrees; b = 30 m; and
W = 15 m.
The path losses predicted by Hata’s model are about 13 to 16 dB lower (see Fig. 12-2) than
those predicted by the Walfisch-Ikegami model. Hata’s model ignores effects from street width,
street diffraction, and scatter losses, which the Walfisch-Ikegami model includes.
• Correction Factor for Attentuation Due to Trees. Weissberger [15] has developed a
modified exponential delay model that can be used where a radio path is blocked by
dense, dry, in-leaf trees found in temperate climates. The additional path loss can be
calculated from the following expression:
0.284 0.588
L t = 1.33 ( f c ) (d f ) dB, for 14 ≤ d f ≤ 400 m (12.14)
0.284
= 0.45 ( f c ) d f dB, for 0 ≤ d f ≤ 14 m (12.15)
where Lt = loss in dB,
fc = frequency in GHz, and
df = tree height in meters.

Table 12-1 A Comparison of Path Loss from Hata and Walfisch-Ikegami Models
Path Loss (dB)
Distance (km) Hata’s Model Walfisch-Ikegami Model
1 126.16 139.45
2 136.77 150.89
3 142.97 157.58
4 147.37 162.33
5 150.79 166.01
238 Ch. 12 • RF Engineering and Network Planning

180
Path Loss in dB

160

140

120
0 1 2 3 4 5 6
Distance from Transmitter (km)
Walfisch-Ikegami (COST231) Model
Hata Model

Figure 12-2 Comparison of Hata and Walfisch-Ikegami Models (Path Loss vs. Distance)

The difference in path loss for trees with and without leaves has been found to be about 3
to 5 dB. For a frequency of 900 MHz, the above equations are reduced to
0.588
L t = 1.291 ( d f ) dB, for 14 ≤ d f ≤ 400 m (12.16)

= 0.437d f dB, for 0 ≤ d f ≤ 14 m (12.17)

12.7.2 Models for Indoor Environment

Experimental studies have indicated that a portable receiver moving in a building experi-
ences Rayleigh fading for obstructed propagation paths and Ricean fading for line-of-sight
(LOS) paths, regardless of the type of building. Rayleigh fading is short-term fading resulting
from signals traveling separate paths (multipath) that partially cancel each other. An LOS path is
clear of building obstructions; in other words, there are no reflections of the signal. Ricean fad-
ing results from the combination of a strong LOS path and a ground path plus numerous weak
reflected paths. A TIA IS-95A mobile station, however, is capable of discerning signals that
travel different paths since a RAKE receiver is incorporated. In fact, TIA IS-95A does not
address equalization of delay spread on the radio link, and thus the mobile station’s receiver
does not have an equalizer.
Quantification of propagation between floors is important for an in-building wireless sys-
tem of multifloor buildings that need to share frequencies within the building. Frequencies are
reused on different floors to avoid cochannel interference. The type of building material, the
aspect ratio of building sides, and types of windows have shown to impact the RF attenuation
Propagation Models 239

between floors. Measurements have indicated that loss between floors does not increase linearly
in dB with increasing separation distance. The greatest floor attenuation in dB occurs when the
receiver and transmitter are separated by a single floor. The overall path loss increases at a
smaller rate as the number of floors increases. Typical values of attenuation between floors is 15
dB for one floor of separation and an additional 6 to 10 dB per floor separation up to four floors
of separation. For five or more floors of separation, path loss will increase by only a few dB for
each additional floor (see Table 12-2).
The signal strength received inside a building due to an external transmitter is important
for wireless systems that share frequencies with neighboring buildings or with an outdoor sys-
tem. Experimental studies have shown that the signal strength received inside a building
increases with height [9]. At lower floors of a building, the urban cluster induces greater atten-
uation and reduces the level of penetration. At higher floors, an LOS path may exist, thus caus-
ing a stronger incident signal at the exterior wall of the building. RF penetration is found to be
a function of frequency as well as height within a building. Penetration loss decreases with
increasing frequency. Measurements made in front of a window showed 6 dB less penetration
loss on coverage than those measurements made in parts of the building without windows.
Experimental studies also showed that building penetration loss decreased at a rate of about 2
dB per floor from ground level up to the 10th floor and then began to increase around the 10th
floor. The increase in penetration loss at the higher floors was attributed to shadowing effects of
adjacent buildings.

Table 12-2 Mean Path-Loss Exponents and Standard Deviations

Type γ σ (dB)
All buildings All locations 3.14 16.3
Same floor 2.76 12.9
Thru 1 floor 4.19 5.1
Thru 2 floors 5.04 6.5
Office building 1 Entire building 3.54 12.8
Same floor 3.27 11.2
West wing 5th floor 2.68 8.1
Central wing 5th floor 4.01 4.3
West wing 4th floor 3.18 4.4
Grocery store 1.81 5.2
Retail store 2.18 8.7
Office building 2 Entire building 4.33 13.3
Same floor 3.25 5.2
240 Ch. 12 • RF Engineering and Network Planning

The mean path loss is a function of distance to the γth power [7].

L 50 ( r ) = L ( r 0 ) + 10 × γ log  ---- dB
r
(12.18)
 r 0

where L50(r) = mean path loss (dB),

L(r0) = path loss from transmitter to reference distance r0 (dB),
γ = mean path-loss exponent,
r = distance from the transmitter (m), and
r0 = reference distance from the transmitter (m).
We choose r0 equal to 1 m and assume L(r0) due to free-space propagation from the transmitter
to a 1-m reference distance. Next, we assume the antenna gain equals the system cable losses*
and get a path loss, L(r0), of 31.7 dB at 914 MHz over a 1-m free-space path.
The path loss was found to be log-normally distributed about Eq. (12.18). The mean path-
loss exponent γ and standard deviation σ are the parameters that depend on building type, building
wing, and number of floors between the transmitter and receiver. The path loss at a transmitter-
receiver (T-R) separation of r meters can be given as

L ( r ) = L 50 ( r ) + X σ dB (12.19)

where L(r) = path loss at a T-R separation distance r meters, and

Xσ = 0 mean log-normally distributed random variable with standard deviation σ dB.
Table 12-2 gives a summary of the mean path-loss exponents and standard deviation about
the mean for different environments [9].
In a multifloor environment, Eq. (12.18) can be modified to emphasize that the mean path-
loss exponent is a function of the number of floors between the transmitter and the receiver. The
value of γ (multifloor) is given in Table 12-2.

L 50 ( r ) = L ( r 0 ) + 10 × γ ( multifloor ) log  ----

r
(12.20)
 r 0

Another path-loss prediction model suggested in [7] uses a Floor Attenuation Factor
(FAF). A constant floor attenuation factor (in dB), which is a function of the number of floors
and building type, was included in the mean path loss predicted by a path-loss model that uses
the same-floor path-loss exponent for the particular building type.

L 50 ( r ) = L ( r 0 ) + 10 × γ ( same-floor ) log  ---- + FAF

r
dB (12.21)
 r 0

where r is in meters and L(r0) = 31.7 dB at 914 MHz.

* This is obviously not always true.

Propagation Models 241

Table 12-3 Average Floor Attenuation Factor

Type FAF (dB) σ

Office building 1 Thru 1 floor 12.9 7.0
Thru 2 floor 18.7 2.8
Thru 3 floor 24.4 1.7
Thru 4 floor 27.0 1.5
Office building 2 Thru 1 floor 16.2 2.9
Thru 2 floor 27.5 5.4
Thru 3 floor 31.6 7.2

Table 12-3 provides the floor attentuation factors and the standard deviation (in dB) of the
difference between the measured and predicted path loss. Values for the floor attenuation factor
in Table 12-3 are an average (in dB) of the difference between the path loss observed at multi-
floor locations and the mean path loss predicted by the simple rγ model (Eq. [12.18]), where γ is
the same-floor exponent listed in Table 12-2 for the particular building structure and r is the
shortest distance, measured in three dimensions, between the transmitter and receiver.

12.7.2.1 Soft-Partition and Concrete-Wall Attenuation Factor Model

The path-loss effects of soft partitions and concrete walls (in dB) between the transmitter
and receiver for the same floor were modeled in [5] and have been given as

L 50 ( r ) = 20 log  --------- + p × AF ( soft-partition ) + q × AF ( concrete-wall )

4πr
(12.22)
 λ 

where p = number of soft partitions between the transmitter and receiver,

q = number of concrete walls between the transmitter and receiver,
λ = wavelength (m),
AF = 1.39 dB for each soft partition, and
AF = 2.38 for each concrete wall.

EXAMPLE 12.1
Use the two models (Eqs. [12.20] and [12.21]) to predict the mean path loss at a distance r = 30 m
through three floors of an office building; assume the mean path-loss exponent for same-floor mea-
surements in the building is γ = 3.27, the mean path-loss exponent for three-floor measurements is γ
= 5.22, and the average FAF is 24.4 dB.
From Eq (12.20):

L 50 ( 30 ) = 31.7 + 10 × 5.22 log  ------ = 108.8 dB

30
 1
From Eq. (12.21):
242 Ch. 12 • RF Engineering and Network Planning

L 50 ( 30 ) = 31.7 + 10 × 3.27 log  ------ + 24.4 = 104.4 dB

30
 1

The results obtained by the two models are fairly close.

12.7.3 IMT-2000 Models

Since ITU IMT-2000 will be the worldwide standard, the proposed models to evaluate
radio transmission technologies consider a broad range of environmental characteristics includ-
ing large and small cities, suburbs, tropical, rural, and desert areas. IMT-2000 operating environ-
ments have been identified by an appropriate subset consisting of indoor office environment,
outdoor-to-indoor and pedestrian environment, and vehicular environment. The key parameters
of each propagation model are:
• delay spread, its structure, and its statistical variation (see section 12.8)
• geometrical path-loss rule and excess path loss
• shadow fading
• multipath fading characteristics (e.g., Doppler spectrum, Rician vs. Rayleigh) for the
envelope of channels
• operating radio frequency

12.7.3.1 Indoor Office Environment Model

This environment is characterized by small cells and low transmit powers. The base sta-
tions and pedestrian users are located indoors. Root-mean-square (RMS) delay spread ranges
from around 35 ns to 460 ns. The path-loss rule varies due to scatter and attenuation by walls,
floors, and metallic structures such as partition and filing cabinets. These objects also produce
shadowing effects. A log-normal shadowing with a standard deviation of 12 dB can be expected.
The fading characteristic ranges from Rician to Rayleigh, with Doppler frequency offsets set by
walking speeds. The path-loss model for this environment is
[ ( F + 2 ) ⁄ ( F + 1 ) – 0.46 ]
L 50 = 37 + 30 log r + 18.3F (12.23)

where r = separation between transmitter-receiver (m), and

F = number of floors in the path.

12.7.3.2 Outdoor-to-Indoor and Pedestrian Environment

This environment is characterized by small cells and low transmit power. Base stations
with low antenna heights are located outdoors; pedestrian users are located on streets and inside
buildings and residences. Coverage into buildings in high-power systems is included in the
vehicular environment discussed in the next section. RMS delay spread varies from 100 to 1800
ns. A geometrical path-loss rule of r–4 is applicable. If the path is a line of sight on a canyon-like
street, the path loss follows an r–2 rule, when there is Fresnel zone clearance. For the region with
longer Fresnel zone clearance, a path-loss rule of r–4 is appropriate, but a range of up to r–6 may
be encountered due to trees and other obstructions along the path. Log-normal shadow fading
Delay Spread 243

with a standard deviation of 10 dB for outdoors and 12 dB for indoors is reasonable. Average
building penetration loss of 18 dB with a standard deviation of 10 dB is appropriate. Rayleigh
and/or Rician fading rates are generally set by walking speeds, but faster fading due to reflec-
tions from moving vehicles may occur sometimes. The following path-loss model has been sug-
gested for use in this environment:

L 50 = 40 log r + 30 log f c + 49 dB (12.24)

where fc = carrier frequency (MHz).

This model is valid for a no-LOS case only and describes worst-case propagation with log-
normal shadow fading with a standard deviation of 10 dB. The average building penetration loss
is 18 dB with a standard of 10 dB.

12.7.3.3 Vehicular Environment

This environment consists of larger cells and higher transmit power. RMS delay spread
from 0.4 ms to about 12 ms may occur on elevated roads in hilly or mountainous terrain. A geo-
metrical path-loss rule of r–4 and log-normal shadow fading with a standard deviation of 10 dB
are used in urban and suburban areas. Building penetration loss averages 18 dB with a standard
deviation of 10 dB. In rural areas with flat terrain, the path loss is lower than that of urban and
suburban areas. In mountainous terrain, if path blockages are avoided by placement of base sta-
tions, the path-loss rule is closer to r–2. Rayleigh fading rates are set by vehicle speeds. Lower
fading rates are appropriate for applications employing stationary terminals. The following
model is used for this environment:
–2
L 50 = 40 ( 1 – 4 × 10 ∆h b ) log r – ( 18 ⋅ log ∆h b ) + 21 ⋅ log f c + 80 dB

where r = base station and mobile separation (km),

fc = carrier frequency (MHz), and
∆hb = base station antenna height measured from average rooftop level (m).

12.8 Delay Spread

The radio signal follows different paths because of multipath reflection. Each path has a different
path length, so the time of arrival for each path is different. The effect, which smears or spreads
out the signal, is called delay spread. As an example, if an impulse is transmitted by the transmit-
ter, by the time this impulse is received at the receiver, it is no longer an impulse but rather a
pulse that is spread (refer to Fig. 12-3). In a digital system, the delay spread causes intersymbol
interference, thereby limiting the maximum symbol rate of a digital multipath channel [8].
The mean delay spread τd is
∞
∫0 tD ( t ) dt
τ d = ------------------------
∞
(12.25)
∫ D ( t ) dt
0
244 Ch. 12 • RF Engineering and Network Planning

Initial Received
Transmitted 3
Pulse
Pulses

1
Base Station
Antenna
4

2 N: 4

N >> 4

t
τd

Figure 12-3 Delay Spread Phenomenon

where D(t) is the delay probability density function, and

∞
∫0 D ( t ) dt = 1.

Some representative delay functions are

• Exponential
t
– -----
1 τd
D ( t ) = ----- e
τd

• Uniform

τ
D ( t ) = ----d- , 0 ≤ t ≤ 2τ d
2

D ( t ) = 0 , elsewhere
The measured data suggest that the mean delay spreads are different in different environ-
ments (refer to Table 12-4).

Table 12-4 Measured Data for Delay Spread

Type of Environment Delay Spread τd (µs)

Open area < 0.2
Suburban area 0.5
Urban area 3
Delay Spread 245

A majority of the time, the RMS delay spreads are relatively small, but, occasionally, there
are worst-case multipath characteristics that lead to much larger RMS delay spreads. Measure-
ments in outdoor environments show that RMS delay spread can vary over an order of magni-
tude within the same environment. Delay spreads can have a major impact on system
performance. To accurately evaluate the relative performance of radio transmission technolo-
gies, it is important to model the variability of delay spread as well as the worst-case locations
where delay spread is relatively large. Three multipath channels are defined by IMT-2000 for
each environment. Channel A represents the low-delay-spread case that occurs frequently; chan-
nel B corresponds to the medium-delay-spread case that also occurs frequently; channel C is the
high-delay-spread case that occurs only rarely. Table 12-5 provides the average RMS delay
spread values for each channel and for each environment.

12.8.1 Coherence Bandwidth

The coherence bandwidth Bc is a statistical measure of the range of frequencies over which
the channel passes all spectral components with approximately equal gain and linear phase. The
coherence bandwidth represents a frequency range for which either the amplitudes or phases of
two received signals have a high degree of correlation. A signal’s spectral components in that
frequency range are affected by the channel in a similar manner—for example, exhibiting fading
or no fading. As an approximation

B c ≈ 1 ⁄ τ dmax (12.26)

where τdmax = maximum delay spread.

The maximum delay spread, τdmax , is not necessarily the best indicator of how any given
system will perform on a channel because different channels with the same value of τdmax may
exhibit very different profiles of signal intensity over the delay span. A more useful measurement
is often expressed in terms of the RMS delay spread, τdrms. An exact relationship between coher-
ence bandwidth and delay spread does not exist. Several approximations have been proposed.
The coherence bandwidth is defined as that bandwidth where the correlation function
|RT (Ω)| = 0.5, for two fading signal envelopes at two frequencies f1 and f2, respectively, where

Table 12-5 Average RMS Delay Spread Values (IMT-2000)

Channel A Channel B Channel C
Environment τrms (ns) % Occurrence τrms (ns) % Occurrence τrms (ns) % Occurrence
Indoor office 35 50 100 45 460 5
Outdoor to indoor 100 40 750 55 800 5
and pedestrian
Vehicular high 400 40 4000 55 12,000 5
antenna
246 Ch. 12 • RF Engineering and Network Planning

∆f = f 1 – f 2 (12.27)

Two frequencies that are farther apart than the coherence bandwidth, Bc , will fade inde-
pendently. This concept is also useful for diversity reception.
The coherence bandwidth for two fading amplitudes of two received signals is

1
∆ f > B c = ------------------ (12.28)
2πτ drms

If coherence bandwidth is defined as the frequency interval over which the channel’s complex
frequency transfer function has a correlation of at least 0.9, the coherence bandwidth is approxi-
mately given as

1
B c ≈ ----------------- (12.29)
50τ drms

For mobile radio, the model generally accepted as a useful model for the urban environ-
ment has an array of radially uniformly spaced scatters, all with equal magnitude reflection coef-
ficients but independent, randomly occurring reflection phase angles. This model is referred to
as the dense-scatter channel model. Using such a model, coherence bandwidth has been defined
for a bandwidth interval over which the channel’s complex frequency transfer function has a cor-
relation of at least 0.5 [10,11], or

0.276
B c ≈ ------------- (12.30)
τ drms

A more popular approximation of Bc corresponding to a bandwidth interval with a correlation of

at least 0.5 is

1
B c ≈ -------------- (12.31)
5τ drms

A channel is a frequency-selective channel if Bc < 1/Ts = Bw , where the symbol rate 1/Ts is
nominally taken to be equal to the signal bandwidth Bw . Frequency-selective fading distortion
occurs when a signal’s spectral components are not all affected equally by the channel.
Frequency-nonselective or flat-fading degradation occurs whenever Bc > Bw . Hence, all of
the signal’s spectral components are affected by the channel in a similar manner. Flat fading
does not introduce channel-induced intersymbol interference (ISI) distortion, but performance
degradation can still be expected due to loss in SNR whenever the signal is fading. In order to
avoid channel-induced ISI distortion, the channel must be a flat-fading channel by ensuring that
1
B c > B w = ----- (12.32)
Ts
Thus, the channel coherence bandwidth sets an upper limit on the transmission rate that
can be used without incorporating an equalizer in the receiver.
Doppler Spread 247

The GSM symbol rate (or bit rate, since modulation is binary) is 271 ksps and the band-
width is Bw = 200 kHz. If the typical RMS delay spread in an urban environment is τdrms = 2µs,
then using Eq. (12.31) the coherence bandwidth Bc ≈ 100 kHz. It is therefore apparent that since
Bc < Bw , the GSM receiver must use some form of mitigation to overcome frequency-selective
distortion. To accomplish this goal, the Viterbi equalizer [1] is typically used.

EXAMPLE 12.2
Assume vehicle speed equal to 60 mph (88 ft/sec), carrier frequency fc = 860 MHz, and delay spread
τdrms = 2 µsec. Calculate coherence time and coherence bandwidth. At a coded symbol rate of 19.2
kbps (IS-95), what kind of symbol distortion will be experienced? What type of fading will be expe-
rienced by the IS-95 channel?
v = 60 mph (= 88 ft/sec)
8
c 9.84 × 10
λ = --- = ------------------------6- = 1.1442 ft/sec
f 860 × 10
v 88
f m = --- = ---------------- = 77 Hz
λ 1.1442
1 1
T c = ------------- = ------------------ = 0.0021 seconds
2π f m 2π × 77
6
10
T s = ---------------- = 52 µsec
19,200
The symbol interval is much smaller than the channel coherence time. Therefore, symbol dis-
tortion is minimal. In this case fading is slow.
1 1
B c ≈ ----------------- = -------------------------------
- = 79.56 kHz
2πτ drms 2π × 2 × 10
–6

This shows that IS-95 is a wideband system in this multipath situation and experiences selec-
tive fading over only 6.5% ( 79.57/1228.8 = 0.0648) of its bandwidth.

12.9 Doppler Spread

The width of the Doppler power spectrum is called spectral broadening or Doppler spread, is
denoted fd, and is sometimes called the fading bandwidth of the channel. The Doppler shift of
each arriving path is generally different from that of another path. The effect on the received sig-
nal is seen as a Doppler spreading or spectral broadening of the transmitted signal frequency
rather than a shift. Doppler spread and coherence time T0 are reciprocally related as

1
T 0 = ----- (12.33)
fd

fd is regraded as the typical fading rate of the channel. When T0 is defined as the time dura-
tion over which the channel’s response to a sinusoid has a correlation greater than 0.5, then
248 Ch. 12 • RF Engineering and Network Planning

9
T 0 ≈ --------------- (12.34)
16π f d

A popular rule of thumb is to define T0 as the geometric mean of Eqs. (12.33) and (12.34)

9 0.423
T0 = ---------------2 = ------------- (12.35)
16π f d fd

The time required to traverse a distance λ/2 (equal to fade interval) when traveling at a
constant velocity v is

λ⁄2 0.5 0.5

T 0' = ---------- = -------------- = ------- ≈ T 0 (12.36)
v (v ⁄ λ) fd

With frequency equal to 900 MHz and velocity equal to 120 km/h, the coherence time is
about 5 ms and Doppler spread is approximately 100 Hz. For a voice-grade channel with a typical
transmission rate of 10 ksps, the fading rate is considerably less than the symbol rate. Under such
conditions, the channel would manifest slow-fading effects. A channel is referred to as fast fading
if the symbol rate 1/Ts is less than the fading rate 1/T0 i.e., the fast fading is characterized by

Bw < f d (12.37a)

or

Ts > T0 (12.37b)

A channel is referred to as slow fading if the signaling rate is greater than the fading rate.
Thus, to avoid signal distortion caused by fast fading, the channel must be made to exhibit slow
fading by ensuring that the signaling rate exceeds the channel fading rate

Bw > f d (12.38a)

or

Ts < T0 (12.38b)

The channel fading rate fd sets a lower limit on the signaling rate that can be used without
suffering fast-fading distortion. A better way to state the requirement for mitigating the effects of
fast fading would be that we desire Bw » fd (or Ts « T0). If this condition is not satisfied, the ran-
dom frequency (FM) due to varying Doppler shifts will limit system performance significantly.
The Doppler effect yields an irreducible error rate that cannot be overcome by simply increasing
Eb /It . This irreducible error rate is most pronounced for any modulation that involves switching
the carrier phase. For voice-grade applications with error rates of 10–3 to 10–4, a large value of
Doppler shift is considered to be on the order of 0.01 Bw . Thus, to avoid fast-fading distortion
and Doppler-induced irreducible error rate, the signaling rate must exceed the fading rate by a
Intersymbol Interference 249

factor of 100 to 200. The exact factor depends on the signal modulation, receiver design, and
required bit error rate.

12.10 Intersymbol Interference

In practical radio systems, the presence of a transmitter band-pass filter is essential to save spec-
trum as much as possible. However, such a band-limited channel could degrade the transmission
performance due to ISI. Therefore, we should reduce the signal bandwidth as much as possible
without producing any ISI.
In a time-dispersive medium, the transmission rate Rb for a digital transmission is limited
by delay spread. If a low bit-error-rate performance is required, then

1
R b < -------- (12.39)
2τ d

In a real situation, Rb is determined based upon the required bit error rate.

12.11 Link Budget and Cell Coverage

The L50 path-loss estimates provide the average signal strength at a given distance from the trans-
mitter. The signal varies from that average by large amounts, both lower and higher. Field data
show that statistically the signal strength at a given distance from the transmitter has a log-normal
distribution with standard deviation equal to 8 to 10 dB. These variations are called shadow
losses and are caused by obstructions between the transmitter and the receiver. Field data also
indicate that an obstruction affects the path loss for an average distance of 500 m in the suburban
environment and 50 m in the urban environment. Fig. 12-4 shows a typical plot of the path loss
vs. distance in meters with a mean path loss (MPL) and an MPL ±10.24 dB. The path loss has a
normal distribution with a standard deviation of ±8 dB. A normal distribution curve has ±32% of
the values outside ±8 dB. This means 16% of the values will be more than 8 dB above the mean.
We must consider this variation when designing a system so that the received signal to interfer-
ence ratio is greater than the minimum value required for an acceptable voice quality over most
of the coverage area. For example, if the system objectives are to provide adequate voice quality
over 90% of the coverage area, the system must be designed with at least a 10.24-dB margin.
To determine the maximum cell range, designers must calculate the maximum allowable
path loss that will provide adequate signal strength at the cell boundary for acceptable voice
quality over 90% of the coverage area. The allowable path loss is the difference between the
transmitter’s effective radiated power and minimum signal strength required at the receiver for
acceptable voice quality. The components that determine the path loss are called the link budget.
Link budgets are used to calculate the coverage and performance for a base station and a
mobile station. The components include propagation factors to calculate path loss and system
parameters (transmitter power, receiver noise figure, antenna gains, receiver bandwidth, process-
ing gain, and interference). Other losses such as power control errors, building penetrations,
body/orientation losses, and interference from other sources are also included.
250 Ch. 12 • RF Engineering and Network Planning

Note: 10% of the path loss will exceed the MPL by more than 10.24 dB.

Figure 12-4 Shadow Margin

For CDMA, the link budget is used to

• Decide an appropriate network loading
• Allocate appropriate power to various forward link channels
The following procedure is used for link budget analysis:
• Identify parameters affecting the forward and reverse links
◆ access technology-specific parameters
◆ product-specific parameters
◆ morphology-based parameters
• Determine the maximum allowable path loss to maintain communication on the for-
ward and reverse links
Link Budget and Cell Coverage 251

• Balance the forward and reverse link

We illustrate the procedure for calculating link budget and determining the range of a base
station.

EXAMPLE 12.3
Refer to Figure 12-5 and use the following parameters to calculate the maximum allowable path
loss:
• Information rate = 9600 bps
• Mobile station’s effective radiated power (Pm ) = 200 mW (23 dBm)
• Base station antenna gain (Gb) = 14 dBi
• Base station receiver antenna cable loss (Lc) = 2.5 dB
• PCS minicell receiver noise figure (Fb ) = 5 dB
• Required margin (Eb /Nt ) = 6.8 dB (with diversity antenna at base station)
• Base station noise floor (N0 ) = –174 dBm/Hz
• Log-normal shadowing margin = 8 dB
• Body/orientation loss = 2 dB
• Building penetration loss = 10 dB
Base station noise floor

N T = N 0 + F b = – 174 + 5 = – 169 dBm/Hz (12.40)

Gb Base Station

RF Amplifier

Lc
Fb

Gb
Mobile
CRM

Lc
Fb

Figure 12-5 Transmission between Mobile and Base Station

252 Ch. 12 • RF Engineering and Network Planning

Minimum bit energy required for specified Eb /N0

( E b ) min = N T + ( E b ⁄ N t ) reqd = – 169 + 6.8 = – 162.2 dBm/Hz (12.41)

Minimum signal strength required

S min = ( E b ) min + 10 log R = – 162.2 + 10 log 9600 = – 122.4 dBm (12.42)

Mean path loss (L50) with Smin

L 50 = P m – S min + G b – L c = 23 + 122.4 + 14 – 2.5 = 156.9 dB (12.43)

To provide margin for shadowing

Path loss = L 50 – 10.2 = 156.9 – 10.2 = 146.7 dB (12.44)

To provide margin for body/orientation loss and building penetration loss

Allowable path loss = 146.7 – 2.0 – 10 = 134.7 dB (12.45)

EXAMPLE 12.4
Using the allowable path-loss value from Example 12.3, determine the coverage for the mini-PCS
cell. Assume the following data:
• PCS frequency (fc ) = 1800 MHz
• Street width (W) = 20 m
• Spacing between buildings (b) = 40 m
• Average roof height of building (hr) = 40 m
• Mobile antenna height (hm) = 2 m
• Base station antenna height (hb) = 40 m
• Street orientation, φ = 90 degrees
We use the COST 231 model
∆h m = h b – h m = 40 – 2 = 38 m

∆h b = h b – h r = 40 – 40 = 0 m

L 0 = 4 – 0.114 ( φ – 55 ) = 4 – 0.114 ( 90 – 55 ) = 0

L bsh = – 18 log 11 + ∆h b = – 18.75 dB

15∆h fc
k a = 54, k d = 18 – --------------b- = 18, k f = 4 + 1.5  --------
- – 1 = 4 + 1.5  ------------ – 1 = 5.42
1800
∆h m  925   925 

L f = 32.4 + 20 log r + 20 log 1800 = 97.5 + 20 log r dB

L rts = – 16.9 – 10 log 20 + 10 log 1800 + 20 log 38 + 0 = 34.25 dB

L ms = – 18 log 11 + 0 + 54 + 18 log r + 5.42 log 1800 – 9 log 40 = 38.47 + 18 log r dB

Dual-Mode CDMA Mobiles 253

∴134.7 = 97.5 + 20 log r + 34.25 + 38.47 + 18 log r

38 log r = – 35.52
log r = – 0.935 , r = 0.116 km or 116 m

12.12 Dual-Mode CDMA Mobiles

The nominal CDMA channel requires 41 contiguous analog channels with a 9-channel spacing
as a guard between the edge of a CDMA channel and the adjacent analog channels. This implies
that 59 contiguous analog channels should be removed from the service to introduce the first
CDMA channel, 100 analog channels for the first and second CDMA channels, and so on. In an
analog system with a reuse factor of 7, 3 analog channels per sector will be removed to add the
first CDMA channel (i.e., [59/7]/3 ≈ 3) and an additional 2 analog channels per sector will be
removed for each additional CDMA channel ([42/7]/3 = 2).
Consider an example where the base station is equipped with 78 channel elements. Four
channel elements on each sector are dedicated to the pilot channel, sync channel, and two paging
channels, leaving 22 channel elements per sector remaining for traffic channels. Assume that 22
channel elements are permanently assigned to each sector. The real system pools all 66 channel
elements, so any channel element can be assigned to any sector; this method is referred to as
dynamic channel assignment. Dynamic channel assignment improves the base station capacity
as it reduces the probability of blocking on a sector since an idle channel element can be used in
any sector. Blocking occurs when the interference from an additional cell will reduce the voice
quality below the acceptable limit. Blocking is determined by the channel capacity and not by
the amount of hardware used. Using the method of permanently assigning channel elements to a
sector gives the lower bound on the total traffic carried by the base station. The traffic carried by
the same channel elements will be higher in the real system.
Consider the case where a 3-sector analog base station is equipped with 57 channels that
are assigned an N = 7 reuse pattern. We remove 3 analog channels to introduce 1 CDMA chan-
nel to serve 22 users. Assume no overflow traffic from CDMA to analog channels, assume 2%
probability of blocking, and model the traffic using Erlang B-statistic. Also assume that each
sector can serve 22 simultaneous calls with the same voice quality as an analog channel. Table
12-6 provides the calculated capacity of the analog base station and the combined CDMA and
analog base station. It may be noted that the traffic capacity of the combined base station is dou-
ble the capacity of the analog base station.

EXAMPLE 12.5
Estimate the cell capacity as a function of CDMA user penetration for an overlay system with 2%
blocking for both analog and digital subscribers. Assume N = 7 as the reuse pattern for the analog
channels and N = 1 for CDMA channels. The average traffic per subscriber is 0.02 Erlangs. See
Table 12-7.
 254
Table 12-6 Comparison of BS Capacity with Analog and Combined CDMA and Analog Channels

Analog Voice CDMA Traffic Analog Traffic per Total Analog CDMA Traffic per Total CDMA Total BS Traffic
Channels Channels Sector (Erlangs) Traffic (Erlangs) Sector (Erlangs) Traffic (Erlangs) (Erlangs)

19 0 12.34 37.0 — — 37.0

16 22 9.83 29.5 14.9 44.7 74.2

Table 12-7 Capacity of a cell vs. CDMA User Penetration

Analog CDMA CDMA

Channels Channels Users % Total
per Cell, per per CDMA Analog Mobile CDMA Mobile Erlangs Analog CDMA Total
Na Cell,Nc Cell, CE Traffic (E) (E) (E) Subscribers Subscribers Subscribers

57 0 0 0 12.34 × 3 = 37.0 0 37.0 1850 0 1850

48 1 66 60.24% 9.83 × 3 = 29.5 14.9 × 3 = 44.7 74.2 1475 2235 3710

Ch. 12 • RF Engineering and Network Planning

42 2 132 80.91% 8.2 × 3 = 24.6 34.68 × 3 = 104.0 128.6 1230 5200 6430
36 3 198 89.3% 6.62 × 3 = 19.9 55.33 × 3 = 166 185.9 995 8300 9295
30 4 264 93.73% 5.084 × 3 = 15.3 76.38 × 3 = 229.1 244.4 765 11,455 12,220
24 5 330 96.43% 3.627 × 3 = 10.9 97.69 × 3 = 293.1 304.0 545 14,655 15,200
18 6 396 98.13% 2.277 × 3 = 6.8 119.2 × 3 = 357.6 364.4 340 17,880 18,220

12 7 462 99.22% 1.092 × 3 = 3.3 140.7 × 3 = 422.2 425.5 165 21,110 21,275
6 8 528 99.86% 0.223 × 3 = 0.7 162.4 × 3 = 487.3 488 35 24,365 24,400
0 9 594 100% 0 184.2 × 3 = 552.5 552.5 0 27,625 27,625
The Transition from an Analog System to a Digital System 255

The impact on introducing CDMA in an analog system results in reducing the cell analog
capacity by 20%, but it doubles the total capacity of the cell. This means that at least 20% of the
offered traffic load must come from dual-mode mobiles before activating CDMA. To realize 100%
capacity increase in the cell, about 60% of the offered traffic load in the cell must be from dual-mode
mobiles.

12.13 The Transition from an Analog System to a Digital System

The transition from an analog system to the CDMA system, as offered by cellular system equip-
ment vendors and envisioned by the cellular system operators, generally falls into three basic
modes:
• An overlay design of two separate, independent systems—one is analog and the other is
CDMA
• A completely integrated system where CDMA and analog service is offered every-
where
• A partially integrated system in which CDMA and analog coverage is provided in part
of the overall service area and only analog coverage exists in the remainder of the ser-
vice area

12.13.1 Overlay Design

In the overlay scenario, a new CDMA system is superimposed over the existing analog
system. The CDMA overlay could require a one-to-one digital base station for each analog base
station, but it may use a smaller number of larger coverage area cells than the existing analog
system to reduce the cost of base stations. Fig. 12-6 shows a CDMA base station coverage area
that is three times larger than the analog base station coverage area. Thus, it requires one-third as
many base stations to provide CDMA coverage over the service area. Such an overlay design
could be operated as two separate, independent systems that are operated by different vendors
from separate mobile switching centers (MSC) or as a single system in which a single MSC con-
trols both types of base stations. Handoffs between the CDMA and analog system are not possi-
ble, and handoffs from analog to CDMA are not allowed by the IS-95 standard. The advantages
and disadvantages of the full overlay design are
• Advantages
◆ It allows independent analog and CDMA systems in the market.
◆ There can be a different vendor rather than the analog system vendor for the
CDMA system if this is desired.
◆ The service provider will be able to advertise that CDMA (digital) service is
available throughout the service area.
◆ Since a smaller number of base stations can be used, the investment in digital
equipment should be lower than for designs that require digital equipment in a
larger number of base stations.
256 Ch. 12 • RF Engineering and Network Planning

A = Analog Base Stations

A/D = Co-located Analog and CDMA Base Stations

A/D A/D

A A A

A A CDMA Overlay at 1/3 of

Cell Sites
A/D A/D A/D

A A

A A A

A/D A/D

Figure 12-6 Overlay Design

• Disadvantages
◆ There is capacity loss due to segmentation of the cellular spectrum since there
will be fewer radio frequencies for the analog subscribers.
◆ The grade of service to analog subscribers, whose numbers may increase if ana-
log terminal dumping occurs, will require investing in additional analog base sta-
tion infrastructure.
◆ There is increased system operational complexity. The engineering, Operation,
Administration, and Maintenance (OA&M) of a two-system CDMA/analog over-
lay is much more complex than for a single system.
◆ The analog-only base stations may require additional RF filters to reduce the
probability that a nearby CDMA mobile will overload the analog base station
receiver.

12.13.2 Integrated Design

In this scenario (Fig. 12-7), the system is designed to support both analog and CDMA cus-
tomers everywhere in the service area. As with the other designs, this would be a transitional
approach containing the capability for high CDMA capacity in the core and lower CDMA
capacity in the noncore areas. Over time, the area with higher CDMA capacity would expand to
include more and more of the system. The advantages and disadvantages of this approach are
The Transition from an Analog System to a Digital System 257

A/D = Co-located Analog and CDMA Base Stations

A/D

A/D A/D

A/D A/D A/D

A/D A/D

A/D A/D A/D

A/D A/D

A/D A/D A/D

A/D A/D

A/D

Figure 12-7 Integrated Design

• Advantages
◆ The entire system would have complete digital coverage.
◆ This scenario avoids receiver overload and other radio problems.
◆ There is no cellular spectrum segmentation.
◆ Full-spectrum efficiency is achieved through high reuse (N = 1) everywhere.
◆ Dual-mode terminals and handoffs between CDMA and analog channels are not
required (except for roaming into analog coverage areas).
◆ OA&M is simplified from that of the overlay system.
◆ The system operator can advertise digital service everywhere in the service area.
• Disadvantages
◆ It requires digital equipment everywhere in the system. Although the investment
costs can be reduced in areas that need only a low-capacity digital service, it is
still a larger investment than for the partial digital system option.

12.13.3 Partial CDMA Coverage, Integrated System

In the partially integrated system, only a part of the system is converted to support analog
and digital traffic (usually the core of the system, most in need of traffic relief from the digital
design). Surrounding the core area of the system, a transition or buffer zone is required to avoid
interference between cochannel analog and digital channels. Cochannels are not allowed in the
buffer zone. Beyond the buffer zone base stations can be assigned analog channels that are
cochannel with digital channels within the core. A dual-mode mobile assigned a digital channel
258 Ch. 12 • RF Engineering and Network Planning

A A

A A A

A A/D

A A/D A/D
A Partial Analog
A A/D

A A A A/D CDMA & Analog

A A
A Full Analog
A

Figure 12-8 Partial CDMA Coverage—Integrated Design

in the core would be handed off to an analog channel as the mobile approaches the edge of the
digital coverage area, and then it would be handed to a cell in the buffer zone. Mobiles assigned
an analog channel in the analog-only cells would not transition to a digital channel once inside
the digital coverage area.
The transition zone can gradually be moved outward from the core, and digital coverage can
be expanded. The rate of expansion can be determined by the mix of terminals in the system, the
need for capacity relief, the strategy for moving the customer base to digital terminals, and eco-
nomics of the system operator. Fig. 12-8 shows an example of a simplified model with a uniform
cell size. The buffer zone between the CDMA/analog cells and analog-only cells may require two
tiers of cells depending on the propagation and relative sizes of the actual cells. If the cells in the
buffer zone are at full capacity, then adding CDMA will reduce the capacity of the buffer zone
cells since frequencies required by CDMA channels cannot be used in the buffer zone.
Fig. 12-9 shows the case where a CDMA base station interferes with the analog mobile. The
performance criterion for analog mobiles is that the carrier signal from the analog base station is
17 dB greater than the total interference on the channel. If the analog system is designed for

S anal
---------
- = 18 dB (12.46)
I coch

and if

S anal
- ≥ 17 dB
------------------------------- (12.47)
I coch + I CDMA
The Transition from an Analog System to a Digital System 259

CDMA Analog
Base Station Base Station

D
Figure 12-9 CDMA Base Station Interference with Analog Mobile

provides adequate voice quality, then

S anal
- > 24 dB
-------------- (12.48)
I CDMA

Table 12-8 lists the relative sizes of CDMA and analog base stations to achieve the criterion
as a function of separation between the CDMA and analog center frequencies. With the recom-
mended 9 guard channels, the CDMA base station has to be only slightly farther away than the

Table 12-8 Required D/Ra Ratios vs. Center Frequency Separation between CDMA and Analog
Center Frequencies: CDMA BS Interferes with Analog MS

Center Frequency Separation (fs) Required D/R

fs < 900 kHz ≥ 2.90
900 ≤ fs < 1980 kHz ≥ 1.33
fs ≥ 1980 kHz ≥ 1.14
a
D = distance between analog and CDMA base stations
R = radius of the analog base station
260 Ch. 12 • RF Engineering and Network Planning

radius of the analog cell to have adequate analog voice-quality performance. If the analog base sta-
tion assigns frequencies within the CDMA channel, then the CDMA base station must be about 3
times the analog cell radius away, or a one cell buffer zone for equal-size analog and CDMA cells.
(R = radius of the analog base station, and D = distance between analog and CDMA base stations.)
Fig. 12-10 shows a case where the CDMA mobile interferes with the analog base station.
The performance criterion for the analog base station is that the received signal be 17 dB greater
than the total received noise. If the analog system is designed for
S anal
---------
- = 18 dB (12.49)
I coch
and if
S anal
- ≥ 17 dB
------------------------------- (12.50)
I coch + I CDMA

provides adequate voice quality, then

S anal
- > 24 dB
-------------- (12.51)
I CDMA

CDMA Analog
Base Station Base Station

D
Figure 12-10 CDMA Mobile Interference with Analog Base Station
The Transition from an Analog System to a Digital System 261

Table 12-9 Required D/Ra Ratio vs. Frequency Separation between CDMA and Analog Center
Frequencies: CDMA MS Interferes with Analog BS

Frequency Separation fs Required D/R

fs < 900 kHz ≥ 1.23
900 ≤ fs < 1980 kHz ≥ 1.05
fs ≥ 1980 ≥ 1.02
a
D = distance between the analog and CDMA base stations
R = radius of the CDMA base station

Table 12-9 lists the required D/R ratios as a function of the separation between the CDMA
and analog center frequencies to achieve the criterion, where D = distance between the analog
and CDMA base station and R = radius of the CDMA base station. With the recommended guard
9 channels, the CDMA mobile can be close to the analog base station before the interference
becomes a problem. This situation is less critical than the previous case.
Both rules (Tables 12-8 and 12-9) must be satisfied when deploying CDMA base stations
to prevent interference to the existing analog cells. Meeting these criteria will insure that the
interference from the analog transmitters to the CDMA receivers is not a problem as well.
The following are the advantages and disadvantages of the partial integrated design:
• Advantages
◆ The CDMA capacity advantage can be placed where it is needed most, i.e., in the
core system. Investment only for the core system will be needed.
◆ OA&M is simpler than the overlay approach.
◆ The design avoids the receiver overload problem of the overlay design.
• Disadvantages
◆ The system operator cannot advertise digital-everywhere service.
◆ Handoff is required between CDMA and analog coverage areas.
◆ IS-95 does not provide analog-to-CDMA handoff, so a call initiated in the analog
area will not provide any digital features available in the digital coverage area.
◆ Voice-quality changes may be perceived during CDMA-to-analog handoffs.

EXAMPLE 12.6
Consider a small-city cellular system that is growing at the predicted growth rate for the next 7 years
(see Table 12.10). The start-up system required 9 omnidirectional-coverage base stations and has
grown to 29 directional analog base stations to provide service for 36,000 Busy Hour Call Attempts
(BHCA). Based on the predictions, this system must expand to provide capacity for 100,000 BHCA
at the end of 7 years. The service provider has chosen to provide CDMA service over the complete
coverage area by overlaying the coverage with 10 minicells. The service provider reduces the analog
262 Ch. 12 • RF Engineering and Network Planning

subscribers gradually as indicated in Table 12.10. The traffic per subscriber during the busy hour is
0.02 Erlangs.
Year 0 (all analog) (see Fig. 12-11)
• Total traffic during busy hour: 36,000 × 0.02 = 720.0 Erlangs
720.0
• Traffic per sector: --------------- = 8.276 Erlangs
29 × 3
• Number of voice channels per sector to provide 2% blocking: 14

Table 12-10 Prediction for Analog and Digital Subscribers

End of Year Analog Subscribers CDMA Subscribers Total Subscribers
0 36,000 0 36,000
1 29,000 13,000 42,000
2 24,000 22,000 46,000
3 16,000 39,000 55,000
4 10,000 52,000 62,000
5 4,000 66,000 70,000
6 0 82,000 82,000
7 0 100,000 100,000

14

1 2 3

11 12 13

15 16 17

4 5 6

18 19

20 21 22

7 8

23 24 25

26 27

9 10

28 29

Figure 12-11 Configuration at Year 0—all analog cells

The Transition from an Analog System to a Digital System 263

Table 12-11 summarizes the results at the end of year 0.

Year 1 (see Fig. 12-12)
• CDMA traffic: 13,000 × 0.02 = 260 Erlangs
• Analog traffic: 29,000 × 0.02 = 580.0 Erlangs
To provide full CDMA coverage, CDMA minicell equipment is added to 10 base stations.

260.0
• CDMA traffic per sector: --------------- = 8.67 Erlangs
10 × 3

One CDMA channel provides 14.9 Erlangs per sector with 2% system blocking to serve the dual-
mode mobiles, so the service provider decides to eliminate all analog channels in the expanded spec-
trum and to use the spectrum for one CDMA channel in the minicell equipment. This reduces the
analog capacity by 3 channels per sector in all 29 base stations.

29, 000 × 0.02

• Analog traffic per sector: ---------------------------------- = 6.67 Erlangs
29 × 3
• Number of analog channels per sector: 14 – 3 = 11

Table 12-11 Number of Analog and CDMA BSs and Capacity at the End of Year 0

No. of BSs No. of RF Channels/Sector Capacity Erlangs/Sector

Analog CDMA Analog CDMA Analog CDMA BHCA
29 0 14 0 8.2 0 36,000

14

1 2 3

11 12 13

15 16 17

4 5 6

18 19

20 21 22
Analog
7 8 Cell Site
23 24 25

26 27

9 10 Analog/CDMA
Cell Site
28 29

Figure 12-12 Configuration at the End of Year 1 and Year 2

264 Ch. 12 • RF Engineering and Network Planning

These 11 channels will carry 6.83 Erlangs of traffic at 4% blocking for the analog subscribers. The
single CDMA channel will carry all the offered traffic load with virtually no blocking for the CDMA
subscribers. If the grade of service for the analog subscribers is an unacceptable business strategy,
then additional analog capacity will have to be provided by adding additional channels in the limited
spectrum band, possibly resulting in additional cochannel interference. The results are summarized
in Table 12-12. In the calculations, we assume an average of 22 CDMA calls per sector with 2%
blocking. As can be seen, the offered CDMA traffic load (8.67 Erlangs per sector) is much lower
than the 2% blocking capacity (14.9 Erlangs per sector). CDMA subscribers will experience virtu-
ally no blocking while analog subscribers will experience 4% blocking.
Year 2 (see Fig. 12-12)
22, 000 × 0.02
• CDMA traffic per sector: ---------------------------------- = 14.67 Erlangs
10 × 3
One CDMA channel on 10 base stations with 2% blocking provides 14.9 Erlangs per sector.
24, 000 × 0.02
• Analog traffic per sector: ---------------------------------- = 5.52 Erlangs
29 × 3
• Number of analog channels per sector: 14 – 3 = 11
These 11 channels will carry 5.84 Erlangs of traffic at 2% blocking for analog subscribers. Table 12-
13 summarizes the results.
Year 3 (see Fig. 12-13)
39, 000 × 0.02
• CDMA traffic per sector: ---------------------------------- = 26.0 Erlangs
10 × 3
Two CDMA channels on 10 base stations with 2% blocking gives 34.68 Erlangs per sector.
16, 000 × 0.02
• Analog traffic per sector: ---------------------------------- = 4.64 Erlangs
23 × 3
We remove 6 analog base stations, leaving 23 base stations to carry analog traffic.
• Number of analog channels per sector: 14 – 5 = 9

Table 12-12 Number of Analog and CDMA BSs and Capacity at the End of Year 1

No. of BSs No. of RF Channels/Sector Capacity Erlangs/Sector

Analog CDMA Analog CDMA Analog CDMA BHCA
a
29 10 11 1 6.83 14.9 42,000
a
4% blocking.

Table 12-13 Number of Analog and CDMA BSs and Capacity at the End of Year 2
No. of BSs No. of RF Channels/Sector Capacity Erlangs/Sector
Analog CDMA Analog CDMA Analog CDMA BHCA
29 10 11 1 5.84 14.9 46,200
The Transition from an Analog System to a Digital System 265

1 2 3

12 13

15 16

4 5 6

18 19

21
Analog
7 8 Cell Site
24 25

26 27

9 10 Analog/CDMA
Cell Site
28 29

Figure 12-13 Configuration at the End of Year 3

These 9 channels will carry 4.35 Erlangs per sector of traffic at 2% blocking for the analog sub-
scribers. Since the offered traffic load is slightly more than the capacity at 2% blocking, the
analog subscribers will experience about 2.5% blocking, whereas the CDMA subscribers will
experience almost no blocking since offered load is less than the capacity. The results are sum-
marized in Table 12-14.
Year 4 (see Fig. 12-14)

52, 000 × 0.02

• CDMA traffic per sector: ---------------------------------- = 34.67 Erlangs
10 × 3

Two CDMA channels on 10 base stations with 2% blocking gives 34.68 Erlangs per sector.

10, 000 × 0.02

• Analog traffic per sector: ---------------------------------- = 3.33 Erlangs
20 × 3
• Number of analog channels per sector: 14 – 5 = 9
We remove 3 analog base stations, leaving 20 base stations to carry analog traffic.
These 9 channels will carry 4.35 Erlangs per sector of traffic at 2% blocking for the analog
subscribers. The results are summarized in Table 12-15.

Table 12-14 Number of Analog and CDMA BSs and Capacity at the End of Year 3
No. of BSs No. of RF Channels/Sector Capacity Erlangs/Sector
Analog CDMA Analog CDMA Analog CDMA BHCA
23 10 9 2 4.35 34.68 55,000
266 Ch. 12 • RF Engineering and Network Planning

1 2 3

12 13

15 16

4 5 6

18 19

21
Analog
7 8 Cell Site
24

26 27

9 10 Analog/CDMA
Cell Site

Figure 12-14 Configuration at End of Year 4

Table 12-15 Number of Analog and CDMA BSs and Capacity at the End of Year 4

No. of BSs No. of RF Channels/Sector Capacity Erlangs/Sector

Analog CDMA Analog CDMA Analog CDMA BHCA
20 10 9 2 4.35 34.68 62,000

Year 5 (see Fig. 12-15)

66, 000 × 0.02

• CDMA traffic per sector: ---------------------------------- = 44.0 Erlangs
10 × 3

Three CDMA channels on 10 base stations with 2% blocking gives 55.33 Erlangs per sector.

4, 000 × 0.02
• Analog traffic per sector: ------------------------------- = 2.67 Erlangs
10 × 3

We remove 10 analog base stations, leaving 10 base stations to carry analog traffic.
• Number of analog channels per sector: 14 – 7 = 7
These 7 channels will carry 2.94 Erlangs per sector at 2% blocking. The results are summarized in
Table 12-16.
Year 6 (all digital) (see Fig. 12-16)

82, 000 × 0.02

• CDMA traffic per sector: ---------------------------------- = 54.67 Erlangs
10 × 3
The Transition from an Analog System to a Digital System 267

1 2 3

4 5 6

7 8

9 10 Analog/CDMA
Cell Site

Figure 12-15 Configuration at End of Year 5

1 2 3

4 5 6

7 8

9 10
CDMA
Cell Site

Figure 12-16 Configuration at End of Year 6 and Year 7

268 Ch. 12 • RF Engineering and Network Planning

Table 12-16 Number of Analog and CDMA BSs and Capacity at the End of Year 5
No. of BSs No. of RF Channels/Sector Capacity Erlangs/Sector
Analog CDMA Analog CDMA Analog CDMA BHCA
10 10 7 3 2.94 55.33 70,000

Table 12-17 Number of Analog and CDMA BSs and Capacity at the End of Year 6
No. of BSs No. of RF Channels/Sector Capacity Erlangs/Sector
Analog CDMA Analog CDMA Analog CDMA BHCA
0 10 0 4 0 76.38 82,000

Table 12-18 Number of Analog and CDMA BSs and Capacity at the End of Year 7

No. of BSs No. of RF Channels/Sector Capacity Erlangs/Sector

Analog CDMA Analog CDMA Analog CDMA BHCA
0 10 0 4 0 76.38 100,000

We eliminate all remaining analog base stations and use 4 CDMA channels on 10 base stations with
2% blocking to give 76.38 Erlangs per sector. The results are summarized in Table 12-17.
Year 7 (see Fig. 12-16)

100, 000 × 0.02

• CDMA traffic per sector: ------------------------------------- = 66.67 Erlangs
10 × 3

Four CDMA channels per sector provides 76.38 Erlangs of traffic at 2% blocking. The results are
given in Table 12-18.

12.14 Facilities Engineering

CDMA technology offers a significant capacity improvement with respect to analog and other
digital technologies. However, in order to fully realize these capacity improvements, the service
provider must properly engineer the CDMA system. This section discusses the engineering of
facilities and its relationship to call capacity.
Facilities encompass terrestrial facilities, radio facilities, transcoders (vocoders), and net-
work facilities. As discussed earlier, transcoders can be physically located at the base station (see
Fig. 12-17) or at the MSC (see Fig. 12-18).
Only the radio facility is related to the capacity of the radio link. Also the service provider
must configure a sufficient number of CDMA channels to support the maximum number of
simultaneous calls (CDMA channels include the pilot channel, sync channel, access channels,
Facilities Engineering 269

PSTN

MSC

Terrestrial
Transcoder Transcoder
Facilities

Radio Radio
Facilities Facilities

Base Station Base Station

Figure 12-17 Basic CDMA Facilities Configuration (Transcoder at Base Station)

PSTN

MSC

Transcoder

Terrestrial
Radio Facilities Radio
Facilities Facilities

Base Station Base Station

Figure 12-18 Basic CDMA Facilities Configuration (Transcoder at MSC)
270 Ch. 12 • RF Engineering and Network Planning

paging channels, and traffic channels). Once this is done, the required frequency spectrum can
be determined. This section provides an analytical discussion of the capacity of the reverse radio
link and a qualitative discussion of the capacity of the forward radio link.
Unlike analog and TDMA technology, CDMA technology does not impose a definite limit
on the radio capacity. Rather, CDMA technology exercises a soft limit, in which mobile sub-
scribers experience a level of degradation that is related to the total interference and the thermal
noise. The capacity is soft since the number of mobile subscribers can be increased if the service
provider is willing to lower the grade of service and thus decrease customer satisfaction. With a
greater number of simultaneous CDMA calls, the noise floor increases. If the noise floor
increases, the probability of receiving a correct frame decreases, i.e., the FER increases. Interfer-
ence is generated by other CDMA mobile stations occupying the same radio spectrum on the
same cell or on a different cell. Moreover, for 800-MHz operation, interference is generated if a
mobile station operating in the analog mode is occupying a portion of the frequency spectrum
that is used by the CDMA channel. However, in a properly engineered CDMA system, analog
operation is restricted to cells sufficiently separated from the cell serving the CDMA mobile.
This significantly reduces the interference to the CDMA mobile.
In order for a CDMA system to achieve the expected capacity enhancement, it is imperative
that power control be properly functioning in both the forward and reverse directions of the radio
channel (refer to chapter 10). Power control is executed only on the traffic channels. However, in
the following analysis, it is not assumed that power control is perfect. This fact is reflected by
assuming that the instantaneous power varies about the desired Eb /It level with a log-normal distri-
bution having a standard deviation σc. Typical values of σc are on the order of 1.5 to 2.5 dB [12].
One method of estimating the capacity is to determine the probability that a CDMA chan-
nel does not have sufficient bandwidth to accommodate a mobile station for a given frame inter-
val and still satisfy the interference constraints. This event is called an outage. During an outage
of the reverse radio channel, the FER can exceed the desired maximum limit. This situation is
not catastrophic but does lead to degraded service. In the following analysis, the desired interfer-
ence is given by I0/N0 < 1/η. Typically, η is between 0.25 and 0.1 which corresponds to power
ratios of I0/N0 between 6 dB and 10 dB.
The explicit formula [12,13]for the normalized average user occupancy (λ/µ in terms of
Erlangs per sector) is given by

λ
--- ν ( 1 + f ) = ∆' r × F ( B, σ c ) (12.52)
µ
where: λ = average call rate for the entire CDMA system,
1
--- = average call duration,
µ
ν = voice activity factor,
–1 2
[ Q ( P out ) ]
B = ------------------------------ ,
∆' r
Facilities Engineering 271

B
------w
R
∆' r = -------- × ( 1 – η ), and
E b0
--------
Nt
3
α c⋅B
F(B, σc) = ----- 1 + ----------------  1 – 1 + -----------------
1 4 
-
αc 2  α c × B
3

in which
2
( βσ c ) ⁄ 2
αc =e
β = (ln[10])/10, and
∞
 [ ( – x2 ) ⁄ 2 ] 
Q(z) = ∫e d x ⁄ ( 2π )
z 
Pout is the probability of outage,
f is the mean interference from neighboring cells, and
Eb0/It is the median of the desired Eb/It.
Using Eq. (12.52), the following configurations are analyzed:
1. Hard handoff only for a CDMA system with a frequency bandwidth of 1.25 MHz and
equipped with 8-kbps transcoders (see Table 12-19), Bw = 1.25 MHz, R = 9.6 kbps, and
Bw /R = 130.
2. Soft handoffs with a maximum of 2 cells with a 1.25-MHz bandwidth and equipped
with 8-kbps transcoders (see Table 12-20).
3. Soft handoffs with a maximum of 3 cells with a 1.25-MHz bandwidth and equipped
with 8-kbps transcoders (see Table 12-21).
4. Hard handoffs only with a 1.25-MHz bandwidth and equipped with 13-kbps transcod-
ers (see Table 12-22), Bw = 1.25 MHz, R = 14.4 kbps, and Bw /R = 87.
5. Soft handoffs with a maximum of 2 cells with a 1.25-MHz bandwidth and equipped
with 13-kbps transcoders (see Table 12-23).
6. Soft handoffs with a maximum of 3 cells with a 1.25-MHz bandwidth and equipped
with 13-kbps transcoders (see Table 12-24).
7. Hard handoffs only with a 10-MHz bandwidth and equipped with 8-kbps transcoders
(see Table 12-25), Bw = 10 MHz, R = 9.6 kbps, and Bw /R = 1042.
8. Soft handoff with a maximum of 2 cells with a bandwidth of 10-MHz and equipped
with 8-kbps transcoders (see Table 12-26).
9. Soft handoffs with a maximum of 3 cells with a 10-MHz bandwidth and equipped with
8-kbps transcoders (see Table 12-27).
Configurations 1, 2, and 3 correspond to wireless systems supporting standards TIA IS-95A
and TIA IS-96A 8-kbps vocoder; configurations 4, 5, and 6 correspond to wireless systems sup-
porting standards TIA IS-95A and Qualcomm’s proprietary 13-kbps vocoder; configurations 7, 8,
272 Ch. 12 • RF Engineering and Network Planning

and 9 correspond to TIA SP-2977 and TIA IS-96A. Furthermore, in all these configurations, it is
assumed that attenuation due to propagation losses decreases as the fourth power of distance and
that the log-normal component has a standard deviation of 8 dB (refer to Eq. [12.2]). The voice
activity factor ν is assumed to equal 0.4, and the standard deviation for power control σc is assumed
to be 2.5 dB. In Tables 12-19 through 12-27, we use f = 2.38 for hard handoffs, f = 0.77 for soft
handoffs with a maximum of 2 cells, and f = 0.57 for soft handoffs with a maximum of 3 cells.
Comparing Tables 12-19 to 12-27, we can make several observations:
1. Hard handoffs reduce the capacity on the reverse radio link by approximately 50% with
respect to two-way soft handoffs (Table 12-19 vis-à-vis Table 12-20).
2. If a system is equipped with 13-kbps transcoders rather than 8-kbps transcoders, the
capacity is reduced approximately 40% (Table 12-21 vis-à-vis Table 12-24). This
observation is consistent with our expectations since the rate is increased from 9.6 kbps
to 14.4 kbps (50% increase).

Table 12-19 Configuration 1—Hard Handoffs Only (f = 2.38),

Bandwidth = 1.25 MHz, 8-kbps Transcoders (Bw /R = 130)
η Eb0/It Pout λ/µ (Erlangs)
0.10 2.51 (4 dB) 0.01 21.63
0.25 2.51 (4 dB) 0.01 17.38
0.10 3.98 (6 dB) 0.01 12.38
0.25 3.98 0.01 9.86
0.10 2.51 0.05 24.18
0.25 2.51 0.05 19.63
0.10 3.98 0.05 14.23
0.25 3.98 0.05 11.48

Table 12-20 Configuration 2—Soft Handoffs (f = 0.77),

Bandwidth = 1.25 MHz, 8-kbps Transcoders (Bw /R = 130)
η Eb0/It Pout λ/µ (Erlangs)
0.10 2.51 (4 dB) 0.01 41.30
0.25 2.51 0.01 33.19
0.10 3.98 (6 dB) 0.01 23.63
0.25 3.98 0.01 18.83
0.10 2.51 0.05 46.17
0.25 2.51 0.05 37.49
0.10 3.98 0.05 27.17
0.25 3.98 0.05 21.92
Facilities Engineering 273

Table 12-21 Configuration 3—Soft Handoffs (f = 0.57),

Bandwidth = 1.25 MHz, 8-kbps Transcoders (Bw /R = 130)
η Eb0/It Pout λ/µ (Erlangs)
0.10 2.51 (4 dB) 0.01 46.56
0.25 2.51 0.01 37.42
0.10 3.98 (6 dB) 0.01 26.64
0.25 3.98 0.01 21.23
0.10 2.51 0.05 52.05
0.25 2.51 0.05 42.27
0.10 3.98 0.05 30.63
0.25 3.98 0.05 24.71

Table 12-22 Configuration 4—Hard Handoffs Only (f = 2.38),

Bandwidth = 1.25 MHz, 13-kbps Transcoders (Bw /R = 87)
η Eb0/It Pout λ/µ (Erlangs)
0.10 2.51 (4 dB) 0.01 13.25
0.25 2.51 0.01 10.46
0.10 3.98 (6 dB) 0.01 7.43
0.25 3.98 0.01 5.86
0.10 2.51 0.05 15.17
0.255 2.51 0.05 12.25
0.10 3.98 0.05 8.79
0.25 3.98 0.05 7.04

Table 12-23 Configuration 5—Soft Handoffs (f = 0.77),

Bandwidth = 1.25 MHz, 13-kbps Transcoders (Bw /R = 87)
η Eb0/It Pout λ/µ (Erlangs)
0.10 2.51 (4 dB) 0.01 25.29
0.25 2.51 0.01 20.17
0.10 3.98 (6 dB) 0.01 14.18
0.25 3.98 0.01 11.19
0.10 2.51 0.05 28.97
0.255 2.51 0.05 23.39
0.10 3.98 0.05 16.79
0.25 3.98 0.05 13.45
274 Ch. 12 • RF Engineering and Network Planning

Table 12-24 Configuration 6—Soft Handoffs (f = 0.57),

Bandwidth = 1.25 MHz, 13-kbps Transcoders (Bw /R = 87)
η Eb0/It Pout λ/µ (Erlangs)
0.10 2.51 (4 dB) 0.01 28.52
0.25 2.51 0.01 22.74
0.10 3.98 (6 dB) 0.01 15.99
0.25 3.98 0.01 12.62
0.10 2.51 0.05 32.66
0.255 2.51 0.05 26.37
0.10 3.98 0.05 18.93
0.25 3.98 0.05 15.17

Table 12-25 Configuration 7—Hard Handoffs Only (f = 2.38),

Bandwidth = 10 MHz, 8-kbps Transcoders (Bw /R = 1042)
η Eb0/It Pout λ/µ (Erlangs)
0.10 2.51 (4 dB) 0.01 221.51
0.25 2.51 0.01 182.21
0.10 3.98 (6 dB) 0.01 134.93
0.25 3.98 0.01 110.62
0.10 2.51 0.05 230.52
0.255 2.51 0.05 190.34
0.10 3.98 0.05 141.87
0.25 3.98 0.05 116.86

Table 12-26 Configuration 8—Soft Handoffs (f = 0.77),

Bandwidth = 10 MHz, 8-kbps Transcoders (Bw /R = 1042)
η Eb0/It Pout λ/µ (Erlangs)
0.10 2.51 (4 dB) 0.01 423.00
0.25 2.51 0.01 347.93
0.10 3.98 (6 dB) 0.01 257.66
0.25 3.98 0.01 211.24
0.10 2.51 0.05 440.21
0.255 2.51 0.05 363.48
0.10 3.98 0.05 270.91
0.25 3.98 0.05 223.16
Facilities Engineering 275

Table 12-27 Configuration 9—Soft Handoffs (f = 0.57),

Bandwidth = 10 MHz, 8-kbps Transcoders (Bw /R = 1042)
η Eb0/It Pout λ/µ (Erlangs)
0.10 2.51 (4 dB) 0.01 476.88
0.25 2.51 0.01 392.28
0.10 3.98 (6 dB) 0.01 290.48
0.25 3.98 0.01 238.15
0.10 2.51 0.05 496.29
0.255 2.51 0.05 409.79
0.10 3.98 0.05 305.42
0.25 3.98 0.05 251.59

3. If the bandwidth increases from 1.25 MHz to 10 MHz (eightfold increase), the capacity
increases approximately 10 times (Table 12-20 vis-à-vis Table 12-26).
We investigate the effect on the reverse radio link capacity if the power control is perfect,
i.e., σc = 0. In this case, F(B, σc) in Eq. (12.52) simplifies to

F ( B, σ c ) = 1 + ---  1 – 1 + ---
B 4
(12.53)
2 B
As an example of applying Eq. (12.53), we can compare the results in Table 12-20 in
which η = 0.10, Eb0/It = 6 dB, and Pout = 0.01. In this case, λ/µ = 23.63 Erlangs. If the power con-
trol is perfect, i.e., σc = 0, λ/µ increases to 27.14 Erlangs or a 15% increase. However, Eq.
(12.53) is based upon Eq. (12.2), which may not adequately model all radio environments. In
such cases, the engineer needs to use either a more complicated mathematical model or to exe-
cute a computer simulation.
The power of the reverse link channels for a specific user is adjusted at a rate of 800 times
per second in order for the received power at the base station to provide the same minimum Eb /It
as required for the specified link quality. The accuracy of the power control affects the reverse
link capacity (see Fig. 12-19).
When determining the capacity of the reverse radio link, one must also include the capac-
ity needed to support the access channels. Both Call Setup and Registration messages are trans-
mitted on the access channel. The decrease of the capacity on the reverse radio link is small due
to supporting the access channels, typically about a 1% reduction of the supportable Erlang traf-
fic. Thus this reduction is ignored in the determination of the radio link capacity.
When designing a CDMA system for a given call load, the engineer must determine the
number of CDMA Channel Modem (CM) circuits that must be supported at each cell. A CM
operates at baseband rather than at RF frequencies. It demodulates the CDMA signal for a given
mobile station and combines signals from multiple sectors of a given cell during softer handoffs
(see Fig. 12-20). The number of CMs is affected by the number of simultaneous soft handoffs
276 Ch. 12 • RF Engineering and Network Planning

30
Change in Cell Capacity (%)

25

20

15

10

0
0 1 2 3 4 5
Std. Dev. Power Control (dB)

Figure 12-19 Effect of Power Control on Reverse Link Capacity

B F

Figure 12-20 Soft-Softer Handoff Configuration

but not upon softer handoffs. In our analysis, we assume that a CM is not dedicated to a particu-
lar sector or CDMA carrier, although this assumption is dependent upon the actual manufac-
turer’s implementation.
The exact distribution of two-way soft handoffs, three-way soft handoffs, softer handoffs,
and soft-softer handoffs is very dependent upon the radio environment and radio configuration.
The service provider needs to tune the system in order to optimize the radio link capacity. This
tuning process includes adjusting the transmitted power of the pilot channels and the threshold
levels that trigger handoffs. TIA IS-95A defines several thresholds: T_ADD, T_DROP,
Facilities Engineering 277

Table 12-28 Handoff Distribution—CDMA Formal Field Test, November 18–23, 1991
Handoff Type Mean (%) Standard Deviation (%)
Softer 62.80 19.14
Two-way soft 6.52 19.05
Soft-softer 4.42 7.90
Three-way soft 0.40 0.82
Three-way softer 0.20 0.42
No handoff 25.66 10.80

T_TDROP, and T_COMP. The base station sends the values of these thresholds in the Systems
Parameter message (which is transmitted on the paging channel) and the Handoff Direction mes-
sage or the Extended Handoff Direction message (which is transmitted on the traffic channel).
During a call, the mobile station measures the strength of the pilot channel of the serving cells
(sectors) and potential candidates. The mobile station sends a Pilot Strength Measurement mes-
sage to the base station in order to initiate a possible handoff for one of the following reasons:
1. Pilot’s signal strength of a serving cell drops below T_DROP for a duration equal to
T_TDROP.* T_TDROP is determined by the service provider and ranges from 0 to 15
seconds.
2. Pilot’s signal strength of a candidate exceeds T_ADD.†
3. Pilot’s signal strength of a candidate exceeds that of a serving cell by T_COMP.‡
As an example, Table 12-28 shows the handoff distribution exhibited during CDMA trials
in San Diego, California.
As a general rule, a CDMA system is tuned so that calls are in some form of soft handoff,
i.e., two-way soft, soft-softer, or three-way soft, for approximately 40–60% of the time. Percent-
ages greater than this often are not justified by the improvement of the call quality. It is interest-
ing to note that the aggregated results of the CDMA Formal Field Test (Table 12-28) indicate the
total soft handoff (the sum of two-way soft, three-way soft, and soft-softer handoffs) is 12% of
the time. This observation may be rationalized by the fact that calls were in total softer handoff
(the sum of softer, soft-softer, or three-way softer) 67% of the time. Also note that the standard
deviation of the handoff distributions is large, particularly for two-way soft and soft-softer hand-
offs. There are two reasons for this:

• The handoff distribution varies with the cells that are serving the call. Cells are tuned
differently in order to optimize the call capacity.

* As defined in Section 6.6.6.2.5.2 of TIA IS-95A, this pilot is contained in the Active Set.
† TIA IS-95A refers to the pilot as being in the Neighbor Set or in the Remaining Neighbor Set.
‡ TIA IS-95A refers to the pilot as being in the Candidate Set.
278 Ch. 12 • RF Engineering and Network Planning

• The handoff distribution varies with the environment of the mobile station. Even
though the mobile station may be served by the same cell, the terrain within the cell’s
domain varies sufficiently to significantly affect the handoff distributions.
When a call is in a two-way soft, soft-softer, or three-way softer handoff, two CMs are
required to support the call. When a call is in a three-way soft handoff, three CMs are needed. Only
one CM is needed for the call during a softer handoff or when no handoff configuration occurs.

EXAMPLE 12.7
Assume an equal call distribution across all cells and sectors with the same handoff distribution. The
handoff distribution is as follows: 40% softer handoff, 20% two-way soft handoff, 10% soft-softer
handoff, 29% no handoff, and 1% three-way soft handoff. There are two CDMA carriers, each hav-
ing a bandwidth of 1.25 MHz. The system is equipped with transcoders that conform to TIA IS-96A,
i.e., 8-kbps voice coding and 9.6 kbps on the physical layer. Assume η = 0.25, Eb0/I0 = 6 dB, and Pout
= 0.01. The system is configured with 10 cells, each having 3 sectors. The average call duration is 90
seconds, and each mobile subscriber generates 0.03 Erlang during the busy hour.
Determine the number of calls that can be supported per hour by the system. Determine the
number of mobile subscribers that can be provided service if 2% blocking is acceptable. Also, deter-
mine the number of CMs that must be equipped to support the calculated number of subscribers.
From Table 12-21, the capacity of the reverse radio link per sector is:
Erlangs
21.23 ------------------ × 2 carriers = 42.46 Erlangs
carrier
Thus, each sector can simultaneously support 42 CDMA channels. This does not equal the
number of simultaneous calls since some of the channels are assigned as the second and third chan-
nels for calls in softer and soft handoff. A CDMA channel is required for each sector configured in
the call (refer to Fig. 12-21).

E
D
T F H
S G
U B I
A
Q C K
P J
R N L
M
O

Figure 12-21 Sector Configuration for Example 12.7

Facilities Engineering 279

In order to determine the number of simultaneous calls that can be supported by the CDMA
system, we must associate each call with one sector even though the given call is being served by
multiple sectors. In this example, we assume that the call is associated with the oldest serving sector,
although there may be other assignments. If complete homogeneity is assumed, we can determine
the total number of simultaneous calls supported by the entire system by multiplying the number of
sectors (10 × 3 = 30) by the number of simultaneous calls supported by each sector. To illustrate this
point, let us determine the number of simultaneous calls supported by sector B as shown in Fig. 12-
21. Softer handoffs are served by sectors B-C and B-A; two-way soft handoffs by B-G and B-F;
three-way soft handoffs by B-G-F; soft-softer handoffs by B-C-G, B-C-J, B-C-F, B-C-N, B-A-U, B-
A-F, B-A-Q, and B-A-G; no handoffs by B. Let the number of simultaneous calls supported by the
sector be x. Then x can be determined by
x + 0.40x + 0.2x + 0.1x + 0.01x = 42
where 40% of the channels are supporting other sectors in softer handoff, 20% of the channels are
supporting other cells in two-way soft handoff, 10% of the channels are supporting other cells in
soft-softer handoff, and 1% of the channels are supporting cells in three-way soft handoff.
Thus, x equals 24 simultaneous calls per sector.
Therefore, the entire CDMA system can serve
sectors calls
10 cells × 3 ---------------- × 24 -------------- = 720 calls
cell sector
In order to determine the number of subscribers that can be supported by this system by each
sector during the busy hour, we use the Erlang-B formula. Each sector can support 24 simultaneous
calls, which is equivalent to trunks or radios. From Erlang-B tables, we find that 16.63 Erlangs per
sector can be supported during the busy hour. Thus, the number of mobile subscribers that the sys-
tem supports is determined by
Erlang Erlang sector
0.03 ------------------------ × N ( subscribers ) = 16.63 ---------------- × 3 -------------- × 10 cells
subscriber sector cell
16.63 × 3 × 10
N = ----------------------------------- subscribers
0.03
N = 16,630 subscribers
Next, we calculate the number of CMs that need to be equipped at each sector. We have
already determined that each sector can support 42 CDMA channels. If a call is in a softer or no
handoff, one CM is needed; for two-way soft handoff and softer-soft handoff, two CMs are needed;
for a three-way soft handoff, three CMs are needed. The number of CMs that need to be equipped at
each sector for supporting the traffic channels is
x + 0.2x + 0.1x + 0.01x = 42 CM
where 20% of the CMs are supporting other cells in soft handoff, 10% of the CMs are supporting
other cells in soft-softer handoff, and 1% of the CMs are supporting other cells in three-way soft
handoff.
x = 32 CM
In addition, CMs must be equipped for the access channel. Even though the access channel
has a negligible effect upon the reverse radio channel (16.31 × 0.01 = 0.02 Erlangs), a CM must be
equipped to support this channel.
280 Ch. 12 • RF Engineering and Network Planning

12.15 Design Considerations at the Boundary of a CDMA System

Standards do not currently support soft handoffs between CDMA systems operated by different
service providers. Thus, if a mobile station moves between these CDMA systems, a hard hand-
off will occur. This results in the reduction of capacity. Comparing Table 12-19 with Table 12-
20, we see that the capacity is degraded by approximately 50%. The service provider has several
options in order to mitigate this problem. First, the boundary cells (sectors) can be located at low
traffic areas. If this is not practical, additional spectrum can be allocated at the boundary sectors.
Neither option is very appealing to the service provider. Consequently, TIA TR45 has developed
an intervendor soft handoff.

12.16 Interfrequency Handoff

The CDMA network may have multiple frequency carriers in each cell, and a hot-spot cell could
have a larger number of frequencies than neighboring cells. Also, in hierarchical cell structures,
microcells have a different frequency than the macrocell overlying the microcells. Therefore, an
efficient scheme is required to hand off between different frequencies. A blind handoff used by
an IS-95A CDMA network does not provide an adequate call quality. The mobile station should
be able to determine the signal strength and quality of another carrier frequency while still main-
taining the connection with the current carrier frequency.
A CDMA transmission is continuous; there are no idle slots for the interfrequency mea-
surements. Therefore, compressed mode and dual receiver have been proposed as a solution to
interfrequency handoff. In the compressed (slotted) mode, measurement slots are created by
transmitting the data frame with a lower spreading ratio during a shorter period, and the rest of
the frame is used for measurements on other carriers. The dual receiver can measure other fre-
quencies without affecting the reception of the current frequency.
The dual-receiver approach is considered suitable if the mobile terminal uses antenna
diversity. During the interfrequency measurements, one receiver branch is switched to another
frequency for measurements, while the other branch keeps receiving from the current fre-
quency. The loss of diversity gain during measurements needs to be compensated for with
higher forward link transmission power. The advantage of the dual-receiver approach is that
there is no break in the current frequency connection. Fast closed-loop power control is running
all the time.
The slotted-mode approach shown in Fig. 12-22 is considered attractive for the mobile sta-
tion without antenna diversity. The information transmitted during a frame is compressed in
time. For a compressed mode, there are a number of alternatives for implementations: variable
spreading factor, code rate increase, multicode, and higher modulation. Variable rate spreading
and code rate increment (puncturing) cause 1.5 to 2.5 dB loss in Eb /It, and higher-order modula-
tion causes even higher loss (~ 5 dB). This is due to a break in power control and less coding. A
further drawback of the variable spreading factor is that simple terminals have to be able to oper-
ate with different spreading ratios.
Summary 281

Regular Frame
Interfrequency measurements are
performed during an idle period.

Figure 12-22 Interfrequency Slotted Mode

12.17 Summary
This chapter discussed principles for engineering a CDMA system including propagation mod-
els, link budgets, and facilities engineering. However, extensive RF measurements and RF mod-
eling are needed in order to plan a real commercial system. The chapter’s intent was to provide
some tools and a better understanding toward achieving this goal. We also observed that the cal-
culated path losses with the Okumura/Hata model are about 13 to 15 dB lower than the path
losses obtained using the Walfisch-Ikegami (COST 231) model. This is because the Okumura/
Hata model neglects several important parameters such as street width and street orientation.
The chapter also covered the concepts of the delay spread, coherence bandwidth, and Dop-
pler spread. When the signal bandwidth is much larger than the coherence bandwidth, the signal
is called wideband signal and the fading is frequency selective. This means that only a portion of
the signal bandwidth fades at any instant of time. With signal bandwidth much smaller than the
coherence bandwidth, flat fading of the entire signal occurs.
With the signal symbol interval being much larger than the coherence time, the channel
fades rapidly compared to the symbol rate. This is called fast fading relative to symbol time, and
frequency dispersion occurs, causing signal distortion. With the signal symbol interval being
much smaller than the coherence time, the channel does not change during the symbol interval;
it is referred to as the slow-fading channel, relative to the symbol time. The chapter concludes
with a discussion of various schemes to transition from an analog system to a digital system.

12.18 References
1. Forney, G. D., “The Viterbi Algorithm,” Proceedings of IEEE 61(3), March 1978, pp. 268–
78.
2. Garg, V. K., and Wilkes, J. E., Wireless and Personal Communications Systems, Prentice
Hall PTR, Upper Saddle River, NJ, 1996.
282 Ch. 12 • RF Engineering and Network Planning

3. Hata, M., “Empirical Formula for Propagation Loss in Land Mobile Radio Services,” IEEE
Transactions on Vehicular Technology 29(3), 1980.
4. Jakes, W. C., ed., Microwave Mobile Communications, John Wiley, New York, 1974.
5. Motley, A. J., and Keenan, J. M., “Radio Coverage in Buildings,” British Telecom Tech.
Journal, Special Issue on Mobile Communications 8(1), January 1990, pp. 19–24.
6. Okumura, Y., Ohmori, E., Kawano, T., and Fukuda, K., “Field Strength and Its Variability in
VHF and UHF Land-Mobile Radio Service,” Rev. Elec. Communication Lab. 16, 1968, pp.
825–73.
7. Rappaport, T. S., Wireless Communications, Prentice Hall PTR, Upper Saddle River, NJ,
1996.
8. Sampei, Seiichi, Applications of Digital Wireless Technologies to Global Wireless Commu-
nications, Prentice Hall PTR, Upper Saddle River, NJ, 1997.
9. Seidel, S. Y., and Rappaport, T. S., “914 MHz Path Loss Prediction Models for Indoor Wire-
less Communications in Multifloored Buildings,” IEEE Trans., Antennas & Propagation
40(2), February 1992.
10. Sklar, B., “Rayleigh Fading Channels in Mobile Digital Communication Systems—Part I:
Characterization,” IEEE Communication Magazine 35(9), September 1997, pp. 136–46.
11. Sklar, B., “Rayleigh Fading Channels in Mobile Digital Communication Systems—Part II:
Mitigation,” IEEE Communication Magazine 35(9), September 1997, pp. 148–55.
12. Viterbi, Andrew J., “CDMA Principles of Spread Spectrum Communication,” Addison-Wes-
ley Publishing Company, Reading, MA, 1995.
13. Viterbi, Audrey M., and Viterbi, Andrew J., “Erlang Capacity of a Power Controlled CDMA
System,” IEEE Journal on Selected Areas in Communications, August 1993.
14. Walfisch, J., and Bertoni, H. L., “A Theoretical Model of UHF Propagation in Urban Envi-
ronment,” IEEE Trans., Antennas & Propagation, to be published.
15. Weissberger, M. A., “An Initial Critical Summary of Models for Predicting the Attenuation
of Radio Waves by Trees,” ESD-TR-81-101, Electromagnet Compat. Analysis Center,
Annapolis, MD, July 1982.
 C H A P T E R 1 3

Reverse and Forward

Link Capacity of IS-95
CDMA System

13.1 Introduction
The CDMA system is an interference-limited system in which link performance depends on the
ability of the receiver to detect a signal in the presence of interference. In order for a CDMA link
to perform satisfactorily, the designer must specify a frame error rate (FER). Field trials help the
designer to establish the required Eb/It values on the reverse link and various channels of the for-
ward link that will maintain the specified FER. The key issue in CDMA network design is to
minimize multiple access interference. Power control is critical to reduce multiaccess interfer-
ence. Designers must include the interference from other cells in the system to determine the
actual reuse factor in the CDMA system.
In the forward direction (BS to MS) a pilot signal is used by the mobile demodulator to
provide a coherent reference that is effective even in a fading environment because the desired
signal and pilot fade together. In the reverse direction (MS to BS), no pilot is used for power effi-
ciency considerations because, unlike the forward direction, an independent pilot would be
required for each signal. A modulation consistent with, and relatively efficient for, noncoherent
reception is used for the reverse link.
The maximum number of mobiles that can be supported on the forward link of a CDMA
system is different from the maximum number that can be supported on the reverse link. Nor-
mally, the capacity of a CDMA system depends upon the reverse link capacity. The forward link
capacity is governed by the total transmitted power of the cell site and its distribution to traffic
channels and other overhead channels including the pilot, paging, and sync channels. If the
power amplifier cannot provide enough power to the forward traffic channels, system capacity
may become forward link limited. Soft handoffs improve the capacity of the reverse link; how-
ever, they also affect the capacity of the forward link—the forward link capacity is reduced by
the number and types of soft handoffs.

283
284 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

This chapter presents procedures for calculating the capacity of the reverse and forward
links of a CDMA system. We establish the relationship for determining the pole point or asymp-
totic cell capacity that can be achieved as the power received at the base station from a mobile
approaches infinity. We also relate the ratio of the received cell power to cell noise with cell
loading. The chapter also covers a procedure to develop a link safety margin parameter for each
of the forward link channels.

13.2 Reverse Link Capacity

Consider an omnidirectional cell site serving a given set of mobiles. We divide mobiles into two
groups—the mobiles that are powered up, and those that are not powered up. The mobiles that
are powered up are further divided into four subgroups:
• Active and transmitting (i.e., mobiles in conversational mode)
• Active, but not transmitting (mobiles in nonconversational mode)
• Idle and transmitting (mobiles in access mode)
• Idle and not transmitting (mobiles in nonaccess mode)
We assume that interference at the cell site by mobiles in the access mode is typically too
small to worry about—it can be accounted for as a source of some degradation in system quality
and system capacity. We focus only on the active mobiles in our analysis. We assume there are M
mobiles that are transmitting at a given time in a cell. In a CDMA environment, for each mobile,
there are (M – 1) cochannel interferers. At the cell site, the average signal power received from
the ith mobile is Sri. This signal power provides a bit energy equal to Eb, i.e.,
Sr
E b = ------i (13.1)
R
where R = mobile transmission rate in bps.
The thermal noise power is N0Bw , where N0 is the thermal noise power spectral density (psd) and
Bw is the spreading bandwidth. The average cochannel interference psd at the base station is
given as
M–1
1
I 0 = ------
Bw ∑ ν f ⋅ Sri (13.2)
i=1

where νf = voice activity factor.

In Eq. (13.2) we assume a perfect power control on the reverse link, and the signals transmitted
from all the mobiles arrive at the base station with the same received power; in other words, Sri =
Sr for all values of i (i.e., 1 ≤ i ≤ M – 1). The total interference and thermal noise psd will be
M–1
1
I t = I 0 + N 0 = ------ ⋅
Bw ∑ ν f ⋅ Sri + N 0 (13.3)
i=1
Reverse Link Capacity 285

Recognizing that Sri = Sr , from Eq. (13.3) we get

( M – 1 ) ⋅ ν f ⋅ Sr
I t = -------------------------------------
- + N0 (13.4)
Bw
Eb /It will be given as

E Bw Sr Sr
-----b- =  ------ ⋅ ---------------------------------------------------------------- = G p ⋅ ---------------------------------------------------------------- (13.5)
It  
R [ N 0 Bw + ( M – 1 ) ⋅ ν f ⋅ Sr ] [ N 0 Bw + ( M – 1 ) ⋅ ν f ⋅ Sr ]

where Gp = processing gain = Bw /R

Next we express the signal strength Sr in dB as
S r = P m + G m + G b + G dv + G sho + L p + M fade + L b + L pent + L c (13.6)

where Gm = transmit antenna gain of the mobile (dB)

Gb = receive antenna gain of the base station (dB)
Gdv = base station antenna diversity (dB)
Gsho = soft-handoff gain
Lb = body loss (dB)
Lc = cable connection loss (dB)
Lp = path loss (dB)
Lpent = penetration loss through a vehicle or building
Mfade = log-normal shadow margin (dB)
Pm = transmit power of the mobile (dB)
Solving Eq. (13.5) for M, we get

1 N 0 ⋅ Bw
M = 1 + G p ⋅ ---------------------------- – -----------------
- (13.7)
( Eb ⁄ I t ) ⋅ ν f Sr ⋅ ν f
and solving Eq. (13.5) for Sr , we get:
( Eb ⁄ I t ) ⋅ N 0
S r = -----------------------------------------------------
-
1 ( M – 1 )ν f ( E b ⁄ I t ) (13.8)
--- – ------------------------------------------- -
R Bw
If we include an interference factor f (see section 13.4) from the other cells, we can rewrite
Eq. (13.5) as
E Sr
-----b- = G p ⋅ --------------------------------------------------------------------------------- (13.9)
It N 0 ⋅ Bw + ( M – 1 ) ⋅ ν f ⋅ Sr ( 1 + f )

We also include an imperfect power control factor, ηc, and rewrite Eq. (13.9) as

E Sr
-----b- = G p ⋅ ---------------------------------------------------------------------------------------------------- (13.10)
It Bw ⋅ N 0 + ( M – 1 ) ⋅ ν f ⋅ ( Sr ⁄ ηc ) ⋅ ( 1 + f )
286 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

Solving Eq. (13.10) for M, we get

ηc N 0 ⋅ Bw ⋅ ηc
M = 1 + G p ⋅ --------------------------------------------------- – ------------------------------------
- (13.11)
( Eb ⁄ I t ) ⋅ ν f ⋅ ( 1 + f ) Sr ⋅ ν f ⋅ ( 1 + f )

Solving Eq. (13.10) for Sr , we get

( Eb ⁄ I t ) ⋅ N 0
S r = ----------------------------------------------------------------------------
-
1 ( M – 1 ) ( ν f ) ( 1 + f ) ( Eb ⁄ I t ) (13.12)
--- – -------------------------------------------------------------------
R Bw ⋅ ηc

From Eq. (13.11) the maximum value of M is given as

ηc
M max = 1 + G p ⋅ ------------------------------------------------------- (13.13)
( Eb ⁄ I t ) ⋅ ( ν f ) ⋅ ( 1 + f )

Mmax is called the pole point or asymptotic cell capacity that is achieved when Sr → ∞. For sim-
plification we neglect 1 and rewrite Eq. (13.13) as

ηc
M max ≈ G p ⋅ ------------------------------------------------------- (13.14)
( Eb ⁄ I t ) ⋅ ( ν f ) ⋅ ( 1 + f )

To gain further insight into the capacity dynamic, we can rewrite Eq. (13.12) as
Sr ⁄ ηc 1
-------------- = ------------------------------------------------------------------ (13.15)
N 0 Bw M max ⋅ ν f ⋅ ( 1 + f ) ⋅ ( 1 – ρ )

where ρ = M/Mmax = cell loading factor

Equation (13.15) is plotted in Fig. 13-1. The required SNR per mobile increases in a nonlin-
ear fashion with ρ. The power per mobile is much larger for a heavily loaded cell than it is for the
lightly loaded cell. Practical capacity limits may be set at a point where the slope becomes too
steep (such as ρ = 0.8). Also, for the smallest practical value of Mmax – M to be equal to 1, Eq.
(13.15) indicates that the maximum received power per mobile is close to the cell-site noise level.
We further adjust Eq. (13.15) to reflect the total received power (interference) equal to Prec
= νf · (1 + f ) · (Sr /ηc) · M. The ratio of Prec to cell-site noise can be expressed in terms of the load-
ing factor ρ. Eq. (13.16) shows that the ratio of interference and noise also rises in a nonlinear
fashion with ρ (see Fig. 13-2).

P rec ρ
------------- = ------------ (13.16)
N 0 Bw 1–ρ

Since total power Ptotal is equal to Prec + N0 Bw , we can rewrite Eq. (13.16) as

P total P rec + N 0 B w ρ 1
------------- = ----------------------------
- = ------------ + 1 = ------------ (13.17)
N 0 Bw N 0 Bw 1–ρ 1–ρ
Reverse Link Capacity 287

0.6

0.5
Signal-to-Noise Ratio (SNR)

0.4

0.3

0.2

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Cell Loading (%) ρ

Figure 13-1 SNR vs. Cell Loading

10

8
Total Received Power-to-Noise Ratio

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Cell Loading (%) ρ

Figure 13-2 Total Received Power-to-Noise Ratio vs. Cell Loading

288 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

Typical average values of different parameters are

νf = 0.5 (0.4 → 0.6)
Eb/It = 6 → 7 dB (4 to 5)
f = 0.67 (0.56 → 1.28) based on path-loss exponent γ = 4 and standard deviation of path
loss of 6 to 10 dB
ηc = 0.8 (0.7 → 0.85)
Using these average values, we estimate the cell capacity as
6
1.23 × 10 0.8
M max = ------------------------- ⋅ ----------------------------- ≈ 31 (with Eb /It = 6 dB)
9600 × 4 0.5 × ( 1.67 )
and
6
1.23 × 10 0.8
M max = ------------------------- ⋅ ----------------------------- ≈ 25 (with Eb /It = 7 dB)
9600 × 5 0.5 × ( 1.67 )
With a three-sector antenna, a practical gain of 2.55 can be achieved; the capacity range
per sector will be

2.55 2.55
M sector = 31 × ---------- ≈ 26 or M sector = 25 × ---------- ≈ 21
3 3
The capacity range of a sector will be 21 to 26. In practice, sector loading is often limited to 0.5–
0.7 of the calculated values, giving an average number of mobiles per sector equal to 13–16.

EXAMPLE 13.1
Calculate the required Eb /It using the following parameters: Bw = 1.23 MHz; R = 9.6 kbps; Pm = 63
mW (18 dBm); Lc = –2 dB; Gm = 0 dB; Lp = –135 dB; Mfade = –8 dB; Gb = 9 dB; F (noise figure) = 5
dB; T = 290 degrees Kelvin (K); kb = Boltzmann’s constant = 1.380662 × 10–23; νf = 0.4; M = 20.
Assume all other parameters to be 0.
The received signal power will be
S r = P m + L c + G m + G b + L p + M fade

S r = 18 + ( – 2 ) + ( 0 ) + ( 9 ) + ( – 135 ) + ( – 8 ) = – 118 dBm

– 23 – 20 – 17
N 0 = FT k b = 3.16228 × 290 × 1.380662 × 10 = 1.266 × 10 W = 1.266 × 10 mW

E B Sr
-----b- =  -----w- ------------------------------------------------------
-
It  R  [ N 0 B w + ( M – 1 )ν f S r ]
6 – 11.8
E 1.23 × 10 10
3
- × -------------------------------------------------------------------------------------------------------------------
-----b- = ------------------------ – 17 6 – 11.8
- dB
It 9.6 × 10 [ 1.266 × 10 × 1.23 × 10 + 19 × 0.4 × 10 ]
Reverse Link Capacity 289

– 11.8 – 0.8
E 128 × 10 128 × 10
-----b- = --------------------------------------------------------
– 11
- = ---------------------------- = 7.345 = 8.66 dB
It 10 [ 1.5557 + 1.2045 ] 2.7602

EXAMPLE 13.2
For the IS-95 CDMA system, a chip rate of 1.2288 Mcps is specified for the data rate of 9.6 kbps
(i.e., 8-kbps vocoder). The required Eb /It is specified as 7.0 dB. Calculate the pole capacity. What is
the average number of mobiles that can be supported by a sector of the three-sector cell?. Assume:
interference factor from the neighboring cells f = 0.55; the voice activity factor νf = 0.5; the power
control accuracy factor = 0.80; the gain due to sectorization = 2.55. Assume all other parameters to
be 0. How much reduction in sector capacity will occur with a 13.0-kbps vocoder provided all other
things remain unchanged?
ηc
M max ≈ G p ⋅ ------------------------------------------------------
-
( Eb ⁄ I t ) ⋅ ( ν f ) ⋅ ( 1 + f )
6
B 1.23 × 10 E
G p = -----w- = ------------------------
3
- = 128 , -----b- = 7 dB = 5.0
R 9.6 × 10 It

128 × 0.80
M max = ---------------------------------------------------- = 26.42
5.0 × 0.5 × ( 1 + 0.55 )
26.42 × 2.55
Average subscriber/sector = ------------------------------ = 22.46 ≈ 22
3
With a 13.0-kbps vocoder ( data rate R = 14.4 kbps), the processing gain will be
6
1.23 × 10
G p = ------------------------3- = 85.4
14.4 × 10
The reduction in sector capacity will be
128 – 85.4
= ------------------------- × 100 = 33.28%
128

EXAMPLE 13.3
A total of 36 equal-power mobiles share a frequency band through a CDMA system. Each mobile
transmits information at 9.6 kbps with a DSSS BPSK-modulated signal. Calculate the minimum
chip rate of the PN code in order to maintain a bit error probability of 10–3. Assume: interference fac-
tor from other cells f = 0.60; voice activity factor νf = 0.5; power control accuracy factor = 0.8.
–E ⁄ I
2E e b t
Bit error probability for BPSK P b = Q  --------b- ≈ ----------------------------- = 10
–3
 It  2 π( E ⁄ I )
b t

E
Required -----b- ≈ 4.8 = 6.8 dB
It
290 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

Gp 1 1
- × ------------ × ----- × 0.8
M = 36 = ------------
Eb ⁄ I t 1 + f v f

G 1 1
------p- × ------- × ------- × 0.8 = 36
4.8 1.6 0.5
∴G p = 172.8
3
Chip rate = 172.8 × 9.6 × 10 = 1.6588 Mcps

EXAMPLE 13.4
Calculate the Erlang capacity of a sector for a three-sector cell using the following parameters: car-
rier bandwidth, Bw = 1.23 MHz; RS2 R = 14.4 kbps; required Eb /It = 7 dB; voice activity factor νf =
0.4; interference due to other cells f = 0.6; three-sector antenna gain α = 2.61; cell loading factor ρ =
0.54; outage or call blocking probability Pout = 2%.
6
B 1.23 × 10
G p = -----w- = ------------------------3- = 85.4
R 14.4 × 10

Gp 1 1 85.4 1 1
M max = 1 + ------------------------- ⋅ ----- ⋅ ------------ ⋅ α = 1 + ------------- ⋅ ------- ⋅ ---------------- ⋅ 2.61 = 70.4
( E b ⁄ I t ) reqd v f 1 + f 5.012 0.4 1 + 0.6

Mmax per sector = 70.4 ⁄ 3 = 23.48 ≈ 24

M = 0.54 × 24 ≈ 13
From the Erlang B table at 2% blocking probability for M = 13 channels, the capacity = 7.4
Erlangs.

13.3 Multicell Network

In a CDMA multiple-cell (multicell) network the same frequency band is used in all cells unlike
other access technologies in which the bandwidth used in a given cell is reused only in cells that
are sufficiently far away to avoid cochannel interference. To compare CDMA with other multi-
ple access schemes, capacity is determined (or measured) as the total number of users in the
multicell network rather than the number of users per bandwidth or per isolated cell.
The CDMA system is an interference-limited system. CDMA link performance depends
on the ability of the receiver to discern a signal in the presence of interference. In order for the
performance of a CDMA link to be satisfactory, an FER of about 1% is recommended. Field tri-
als are conducted to establish the required Eb /It value on the reverse link and various channels of
the forward link to maintain the recommended FER. The link budget is established to achieve
the value of Eb /It. The required value of Eb /It depends upon the propagation environment and the
speed of the mobile. Based on field trials, the following values of Eb /It are suggested:
Intercell Interference 291

• Low-speed mobiles, speed ≤ 5 mph: 5 dB. In this case, the duration of fades is much
larger than the time between power control updates for a mobile. Thus, the effect of any
fade is compensated by a quick response of the power control mechanism.
• Medium-speed mobiles, speed ≈ 30 mph: 7 dB. The advantages of high or low speed
are not applicable; therefore, the required Eb /It is somewhat higher.
• High-speed mobiles, speed ≥ 60 mph: 6 to 6.5 dB. In this case, the fade duration is
smaller compared to chip length. Thus, only burst errors occur on the links that are cor-
rected by interleaving and Viterbi decoding. Therefore the required Eb /It is low.
The key issue in a CDMA network design is to minimize multiple access interference.
Power control is critical to multiaccess interference. Each cell controls the transmit power of its
own mobiles. However, a serving cell is unable to control the power of mobiles in the neighbor-
ing cells. The mobiles in the neighboring cells introduce additional interference, thereby reduc-
ing the capacity of the reverse link. In Eq. (13.10) we include this effect by a factor f. The
interference from other cells determines the actual reuse factor of the CDMA system. CDMA
networks are designed to tolerate a certain amount of interference and, therefore, have a capacity
advantage over TDMA or FDMA in this regard.

13.4 Intercell Interference

The intercell interference factor f is difficult to evaluate because the serving cell does not have
control over the power received from mobiles in other cells. The f depends on the geometry of
the serving cell and neighboring cells. It will be small if the serving cell radius is large, if the
path-loss slope has a higher value, or if the standard deviation of path loss is small.
For γ = 4 and a standard deviation of path loss σ = 8 dB, the upper bound on f is 0.77.
Table 13-1 lists the value of f with two-way and three-way soft handoff for different values of σ
and γ = 4 [5].
The cell loading is a measure of the total interference I0 allowed in the system in reference
to thermal noise.

Table 13-1 Intercell Interference factor f for γ = 4

Other Cell Interference Factor f
σ dB Two-Way Soft Handoff Three-Way Soft Handoff
0 0.44 0.44
2 0.43 0.43
4 0.47 0.45
6 0.56 0.49
8 0.77 0.57
10 1.28 0.75
12 2.62 1.17
292 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

M I0
ρ = ------------ ≈ ----------------- (13.18)
M max I 0 + N 0

where M = active number of mobiles in the cell, and

Mmax = maximum possible mobiles in the cell.
ρ equal to 0.5 implies that the interference in the system is equal to the thermal noise level.
ρ < 0.5 implies that the system is noise limited, whereas ρ > 0.5 indicates that the system is
interference limited. Typically, a value of ρ between 0.5 and 0.7 is used.

13.5 Erlang Capacity of a Single Cell

To calculate the Erlang capacity of a single cell in a CDMA system, we assume that the number
of active users M can be modeled by Poison distribution.
M
(λ ⁄ µ) –λ ⁄ µ
p m = ------------------- ⋅ e (13.19)
M!
where λ/µ = offered average traffic load in Erlang,
λ = average arrival rate of users, and
1/µ = average time per call.
The call service time τ per user is assumed to be exponentially distributed, so that the probability
that τ exceeds T is given as
– µT
pr ( τ > T ) = e T>0 (13.20)

Using these assumptions, it has been shown in reference [5] that the blocking or outage proba-
bility pout is
∞
– ( λv f ) ⁄ µ v f λ K 1
 --------  ∆ r' – ( v f λ ) ⁄ µ
p out = e ⋅ ∑  µ  K!
⋅ ------ ≈ Q  ---------------------------------
 (v λ) ⁄ µ 
(13.21)
K ∆ r' f

G p(1 – η)
where ∆' r = ------------------------ , and
( E b ⁄ I t ) sp
1
--- = the ratio of total interference plus thermal noise power to thermal noise power.
η

By taking into account the interference from other cells and an imperfect power control,
Eq. (13.21) can be modified as
2
( βσ c ) ⁄ 2
∆ r' – v f ( λ ⁄ µ ) ( 1 + f )e
p out ≈ Q -----------------------------------------------------------------------2- (13.22)
( βσ c )
v f ⋅ (λ ⁄ µ) ⋅ (1 + f ) ⋅ e
Erlang Capacity of a Single Cell 293

where β = (ln10)/10, and

σc = standard deviation of power control.
We may invert the approximate expression for blocking probability Eq. (13.22) by solving a
quadratic equation, to obtain the explicit formula for normalized average user occupancy, λ/µ, in
terms of Erlangs per sector [3,6] as

( λ ⁄ µ ) ⋅ v f ⋅ ( 1 + f ) = ∆ r' ⋅ F ( B, σ c ) (13.23)

–1 2
[ Q ( P out ) ]
where B = ------------------------------ , and
∆ r'
3
1 αc B  4 
F(B, σc ) = ----- ⋅ 1 + ----------  1 – 1 + ---------
- in which
αc 2  α B
3
c
2
( βσ c ) ⁄ 2
αc =e ; β = ( 1n10 ) ⁄ 10 = 0.2303

EXAMPLE 13.5
Find the Erlang capacity of a CDMA cell assuming
• Blocking or outage probability (Pout) = 1%
• Log-normal shadowing margin (Mfade) = 8 dB
• Path-loss exponent (γ) = 4
• Voice activity factor (vf) = 0.4
• Other cell interference factor (f) = 0.55
• Spreading bandwidth (Bw ) = 1.23 MHz
• Data rate (R) = 9.6 kbps
• (Eb /It)sp = 7 dB = 5.0
• 1/η = 10
• σc = 2 dB = 1.5849
6
–1 Gp 1.23 × 10 ( 1 – 0.1 )
- ⋅ ( 1 – η ) = ------------------------
Q ( 0.01 ) = 2.33 ; ∆ r' = --------------------- - ⋅ --------------------- = 23.04
( E b ⁄ I t ) sp 9.6 × 10
3 5
2
( 0.2322 × 1.5849 ) ⁄ 2
αc = e = 1.0701
2
( 2.33 )
B = ----------------- = 0.2356
23.04
3
( 1.0701 ) × 0.2356
F ( B, σ c ) = ---------------- 1 + ----------------------------------------------  1 – 1 + ---------------------------------------------
-
1 4
= 0.5494
1.0701 2  ( 1.0701 ) × 0.2356
3

λ 23.04 × 0.5494
--- = ------------------------------------ = 20.42 Erlangs
µ 0.4 ⋅ ( 1 + 0.55 )
The number of users from the Erlang B table at 1% blocking ~ 30.
294 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

13.6 Forward Link Capacity

An important feature of CDMA that contributes to the added capacity on the reverse link is soft
handoff [2]. In a CDMA network, a mobile can be served by multiple cells simultaneously.
However, that same feature puts an additional burden on the forward link. Since multiple cells
have to provide service to the same mobile, additional resources are allocated on the forward
link. Forward link performance differs vastly from that of the reverse link because
• Access is one to many instead of many to one.
• Synchronization and coherent detections are facilitated by use of a common pilot
channel.
• The interference is received from a few concentrated large sources (cells) rather than
many distributed small ones (mobiles).
To maximize the capacity of the forward link, it is essential to control the power of the cell
so that power can be allocated to individual mobiles according to their needs. More power is
provided to those mobiles that receive the highest interference from neighboring cells. Mobiles
on the boundaries may be in soft handoff, in which case they also receive signal power from two
or more cells. Power control on the forward link is accomplished by measuring the mobile
power received from its serving cell and the total received power. The information about these
two power values is transmitted to the serving cell.
For the forward link a figure of merit is defined for various channels. The figure of merit is
the difference between the received (rec) and specified (sp) Eb /It . The link safety margin param-
eter for each of the channels on the forward link is defined as
M pilot = ( E c ⁄ I t ) rec – ( E c ⁄ I t ) sp > 0 (13.24a)

M traffic = ( E b ⁄ I t ) rec – ( E b ⁄ I t ) sp > 0 (13.24b)

M sync = ( E b ⁄ I t ) rec – ( E b ⁄ I t ) sp > 0 (13.24c)

M paging = ( E b ⁄ I t ) rec – ( E b ⁄ I t ) sp > 0 (13.24d)

Note for the pilot channel Ec /It is used instead of Eb /It —this is because the pilot channel does not
carry any information. Energy per chip, Ec , is used, the chip rate being 1.2288 Mcps.
The forward link budget is used to confirm that quantities in Eqs. (13.24a–d) are positive
and that there is sufficient margin for the forward link to perform efficiently. Of Mpilot, Mtraffic,
Msync, and Mpaging, the first two are more critical. If these two are positive, then the other two are
also likely to be positive. For perfect link balance, all margin parameters should be 0, particu-
larly Mpilot and Mtraffic. The suggested values for the specified Eb /It and Ec /It parameters are
• Pilot channel: (Ec /It )sp = –15 dB
• Traffic channel: (Eb /It )sp = 7 dB
• Sync channel: (Eb /It )sp = 7 dB
• Paging channel: (Eb /It )sp = 7 dB
Forward Link Capacity 295

We use the following assumptions for the CDMA forward link budget:
1. All mobiles are
◆ at the cell edge
◆ at least in two-way soft handoff
◆ traveling at a medium speed
◆ (Eb /It ) = 7 dB for 1% FER
2. Power control is working perfectly for all mobiles.
3. Total forward link traffic channels’ power is equally divided among all mobiles.
Forward link capacity depends on the power that is available for the traffic channels. The
power allocation to each overhead channel (i.e., Ppilot, Psync, and Ppaging) is determined from field
tests. The suggested power allocations for the forward link channels are
• Ppilot = 15–20% Pcell-site
• Psync = 10% of Ppilot = 1.5–2% Pcell-site
• Ppaging = 30–40% of Ppilot = 7% Pcell-site
• Ptraffic = [1 – (0.2 + 0.02 + 0.07)] = 71–76.5% Pcell-site
Note that Ppaging and Ptraffic represent the total allocated power for all the paging and traffic chan-
nels, respectively, and Pcell-site is the total transmit power of the cell site.

P ( traffic ) ⁄ ( mobile ) = P traffic ⁄ ( M total ⋅ α chan ) (13.25)

M total = M ( 1 + ξ co ) (13.26)

where M = number of active mobiles per sector,

ξco = channel overhead factor for extra traffic channels required for mobiles in differ-
ent types of soft handoffs (see Table 13-2), and
αchan = channel activity factor.

P ( paging ) ⁄ ( channel ) = P paging ⁄ N p (13.27)

where Np = number of paging channels.

P(traffic)/(mobile) is a nominal value. Actual power allocated for each mobile can be up to ±4 dB
around this value depending on the forward link power control for each mobile. On the forward
link, extra traffic channels are required for the mobiles in various types of soft handoffs. The per-
centage of the coverage area in handoff is a design criterion. The extra number of traffic channels
in handoff can be related to the area in handoff. Table 13-2 provides the suggested values.

13.6.1 Pilot Channel

The mobile measures the Ec /It of the pilot channel continuously and compares it against
threshold values of the handoff parameters T_ADD and T_DROP (Ec is the energy per chip and
296 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

Table 13-2 Channel Overhead Factor for Various Types of Soft Handoffs
Type of Handoff % Area in Handoff ξco

Soft 25% 0.25

Softer 20% 0.20
Soft-soft 10% 0.20
Soft-softer 10% 0.20
Total ξco — 0.85

It is the interference plus noise density measured on the pilot channel). The mobile reports the
results of these comparisons to the serving cell. The serving cell decides whether or not the
mobile needs handoff. The Ec /It of the pilot channel is important for determining whether or not
the mobile is within the coverage area of the particular cell. The pilot signal from a cell is trans-
mitted at a relatively higher power than those of other forward link logical channels (i.e., paging,
sync, traffic). In order to set up a call, the mobile must receive the pilot signal successfully. The
pilot channel acts as a coherent carrier phase reference for demodulation of other logical chan-
nels on the forward link. Since the Ec /It effectively determines the coverage area of a cell or sec-
tor, it is essential that the Ec /It be sufficiently large.

13.6.2 Traffic Channel

Let (S1)m equal the power received by the mth mobile from the cell/sector providing maxi-
mum power (i.e., serving cell), and let (S2 )m … (SQ)m equal the power received by the mth
mobile from neighboring cells.
Thus

( S1 )m > ( S2 )m … > ( SQ )m > 0 (13.28)

We assume that the power received from Q cells or sectors is significant and that all other
cells’ power is negligible. We assume that all cell sites beyond the second ring around a serving
cell contribute negligible received power, so that Q ≤ 18. The received bit energy-to-interference
plus thermal noise for the mth mobile will be [4]

 
 Bw ωm ( S1 )m 
E b Φ ⋅ ------ ⋅ ----------------------------------------------
 t R Q ( S j )m - 
 -----
- ≥
 I t  m   (13.29)
 ∑ ( S j ) m + N 0 B w

 j=1 
where Φt = fraction of total cell-site power assigned to traffic channels,
(1 – Φt) = fraction of total cell power that is assigned to transmission of overhead chan-
nels (pilot, sync, and paging channels),
Forward Link Capacity 297

N0 = thermal noise density,

Bw = spreading bandwidth,
R = data rate,
Gp = processing gain = Bw /R,
ωi = fraction of total power allocated to the ith mobile,
Note: the weighting factor ωi is proportional to the total sum of other base
station powers, S2, S3, …, SQ, relative to the mobile’s own base station
power S1
M = number of users in mobile’s own cell or sector.
M

∑ ωi ≤ 1 (13.30)
i=1

From Eq (13.29) the weighting factor ωm is given as

 Q 

( Eb ⁄ I t )m ∑ j  + ------------
ω m ≤ --------------------- 1 +  j = 2
( S ) σn
2
- (13.31)
Φt G p  ------------------- ( S1 )m
 ( S1 ) 
 m

where σn2 = thermal power.

Since Φt S1 is the maximum total power allocated to the cell/sector containing the given
mobile and M is the total number of mobiles in the cell/sector, we define the relative received
cell power as

 Q 
 
= 1 +  ∑ j
S
fm (13.32)
 -------------
j=2 
 S -
 1 m
Next we combine Eq. (13.30) and Eq. (13.31) to get
M M 2
G p Φt σn
∑ f i ≤ -------------
Eb ⁄ I t
– ∑ ----------
( S1 )i
- = ∆f (13.33)
i=1 i=1

Generally the background noise is well below the total largest received cell-site signal power
and the second term in Eq. (13.33) is typically negligible relative to the first term. The capacity
can be estimated from the outage or blocking probability, defined as
P out = p r [ BER > ( BER ) sp ] (13.34)

where (BER)sp = specified bit error rate for which Eb /It is equal to (Eb / It)sp.
We compute ∆ 'f for (Eb /It )sp from Eq. (13.33) and express the outage or blocking probability as
298 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

M
p out = p r ∑ f i > ∆ f' (13.35)
i=1
M
The distribution of ∑ f i cannot be expressed in a closed form. The simulation results for
i=1
the blocking probability for the forward link of IS-95 are shown in Fig. 13-3 for Gp = 128, with
20% of the transmitted power in the cell/sector to the pilot channel and with the required Eb /It =
5 dB for the traffic channel to ensure BER ≤ 10–3 [2]. The reduction of 2 dB relative to reverse
link is justified by the coherent reception using the pilot as a reference, as compared to the non-
coherent detection in the reverse link. In the simulation, powers from base stations were repre-
sented as the product of the fourth order of distance and a log-normally distributed attenuation.
With these parameters, the forward link can support the BER of 10–3 for more than 99% of the
time for 38 mobiles per sector or 114 mobiles per cell.

EXAMPLE 13.6
Using the following parameters of an IS-95 CDMA system, estimate the sector capacity based on
reverse and forward link performance.
• Spreading bandwidth (Bw ) = 1.23 MHz
• Data rate (R) = 9.6 kbps
• 1/η = 10
• Outage or blocking probability (Pout) = 1%
• (Eb /It)sp for the reverse link = 7 dB = 5.0
• (Eb/It)sp for the forward link = 5 dB

0.1
Bit Error Rate Probability
Pr (BER) > 0.001)

0.01

0.001

0.0001
30 35 38 40 45
Number of Users/Sector
Figure 13-3 Forward Link Capacity of a Cellular CDMA System
Forward Link Capacity 299

• Standard deviation of power control (σc) = 2.5 dB = 1.7783

• Voice activity factor (vf ) = 0.5
• Interference from other cells (f ) = 0.65
• Path-loss exponent (γ) = 4
• Standard deviation for shadow margin = 8 dB
• Channel overhead factor for soft handoffs (ξco) = 0.85
Reverse Link
–1
Q ( 0.01 ) = 2.33
6
1.23 × 10 ( 1 – 0.1 )
∆ r' = ------------------------
3
- × --------------------- = 23.04
9.6 × 10 5
2
B = ( 2.33 ) ⁄ 23.04 = 0.2356
2
( 0.2303 × 1.7783 ) ⁄ 2
αc = e = 1.8075

3
( 1.0875 ) × 0.2356
F ( B, σ c ) = ---------------- 1 +  ----------------------------------------------  1 – 1 + ---------------------------------------------
-
1 4
= 0.5339
1.0875  2  ( 1.0875 ) × 0.2356
3

0
For perfect power control α c = e = 1

F ( B, σ c ) = 1 + ----------------  1 – 1 + ---------------- = 0.6183

0.2356 4
2  0.2356
0.5339
Efficiency of power control: η c = ---------------- = 0.8635
0.6183
λ 23.04 × 0.5339
--- = -------------------------------------- = 14.91 Erlangs
µ 0.5 × ( 1 + 0.65 )

 --λ- 14.91
= ---------------- = 17.27 Erlangs
 µ perfect 0.8635
From the Erlang B table at 1% blocking with 14.91 Erlangs, ≈ 23 mobiles can be supported per sec-
tor, whereas with perfect power control we can support ≈ 27 mobiles. We lose about 15% of the sec-
tor capacity due to imperfect power control. With a loading factor of about 70%, the reverse link
capacity will be about 16 mobiles per sector.
Forward Link
From Fig. 13-3 we can see that the forward link can support 38 mobiles per sector with perfect
power control. If we assume the same accuracy of the power control as on the reverse link, the sector
capacity will be reduced to
No. of mobiles per sector ≈ 38 × 0.8635 = 33
Next we consider the effect of soft handoffs on forward link capacity. If the channel overhead
factor is ξco = 0.85, the sector capacity based on the performance of the forward link will be
33
Sector capacity = ---------- ≈ 18
1.85
300 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

In this example, the sector capacity is controlled by forward link performance. In practice, a
loading factor of about 70% is often suggested; this will give the sector capacity about 13 mobiles.
Note that the sector capacity of the forward link is significantly affected by total percentage and dis-
tribution of soft handoffs.

13.7 CDMA Cell Size

For CDMA systems with AWGN channels, the signal-to-interface ratio (SIR)* is an accurate
measure of performance [1]. The SIR, the number of users in the cell/sector M, and the maxi-
mum mobile transmit power establish the size of the cell on the reverse link.
In a noise- (coverage-) limited system, cell size is the main concern rather than capacity.
Receiver sensitivity is used to calculate the size of the cell. Fading margin, propagation path
loss, and the minimum received signal level needed to achieve an acceptable performance are
the key factors in establishing the cell size. Propagation path losses are determined either from
statistical propagation models or from actual measurements.
In an interference- (capacity-) limited system, the size of the cell is determined mainly by
the level of the interference from other users. On the reverse link, the maximum path loss that
the cell can tolerate is determined by the SIR, the number of simultaneous users, and the maxi-
mum power that the mobile can transmit. The maximum cell size should be such that the mobile
can close the link.
In a noise-limited system, the minimum SIR can be translated to a minimum signal
strength requirement on the forward link. The forward link cell boundary is defined by pilot Ec /It.
The maximum cell size should be such that, within the coverage area, the received pilot Ec /It
should be above a predefined threshold.
• Reverse link cell size. At the cell site, the SIR per antenna can be given by

p m ⋅ L p ( r ) ⋅ G' b ⋅ G' m
SIR ( r ) = ---------------------------------------------------------------------------------------------------------------- (13.36)
[ N 0 B w ] cell + [ M ⁄ f r – 1 ] ( v f ⋅ p m )L p ( r )G' b G' m

where pm = mobile’s power amplifier output,

G'b = cell antenna gain including cable losses,
G'm = mobile antenna gain including cable losses,
Lp(r) = reverse link path loss,
M = number of users in a cell/sector,
vf = voice activity factor,
fr = frequency reuse factor, and
[N0 Bw ]cell = thermal noise of the cell.
The quantity [N0 Bw ]cell + [M/fr – 1] vf pm(r) Lp(r) G'b G'm } depends only on system loading. It has
been shown that [7]

* SIR includes both interference and thermal noise.

CDMA Cell Size 301

[ M ⁄ f r – 1 ] ( [ v f ⋅ p m ] ⋅ L p ( r )G' b G' m ) 1
1 + -----------------------------------------------------------------------------------------
- = ------------ (13.37)
N 0 Bw 1–ρ

where ρ = system loading factor.

The maximum path losses that the mobile can tolerate are given by

( SIR ) [ N 0 B w ] cell  ------------

1
 1 – ρ
L p ( r ) = ---------------------------------------------------------- (13.38)
p m G' b G' m

We express Eq. (13.38) in dB to get

L p ( r ) = ( SIR ) min + [ N 0 B w ] cell – p m – G' b – G' m – 10 log ( 1 – ρ ) (13.39)

The maximum transmission loss will be

T ( r ) = L p ( r ) + G' b + G' m (13.40)

• Forward link cell size. On the forward link, the parameter that determines the cell
size is the pilot Ec /It which is given as

E φ p ⋅ p c ⋅ L p ( r ) ⋅ G' b ⋅ G' m φ p ⋅ pc ⋅ T ( r )
-----c = -------------------------------------------------------------------------------
- = -------------------------------------------------------------------------
-
It [ N 0 B w ] mob + I oc ( r )B w + I 0 ( r )B w [ N 0 B w ] mob + I 0 ( r )B w ⋅ ( 1 + ξ )
φ p ⋅ pc ⋅ T ( r )
= ------------------------------------------------------------------
- (13.41)
[ N 0 B w ] mob + p c T ( r ) ( 1 + ξ )

Solving for T(r) we get

( E c ⁄ I t ) ( N 0 B w ) mob
T ( r ) = ---------------------------------------------------------
- (13.42)
pc [ φ p – ( Ec ⁄ I t ) ( 1 + ξ ) ]

We express Eq. (13.42) in dB to get

( Ec ⁄ I t ) ⁄ 10
ξ ⁄ 10
T ( r ) = ( E c ⁄ I t ) min + ( N 0 B w ) mob – p c – 10 log [ φ p – ( 10 ) ( 1 + 10 )]
min
(13.43)

( Ec ⁄ I t ) ⁄ 10
ξ ⁄ 10
L p ( r ) = ( E c ⁄ I t ) min + ( N 0 B w ) mob – p c – 10 log [ φ p – ( 10 ) ( 1 + 10 )]
min

– G' b – G' m (13.44)

where φp = fraction of the cell power allocated to the pilot,

pc = cell output,
G'b = cell antenna gain including cable losses,
G'm = mobile antenna gain including cable losses,
Ioc(r) = other cell interference power spectral density,
302 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

Io(r) = serving cell interference power spectral density,

ξ = Ioc /Io,
[N0 Bw ]mob = thermal noise at the mobile, and
(Ec /It )min = minimum required value for pilot.

EXAMPLE 13.7
Calculate transmission loss vs. cell loading for a 200-mW mobile unit. Assume: (Eb /It )min = 7 dB;
processing gain at the cell site = 21 dB; cell noise figure = 5 dB.
(SIR)min + processing gain = (Eb /It )min
(SIR)min = 7 – 21 = –14 dB
– 23 6 3
( N 0 B w ) cell = [ 3.1622 × 290 × 1.38066 × 10 × 1.2288 × 10 ] × 10 mW = –108 dBm

T ( r ) = ( SIR ) min + ( N 0 B w ) cell – p m – 10 log ( 1 – ρ ) = – 14 – 108 – 23 + 10 log ( 1 – ρ )

= – 145 – 10 log ( 1 – ρ ) dB
Fig. 13-4 shows a plot of T(r) vs. ρ.

EXAMPLE 13.8
Plot maximum transmission loss (dB) vs. % of power allocated to the pilot channel. Assume:
(Ec /It )min = –15 dB; mobile noise figure = 8 dB; Ioc /Io ~ 2.5 dB; cell-site output = 44 dBm.
( E c ⁄ I t ) min ⁄ 10 I oc ⁄ I o
T ( r ) = ( E c ⁄ I t ) min – p c + ( N 0 B w ) mob – 10 log [ φ p – ( 10 ) ( 1 + 10 )]

T ( r ) = – 15 – 44 – 105 – 10 log [ φ p – 0.03162 × ( 1 + 1.7783 ) ]

T ( r ) = – 164 – 10 log [ φ p – 0.08785 ] dB

Fig. 13-5 shows a plot of T(r) vs. φp.

13.8 Forward and Reverse Link Balance

A more powerful forward link results in extra interference to mobiles in other cells. On the other
hand, a more powerful reverse link sacrifices capacity. It is desirable then to design the system so
that the two boundaries coincide. Balanced links minimize interference and eliminate associated
handoff problems. The boundary of the cell on the reverse link is determined by cell loading, and
the boundary on the forward link is obtained by a minimum pilot (Ec /It ). In order to keep the two
boundaries close, we need to equate the path loss on both links. From Eqs (13.39) and (13.44)
we define the balance factor Bf as

B f = [ ( SIR ) min – ( E c ⁄ I t ) min ] + [ N 0 F cell – N 0 F mob ] ⋅ B w + [ p c – p m ]

( Ec ⁄ I t ) ⁄ 10 ( I oc ⁄ I o ) ⁄ 10
 ( φ p – 10 ) ( 1 + 10 )
min

+ 10 log  -------------------------------------------------------------------------------------------- (13.45)

 1 – ρ 
Forward and Reverse Link Balance 303

-134

-136
Reverse Link Max. Transmission Loss (dB)

-138

-140

-142

-144

-146
0 0.1 0.2 0.3 0.4Cell Loading
0.5 (%) P 0.6 0.7 0.8 0.9 1

Cell Loading (%) ρ

Figure 13-4 Reverse Link Maximum Transmission Loss vs. Cell Loading

-144

-146

-148
Max. Transmission Loss (dB)

-150

-152

-154

-156

-158

-160
0 5 10 15 20 25 30 35 40 45
% of Power Allocated to Pilot Channel φp
% of Power Allocated to Pilot Channel φp

Figure 13-5 Transmission Loss vs. Percent of Power Allocated to the Pilot Channel
304 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

Based on Bf , the system designer can decide which link is the limiting factor. If Bf < 0, the sys-
tem is forward link limited; Bf = 0, links are balanced; if Bf > 0 the system is reverse link limited.
A more realistic rule is
Bf < – δ: the system is forward link limited
|Bf | ≤ δ: the two links are balanced
Bf > δ: the system is reverse link limited
δ is the parameter that takes into account the tolerance in all factors involved in calculating Bf . A
good system design should insure that the two links are balanced. This makes handoff transition
smoother and reduces the amount of interference.

EXAMPLE 13.9
Using the following data, calculate the power allocated to the pilot channel. What is the allocated
power for the pilot channel to balance the forward and reverse link?
• Maximum mobile power (pm ) = 200 mW (23 dBm)
• Maximum cell-site power (pc) = 10 W (40 dBm)
• Voice activity factor (vf ) = 0.4
• Processing gain at the base station = 21 dB
• Number of users per cell (M) = 20
• Cell loading factor (ρ) = 0.5
• Cell noise figure (Fcell ) = 5 dB
• Mobile noise figure (Fmob)= 8 dB
• Cell-site antenna gain including cable loss (G'b ) = 6 dB
• Mobile antenna gain including cable loss (G'm ) = 0 dB
• Ioc /Io = 2.5 dB
• (Ec /It )min = –15 dB
• (Eb /It )min = 7 dB
(SIR)min + processing gain = (Eb /It )min
(SIR)min = –21 + 7 = –14 dB
Using Eq. (13.39), the path loss (cell site) on the reverse link is
L p ( r ) = – 14 – 108 – 23 – 6 – 10 log ( 1 – 0.5 ) = – 148 dB

T ( r ) = – 142 dB
From Eq. (13.42)
( I ⁄ I ) ⁄ 10
( E c ⁄ I t ) min [ ( N 0 B w ) mob + p c ⋅ T ( r ) ( 1 + 10 oc o ) ]
φ p = ---------------------------------------------------------------------------------------------------------------------------
-
pc ⋅ T ( r )
pc = 104
T(r) = 10–14.2
(N0Bw )mob = –105 dB = 10–10.5
(Ec /It )min = –15 dB = 10–1.5 = 0.03163
Forward Link Budget 305

– 10.5 4 – 14.2
0.03163 ⋅ [ 10 + 10 ⋅ 10 ⋅ ( 1 + 1.7783 ) ]
φ p = -----------------------------------------------------------------------------------------------------------------
4 – 14.2
- = 0.1078 = 10.78%
10 ⋅ 10
For the balanced condition, we use Eq. (13.45) with Bf = 0 and solve for φp
– 1.5
φ p – 10 ( 1 + 1.7783 )
0 = [ – 14 – ( – 15 ) ] + [ – 108 – ( – 105 ) ] + [ 40 – 23 ] + 10 log  -------------------------------------------------------
-
 0.5 
∴φ p = 0.1037 = 10.37%

EXAMPLE 13.10
Consider a minicell in which the maximum transmission loss is equal to –112 dB. The other data
include
• [N0Bw ]mob = –105 dB
• [N0Bw ]cell = –108 dB
• [Eb /It]min = 7 dB
• [Ec /It ]min = –15 dB
• Cell loading = 50%
• Ioc /Io = 2.5 dB
Find the output of the mobile and cell site.
From Eq. (13.39)
–112 = –14 – 108 – pm – 10 log (1 – 0.5)
pm = –7 dBm (= 0.2 mW)
For the minicell, the pilot strength can be determined as
E c
 ---- φp
- = -----------------------
-
 I t  min 1 + I oc ⁄ I o
E
∴φ p =  -----c ( 1 + I oc ⁄ I o ) = 10 ⋅ ( 1 + 10 ) = 0.0879 ~ 9%
– 1.5 0.25
 I t  min
From Eq. (13.43)
E ( E c ⁄ I t ) min ⁄ 10
T ( r ) =  -----c – p c + [ N 0 B w ] mob – 10 log [ φ p – ( 10
( I ⁄ I ) ⁄ 10
) ( 1 + 10 oc o ) ]
 I t  min
– 1.5 0.25
∴– 112 = – 15 – p c + ( – 105 ) – 10 log [ 0.09 – ( 10 ) ( 1 + 10 )]
p c = – 15 + 112 – 105 + 26.7 = 18.7 dBm (~ 74 mW)

13.9 Forward Link Budget

The forward link budget is calculated to confirm that the link safety margin parameters (Eq.
13.46) are positive and that there is sufficient margin for the forward link to work efficiently. Mpi-
lot and Mtraffic are more critical. If both of these are positive, then the other two are also generally
positive.
306 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

Ec Ec
M pilot =  ----- –  ----- (13.46a)
 I t  rec  I t  sp

Eb Eb
M traffic =  ------ –  ------ (13.46b)
 I t  rec  I t  sp

Eb Eb
M sync =  ------ –  ------ (13.46c)
 I t  rec  I t  sp

Eb Eb
M paging =  ------ –  ------ (13.46d)
 I t  rec  I t  sp

where rec means received value and sp means required value.

We use the following procedure to determine link safety margins.
The total cell-site power Ptotal will be
0.1P traffic 0.1P pilot 0.1P sync 0.1P paging
P total = 10 log ( 10 + 10 + 10 + 10 ) dBm (13.47)

where Ptotal = total cell-site effective radiated power (ERP) (dBm),

Psync = ERP of sync channel (dBm),
Ppilot = ERP of pilot channel (dBm),
Ppaging = ERP of paging channel (dBm), and
Ptraffic = ERP of all traffic channels (dBm).

P traffic
- = P traffic – 10 log α chan – 10 log M total dBm
P ( traffic ) ⁄ ( user ) = --------------------------------- (13.48)
( M total ⋅ α chan )

where P(traffic)/(user) = ERP of a traffic channel (dBm),

αchan = channel activity factor,
Mtotal = M(1 + ηco), and
ηco = traffic channel overhead percentage due to soft handoff.
The received power at the mobile on each channel from the serving cell site will be

P r, total = P total + GL (13.49)

P r, pilot = P pilot + GL (13.50)

P r, ( traffic ) ⁄ ( user ) = P ( traffic ) ⁄ ( user ) + GL (13.51)

P r, sync = P sync + GL (13.52)

P r, paging = P paging + GL (13.53)

where GL = G m + L c + L b + L pent + M fade + L p + G b,

Forward Link Budget 307

in which Lp = average propagation path loss between cell site and mobile (dB),
Lpent = penetration loss (dB),
Lb = body/orientation loss (dB),
Lc = cell-site feeder loss (dB),
Mfade = margin for log-normal shadowing (dB),
Gm = mobile antenna gain (dB), and
Gb = cell-site antenna gain (dB).
In-cell interference is caused by other users radiated from the same cell and is given as
0.1P r, total 0.1P r, ch
I ss-ch = 10 log ( 10 – 10 ) – 10 log B w dBm/Hz (13.54)

where ch is the pilot, paging, sync, or traffic/user, and

Bw = bandwidth.
Out-of-cell interference is caused by users in other cells and is given as

I os-ch = I ss-ch + 10 log ( 1 ⁄ f r – 1 ) dBm/Hz (13.55)

where fr = reuse factor.

Total interference will be
0.1 I ss-ch 0.1 I os-ch
I ch = 10 log ( 10 + 10 ) dBm/Hz (13.56)

The thermal noise density will be

– 23
N 0 = 10 log ( 290 × 1.38 × 10 ) + N f + 30 dBm/Hz (13.57)

where Nf = noise figure for the mobile.

The energy per bit for a channel will be

E bch = P r, ch – 10 log R ch (13.58)

where Rch = data rate for the channel.

The Eb /It for a channel can be calculated as

E bch 0.1 N 0 0.1 I ch

- = P r, ch – 10 log R ch – 10 log ( 10
------------------ + 10 ) dB (13.59)
N 0 + I ch

Using Eq. (13.59) we can write

E b
 ----- 0.1 N 0 0.1 I pilot
- = P r, pilot – 10 log B w – 10 log ( 10 + 10 ) dB (13.60)
 I t  rec, pilot
308 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

E b
 ----- 0.1 N 0 0.1 I paging
- = P r, paging – 10 log R paging – 10 log ( 10 + 10 ) dB (13.61)
 I t  rec, paging

E b
 ----- 0.1 N 0 0.1 I sync
- = P r, sync – 10 log R sync – 10 log ( 10 + 10 ) dB (13.62)
 I t  rec, sync

E b
 ----- 0.1 N 0 0.1 I traffic
- = P r, ( traffic ) ⁄ ( user ) – 10 log R traffic – 10 log ( 10 + 10 ) dB (13.63)
 I t  rec, traffic

EXAMPLE 13.11
Use the following data to calculate link safety margin parameters for the forward link channels of a
CDMA system.
• Pilot channel ERP (Ppilot) = 33.8 dBm
• Sync channel ERP (Psync) = 23.8 dBm
• Paging channel ERP (Ppaging) = 29.5 dBm
• Traffic channels ERP (Ptraffic) = 41.0 dBm
• Number of users per sector on the reverse link = 13
• Channel overhead due to soft handoff (ηco) = 0.85
• Path loss between cell site and mobile (Lp) = –130.2 dB
• Penetration loss (Lpent) = –15 dB
• Body/orientation loss (Lb) = –2 dB
• Fade margin (Mfade) = –10.3 dB
• Mobile antenna gain (Gm) = 2 dB
• Cell-site antenna gain (Gb) = 13 dB
• Cable losses (Lc) = –1.5 dB
• Channel activity factor (αchan)= 0.42
• Bandwidth (Bw ) = 1.2288 MHz
• Traffic channel rate = 9600 bps
• Sync channel rate = 1200 bps
• Paging channel rate = 4800 bps
• Cell reuse factor (fr ) = 0.65
4.1 3.38 2.95 2.38
P total = 10 log [ 10 + 10 + 10 + 10 ] = 42.0734 dBm
Mtotal = 13 (1 + 0.85) = 24
P ( traffic ) ⁄ ( user ) = 41.0 – 10 log ( 0.42 ) – 10 log 24 = 30.9654 dBm

GL = 2 + 13 – 1.5 – 2.0 – 15 – 10.3 – 130.2 = – 144 dB

P r, total = 42.0734 – 144 = – 101.9266 dBm

P r, pilot = 33.8 – 144 = – 110.2 dBm

P r, sync = 23.8 – 144 = – 120.2 dBm

P r, paging = 29.5 – 144 = – 114.5 dBm

Forward Link Budget 309

P r, ( traffic ) ⁄ ( user ) = 30.9654 – 144 = – 113.0346 dBm

– 10.19266 – 11.02 6
I ss-pilot = 10 log ( 10 – 10 ) – 10 log ( 1.2288 × 10 ) = – 163.5211 dBm/Hz

I os-pilot = – 163.5211 + 10 log ( 1 ⁄ 0.65 – 1 ) = – 166.2096 dBm/Hz

– 16.35211 – 16.62096
I pilot = 10 log ( 10 + 10 ) = – 161.65 dBm/Hz
– 10.19266 – 12.02
I ss-sync = 10 log ( 10 – 10 ) – 60.9 = – 162.8865 dBm/Hz

I os-sync = – 162.8865 – 2.6885 = – 165.575 dBm/Hz

– 16.28865 – 16.5575
I sync = 10 log ( 10 + 10 ) = – 161.0157 dBm/Hz
– 10.19266 – 11.45
I ss-paging = 10 log ( 10 – 10 ) – 60.9 = – 163.0736 dBm/Hz

I os-paging = – 163.0736 – 2.6885 = – 165.7621 dBm/Hz

– 16.30736 – 16.57621
I paging = 10 log ( 10 + 10 ) = – 161.2030 dBm/Hz
– 10.19266 – 11.30346
I ss-(traffic)/(user) = 10 log ( 10 – 10 ) – 60.9 = – 163.1769 dBm/Hz

I os-(traffic)/(user) = – 163.1769 – 2.6885 = – 165.8654 dBm/Hz

– 16.31769 – 16.58654
I (traffic)/(user) = 10 log ( 10 + 10 ) = – 161.306 dBm/Hz
– 23
N 0 = 10 log ( 290 × 1.38 × 10 ) + 30 + 8 = – 165.9772 dBm/Hz

E c
 ---- 6 – 16.597 – 16.165
- = – 110.2 – 10 log ( 1.2288 × 10 ) – 10 log ( 10 + 10 ) = – 10.82 dB
 I t  rec, pilot

E b
 ----- – 16.597 – 16.10157
- = – 120.2 – 10 log ( 1200 ) – 10 log ( 10 + 10 ) = 8.83 dB
 I t  rec, sync

E b
 ----- – 16.597 – 16.12
- = – 114.5 – 10 log ( 4800 ) – 10 log ( 10 + 10 ) = 8.64 dB
 I t  rec, paging

E b
 ----- – 16.597 – 16.1306
- = – 113.0346 – 10 log ( 9600 ) – 10 log ( 10 + 10 ) = 7.18 dB
 I t  rec, traffic

M pilot = – 10.82 – ( – 15 ) = 4.18

M traffic = 7.18 – 7.0 = 0.18

M paging = 8.64 – 7.0 = 1.64

M sync = 8.83 – 7.0 = 1.83

Since all the link safety parameters are positive, the power allocations on forward link chan-
nels are satisfactory.
Note that link budget calculations can be easily performed using Microsoft Excel or any other
program as shown in Table 13-3.
310 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

Table 13-3 Link Budget Calculations

channel total pilot sync paging fwd traffic
total trfc pwr 57.00
voice activ 0.42
# channels 22.00
% users in soft handoff 25.00
power output 58.49 51.50 41.50 46.94 46.37
path loss 146.00 146.00 146.00 146.00 146.00
fade margin 6.20 6.20 6.20 6.20 6.20
rx ant gain –3.00 –3.00 –3.00 –3.00 –3.00
rx power –96.71 –103.70 –113.70 –108.26 –108.83
in cell I –158.58 –157.69 –157.92 –157.88
freq reuse factor 0.65 0.65 0.65 0.65
out cell I –161.26 –160.38 –160.61 –160.57
total I –156.71 –155.82 –156.05 –156.01
noise figure 8.00 8.00 8.00 8.00
noise –165.98 –165.98 –165.98 –165.98
I/N 9.27 10.15 9.93 9.97
N+I –156.22 –155.42 –155.63 –155.59
data rate 1.23E+06 1200 9600 9600
eb/(I+N) –8.37 10.93 7.55 6.95

channel rvse trfc

power output 23.00
path loss 146.00
fade margin 6.20
rx ant gain 14.00
bs cable loss 2.50
rx power –117.70
voice activ 0.40
in cell I –169.35
freq reuse factor 0.60
out cell I –171.11
total I –167.13
noise figure 5.00
Summary 311

Table 13-3 Link Budget Calculations (Continued)

channel total pilot sync paging fwd traffic
noise –168.98
I/N 1.84
N+I –164.95
data rate 9600
eb/(I+N) 7.43

13.10 Summary
In this chapter, we developed necessary equations to calculate the reverse and forward link
capacity of an IS-95 CDMA system. We found that the maximum number of mobiles that can be
supported on the forward and reverse links of a CDMA system is different. Reverse link capacity
improves with soft handoffs; however, soft handoffs affect the capacity of the forward link.
In a noise-limited system, cell size rather than capacity is the main concern. Receiver sen-
sitivity is used to calculate the size of the cell. In an interference-limited system, the size of the
cell is determined mainly by the level of the interference from other users.
The chapter includes several numerical examples to illustrate the procedure and demon-
strate the importance of several system parameters in capacity calculations. It concludes with
link safety margin parameters for the forward link channels.

13.11 References
1. Borth, D. E., and Pursley, M. B., “Analysis of Direct Sequence Spread Spectrum Multiple
Access Communication over Rician Fading Channels,” IEEE Transactions on Communica-
tions COM-27 (10), October 1979, pp. 1566–77.
2. Gilhousen, K., Jacobs, I., Padovani, R., Viterbi, A., Weaver, L., and Whearley, C., III, “On
the Capacity of a Cellular CDMA System,” IEEE Transactions on Vehicular Technology 40
(2), May 1991, pp. 303–12.
3. Gilhousen, K. S., Jacobs, I. M., Padovani, R., and Weaver, L. A., “Increased Capacity Using
Satellite Communications,” IEEE Transactions on Select Areas in Communications JSAC-8
(4), May 1990, pp. 503–14.
4. Glisic, S., and Vucctic, B., Spread Spectrum CDMA Systems for Wireless Communications,
Artech House, Inc., Boston, 1997.
5. Viterbi, A. J., CDMA, Addison-Wesley Publishing Company, New York, 1995.
6. Viterbi, A. M., and Viterbi, A. J., “Erlang Capacity of a Power Controlled CDMA System,”
IEEE Journal of Selected Areas in Communications 11 (6), 1993, pp. 892–900.
7. Weber, C. L., et al., “Performance Considerations of CDMA Systems,” IEEE Transactions
on Vehicular Technology, February 1981, pp. 3–9.
312 Ch. 13 • Reverse and Forward Link Capacity of IS-95 CDMA System

13.12 Problems
1. A chip rate of 1.2288 Mcps is used for IS-95 RS2 (14.4 kbps with a 13-kbps vocoder).
The required Eb /It is 7 dB. Calculate the average number of mobiles supported by a sec-
tor of the three-sector cell; also calculate the Erlang capacity of a sector at 2% block-
ing. Assume: interference from other cells f = 0.6; cell loading ρ = 0.6; voice activity
factor vf = 0.4; power control accuracy factor ηc = 0.90; gain due to sectorization α =
2.61.
2. Calculate the pole capacity of the IS-95 CDMA system with a chip rate of 1.2288 Mcps
and RS2 (14.4 kbps). Assume: interference factor due to neighboring cells f = 0.67;
voice activity factor vf = 0.6; power control accuracy factor ηc = 0.80; gain due to three-
sector antenna α = 2.55.
3. A total of 20 equal-power mobiles share a frequency band through a CDMA system.
Each mobile transmits data at 16 kbps with a DSSS BPSK-modulated signal. Calculate
the minimum chip rate of the PN sequence in order to maintain a bit error probability of
10–6. Assume: interference factor due to other cells f = 0.6; power control accuracy fac-
tor ηc = 0.8; gain due to three-sector antenna α = 2.55.
4. Calculate the Erlang capacity of a sector for the three-sector cell and the number of
users per sector using the following parameters: carrier bandwidth = 1.23 MHz, RS1,
R = 9.6 kbps; (Eb /It )min = 7 dB; voice activity factor vf = 0.5; interference due to other
cells f = 0.67; cell loading factor ρ = 0.5; call blocking probability, Pout = 1%; three-
sector antenna gain α = 2.55; 1/η = 10; standard deviation of power control σc = 2 dB.
5. Calculate total transmission loss in dB for the forward link of a CDMA system using
the following data:
◆ Cell output power = 40 dBm
◆ Allocated power for pilot channel = 15% of cell output
◆ Mobile noise figure = 8 dB
◆ (Ec /It )min = –13 dB
◆ Ioc /Io = 2.5 dB
6. Calculate total transmission loss in dB for the reverse link of a CDMA system using the
following data:
◆ Mobile output = 200 mW (23 dBM)
◆ (Eb /It ) = 7 dB
◆ Cell noise figure = 5 dB
◆ Cell loading ρ = 60%
7. Using the data given in Problems 5 and 6, find the allocated power of the pilot channel
for balancing the forward and reverse links.
 C H A P T E R 1 4

Wireless Data

14.1 Introduction
This chapter covers data communication services and OSI upper layers and presents wireless
data systems including wide-area systems and high-speed Wireless Local Area Networks
(WLANs). We describe activities for wireless data standards and outline the error-control meth-
ods used by the standards. Also included are packet radio protocols and their channel efficiency
formulas. The contention function of packet radio models the mechanism where mobile stations
access the network on the access channel. Packet services are one of four data services sup-
ported in CDMA. The other three are asynchronous data, facsimile, and short message services
(similar to paging).
We discuss the standards for data services supported by CDMA cellular/PCS systems and
present highlights of the TIA IS-99, TIA IS-637, and TIA IS-657 standards. The chapter
includes the architecture for each of the four data services and the protocol stacks that are sup-
ported by the services.
We include both sets of standards (CDMA and non-CDMA) for two reasons: first, the
WLANs all use some form of spread spectrum communications, either frequency hopping or
direct sequence spreading; second, the two methods (WLANs and CDMA) are part of a larger
wireless network that many companies are constructing. With the phenomenal growth of laptop
personal computers and the Internet, wireless data is no longer limited to just e-mail or faxes. It
encompasses the ability to send and receive data any time from any place in the world. It gives a
user at a remote location full access to all of the desktop services that would normally be avail-
able at an office PC. Data services are delivering the same promise that voice services have
recently delivered: any time, any where communications.

313
314 Ch. 14 • Wireless Data

14.2 Data Communication Services

End-to-end communication services are classified as either synchronous (sync) or asynchronous
(async). A sync communication service delivers a bit stream with a fixed delay and a given bit
error rate. Voice communication is an example of sync communication service. Sync delivery of
a 64-kbps voice bit stream can be implemented by dividing the bit stream into packets that are
received with random delays and are stored in a buffer to hold the bits until they are delivered.
This implementation of a sync transmission service is called packetized voice. In packetized
voice, a buffer is used to absorb the random fluctuations in the packet transmission delays.
Another implementation of the sync transmission of the bit stream is to use a dedicated coaxial
cable that propagates the bits one after the other, all with the same delay.
In an async communication service, the bit stream to be transferred is divided into packets.
The packets are received by the destination with varying delays, and a fraction of them may not
be received correctly at the destination. An async communication service is evaluated by its
Quality of Service (QoS). QoS deals with parameters, such as the packet error rate, delay,
throughput, reliability, and security of the communication.
There are two classes of async communication services: connection oriented and connec-
tionless. A connection-oriented communication service delivers the packet in sequence—i.e., in
correct order—and confirms the delivery. Depending on the QoS requirements, the delivery may
be guaranteed to be free of errors. Thus, connection-oriented service looks from end to end like
a dedicated link, which may be noiseless or noisy. A connectionless communication service
delivers the packets individually. The packets can be delivered out of order, and some may con-
tain errors while others may be lost. Some connectionless services provide an acknowledgment
(ACK) of correctly delivered packets. Thus connectionless services are similar to mail service
provided by the post office: letters may be delivered out of order; normal mail delivery does not
guarantee the delivery. Yet another class of communication service called expedited data is used
in some applications. It corresponds to a potentially faster delivery of packets, usually by mak-
ing them jump to the head of the queues of packets that are waiting to be transmitted.
Communication services are implemented by transporting bits over the network. One
essential objective of the bit transport is connectivity, where one network user should be able to
exchange information with many other users. It should be possible to route the bits of one user
to any one of a large number of other users. The property to vary the path followed by the bits
is called switching. There are three basic methods used for switching bits in communication
networks:
• Circuit switching
• Virtual-circuit packet switching
• Datagram packet switching

In circuit switching, the switch connects transmission paths to establish a circuit between
the transmitter and receiver. Circuit switching is quite suitable for continuous data transmission
services.
OSI Upper Layers 315

A packet-switched network uses another scheme. The nodes of the network, packet-
switching nodes, play a role similar to that of switches in a circuit-switched network. Packet-
switched networks can use two different methods for selecting the path followed by packets:
virtual circuit (VC) and datagram. In the VC transport, the different packets that are part of
the same information transfer are sent along the same path. The packets follow one another as
if they were using a dedicated circuit even though they may be interleaved with other packet
streams. Some implementations of VC perform an error control on each link between succes-
sive nodes. Thus, not only are the packets delivered in sequence by each node to the next
node along the path, but they are also transmitted without errors. This is implemented by each
node checking the correctness of the packets it receives and asking the previous node along
the path to retransmit incorrect packets. VC packet switching does not need a buffer at the
destination.
Since multiple virtual circuits may exist between the source-destination pair, routing can-
not be done on the basis of source-destination address only. Data packets must carry an indica-
tion of VC identification as well. Routing is done based on explicit route number and destination
address. An explicit routing table at each node associates an appropriate outgoing transmission
group with the destination address and explicit route number. Changing the explicit route num-
ber for a given destination will cause a new path to be followed. This introduces alternative route
capability. If a link or node along the path becomes inoperative, any session using that path can
be reestablished on an explicit route by bypassing the failed element. Explicit routes can also be
assigned on the basis of type of traffic, type of physical media along the path (satellite or terres-
trial, for example), or other criteria. Routes could also be listed on the basis of cost, the lowest-
cost route being assigned first, then the next-lowest-cost route, and so forth.
In datagram packet switching, the bits are grouped as packets. Each packet is labeled with
the address of its destination. The packets are routed independently of one another and arrive at
destination out of sequence. Datagram packet switching requires buffers at the source and the
destination. In datagram packet switching networks, each network node keeps a complete (glo-
bal) topological database that is updated regularly as topological changes occur. Generally, the
routing philosophy of datagram networks is to route packets (datagram) along paths of minimum
time delay.

14.3 OSI Upper Layers

This section summarizes the role of the OSI upper layers (transport, session, presentation, and
application).
The transport layer segments the messages into packets of acceptable size and performs
the reassembly at the destination. It may multiplex many low-rate transmissions onto one virtual
circuit or divide a high-rate transmission into parallel virtual circuits. The transport layer con-
trols transmission errors and requests retransmission of packets corrupted by transmission
errors. In addition, the flow may be controlled by some mechanism to prevent one host from
sending data faster than the destination host can handle.
316 Ch. 14 • Wireless Data

The session layer sets up the call and takes care of the authentication of the user and of
billing. The session layer supervises the synchronization (packet numbering) and the recovery in
case of failures. It also closes the session at the end of the transmission.
The presentation layer asks the session layer to set up a call. It specifies the destination’s
name and the type of transmission (e.g., datagram, high priority). The presentation layer trans-
lates between the local syntax used by the application process and the transfer syntax, as well as
performing the required encryption and data compression.
The application layer provides information transfer services for user application pro-
grams. The user interacts with the application layer through a user interface. The application
layer is composed of Specific Application Service Elements (SASEs) that use the services of
Common Application Service Elements (CASEs). A CASE establishes the association between
SASEs and may include an Association Control Service Element (ACSE), a Remote Operation
Service Element (ROSE), and a Commitment Concurrency and Recovery (CCR) element.

14.4 Wireless Data Systems

We can classify wireless data systems into two basic categories: wide-area wireless data sys-
tems, and high-speed wireless local area networks. WLANs and wide-area wireless data systems
serve different categories of user applications and, therefore, have different system design objec-
tives. Wireless data services are used for transaction processing and for interactive, broadcast,
and multicast services. Transaction processing is used for credit card verification, paging, taxi
calls, vehicle theft reporting, and notice of voice or electronic mail. Interactive services include
database access and remote LAN access. Broadcast services are general information services,
weather and traffic advisory services, and advertising. Multicast services are similar to sub-
scribed information services, law enforcement communications, and private bulletin boards.
In the following sections, we briefly describe wide-area wireless data systems and
WLANs that have been deployed in the United States.

14.4.1 Wide-Area Wireless Data Systems

Wide-area wireless data systems are designed to provide high mobility, wide-area cover-
age, and low data rate digital data communications to both vehicles and pedestrians. The techni-
cal challenge is to design a system that efficiently uses the available bandwidth to serve large
numbers of users distributed over wide geographical areas. Table 14-1 gives the details of wide-
area wireless packet data systems deployed in the United States—there are Specialized Mobile
Radio Services (SMRS) allocations centered around 450 MHz and 900 MHz.
The ARDIS data network was developed by Motorola as a joint venture between Motorola
and IBM to support IBM field service repair people. It is now a public service offering and is
solely owned by Motorola. RAM Mobile Data is another public offering that uses the Ericsson
Mobitex technology. Both the ARDIS and RAM networks are evolving to data rates of 19.2
kbps. They have been designed to make use of standard, two-way voice, land mobile-radio chan-
nels, with 12.5 or 25 kHz channel spacing.
Wireless Data Systems 317

Table 14-1 Wide-Area Wireless Packet Data Systems

RAM Mobile
(Mobitex) ARDIS (KDT) Metricom (MDN) CDPD
Data rate 19.2 kbps 19.2 kbps 76 kbps 19.2 kbps
Channel spacing 12.5 kHz 25 kHz 160 kHz 30 kHz
Access slotted ALOHA FHSS (ISM) unused AMPS
CSMA channels
Frequency (MHz) fc ~ 900 fc ~ 800 fc ~ 915 fc ~ 800
Transmit power (W) 0.16 to 10 under 40 1 1.6
power control
Modulation GMSKa GMSK GMSK GMSK BTb = 0.5
a
GMSK = Gaussian minimum shift keying.
b
BT = channel width × bit duration.

The CDPD technology shares the 30-kHz spaced 800-MHz voice channels used by the
AMPS systems. The data rate is 19.2 kbps. The CDPD base station equipment shares cell sites
with the voice cellular radio system. The aim is to reduce the cost of providing packet data service
by sharing the resources with voice cellular systems. This strategy is similar to one that has been
used by nationwide fixed wireline packet data networks to provide an economically viable data
service by using a small portion of the capacity of the networks designed mainly for voice traffic.
Another approach used in wide-area wireless packet data networks is based on the micro-
cell concept of providing coverage in smaller areas. The microcell data networks are designed
for stationary or low-speed users. The basic aim is to reduce the cost of providing wireless data
service by using small and inexpensive base stations that can be installed on utility poles, the
sides of buildings, and inside buildings. The strategy is similar to the one being proposed for
Personal Communications Networks (PCNs). BS-to-BS wireless links are used to reduce the
cost of interconnecting in a data network. A large microcell network of small, inexpensive base
stations has been installed in the lower San Francisco Bay area by Metricom. The slow Fre-
quency Hopping Spread Spectrum (FHSS) in the 902–928 MHz U.S. Industrial Scientific Medi-
cal (ISM) band has been used. Transmitter power is 1 W maximum. Power control is used to
minimize interference and maximize battery life.

14.4.2 High-Speed Wireless Local Area Networks

A WLAN typically supports a limited number of users in a well-defined indoor area. Sys-
tem aspects such as bandwidth efficiency and product standardization are not crucial. The maxi-
mum achievable data rate is an important consideration in the selection of a WLAN. The
transmission channel characteristics and signal processing techniques are important.
WLANs are used to extend wired LANs for convenience and mobility. Three different
approaches have been used for connectivity of WLANs. The first approach includes access to
318 Ch. 14 • Wireless Data

Wide-Area Networks (WANs) and Metropolitan Area Networks (MANs). In the wide area, the
network transmission systems use the cellular arrangement and the wired long-distance network.
The data is packetized to meet the immediate demands of the users’ community. Data must be in
a proper form and format to prevent excessive overhead and consequent latency in transport. The
second approach deals with localized communications services for the added convenience of
connections between building floors and desktops in a dynamic environment. Flexibility to pro-
vide quick connections for moves, additions, and changes gives the organization significant
improvement over the basic wired LAN. The third approach is the flexible mobile LAN arrange-
ment to access a company intranet. This form of connectivity is becoming important in all walks
of life and business communities. As the workforce becomes more mobile, the need to provide
untethered connectivity is increasing exponentially.
Two different technical approaches exist with the WLANs. These are based on radio and
optical technologies. In the radio-based technology, there are two solutions: the licensed micro-
wave radio frequency range (18–23 GHz) or the unlicensed radio frequency range (902–928
MHz, 2.4–2.4835 GHz, and 5.75–5.825 GHz). In the unlicensed radio frequency, there are two
options. The first option uses Frequency Hopping Spread Spectrum (FHSS) technology,
whereas the second option uses Direct Sequence Spread Spectrum (DSSS) technology. The
902–928 MHz frequency band is an unlicensed ISM band that allows manufacturers to supply
products with very limited constraints. Newer products are also emerging that use the 2.4-GHz
band. The following are the major limitations of the unlicensed frequency-band WLANs:
• The system is restricted to 100 mW of output.
• The system must not interfere with other radio-frequency equipment in the same area.
• The system must go through an FCC-type acceptance process (in the international sec-
tor, this is called homologation or type acceptance, and the frequencies may be differ-
ent) using either 902–928 MHz, 2.4–2.4835 GHz, or 5.75–5.825GHz frequencies in
various ISM bands.

14.4.2.1 Spread Spectrum Radio-Based WLANs

WLANs use spread spectrum techniques to allow flexibility and minimize interference
while not being license bound. WLANs have been produced using FHSS and DSSS approaches
with different speeds. The motivation to use spread spectrum for packet radio systems comes
from improved multipath resistance, the ability to coexist with other systems, and the antijam-
ming nature of the code. In an office environment, spread spectrum is a promising choice
because it reduces the effects of multipath caused by reflections from the walls and increases the
mobility of the terminals within the office environment. The low spectral power density per user
of spread spectrum permits an overlay with certain existing systems and reduces the concerns
about health-related issues in high-power transmission. Spread spectrum offers the potential for
greater range and higher data rates compared with optical technology. It improves interception
resistance and provides data privacy.
Wireless Data Systems 319

Table 14-2 provides a partial list of WLANs available in the United States, two of which
we will describe briefly—AT&T WaveLAN, based on the FHSS technology, and Telesystem
ARLAN, based on the DSSS technology [1].

• AT&T WaveLAN. This system supports speeds of up to 20 Mbps and works with var-
ious network operating systems. WaveLAN uses a DS Quadrature Phase Shift Keying
(QPSK) multiplexing scheme to transmit across the entire broadband at higher signal
rates. Through multiplication of the original narrowband signal with PN sequence, the
code is spread across several frequencies. WaveLAN offers better security because the
conventional radio receiver cannot decode the signal without knowing the actual
spreading pattern. WaveLAN can operate up to 800 feet with a power output of 250
mW. It works in any laptop, notebook, or palmtop PC that is equipped for a Personal
Computer Memory Card International Association (PCMCIA) card. WaveLAN allows
users to operate in a cellular network for LANs. Each WaveLAN is assigned its own
identification code and can receive data only if its code corresponds to that of the cell it
occupies. Users can move anywhere within their assigned cell and still be able to com-
municate intracell. If users need to move between cells, they must first stop the applica-
tion from running, then reconfigure their address ID to match with the cell they are
moving into. With roaming this is automatic. WaveLAN is capable of interfacing
directly with the backbone cable systems at standard LAN cable speeds.
• Telesystems Advanced Radio LAN (ARLAN). ARLAN uses DSSS technology.
Using a conventional cable system, ARLAN devices called access points are attached
to the cable to allow for a full range of interconnections. A microcell can be configured
from the backbone network by setting an access point to act like a wireless repeater.
Telesystems Micro-cellular Architecture (TMA) allows the network to cover various
applications and various-sized facilities. With multiple base station antennas, the net-
work can be extended to create microcells, each with its own operating area and
devices. TMA is supported by firmware in each of the ARLAN devices. It supports
multiple overlapping cells, creating a seamless network within the building. Handoff
from cell to cell is a part of the network concept that allows for LAN connectivity of
users who need to move freely throughout departments or floors within the building.
Using SS technology, the system can select various center frequencies and allows for
the coexistence of multiple devices operating within the same area but serving different
needs. ARLAN 600 was designed for high-noise, industrial applications and uses a
spreading ratio of up to 100. It offers a full range of interfaces for async and sync data
transfer from terminals and hosts. The system operates in the 915-MHz and 2.4-GHz
frequency ranges and uses packet burst duplex transmission capabilities. Access to the
ARLAN network is packet-switched Carrier-Sense Multiple Access with Collision
Avoidance (CSMA/CA) (see section 14.6.2.3 for details on CSMA). Power output for
these devices is up to 1 W for distances of up to 500 feet diameter in an office environ-
ment and up to 3000 feet diameter in factories or open-plan offices indoors. For line-of-
Table 14-2 Partial List of WLAN Products

320
No. of Chan. or Mod/ Power Network
Product Freq. Link Rate User Rate Protocol Access Spread Factor Coding (mW) Topol.
Altair Plus 18–19 15 Mbps 5.7 Mbps Ethernet 4-level 25 peak 8 devices per
Motorola GHz FSKa radio
WaveLAN 902–928 2 Mbps 1.6 Mbps Similar to DSSS DQPSKb 250 peer to peer
AT&T MHz Ethernet
AirLAN 902–928 2 Mbps Ethernet DSSS DQPSK 250 PCMCIA
Solectek MHz with antenna
Freeport 902–928 16 Mbps 5.7 Mbps Ethernet DSSS 32 chips per bit 16 PSKc/ 650 hub
Windata MHz Trellis
Intersect 902–928 2 Mbps Ethernet; DSSS DQPSK 250 hub
Persoft Inc. MHz Token-ring
LAWN 902–928 38.4 kbps AX.25 SS 20 users per channel; 20 peer to peer
O’Neill Comm. MHz max. 4 channels
WiLAN 902–928 20 Mbps 1.5 Mbps Ethernet; CDMA/ 3 channels; 10–15 Unconv. 30 peer to peer
WiLAN Inc MHz per channel Token-ring TDMA links each
Radio Port 902–928 242 kbps Ethernet SS 100 peer to peer
ALPS Electric MHz
ARLAN 600 902–928 1.35 Mbps Ethernet DSSS 100 PCs with
Telesys. 2.4 GHz Max. antennas
Radio Link 902–928 250 kbps 64 kbps FHSS 250 ms/hop hub
Cal. Microwave 2.4 GHz 500-kHz space
RangeLAN 902–928 242 kbps Ethernet; DSSS 3 channels 100
Proxim, Inc. MHz Token-ring
RangeLAN 2 2.4 GHz 1.6 Mbps 50 kbps Ethernet; FHSS 10 channels @ 5kbs; 100 peer to peer
Proxim Inc. Token-ring 15 subchannels each bridge
Netwave 2.4 GHz 1 Mbps per Ethernet; FHSS 82 1-MHz channel or hub

Ch. 14 • Wireless Data

Xircom adopter Token-ring hops
Freelink 2.4 and 5.7 Mbps Ethernet DSS 32 chips per bit 16 PSK 100 100
Cabletron System 5.8 GHz trellis
a
FSK = frequency-shift keying.
b
DQPSK = differential quadratic phase-shift keying.
c
PSK = phase-shift keying.
WLAN Standards 321

sight building-to-building communications, the system can achieve distances of 6

miles. With microcell architecture, each cell is capable of handling up to 1 Mbps. The
ARLAN 655 and 670 are complete wireless network interface cards that are mounted
inside a PC, workstation (WS), or other device. They provide the same functionality as
a conventional LAN adapter card and can support multiple topologies in conjunction
with the network operating systems.

14.5 WLAN Standards

All standards for WLANs employ unlicensed bands. There are two approaches that can be used
to regulate an unlicensed band. One approach is based on a standard to allow different vendors
to communicate with one another using a set of interoperable rules. This approach is taken by
IEEE 802.11 and ETSI’s RES 10, HIPERLAN. In the second approach, a minimum set of rules
or “spectrum etiquette” is established to allow terminals designed by different vendors to have a
fair share of the available channel frequency/time resources and to coexist in the same band.
This approach does not preclude the first approach. The second approach has been pursued by
WINForum. In a coexisting environment, a vendor can be interoperable with another vendor by
using the same protocol and transmission scheme.
The three major standard activities for WLANs are IEEE 802.11, HIPERLAN, and WIN-
Forum. IEEE 802.11 developed a standard for DSSS, FHSS, and infrared light technology using
the ISM bands as the radio channel. The HIPERLAN standard is aimed at the 5.2- and 17.1-GHz
bands in European countries. WINForum’s goal is to obtain a PCS band for unlicensed data and
voice applications and to develop spectrum etiquette for them.

14.5.1 IEEE 802.11

IEEE 802.11 addresses the physical and media access (MAC) protocol layers for peer-to-
peer and peer-to-centralized communications topologies using DSSS or FHSS over radio or
infrared light technology. Both SS systems operate in the 2.4–2.4835 GHz ISM band. This band
has been selected over the 902–928 MHz and 5.725–5.85 GHz ISM bands because it is widely
available in most countries. In the 2.4–2.4835 GHz band, more than 80 MHz of bandwidth is
available that is suitable for high-speed data communication. Also, implementation in this band is
more cost effective as compared with the implementation in the higher frequencies. IEEE 802.11
supports DSSS with Binary Phase-Shift Keying (BPSK) and Quadrature Phase-Shift Keying
(QPSK) modulation for data rates of 1 and 2 Mbps, respectively, as well as FHSS with Gaussian
Frequency-Shift Keying (GFSK) modulation and two hopping patterns with data rates of 1 and 2
Mbps. For DSSS the band is divided into five overlapping 26-MHz subbands centered at 2.412,
2.442, 2.470, 2.427, and 2.457 GHz, with the last two overlapping the first three. This set-up pro-
vides five orders of frequency selectivity for the user. It is quite cost effective in improving the
transmission reliability in the presence of interference or severe frequency-selective multipath
fading. For FHSS, the channel is divided into 79 subbands, each with a 1-MHz bandwidth, and
three patterns of 22 hops are user options. A minimum hop rate of 2.5 hops/second is assigned to
322 Ch. 14 • Wireless Data

provide slow frequency hopping in which each packet is sent in one hop and, if it is destroyed,
the following packet is sent from another hop for which the channel condition would be different.
This approach provides a very effective time-frequency diversity and takes advantage of a
retransmission scheme to provide a robust transmission. The IEEE 802.11 standard avoids rigid
requirements and leaves room for vendors to maneuver in the following areas:
1. Multiple physical media—FHSS and DSSS radio, as well as infrared light; additional
media as approved in the future.
2. Common MAC layer regardless of physical layer—all IEEE 802.11-compliant
WLANs use CSMA/CA algorithm similar to Ethernet’s Carrier-Sense Multiple Access
with Collision Detection (CSMA/CD) MAC layer.
3. Common frame format—frames including headers and error protection fields are the
same, regardless of whether the attached wired LAN is 802.3 Ethernet or 802.5 token
ring; the access point handles conversion of 802.11 frames to wireline frame format.
4. Multiple on-air data rates—1 or 2 Mbps, with the possibility of higher rates in the
future.
5. Power limit—a maximum power of 1 W (or +30 dBm), as mandated by the FCC; there
is no minimum power requirement, which leaves open the possibility of low-power
implementations.
The standard defines the basic media and configuration issues, transmission procedures,
throughput requirements, and range characteristics for WLAN technology, focusing more on
access applications that involve the use of personal digital assistants (PDAs) and portable PCs
rather than trunk applications (see Figs. 14-1 and 14-2). Trunk applications use wireless as part
of the enterprise backbone for transmitting data from building to building, whereas access appli-
cations allow users of portable PCs, PDAs, and other wireless devices to tap into corporate
LANs from anywhere in an office or on a factory floor.
The radio transmitter in each user end-station is always listening for activity on the
WLAN. If one end-station is transmitting, another will not. The system has a preset time-out to
block a user from dominating the network, to avoid unnecessary transmission collisions, and to
allow priority traffic through. This is the function of the CSMA/CA access control mechanism.
Once it is determined that the network is free, the end-station ramps up to full power and sends a
preamble (a standard signaling message) to the access point. The preamble is a repeated bit pat-
tern followed by a special bit sequence. It allows the access point to lock onto the signal before
the data is sent. After the link is established, the end-station sends address and protocol informa-
tion. The header is followed by the data, which is transmitted at the on-air data rate. After the
error-check word is sent, the end-station listens for acknowledgment from the destination. If no
acknowledgment is received, the data is re-sent. The sequence is repeated until all the data has
been sent and acknowledged.
The IEEE 802.11 committee has specified that data rates for wireless systems must be
either 1 or 2 Mbps. Either the user chooses the rate, or the system selects the best one according
to the conditions. The on-air data rate includes message headers, retransmissions, and latency
WLAN Standards 323

N
LA
ding
ARLAN 670
B uil
ARLAN 610

ARLAN 655

ARLAN 610

Wireless
Repeater

ARLAN 655 ARLAN 655

PCMCIA

Figure 14-1 Access Application for WLAN

LAN
lding
Bui

ARLAN 670

ARLAN 620 Up to 6 Miles

ARLAN 620
ARLAN 670

AN
gL
ldin
Bui

Figure 14-2 Access Application for WLAN between Two Buildings with Line of Sight
324 Ch. 14 • Wireless Data

(the time between when a network station begins to seek access to a transmission channel and
when that access is granted). Header overhead and retransmissions primarily affect performance
of large data transfers, whereas latency has the greatest effect on short, bursty data transfers; this
is because the latency involved in setting up a transmission introduces more delay than the trans-
mission of message overhead or the retransmissions. Therefore, throughput on a WLAN is lower
for short messages than for longer messages. The actual throughput of an IEEE 802.11 system
on an on-air data rate of 2 Mbps is about 1.0–1.5 Mbps for long messages and 0.5–1.0 Mbps for
short messages. Throughput is also affected by the range of the system. In a typical office envi-
ronment, the range of an IEEE 802.11 WLAN is 200–300 feet, which is sufficient to cover most
partitioned areas and an outside rim of walled offices.
Sensitivity of the system is crucial because signal power can be affected drastically by
obstacles. Sensitivity figures are the smallest amounts of received power that the radio can use.
The IEEE 802.11 standard requires a sensitivity of less than –80 dBm. One issue that is not
addressed by the standard is roaming capability. Roaming is made possible with overlapping
WLAN cells in a configuration similar to that used for analog cellular phones. Roaming is con-
sidered to be part of the application- or driver-level technology, so vendors will likely resort to
different schemes to achieve it.

14.5.2 Wireless Information Networks Forum

The Wireless Information Networks Forum (WINForum) addresses WLAN and wireless
Private Branch Exchange (WPBX) services and focuses on spectrum etiquette to provide fair
access to an unlicensed band widely used for different applications and devices. The etiquette
does not preclude any common air interface standards or access technologies. It demands Lis-
ten-Before-Talk (LBT); thus, a device may not transmit if the spectrum it will occupy is already
in use within its range. The power is limited to keep the range short and allow operation in
densely populated office areas. The power and connection time are related to the occupied band-
width to equalize the interference and provide a fair access to frequency/time resources. In the
view of WINForum, the asynchronous transmission used in WLAN applications is bursty,
begins transmission within milliseconds, uses short bursts that contain large amounts of data,
and releases the link quickly. On the other hand, the isochronous transmission, typified by voice
services such as a WPBX, uses long holding time, periodic transmission, and flexible link access
times that may be extended up to a second. The asynchronous subbands may range from 50 kHz
to 10 MHz, whereas the isochronous subbands may be divided into 1.25-MHz segments. The
two types are technically contrasting and cannot share the same spectrum.

14.5.3 High-Performance Radio Local Area Network

ETSI’s subtechnical committee, RES 10, has been assigned the task of developing a stan-
dard for High-Performance Radio Local Area Network (HIPERLAN). The committee secured
two bands at 5.12–5.30 GHz and 17.1–17.3 GHz for HIPERLAN to operate at a minimum use-
ful bit rate of 20 Mbps for point-to-point applications with a range of 50 m. It is expected that, at
Access Methods 325

this rate and range, a data rate of 500–1000 Mbps—comparable to FDDI—can be achieved for a
standard building floor of approximately 1000 m square. RES 10 is responsible for defining a
radio transmission technique, including type of modulation, coding, and channel access, as well
as the specific protocols.

14.5.4 U.S. Advanced Research Project Agency

The U.S. Advanced Research Project Agency (ARPA) has sponsored WLAN projects at
the University of California at Berkeley (UCB) and University of California at Los Angeles
(UCLA). The UCB Infopad project is based on a coordinated network architecture with fixed
coordinating nodes and DSSS (CDMA), whereas the UCLA project is for peer-to-peer networks
and uses FHSS. Both ARPA-sponsored projects are concentrated on the 900-MHz ISM band.

14.6 Access Methods

14.6.1 Fixed-Assignment Access Methods
In the fixed-assignment access method, a fixed allocation of channel resources (frequency
or time, or both) is made on a predetermined basis to a single user. The three basic access meth-
ods—FDMA, TDMA, and CDMA—are examples of the fixed-assignment access method. In
this section, we discuss only the CDMA method. With CDMA, multiple users operate simulta-
neously over the entire bandwidth of the frequency/time signal domain, and the signals are kept
separate by their distinct user-signal codes. As we discussed in chapter 2, the number of the
users that can be supported simultaneously by a DS-CDMA system is

Gp 1 1
M = ---------------- × ------------ × η c × ----- × α (14.1)
Eb ⁄ N 0 1+ f vf

where Gp = processing gain = Bw /Rs,

Bw = bandwidth,
R = information rate,
Eb /N0 = bit energy-to-noise density,
f = interference from other cells,
ηc = power control factor,
vf = voice activity factor (= 1 for data service), and
α = gain due to sector antenna.

EXAMPLE 14.1
Consider a CDMA system that uses QPSK modulation and convolutional coding. The system has a
bandwidth of 1.25 MHz and transmits data at 9.6 kbps. Find the number of users that can be sup-
ported by the system and the bandwidth efficiency. Assume a three-sector antenna with an effective
gain = 2.6, ηc = 0.9, and an interference factor f = 0.5. A BER of 10–3 is required.
6
1.25 × 10
G p = ------------------------
3
- = 130.2
9.6 × 10
326 Ch. 14 • Wireless Data

–3 1 E
P b = 10 = --- erfc -----b-
2 N0

E
-----b- ≈ 7 dB (5)
N0

130.2 1
M = ------------- × ---------------- × 2.6 × 0.9 = 40.6 ≈ 40
5 1 + 0.5
40 × 9.6
η bw = ------------------------3- = 0.307 bit/sec/Hz
1.25 × 10

EXAMPLE 14.2
Consider a QPSK/DSSS WLAN that is designed to transmit in the 902–928 MHz ISM band. The
symbol transmission rate is 0.1 mega symbols per second (Msps). An orthogonal code with 4 sym-
bols is used. A BER of 10–5 is required. How many users can be supported by the WLAN? A three-
sector antenna with a gain = 2.6 is used. Assume an interference factor f = 0.5 to account for the
interference from users in other cells and ηc = 0.9. What is the bandwidth efficiency of the system?
Band width Bw = 928–902 = 26 MHz
2
Data rate R = R s log 24 = 0.1 log 22 = 0.2 Mbps

B 26
G p = -----w- = ------- = 130
R 0.2

–5 1 E
P b = 10 = --- erfc -----b-
2 N0

E
-----b- ≈ 10 dB (10)
N0

130 1
M = --------- × ---------------- × 2.6 × 0.9 = 20.3 ≈ 20
10 1 + 0.5
20 × 0.2
η bw = ------------------- = 0.154 bit/sec/Hz
26

14.6.2 Random-Access Methods

When each user has a steady flow of information to transmit (for example, a data file
transfer or a facsimile transmission), fixed-assignment access methods are useful because they
make efficient use of communication resources. However, when the information to be transmit-
ted is bursty in nature, fixed-assignment access methods result in a waste of communication
resources. Furthermore, in a cellular system where subscribers are charged based on channel
connection time, fixed-assignment access methods may be too expensive to use for transmitting
Access Methods 327

short messages. Random-access protocols provide flexible and efficient methods for managing
channel access when transmitting short messages. Random-access methods give each user free-
dom to gain access to the network whenever the user has information to send. Because of this
freedom, these schemes result in contention among users trying to access the network simulta-
neously. Contention may cause collisions resulting in the need to retransmit the information.
The commonly used random-access protocols are pure ALOHA, slotted ALOHA, and CSMA/
CD. The following sections briefly describe details of each of these protocols and provide nec-
essary throughput expressions.

14.6.2.1 Pure ALOHA

In the pure ALOHA scheme, users transmit information whenever they have information
to send. A user sends information in packets. After sending a packet, the user waits a length of
time equal to the round-trip delay for an acknowledgment (ACK) of the packet from the receiver.
If no ACK is received, the packet is assumed to be lost in a collision, and it is retransmitted with
a randomly selected delay to avoid repeated collisions.* The normalized throughput S (average
packet arrival rate divided by the maximum throughput) of the pure ALOHA protocol is given as
– 2G
S = Ge (14.2)
where G = normalized offered traffic load.
Note from Eq. (14.2) that the maximum throughput occurs at traffic load G = 50% and is S = 1/2e.
This is about 0.184. Thus, the best channel utilization with pure ALOHA protocol is only 18.4%.

14.6.2.2 Slotted ALOHA

In the slotted ALOHA system, the transmission time is divided into time slots. Each time
slot is made exactly equal to the packet transmission time. Users are synchronized to the time
slots so that, whenever a user has a packet to send, the packet is held and transmitted in the next
time slot. With the synchronized time slots scheme, the interval of a possible collision for any
packet is reduced to one packet time from two packet times, as in the pure ALOHA scheme. The
normalized throughput S for the slotted ALOHA protocol is given as
–G
S = Ge (14.3)
where G = normalized offered traffic load.
The maximum throughput for slotted ALOHA occurs at G = 1.0 (Eq. [14.3]), and it is
equal to 1/e or about 0.368. This implies that, at the maximum throughput, 36.8% of the time
slots carry the successfully transmitted packets, whereas the same percentage of the time slots
remain empty.

* Note that the protocol on CDMA access channels as implemented in TIA IS-95A is based upon the pure ALOHA
approach. The mobile station randomizes its attempt for sending a message on the access channel and may retry if an
acknowledgment is not received from the base station. For further details, reference Section 6.6.3.1.1.1 of TIA IS-95A.
328 Ch. 14 • Wireless Data

14.6.2.3 Carrier-Sense Multiple Access

CSMA protocols have been widely used in both wired and wireless LANs. These proto-
cols provide enhancements over pure and slotted ALOHA protocols. The enhancements are
achieved through use of additional capability at each user station to sense the transmissions of
other user stations. The carrier-sense information is used to minimize the length of collision
intervals. For carrier sensing to be effective, propagation delays must be less than packet trans-
mission times. There are two general classes of CSMA protocols: nonpersistent and p-persistent.
Each of these classes can be used with the slotted or unslotted operation.
• Nonpersistent CSMA. A user station does not sense the channel continuously while it
is busy. Instead, after sensing the busy condition, it waits for a randomly selected inter-
val of time before sensing again. The algorithm works as follows: if the channel is
found to be idle, the packet is transmitted; if the channel is sensed to be busy, the user
station backs off to reschedule the packet to a later time. After backing off, the channel
is sensed again, and the algorithm is repeated.
• p-persistent CSMA. The slot length is typically selected to be the maximum propaga-
tion delay. When a station has information to transmit, it senses the channel. If the
channel is found to be idle, it transmits with probability p. With probability q = 1 – p,
the user station postpones its action to the next slot, where it senses the channel again.
If that slot is idle, the station transmits with probability p or postpones again with prob-
ability q. The procedure is repeated until either the frame has been transmitted or the
channel is found to be busy. When the channel is detected busy, the station then senses
the channel continuously and, when it becomes free, it starts the above procedure again.
If the station initially senses the channel to be busy, it simply waits one slot and applies
the above procedure.
◆ 1-persistent CSMA: 1-persistent CSMA is the simplest form of p-persistent
CSMA. It signifies the transmission strategy, which is to transmit with probabil-
ity 1 as soon as the channel becomes idle. After sending the packet, the user sta-
tion waits for an ACK, and, if it is not received within a specified amount of time,
the user station waits for a random amount of time and then resumes listening to
the channel. When the channel is again found to be idle, the packet is retransmit-
ted immediately.
For more details on CSMA, refer to references [3, 21].
The throughput expressions for the CSMA protocols are as follows:
• Unslotted nonpersistent CSMA
– aG
Ge
S = -----------------------------------------
– aG
- (14.4)
G ( 1 + 2a ) + e
• Slotted nonpersistent CSMA
Access Methods 329

– aG
aGe
S = ----------------------------
– aG
- (14.5)
1–e +a
• Unslotted 1-persistent CSMA
– G ( 1 + 2a )
G [ 1 + G + aG ( 1 + G + ( aG ) ⁄ 2 ) ]e
S = --------------------------------------------------------------------------------------------------------
– aG –G ( 1 + a )
- (14.6)
G ( 1 + 2a ) – ( 1 – e ) + ( 1 + aG )e
• Slotted 1-persistent CSMA
–G ( 1 + a ) – aG
Ge [1 + a – e ]
S = -----------------------------------------------------------------------
– aG –G ( 1 + a )
- (14.7)
( 1 + a ) ( 1 – e ) + ae
where S = normalized throughput,
G = normalized offered traffic load,
τ
a = ------ ,
Tp
τ = propagation delay, and
Tp = packet transmission time.

EXAMPLE 14.3
Consider a WLAN installation in which the maximum propagation delay is 0.4 µsec. The WLAN
operates at a data rate of 10 Mbps, and packets have 400 bits. Calculate the throughput with: (1) an
unslotted nonpersistent, (2) a slotted persistent, and (3) a slotted 1-persistent CSMA protocol.
400
T p = --------- = 40 µsec
10
τ 0.4
a = ------ = ------- = 0.01
Tp 40
–6 6
40 × 10 × 10 × 10
G = ------------------------------------------------- = 1
400
• Unslotted nonpersistent
– 0.01
1×e
S = ------------------------------------------
– 0.01
- = 0.493
( 1 + 0.02 ) + e
• Slotted nonpersistent
– 0.01
0.01 × 1e
S = -------------------------------------
– 0.01
= 0.496
1–e + 0.01
• Slotted 1-persistent
– 1.01 – 0.01
e ( 1 + 0.01 – e )
S = -------------------------------------------------------------------------------
– 0.01 – 1.01
- = 0.531
( 1 + 0.01 ) ( 1 – e ) + 0.01e
330 Ch. 14 • Wireless Data

14.7 Error Control Schemes

Channel coding and Automatic Repeat Request (ARQ) schemes are used to increase the perfor-
mance of mobile communication systems. In the physical layer of the DS-CDMA system, error
detection and correction techniques such as Forward Error Correction (FEC) schemes are used.
For some of the data services, higher-layer protocols use ARQ schemes to enable retransmission
of any data frames in which an error is detected. The ARQ schemes are classified as follows
[21,22]:

• Stop and Wait. The sender transmits the first packet numbered 0 after storing a copy of
that packet. The sender then waits for an ACK numbered 0 (ACK0) of that packet. If
the ACK0 does not arrive before a time-out, the sender makes another copy of the first
packet, also numbered 0, and transmits it. If the ACK0 arrives before a time-out, the
sender discards the copy of the first packet and is ready to transmit the next packet,
which it numbers 1. The sender repeats the previous steps, with numbers 0 and 1 inter-
changed. The advantages of the Stop and Wait protocol are its simplicity and its small
buffer requirements. The sender needs to keep only a copy of the last packet transmit-
ted, and the receiver does not need to buffer packets at the data link layer. The main dis-
advantage of the Stop and Wait protocol is that it does not use the communication link
very efficiently and it is slow.
The total time taken to transmit a packet and to prepare for transmitting the next
one is

T = T p + 2T prop + 2T proc + T a (14.8)

The protocol efficiency without error is

T
η ( 0 ) = -----p- (14.9)
T

where T = total time for transmitting a packet,

Tp = transmission time for a packet,
Tprop = propagation time of a packet or an ACK,
Tproc = processing time for a packet or an ACK, and
Ta = transmission time for an ACK.
If p is the probability that a packet or its ACK is corrupted by transmission errors,
and a successful transmission of a packet and its ACK takes T seconds and occurs with
probability 1 – p, the protocol efficiencies for full-duplex (FD) and half-duplex (HD)
operation are given as

( 1 – p )T p
η FD = ------------------------------------
- (14.10)
( 1 – p )T + pT p
Error Control Schemes 331

( 1 – p )T
η HD = -----------------------p (14.11)
T
• Selective Repeat Protocol (SRP). The data link layer in the receiver delivers exactly
one copy of every packet in the correct order. The data link layer in the receiver may get
the packets in the wrong order from the physical layer. This occurs, for example, when
transmission errors corrupt the first packet and not the second one. The second packet
arrives correctly at the receiver before the first packet gets there. The data link layer in
the receiver uses a buffer to store the packets that arrive out of order. Once the data link
layer in the receiver has a consecutive group of packets in its buffer, it can deliver them
to the network layer. The sender also uses a buffer to store copies of the unacknowl-
edged packets. The number of the packets that can be held in the sender/receiver buffer
is a design parameter. Let W = the number of packets that the sender and receiver buff-
ers can each hold and SRP = number of packets modulo-2 W. The protocol efficiency
without any error and with an error probability of p is given as

WT p 
η ( 0 ) = min  -----------, 1  (14.12)
 T 
For very large W, the protocol efficiency is

η( p) = 1 – p (14.13)
where T = time-out = WTp

2 + p(W – 1)
η ( p ) = ----------------------------------- (14.14)
2 + p ( 3W – 1 )
SRP is very efficient, but it requires buffering packets at both the sender and the
receiver.
• Go-Back-N (GBN). The Go-Back-N protocol allows the sender to have multiple unac-
knowledged packets without the receiver having to store packets. This is done by not
allowing the receiver to accept packets that are out of order. When a time-out timer
expires for a packet, the transmitter resends that packet and all subsequent packets. The
Go-Back-N protocol improves on the efficiency of the Stop and Wait protocol, but is
less efficient than SRP. The protocol efficiency is given as

1
η FD = ---------------------------------
(14.15)
1 +  ------------ W
p
 1 – p

• Window-Control Operation Based on Reception Memory (WORM) ARQ. In digi-

tal cellular systems, bursty errors occur due to multipath fading, shadowing, and hand-
offs. The BER fluctuates from 10–1 to 10–6. Therefore, the conventional ARQ schemes
332 Ch. 14 • Wireless Data

do not operate well in a digital cellular system. WORM ARQ has been suggested for
control of dynamic error characteristics. It is a hybrid scheme that combines SRP with
GBN. GBN protocol is chosen in severe error conditions, whereas SRP is selected in
normal error conditions.
• Variable Window and Frame Size GBN and SRP. Since CDMA systems have bursty
error characteristics, the error control schemes should have a dynamic adaptation to a
bursty channel environment. SRP and GBN with variable window and frame size have
been proposed in [22] to improve error control in CDMA systems. Table 14-3 provides
the window and frame sizes for different bit error rates. If the error rate increases, the
window and frame size are decreased. If the error rate is small, the window and frame
size are increased. The optimum threshold values of BER and window and frame size
were obtained through computer simulation.

In CDMA systems, the forward link consists of pilot, sync, paging, and traffic channels.
System information sent on the sync and paging channels allows each mobile station to evaluate
the BER easily by measuring the ratio of the number of retransmitted frames to the number of
transmitted frames over a 2-second period. Thus, the mobile station can change the window and
frame sizes according to the BER.

Table 14-3 Bit Error Rate vs. Window and Frame Size

BER Window Size (W) Frame Size (bits)

BER ≤ 10–4 32 172
10–4 < BER < 10–3 8 80
10–3 < BER < 10–2 4 40
10–2 < BER 2 16

EXAMPLE 14.4
Consider a WLAN in which the maximum propagation delay is 4 µsec. The WLAN operates at a data
rate of 10 Mbps. The data and ACK packet are of 400 and 20 bits, respectively. The processing time
for a data or ACK packet is 1 µsec. If the probability p = 0.01 that a data packet or its ACK can be
corrupted during transmission, find the data link protocol efficiency with (1) Stop and Wait protocol,
full duplex, (2) SRP with window size W = 8, and (3) Go-Back-N protocol with window size W = 8.
400
T p = --------- = 40 µsec
10
20
T a = ------ = 2 µsec
10

T prop = 4 µsec

T proc = 1 µsec
Data Services in IS-95 333

T = 40 + 2 × 4 + 2 × 1 + 2 = 52 µsec
Stop and Wait:
( 1 – 0.01 ) × 40
η = ------------------------------------ = 0.762
52
SRP:
2 + 0.01 ( 8 – 1 )
η = ---------------------------------------- = 0.954
2 + 0.01 ( 24 – 1 )
Go-Back-N:
1
η = ------------------------------------- = 0.925
1 + 8 -------------------
 0.01
 1 – 0.01

14.8 Data Services in IS-95

The data services in IS-95 include circuit-switched async data, circuit-switched digital fax,
packet data, and analog fax. The current data standards are the following:
• IS-658
◆ Defines the interface between the MSC and an external data interworking func-
tion (IWF)
◆ Applies to circuit-switched async data, digital fax, and packet data services
• Medium Data Rates (MDR)
◆ Included as part of IS-95B standardization
• IS-707
◆ RS1 Service Options (SOs)
SO 4100: Async Data
SO 4101: Digital fax
SO 4103: IP/Mobile IP
SO 4104: CDPD
◆ RS2 SOs
SO 12: Async Data
SO 13: Digital fax
SO 15: IP/Mobile IP
SO 16: CDPD
The cdmaOne* (family of IS-95 standards) packet data technology uses a TCP/IP-compliant
CDPD protocol stack to provide seamless connectivity with enterprise networks and expedite
third-party application development. Adding data to the cdmaOne network allows an operator to

* Trade name used by CDMA Users Group.

334 Ch. 14 • Wireless Data

continue using its existing radios, backhaul facilities, infrastructure, and handsets while merely
implementing a software upgrade with an IWF.
TIA IS-95B allows for code or channel aggregation to provide data rates of 64–115 kbps,
as well as offering improvements in soft handoffs and interfrequency hard handoffs. To achieve a
114-kbps rate, up to eight CDMA traffic channels, each offering 14.4 kbps, are needed to be
aggregated.

14.9 Asynchronous Data and Group-3 Facsimile

The general approach taken in TIA IS-95A [14] for data services reuses the previously specified
physical layer of the IS-95A protocol stack as the physical layer. Fig. 14-3 shows the air inter-
face (Um ) protocol stack.
IS-95A asynchronous data has been structured as a circuit-switched service, in which a
dedicated path is established between the data devices for the duration of the call. It is used for
connectivity through the PSTN when point-to-point communication to a PC or fax user is
required. For example, for a file transfer involving PC-to-PC communications, asynchronous
data service is the preferred cellular service mode.

Async Data Fax

Application Interface

TCP ICMP
IP

SNDCF IPCP LCP

PPP
RLP

IS-95A
TCP: Transmission Control Protocol
ICMP: Internet Control Message Protocol
IP: Internet Protocol
SNDCF: Subnetwork Dependent Convergence Function
IPCP: Internet Protocol Control Protocol
LCP: Link Control Protocol
PPP: Point-to-Point Protocol
RLP: Radio Link Protocol

Figure 14-3 The Um Protocol Stack

Asynchronous Data and Group-3 Facsimile 335

The Radio Link Protocol (RLP) employs ARQ, Forward Error Correction (FEC), and flow
control. Flow control and retransmission of data blocks with errors are used to provide improved
performance in the mobile segment of the data connection at the expense of variations in
throughput and delay. Typical raw channel data error rates for cellular transmission are approxi-
mately 10–2. However, an acceptable data transmission usually requires a BER of about 10–6. In
order to achieve this, it requires the design of efficient ARQ and error correction codes to deal
with error characteristics in the mobile environment.
The CDMA protocol stack for data and facsimile (Figure 14-3) has the following layers:

• Application Interface Layer. This layer includes an application interface between the
data source/destination in the mobile terminal (MT0) or terminal equipment (TE2) and
the transport protocol layer. In the base station, the application interface resides
between the data source/destination on the network (Ai interface) side and the transport
protocol layer. The application interface provides: modem control, AT command pro-
cessing,* negotiation of air interface data compression, and data compression over the
air interface (optional).
• Transport Layer. The transport layer for CDMA asynchronous data and fax services is
based on Internet transport layer protocol known as Transmission Control Protocol
(TCP) [7]. The implementation complies with the requirements for TCP with modifica-
tions as described in IS-95 [14]. If the modified procedure is disabled, there is no max-
imum number of retransmission attempts during synchronization, and an established
TCP connection remains open until explicitly closed by the mobile station or base sta-
tion. The application interface sets the value of R2 in the protocol. The base station fol-
lows either the procedure of the Internet Control Message Protocol (ICMP) [6] or the
procedure given above for TCP.
• Network Layer. The network layer for CDMA async data and fax services is based on
Internet network layer protocol known as the Internet Protocol (IP) [5]. The network
layer includes the ICMP [6]. The implementation complies with the requirements of
the IP [5] and requirements for Internet hosts [8] with modifications as described in IS-
95 [14]. The interface between the network and transport layer complies with the
requirements of the ICMP.
• Sub-Network Dependent Convergence Function (SNDCF). The SNDCF performs
header compression on the headers of the transport and network layers. This function is
negotiated using Point-to-Point Protocol (PPP) and Internet Protocol Control Protocol
(IPCP) [10]. Mobile stations support Van Jacobson TCP/IP header compression. A min-
imum of one compression slot is negotiated. Base stations support TCP/IP header com-
pression compatible with that required for mobile stations. Negotiation of the

* The AT commands were originally defined by Hayes Microcomputer Company for its wireline modems. The com-
mand set has now been adopted by most wireline and wireless modems. The name AT is derived from the use of an
“AT” to preface all commands to the modem.
336 Ch. 14 • Wireless Data

parameters of header compression is carried out using IPCP. The SNDCF sublayer
accepts the network-layer datagram from the network layer, performs header compres-
sion as required, and passes the datagram to the PPP layer, indicating the appropriate
PPP protocol identifier. The SNDCF sublayer receives network-layer datagrams with
compressed or uncompressed headers from the PPP layer, decompresses the datagram
header as necessary, and passes the datagram to the network layer.
• Data Link Layer. This layer uses PPP [11]. The PPP Link Control Protocol (LCP) is
used for initial link establishment and for the negotiation of optional link capabilities.
The data link layer uses the PPP IPCP to negotiate IP addresses and TCP/IP header
compression. The data link layer accepts network-layer datagrams from the SNDCF
and encapsulates them in the PPP information field. The packet is framed using the
octet synchronous framing protocol, except that there is no interframe fill. No flag
octets are sent between a flag octet that ends one PPP frame and the flag octet that
begins the subsequent PPP frame. The framed PPP packets are passed to the RLP layer
for transmission. The data link layer accepts received octets from the RLP layer and
reassembles the original PPP packets. The PPP process discards any PPP packet for
which the received Frame Check Sequence (FCS) is not equal to the computed value.
• Internet Protocol Control Protocol Sublayer. This sublayer supports negotiation of
the IP-address (type = 3) and IP-compression (type = 2) protocol parameters. IPCP
negotiates a temporary IP address for the mobile station whenever a transport layer
connection is actively opened. Mobile stations maintain the temporary IP address only
while a transport layer connection is open or being opened, and discards the temporary
IP address when the transport layer connection is closed.
• Link Control Protocol. If the protocol identifier is 0xC021, the PPP layer processes
the packet according to PPP LCP. For other supported protocol identifiers, the PPP
layer removes the PPP encapsulation and passes the datagram and protocol identifier to
the SNDCF. For unsupported protocol identifiers, the LCP Protocol-Reject is passed to
the RLP layer for transmission. The mobile station supports the PPP LCP Configure-
Request, Configure-ACK, Configure-Negative Acknowledgment (NAK), Configure-
Reject, Terminate-Request, Terminate-ACK, Code-Reject, and Protocol-Reject. Other
LCP packet types may also be supported. The PPP LCP negotiates the following con-
figuration options:

1. Async control character map: The mobile station does not require any mapping
of control characters. The base station may negotiate mapping of control charac-
ters.
2. Protocol field compression: this applies when the protocol number is less than
0xFF.
3. Address and control field compression: this applies when the protocol number is
not 0xC021.
Asynchronous Data and Group-3 Facsimile 337

The mobile station may also support other configuration options (such as maxi-
mum receive unit, authentication protocol, link quality protocol, or magic number).
When an option is received that is not supported, Configure-Reject is sent as an indica-
tion to the peer.
• Radio Link Protocol Layer. The RLP layer provides an octet stream service over the
forward and reverse traffic channels and substantially reduces the error rate typically
exhibited by these channels. This service is used to carry the variable-length data pack-
ets on the PPP layer. The RLP divides the PPP packets into TIA IS-95A traffic channel
frames for transmission. There is no direct relationship between PPP packets and traffic
channel frames. A large packet may span multiple traffic channel frames, or a single
traffic channel frame may contain all or part of several small PPP packets. The RLP is
unaware of higher-layer framing; it operates on a featureless octet stream, delivering
the octets in the order received from the PPP layer. For service options supporting an
interface with multiplex option 1, RLP frames may be transported as primary or sec-
ondary traffic or as signaling traffic via data burst messages. For the primary or second-
ary traffic, the RLP generates and supplies exactly one frame to the multiplex sublayer
every 20 ms. The frame contains the service option information bits. The multiplex
sublayer in the mobile station categorizes every received traffic frame and supplies the
frame type and accompanying bits, if any, to the RLP layer. The frame type and frame
category for primary and secondary traffic are given in Tables 14-4 and 14-5. A blank
frame is used for blank-and-burst transmission of signaling traffic.
The signaling subchannel may carry frames from multiple RLPs, with each RLP
having a distinct BURST_TYPE. Each service option defines a unique BURST_TYPE
used for RLP. The primary and secondary multiplex subchannels each carry at most a sin-
gle RLP layer. RLP data frames sent on one multiplex subchannel are not to be transmit-
ted on another subchannel. RLP frames are not sent on the access and paging channels.
• Radio Interface. The mobile station and base station support the physical layer, mul-
tiplex sublayer, radio link management, and call control as defined in TIA IS-95A.
They use service option 4 for async data services and service option 5 for Group-3 fax

Table 14-4 RLP Frame with Primary Traffic

RLP Frame Type Bits/Frame Multiplex Option 1 Frame Categories
Full rate 171 1
Half rate 80 2, 6, 11
Erasure 0 all others
1/8 rate 16 4, 8, 13
Blank 0 5, 14
Erasure 0 all others
338 Ch. 14 • Wireless Data

Table 14-5 RLP Frame with Secondary Traffic

RLP Frame Type Bits/Frame Multiplex Option 1 Frame Categories
Rate 1 168 14
Rate 7/8 152 13
Rate 3/4 128 12
Rate 1/2 88 11
Blank 0 1–8
Erasure 0 9, 10

services, but they do not transmit 1/4-rate frames when service option 4 or service
option 5 is active. Service options 4 and 5 support an interface with multiplex option 1.
RLP frames for service options 4 and 5 are transported only as primary traffic or signal-
ing traffic. The mobile station and the base station perform service option negotiation for
service options 4 and 5 as described in TIA IS-95A (sections 6.6.4.1.2 and 7.6.4.1.2).
Initialization and connection in the mobile station and the base station to accept service
option 4 or service option 5 in response to a Service Option Request order is performed
according to the specifications in TIA IS-99 (sections 3.8.4.1 and 3.8.4.2).

14.10 Short Message Service

The Short Message Service (SMS) [17] allows the exchange of short alphanumeric messages
between a mobile station and the cellular system and between the cellular system and an external
device capable of transmitting and optionally receiving short messages. The external device may
be a voice telephone, a data terminal, or a short message entry system. The SMS consists of
message entry features, administration features, and message transmission capabilities. These
features are distributed between a cellular system and the SMS Message Center (MC) which
together make up the SMS system. The MC may be either separate from or physically integrated
into the cellular system.
Short message entry features are provided through interfaces to the MC and the mobile
station. Senders use these interfaces to enter short messages, intended destination addresses, and
various delivery options. MC interfaces may include features such as audio response prompts
and Dual-Tone Multifrequency (DTMF) reception for dial-in access from voice telephones, as
well as appropriate menus and message entry protocols for dial-in or dedicated data terminal
access. Mobile station interfaces may include keyboard and display features to support message
entry. Also, a cellular voice service subscriber can use normal voice or data features of the
mobile station to call an SMS system to enter a message.
An SMS teleservice can provide the option of specifying priority level, future delivery
time, message expiration interval, or one or more of a series of short, predefined messages. If
supported by the teleservice, the sender can request acknowledgment that the message was
received by the mobile station. An SMS recipient, after receiving a short message, can manually
Short Message Service 339

acknowledge the message. Optionally, the recipient can specify one of a number of predefined
messages to be returned with acknowledgment to the sender.
SMS administration features include message storage, profile editing, verification of
receipt, and status inquiry capabilities. The SMS transmission capabilities provide for the trans-
mission of short messages to or from an intended mobile station as well as return of acknowl-
edgments and error messages. These messages and acknowledgments are transmitted to or from
the mobile station whether it is idle or engaged in a voice or data call. The cellular service pro-
vider may offer SMS transmission to its cellular voice and data customers only or may provide
an SMS-only service without additional voice or data transmission capability. All available
mobile stations on a CDMA paging channel can receive a broadcast message. A broadcast mes-
sage is not acknowledged by the mobile station. Broadcast messaging services may be made
available to mobile stations on CDMA paging channels as well as to mobile stations in a call on
a CDMA traffic channel.
Fig. 14-4 shows the network reference model for SMS. The base station (BS) contains the
transceiver equipment, the MSC, and any IWF required for network connection. These elements
are grouped together because there is no need to distinguish them. The MC element in the model
represents a generic SMS MC function. The N reference point represents one or more standard-
ized interfaces between an SMS MC and a BS. The TE is voice or data equipment connected
either directly or indirectly to the MC. It is possible for the MC to be included in or co-located
with a BS. In this case, the N interface is internal to the BS.

Um Ai W

MS BS PSTN TE

MC TE

MS: Mobile Station

BS: Base Station
MC: Message Center
PSTN: Public Switching Telephone Network
TE: Terminal Equipment
Figure 14-4 Simplified SMS Reference Model
340 Ch. 14 • Wireless Data

The SMS protocol stack for the CDMA mode of operation is shown in Fig. 14-5. The SMS
bearer service is the portion of the SMS system responsible for delivery of messages between
the MC and mobile user equipment. The bearer service is provided by the SMS transport and
relay layers.
The SMS transport layer is the highest layer of the bearer service protocol. The transport
layer manages the end-to-end delivery of messages. In an entity serving as a relay point, the
transport layer is responsible for receiving SMS transport layer messages from an underlying
SMS relay layer, interpreting the destination address and other routing information, and for-
warding the message via an underlying SMS relay layer. In entities serving as end points, the
transport layer provides the interface between the SMS bearer service and SMS teleservice.
SMS uses the following layers:

• SMS Relay Layer. The SMS relay layer provides the interface between the transport
layer and the link layer used to carry short message traffic. On the Um interface, the
SMS relay layer supports the SMS transport layer by providing the interface to the
TIA IS-95A transmission protocols required to carry SMS data between CDMA
mobile stations and the base stations. On the N interface, the SMS relay layer supports
the SMS transport layer by providing the interface to the network protocols required
to carry SMS data between the MC and TIA IS-95A base stations. The N reference
point is assumed to be an intersystem network link with connectivity to the MC. Inter-
system links can use a variety of public and private protocols. SMS protocols and mes-
sage formats on intersystem links may differ from those used on the CDMA air
interface. The N interface relay layer is responsible for formatting and parsing SMS

Mobile Station Base Station Message Center

(Bearer service end point) (Bearer service (Bearer service end point)
end point)

SMS Teleservice
Layer

SMS Transport
Layer

SMS Relay
Layer

IS-95A IS-95A Link Layer

Figure 14-5 SMS Protocol Stack

Short Message Service 341

messages as necessary when transmitting and receiving messages on the intersystem

links. The SMS relay layer performs the following functions:
1. Accepting transport layer messages and delivering them to the next indicated
relay point or end point
2. Providing error indications to the transport layer when messages cannot be deliv-
ered to the next relay point or end point
3. Receiving messages and forwarding them to the transport layer
4. Interfacing to and controlling the link layer used for message relay
5. Formatting messages according to SMS standards and/or other message stan-
dards, as required by the link layer and/or peer SMS layer
• SMS Transport Layer. This layer resides in SMS bearer service end points and relay
points. In a bearer service end point, the SMS transport layer provides the means of
access to the SMS system for teleservices that generate or receive SMS messages. In a
bearer service relay point, the transport layer provides an interface between relay lay-
ers. The SMS transport layer uses relay layer services to originate, forward, and termi-
nate SMS messages sent between mobile stations and MCs. It is assumed that the link
layers used by the relay layers support message addressing, so that certain address
parameters can be inferred by the relay layer from link layer headers and are therefore
not required in transport layer messages. It is assumed that an SMS point-to-point mes-
sage does not require certain address parameters because the link layers will provide
this address. On the CDMA paging channel, for example, it can be assumed that the
relay layer can extract the address from the ADDRESS field of the TIA-IS-95A data
burst message. SMS transport layers have different functions in SMS bearer service end
points and relay points. In an SMS bearer service end point, the transport layer provides
the following functions:
1. Receiving message parameters from SMS teleservices, formatting SMS transport
layer messages, and passing the message to the relay layer using the appropriate
relay layer service primitives
2. Informing relay layer when all expected acknowledgments of submitted mes-
sages have been received
3. Informing the teleservices when relay layer errors are reported
4. Receiving SMS messages from the relay layer and passing the messages to the
SMS teleservice
5. In mobile stations, performing authentication calculations
In an SMS bearer service relay point, the transport layer provides the following func-
tions:
1. Receiving SMS messages from a relay layer, reformatting the SMS transport
layer message if necessary, and passing the message to another relay layer using
the appropriate relay layer service primitives
342 Ch. 14 • Wireless Data

2. Passing confirmations or error reports between the relay layers if requested

3. In the TIA IS-95A base stations, performing authentication calculations or inter-
facing with the entities performing authentication calculations
The transport layer requires the following services from the relay layer:
1. Accepting transport layer messages and delivering them to the next indicated
relay point or end point
2. Returning confirmations or error reports for messages sent
3. Receiving messages and forwarding them to the transport layer with the appro-
priate parameters
• SMS Teleservice Layer. The teleservice layer resides in a bearer service end point and
supports basic SMS functions through a standard set of subparameters of the transport
layer’s bearer data parameter. When a mobile station sends an SMS User Acknowledg-
ment message, the teleservice layer performs the following:
1. The teleservice layer supplies the destination address parameter to the transport
layer and sets the destination address parameter equal to the address contained in
the originating address field of the SMS message being acknowledged.
2. The MESSAGE_ID field of the message identifier subparameters is set to the
value of MESSAGE_ID field in the SMS message being acknowledged.
Broadcast Messaging Service Teleservice messages are sent using the SMS Deliver
message. For more details refer to TIA IS-637.

14.11 Packet Data Services for CDMA Cellular/PCS Systems [18]

14.11.1 Network Reference Model
The network reference model for packet data services and protocol options specified in
TIA IS-657 is shown in Fig. 14-6.

14.11.2 Network Element

The reference model elements are
• Terminal Equipment 2 (TE2). A TE2 is a data terminal device that has a non-ISDN
user-network interface.

Rm Um L Pi

TE2 MT2 BS/MSC IWF PPDN

Figure 14-6 Network Reference Model

Packet Data Services for CDMA Cellular/PCS Systems [18] 343

• Mobile Terminal 0 (MT0). An MT0 is a self-contained mobile terminal that does not
support an external interface.
• Mobile Terminal 2 (MT2). An MT2 provides a non-ISDN (Rm ) user interface.
• Base Station (BS). A base station represents the equipment on the land side of the Um
interface, including radio processing and management and protocol processing and
management.
• Mobile Switching Center (MSC). The MSC represents the functions provided by the
cellular switch including circuit-switched call management, mobile location, and
mobile management.
• Interworking Function (IWF). An IWF provides functions needed for terminal equip-
ment connected to a mobile termination to interface with other networks such as PSTN
or CDPD network. A CDPD Mobile Data Intermediate System (MD-IS) is an example
of an IWF.
• Public Packet Data Network (PPDN). A public packet-switched data network (such
as the Internet) provides a transport mechanism for packet data between processing ele-
ments capable of using such service.

14.11.3 Network Reference Points

The reference points are
• Reference Point Rm . A physical interface connecting a TE2 to an MT2.
• Reference Point Um . A physical interface connecting an MT0 or MT2 to a BS/MSC.
This is an air interface.
• Reference Point L. A physical interface connecting a BS/MSC to an IWF.
• Reference Point Pi . A physical interface connecting an IWF to a PPDN.

14.11.4 Protocol Options

TIA IS-657 defines the requirements for communication protocols on the links between a
mobile and an IWF, including requirements for Rm , Um , and L interfaces. The relay layer pro-
vides lower-layer communication and packet framing between the entities of the packet data ser-
vice reference model. Over the Rm interface between the TE2 and the MT2, the relay layer is a
simple EIA/TIA 232E interface. Over the Um interface, the relay layer is a combination of RLP
(defined in TIA IS-99 [15]) and TIA IS-95A [14] protocols. On the L interface, the relay layer
uses the protocols defined in TIA IS-687 [19]. The two protocol stack options are discussed in
the following sections.

14.11.4.1 Relay Layer Rm Interface Protocol Option

The relay layer Rm interface protocol option supports TE2 applications in which the TE2 is
responsible for all aspects of packet data service mobility management and network address
management (e.g., IPCP and CDPD registration and authentication protocols). The link layer is
implemented using PPP. When using the relay layer Rm interface protocol option, the link layer
344 Ch. 14 • Wireless Data

connection is between the TE2 and the IWF. The network layer includes protocols such as IP
and CLNP, and packet data network registration and authentication protocols such as MNRP
(see Fig. 14-7).

14.11.4.2 Network Layer Rm Interface Option

The network layer Rm interface protocol option supports TE2 applications in which the
MT2 is responsible for all aspects of packet mobility management and network address manage-
ment (e.g., IPCP and CDPD registration and authentication protocols). In this option, there are
independent link layer connections between the TE2 and the MT2, as well as between the TE2
and IWF. The IWF link layer (between the MT2 and IWF) is implemented using Interned PPP.
The Rm link layer (between the MT2 and TE2) can be implemented using the Interned PPP pro-
tocol to support the IP network layer protocol. For this option, the network layer also provides
independent service between the TE2 and MT2 and between the MT2 and IWF. The TE2
includes routing protocols and operates as if locally connected to a network routing server. The
MT2 includes both routing and packet data network registration and authentication protocols
(see Fig. 14-8).

14.11.4.3 Radio Link Protocol (RLP)

The RLP employs the link layer retransmission approach to improve TCP performance.
Each TCP packet is segmented into several 20-msec radio link frames (192 bits at a 9.6-kbps
transmission rate). The RLP improves the FER seen by the TCP layer. The RLP is a NAK-based
selective repeat ARQ scheme and performs partial link recovery through a limited number of
retransmissions, n, in case of frame error. The RLP frame size is 20 bytes. About 30 RLP frames
per TCP segment are used.

Mobile Station

TE2 MT2 BS/MSC IWF

Upper Upper
Protocol Protocol
Layers Layers

Network Network
Network Layer Layer
Layer Protocols Protocols
IP or CLNP IP or CLNP

Link Layer PPP PPP

RLP RLP Relay Relay

Relay Layer EIA-232 EIA-232
IS-95A IS-95A Layer Layer

Rm Um L

Figure 14-7 Relay Layer Rm Interface Protocol Option

Packet Data Services for CDMA Cellular/PCS Systems [18] 345

Mobile Station

TE2 MT2 BS/MSC IWF

Upper Upper
Protocol Protocol
Layer Layer

Network Network Network

IP or IP or
Layer Layer Layer
CLNP CNLP
Protocols Protocols

Link Layer PPP PPP PPP PPP

RLP RLP Relay Relay

Relay Layer EIA-232 EIA-232
IS-95A IS-95A Layer Layer

Rm Um L

Figure 14-8 Network Layer R Interface Protocol Option

14.11.5 Applicable Mobile Type

The CDMA-CDPD feature applies to mobiles that comply with IS-657 and support ser-
vice negotiation and configuration. The mobiles must support the optional RLP data frame
encryption as specified in IS-657.
Two mobile configurations are defined in IS-657: the relay layer Rm protocol option and
the network layer Rm protocol option. From the system’s perspective, both mobile configurations
are supported equally and transparently. However, the network layer option has an advantage
over the relay layer option—since its mobility management protocols reside with the mobile unit
(MT2), only the standard Internet protocol stacks run on laptop (TE2).

14.11.6 Packet Data Protocol States

The IWF and the mobile use a link layer connection to transmit and receive packets if the
link layer connection is opened when a packet data service option is first connected. Once an
IWF link layer connection is opened, bandwidth (in the form of traffic channel assignment) is
allocated to the connection on an as-needed basis. The IWF link layer connection can be in any
of the following states:
• Closed. The IWF link layer connection is closed when the IWF has no link layer con-
nection state information for the mobile. In this state the mobile does not provide
packet data service.
• Opened. The IWF layer connection is opened when the IWF has link layer connection
state information for the mobile. The opened state has two substates:
1. Active: An open IWF link layer connection is active when there is an L interface
346 Ch. 14 • Wireless Data

virtual circuit for the mobile and the mobile is on a traffic channel with the packet
data service option connected.
2. Dormant: An open IWF link connection is dormant when there is no L interface
virtual circuit for the mobile and the mobile is not on a traffic channel with the
packet data service option connected.
The BS/MSC and IWF maintain the state of link layer connection. The mobile maintains
the state of PPP Link Control Protocol (LCP) and manages the IWF link layer connection using
LCP opening and closing procedures.
With IWF link layer connection in a dormant state, if either the mobile or BS/MSC has
data to send, it is not necessary to reopen the link layer connection or to reinitialize any upper
layer protocols, provided the packet data service type has not changed since the link layer last
entered the dormant state. The mobile and BS/MSC can freely mix packet data service requests
using any supported rate set within a service type.

14.11.6.1 MS Packet Data Service States

MS packet data service states are
• Inactive State. The mobile does not provide packet data service.
• Active State. The mobile provides packet data service.

14.11.6.2 MS Packet Data Service Call Control Functions

The mobile performs the packet data service call control function using the following
states (see Fig. 14-9):
• Null State. The packet data service call control function is in this state when packet
data service has been activated.
• Initialization/Idle State. The mobile attempts to establish a traffic channel for the pur-
pose of initiating packet data service.
• Initialize/Traffic State. The mobile communicates with the BS/MSC on a traffic chan-
nel and attempts to connect a packet data service option for the purpose of initiating
packet data service.
• Connected State. A packet data service option is connected. The mobile can transfer
packet data.
• Dormant/Idle State. The mobile is not on a traffic channel. The mobile cannot transfer
packet data.
• Dormant/Traffic State. The mobile is communicating with the BS/MSC on a traffic
channel, but the packet data service option has been disconnected. The mobile cannot
transfer packet data.
• Reconnect/Idle State. The mobile attempts to establish a traffic channel.
• Reconnect/Traffic State. The mobile communicates with the BS/MSC on a traffic
channel and attempts to connect a packet data service option.
 Service option rejected Null
or released; deactivates Service option
packet data service
State
rejected; deactivates
packet data service
Receives release
reorder, etc.; deactivates
packet data service
Reorder
Service is activated; mobile
is not on a traffic channel;
sends origination Service
Initialization/ Reconnect/
Idle State Idle State
Service is activated,
Traffic channel is initialized; Traffic channel is
mobile is on a traffic
requests packet data initialized; requests
channel; requests packet Release
service option packet service option
service option
Packet Data Services for CDMA Cellular/PCS Systems [18]

Initialization/ Reconnect/
Traffic State Service option Service option Traffic State
connected
Connected connected
State

Figure 14-9 MS Packet Data Service Call Control Functions

Service option
Service option disconnected
connected
Data to send;
requests packet
Release
Dormant/ service option
Traffic State
Traffic channel Serving system
initialized Release
change or data
to send; sends
origination
Dormant/
Idle State
347
348 Ch. 14 • Wireless Data

14.11.6.3 BS/MSC Packet Data Service States

Packet data service processing in the BS/MSC consists of the following states (see Fig.
14-10):
• Inactive State. The BS/MSC does not provide packet data service to the mobile.
• Active State. The BS/MSC provides packet data service to the mobile.

14.11.6.4 BS/MSC Packet Data Service Call Control Functions

The BS/MSC performs the packet data service call control functions which consist of the
following states (see Fig. 14-10):
• Null State. The BS/MSC has no connection of a packet data service option to the
mobile.
• Paging State. The IWF has requested that the BS/MSC connect a packet data service
option to the mobile for delivery of packet data, and the BS/MSC pages the mobile.
• Initialization/Idle State. The BS/MSC is awaiting initialization of a traffic channel
with the mobile.
• Initialization/Traffic State. The mobile is on a traffic channel. The BS/MSC awaits
connection of a packet data service option.
• Connected State. A packet data service option has been connected. Packet data is
exchanged with the mobile.

14.11.7 Packet Mode Data Service Features

The following are packet mode data service features.
• Dual 9.6- and 14.4-kbps Speed. The CDMA packet data services support both 9.6-
and 14.4-kbps data rates.
• Bandwidth Management. The dormant mode is supported to insure that bandwidth is
not wasted if an end point enters an inactive state.
• Mobility. Mobility support for packet mode data services allows the user’s data appli-
cation to continue correct operation during movement of the mobile. Subscriber mobil-
ity is supported for intra-MSC and inter-MSC. Inter-MSC support requires that a
unique System Identification/Network Identification (SID/NID) be assigned for each
MSC. This is because a new IWF will provide the packet data service when a mobile
moves into a new MSC serving area, so an IWF link layer connection transfer must
take place to maintain the user’s packet data connectivity. According to IS-657, the
mobile is required to reopen the link layer connection when it detects a change in the
SID or NID of the serving system. Inter-MSC packet data mobility is easily supported
in a network where each MSC has a unique SID/NID. Intersystem mobility is sup-
ported if both the CDMA and CDPD have roaming agreements and the systems share
Shared Secret Data (SSD).
 Null
Receives link layer
reactivation request, State
MS is not on traffic
channel; pages MS
Receives service option
Receives request for packet service
Service option origination; Service option from MS on traffic channel
rejected; requests IWF initiates link rejected; or receives link layer
link layer closure layer open or requests IWF reactivation request while MS
reactivation layer closure on traffic channel negotiates
packet service option

Paging Service option

Receives page Initialization/
State response; assigns rejected;
traffic channel Idle State requests IWF
link layer closure
Packet Data Services for CDMA Cellular/PCS Systems [18]

Traffic channel
initialized; negotiates
packet service option

Initialization/

Figure 14-10 Packet Data Service Call Control States in BS/MSC

Traffic State

Service option
connected

Service option
Connected disconnected
State
349
350 Ch. 14 • Wireless Data

• Fraud Containment. The CDPD authentication is performed as a part of CDPD regis-

tration procedures when a mobile requests the CDPD service. The mobile user is vali-
dated with the home system based on a triplet formed by the Network Entity Identifier
(NEI), the Authentication Sequence Number (ASN), and the Authentication Random
Number (ARN) assigned to the mobile. The NEI authentication procedure applies to
both intra- and intersystem service requests. The user data sent over the air interface are
encrypted.
• Accounting/Billing Interface. CDPD accounting records consist of information in
terms of data packet count, data octet count, control packet count, control octet count,
discarded packet count, and packet data service connection time. The packet data ser-
vice connection time starts when the mobile first registers with the CDPD network and
ends when the mobile deregisters from the CDPD network. The accounting information
is collected at the serving IWF and distributed to the home IWF in real time. The
CDMA network will continue to collect the accounting records in the same way as the
voice and circuit-mode data calls. It collects only air link connection time, not the data
volume. Additional information may be added to AMA records for packet data calls to
keep track of active and dormant links.
• CDMA Data Encryption. CDMA data encryption is performed on mobile RLP data
frames carried on the digital traffic channel for mobiles conforming to IS-657 or IS-99
according to “Common Crytographic Algorithms, Revision A.1,” defined in TIA
IS95.0.A. The data encryption feature increases the security of user data transmitted
over the air interface. It is mandatory for CDMA-CDPD services (i.e., service options 7
and 15). The feature requires activation of the CDMA authentication feature with glo-
bal authentication because the DataKey and L table used in the encryption process is
calculated for each data service connection only after a successful CDMA authentica-
tion. CDMA authentication is necessary to insure that the authentication center or Vis-
ited Location Register (VLR) has the same SSD and, therefore, the same DataKey and
L table as the mobile.
• Voice or Circuit-Mode Data Service while Packet Data Dormant. The feature
allows subscribers with a multifunction packet data mobile device with voice and/or
circuit data capabilities to use voice or circuit data services while the packet data is in
the dormant mode.The mobile maintains the packet data service call control state infor-
mation while handling the voice call and restarts the packet data protocol stack when
the dormant link layer connection is reactivated. It is assumed that the mobile is set up
either manually or automatically in the proper mode for the requested service. The sys-
tem is not involved in setting up the service mode on the mobile. The dormant link
layer connection is supported with some limitations during a voice call. The limitation
is that a request to reactivate the dormant link layer connection will be rejected if there
is a voice call. When a packet data call is rejected, the IWF will gracefully discard the
data received. This may eventually cause the user data application to be terminated if
the voice call continues longer than is allowed by the user application. When a link
Summary 351

layer connection enters the dormant state, subscriber call features that were deactivated
for packet data calls will be reactivated. Tones and announcements will be played for
the voice call. When a packet data call is established, subscriber call features that are
not supported for packet data calls will be deactivated, and tones and announcements
will be blocked.
• Simultaneous Voice and Data Support. This feature allows voice and packet data
traffic to be carried simultaneously within a single traffic channel. Subscribers must
have a mobile device that is capable of simultaneous voice and data service. Normal
voice calls are allowed and carried as primary traffic at any time, whether the mobile is
connected to a packet data call or not. If a new voice call is requested while the packet
data call is connected, the packet data call will be dropped first to start the voice call.
The packet data call will then be added using service negotiation procedures to provide
a service configuration containing voice as primary traffic and packet data as secondary
traffic. The transition of packet data from primary to secondary traffic may cause loss
of user data, but this will be detected and retransmitted as necessary by the upper layer
protocols without disruption to the user application. Thus, during a voice call, a link
layer connection may be opened and a packet data call established and data carried as
secondary traffic without disrupting the voice call. Both 9.6- and 14.4-kbps data rates
are supported for this feature.
• Service Options 4103 and 15. Service option 4103 provides the standard Internet pro-
tocol packet-mode data services instead of the CDPD data service. Service option 4103
is for a data rate of 9.6 kbps, whereas service option 15 is for the same service at 14.4
kbps. Service options 4103 and 15 may be provided with or without mobility support.
Mobility management for service options 4103 and 15 will be based on mobile IP pro-
tocol. Service options 4103 and 15 without mobility is suitable for users requiring only
portable services (i.e., users not moving while a data application is running).

14.12 Summary
Since wireless data networks do not operate without interconnection to other networks, this
chapter covered a variety of wireless data systems including the wide-area wireless data sys-
tems, high-speed Wireless Local Area Networks (WLANs), and the specific systems supported
by CDMA. We examined the various standards being adopted by the IEEE, WINForum, ETSI
(in Europe), and ARPA for wireless LANs. Since packet networks are an important part of wire-
less networks, the chapter briefly listed the characteristics of the access methods in common use
and defined their throughput equations. The common packet protocols are ALOHA, slotted
ALOHA, and Carrier-Sense-Multiple Access (CSMA) ALOHA.
We then presented the methods used to control errors for wireless data systems and con-
cluded with the highlights of the TIA IS-99, TIA IS-637, and TIA IS-657 standards for CDMA
cellular systems. CDMA supports asynchronous data, facsimile, packet data, and short message
service (SMS) to end points in another wireless network or to the wireline network. We exam-
ined the reference models and protocol stacks for each of these data services.
352 Ch. 14 • Wireless Data

14.13 References
1. Bates, R. J., Wireless Networked Communication, McGraw-Hill, Inc., NewYork, 1994.
2. Habab, I. M., Kavehrad, M., and Sundberg, C. E. W., “ALOHA with Capture over Slow and
Fast Fading Radio Channels with Coding and Diversity,” IEEE Journal of Selected Areas of
Communications 6, 1988, pp. 79–88.
3. Hammond, J. L., and O’Reilly, J. P., Performance Analysis of Local Computer Networks,
Addison-Wesley Publishing Company, Reading, MA, 1986.
4. Pahlavan, K., and Levesque, A. H., Wireless Information Networks, John Wiley and Sons,
Inc., New York, 1995.
5. RFC 791, “Internet Protocol.”
6. RFC 792, “Internet Control Message Protocol.”
7. RFC 793, “Transmission Control Protocol.”
8. RFC 1122, “Requirements for Internet Hosts—Communication Layers.”
9. RFC 1144, “Compressing TCP/IP Headers for Low-Speed Serial Links.”
10. RFC 1332, “The PPP Internet Protocol Control Protocol (IPCP).”
11. RFC 1661, “The Point-to-Point Protocol (PPP).”
12. RFC 1700, “Assigned Numbers.”
13. Skalar, B., Digital Communications—Fundamentals and Applications, Prentice Hall, Engle-
wood Cliffs, NJ, 1988.
14. TIA IS-95A, “Mobile Station–Base Station Compatibility Standard for Dual-Mode Wide-
band Spread Spectrum Cellular System.”
15. TIA IS-99, “Data Services Option Standard for Wideband Spread Spectrum Digital Cellular
System.”
16. TIA 232E, “Interface between DTE and DCE Employing Serial Binary Data Interchange.”
17. TIA IS-637, “Short Message Services for Wideband Spread Spectrum Cellular Systems.”
18. TIA IS-657, “Packet Data Services Option for Wideband Spread Spectrum Cellular System.”
19. TIA IS-687, “Data Services Inter-Working Function Interface Standard for Wideband Spread
Spectrum Digital Cellular System.”
20. Viterbi, A. J., and Padovani, Roberto, “Implications of Mobile Cellular CDMA,” IEEE
Communication Magazine 30 (12), 1992, pp. 38–41.
21. Walrand, J., Communications Networks: A First Course, Irwin, Homewood, IL, 1991.
22. Woo, Ill, and Cho, Dong-Ho, “A Study on the Performance Improvements of Error Control
Schemes in Digital Cellular DS/CDMA Systems,” IEICE Trans. Communications E77-B
(7), July 1994.
 C H A P T E R 1 5

cdma2000 System

15.1 Introduction
The International Telecommunications Union-Radio Communication (ITU-R) standardization
sector developed specifications for International Mobile Telecommunications—2000 (IMT-
2000). IMT-2000 has greatly expanded the range of service capabilities and covers a wide range
of environments. IMT-2000 [5] specifications are aimed to facilitate the introduction of new
capabilities and to provide a seamless evolution from the substantially installed second-genera-
tion (2G) telecommunications base by the year 2000+. The third-generation (3G) telecommuni-
cations systems based on IMT-2000 specifications will be introduced in service in the years
2000–2002. The 3G systems will offer a plethora of telecommunications services including
voice, low- and high-bit-rate data, multimedia, and video to mobile users via a range of mobile
terminals, operating in both public and private environments (office areas, residential areas,
transportation media, etc.).
The ITU World Administration Radio Conference in 1992 (WARC-92) identified 230
MHz in a 2-GHz band for use on a worldwide basis for the satellite and terrestrial components
of IMT-2000. The WARC-95 revised 2-GHz frequency allocations for mobile satellite services
(MSS) to provide a satellite component of IMT-2000 (see Fig. 15-1).
Table 15-1 provides a summary of the current IMT-2000 3G air interface proposals to ITU
and their network interfaces.
The cdma2000 Radio Transmission Technology (RTT) is a wideband, spread spectrum
radio interface that uses CDMA technology to satisfy the needs of 3G wireless communication
systems. This RTT meets all requirements specified in the ITU circular letter and the corre-
sponding documents of the IMT-2000. The service requirements are satisfied for indoor office,
indoor-to-outdoor/pedestrian, and vehicular environments. The cdma2000 system will also be
backward compatible with the current cdmaOne (IS-95) family of standards.

353
 354

ITU IMT-2000
MSS MSS MSS MSS
1885 1980 2010 2025 2110 2160 2170 2200 MHz

Europe DECT
IMT- IMT-2000 MSS MSS IMT-2000 MSS
1880 1900 1980 2010 2025 2110 2170 2200 MHz

Japan PHS IMT-2000 MSS IMT-2000 MSS

1885 1895 1918.1 1980 2010 2025 2110 2170 2200 MHz

Figure 15-1 ITU Frequency Allocation Worldwide for 3G

CDMA FDD TDD CDMA FDD
China WLL WLL WLL MSS MSS
1865 1880 1900 1920 1945 1960 1980 2010 2025 2110 2170 2200 MHz

USA A D B E F C PCS A D B E F C MSS Broadcast Auxiliary Reserve MSS

1850 1910 1930 1990 2025 2110 2150 2185 2200 MHz
Ch. 15 • cdma2000 System
Introduction 355

Table 15-1 Current IMT-2000 Proposals

cdma2000 ARIB/DOCOMO UMTS UWC-136
2G System IS-95/cdmaOne PDC GSM IS-136
3G Air cdma2000 W-CDMA UTRA (W-CDMA/ IS-136+/
Interface TD-CDMA) 136-HS/
136+HS
3G Network Evolved ANSI-41 Evolved GSM Evolved GSM Evolved ANSI-41
Interface MAP MAP
Standards TIA TR-45 ARIB ETSI TIA TR-45
Bodies (supported by CDG) (supported by UWCC)
UMTS = Unversal mobile telecommunications system
UWC = Universal wireless communication
PDC = Pacific digital cellular
HS = High speed
UTRA = UMTS terrestrial radio access
UWCC = UWC Consortium
ARIB = Association of Radio Industries Business (Japan)

The cdma2000 system provides a wide range of implementation options to support data
rates (both circuit switched and packet switched) starting from the TIA IS-95B-compatible rate
of 9.6 kbps up to greater than 2 Mbps. The cdma2000 system provides maximum flexibility to
carriers in making engineering trade-offs between
• Channel sizes of 1, 3, 6, 9, and 12 × 1.25 MHz
• Support for advanced antenna technologies
• Cell sizes (e.g., the cdma2000 system’s increased performance can be realized in terms
of increased range to permit carriers to reduce the total number of cell sites)
• Higher data rates that can be supported in all channel sizes
• Support for advanced services possible or practical in other systems (e.g., high-speed
circuit data, B-ISDN, or H.224/223 teleservices)
The cdma2000 system can be operated economically in a wide range of environments,
including
• Outdoor megacells (cell >35 km radius)
• Outdoor macrocells (cell 1–35 km radius)
• Indoor/outdoor microcells (up to 1 km radius)
• Indoor/outdoor picocells (<50 m radius)
The cdma2000 system can be deployed in
• Indoor/outdoor environment
• Wireless local loop (WLL)
356 Ch. 15 • cdma2000 System

• Vehicular environment
• Mixed vehicular and indoor/outdoor environment
The cdma2000 system’s mobility is variable, ranging from fixed wireless to high speeds of
up to 300 mph. cdma2000 provides a layered structure to support the integration of the bottom
two layers of the RTT into systems that implement any network standards (e.g., ITU-T-defined
signaling services). It also provides backward compatibility to TIA IS-95B signaling and call
control models. An extended cdma2000 upper layer signaling structure is capable of supporting
a wide range of advanced services (e.g., multimedia) in an optimized and efficient manner.
cdma2000 supports the 3G wireless intelligent networking (WIN) services and services
defined by the ITU or other international standards organizations and provides a graceful evolu-
tion from existing 2G TIA IS-95B technology. It includes the following features:
• Support for overlay configurations
• Support for backward compatibility to TIA IS-95B signaling and network
• Support for graceful and gradual upgrade from 2G system to 3G system
• Sharing of common channels with an underlay TIA IS-95B system during transition
periods
cdma2000 provides an evolutionary path by reusing existing TIA IS-95B standards,
including
• TIA IS-95B: Mobile Station and Radio Interface Specifications
• IS-707: Data Services (Packet, Async, and Fax)
• IS-127: Enhanced Variable-Rate Codec (EVRC) 8.5-kbps speech coder
• IS-733: 13-kbps speech coder
• IS 637: Short Message Service (SMS)
• IS 638: Over-the-Air Activation and Parameter Administration (supporting the configu-
ration and service activation of mobile stations over the radio interface)
• IS-97 and IS-98 (Minimum Performance Specifications)
• The basic TIA IS-95B channel structure
• Extensions to TIA IS-95B Fundamental/supplemental channel structure, multiplex
layer, and signaling to support higher rate operation, common broadcast channels
(pilot, paging, and sync)
• IS-634A: no significant changes expected for cdma2000; the layered structure of
cdma2000 integrates smoothly with the component structure of IS-634A
• TIA IS-41D: No significant changes needed for cdma2000; the layered structure of
cdma2000 offers the potential for easy integration with enhanced network services (WIN)

15.2 cdma2000 Layering Structure

15.2.1 Upper Layer
Fig. 15-2 shows the layer structure of cdma2000. The upper layers contain three basic
services:
cdma2000 Layering Structure 357

IS-95 cdma2000 Other Upper Packet Voice Circuit Data

OSI Layer Circuit Upper Layer Data Services Service
Layers Upper Signaling Layer Signaling Service
(3–7) Layers
Signaling

IS-95 cdma2000 Other Signaling Packet Data Null Circuit Data LAC
Signaling Signaling Layer 2 Layer 2 Layer 2 Layer 2 Sub-
Layer 2 Layer 2 layer
PLICF for MAC Instance 1 PLICF for MAC Instance 2
OSI Link (e.g., Signaling) (e.g., Packet or Circuit Data Service)
Layer 2 Layer MAC
PLDCF Specific to Instance 1 PLDCF Specific to Instance 2 Sub-
layer
PLDCF MUX and QoS Sublayer

OSI Physical
cdma2000 Physical Layer
Layer 1 Layer

PLICF = Physical-Layer-Independent Convergence Function

PLDCF = Physical-Layer-Dependent Convergence Function
QoS = Quality-of-Service

Figure 15-2 cdma2000 Layering Structure

• Voice services. Voice telephony services, including PSTN access, mobile-to-mobile

voice services, and Internet telephony.
• End-user data-bearing services. Services that deliver any form of data on behalf of
mobile end user, including packet data (e.g., IP service), circuit data services (e.g., B-
ISDN emulation services), and SMS. Packet data services conform to industry standard
connection-oriented and connectionless packet data including IP-based protocols (e.g.,
TCP and UDP) and ISO/OSI Connectionless Interworking Protocol (CLIP). Circuit
data services that emulate international-standards-defined, connection-oriented ser-
vices such as asynchronous (async) dial-up access, fax, V.120 rate-adapted ISDN, and
B-ISDN services.
• Signaling. Services that control all aspects of operation of the mobile.

15.2.2 Link Layer

The link layer provides varying levels of reliability and QoS characteristics according to
the needs of the specific upper layer service. It gives protocol support and control mechanisms
for data transport services and performs all the functions necessary to map the data transport
needs of the upper layers into specific capabilities and characteristics of the physical layer. The
link layer is subdivided into sublayers.
• Link Access Control (LAC) sublayer (see Fig. 15-3)
• Media Access Control (MAC) sublayer (see Fig. 15-4)
The LAC sublayer manages point-to-point communication channels between peer upper
layer entities and provides framework to support a wide range of different end-to-end reliable
link layer protocols.
358 Ch. 15 • cdma2000 System

OSI Layers
(3–7)
Upper Layer Entities

LAC IS-95 Signaling cdma2000 Other Signaling Packet Data Null Circuit Data
Sublayer Layer 2 Signaling Layer 2 Layer 2 Layer 2 Layer 2
Layer 2 Link
Layer
(OSI 2)

Media Access Control (MAC)

OSI Layer 1 cdma2000 Physical Layer

Figure 15-3 Link Layer of cdma2000

IS-95 2G cdma2000 Other

Upper Layers Packet Circuit
Layer 3 upper layer upper layer Voice
(OSI 3–7) Data
Signaling signaling signaling Data Services
Service Service

LAC
Sublayer LAC Sublayer

PLICF for MAC Instance 1 PLICF for MAC Instance 2 Link

(e.g., signaling) (e.g., packet or circuit data service) Layer
MAC (OSI 2)
Sublayer PLDCF Specific to Instance 1 PLDCF Specific to Instance 2

PLDCF MUX and QoS Sublayer

Physical
cdma2000 Physical Layer
Layer (OSI 1)

Figure 15-4 cdma2000 MAC Sublayer

The cdma2000 system includes a flexible and efficient MAC sublayer that supports multi-
ple instances of an advanced-state machine, one for each active packet or circuit data instance.
Together with the QoS control entity, the MAC sublayer realizes the complex multimedia, multi-
service capabilities of 3G wireless systems with QoS management capabilities for each active
service. The MAC sublayer provides three important functions:
• MAC control state. Procedures for controlling the access of data services (packet and
circuit) to the physical layer (including contention control between multiple services
from a single user as well as between competing users).
• Best-effort delivery. Reasonably reliable transmission over the radio link with radio
link protocol (RLP) providing a best-effort level of reliability.
cdma2000 Layering Structure 359

• Multiplexing and QoS control. Enforcement of negotiated QoS levels by mediating

conflicting requests from competing services and appropriately prioritizing access
requests.
The MAC sublayer provides differing QoS to the LAC sublayer (e.g., different modes of
operation). It may be constrained by backward compatibility (e.g., for IS-95B signaling layer 2),
and it may have to be compatible with other link layer protocols (e.g., for compatibility with
non-IS-95 air interfaces or for compatibility with future ITU-defined protocol stacks). The MAC
sublayer is subdivided into
• Physical-Layer-Independent Convergence Function (PLICF)
• Physical-Layer-Dependent Convergence Function (PLDCF) which is further subdi-
vided into
◆ Instance-specific PLDCF
◆ PLDCF MUX and QoS sublayer
PLICF provides service to the LAC sublayer and includes all MAC operational proce-
dures and functions that are not unique to physical layer. Each instance of PLICF main-
tains service status for the corresponding service. PLICF uses services provided by
PLDCF to implement actual communications activities in support of MAC layer ser-
vice. Services used by PLICF are defined as a set of logical channels that carry differ-
ent types of control or data information. The PLICF data service consists of the
following states/substates (see Figs. 15-5 and 15-6):
• Null state
• Initialization state
• Control hold state
◆ Normal substate
◆ Slotted substate
• Active state
• Suspended state
◆ Virtual traffic substate
◆ Slotted substate
• Dormant state
◆ Dormant/idle substate
◆ Dormant/burst substate
The null state is considered to be the default state prior to activation of packet data service.
After the packet service is invoked, a transition to the initialization state occurs during which an
attempt is made to connect the packet service.
360 Ch. 15 • cdma2000 System

TIA/EIA-95B
MAC
Big Time-Out

Active Dormant
State State

Traffic

IS-95 3G
MAC
Time-Out Time-Out Time-Out
Active Control Suspended Dormant
State Hold State State State
Traffic Traffic

Traffic

Release

• Traffic, power control, • Power control & control • No dedicated channels • No dedicated channels
& control channels channels assigned • RLP & PPP state • No BS, MSC resources
assigned • Very fast traffic channel maintained • PPP state maintained
reassignment • Virtual active set • Small data bursts
• Slotted submode

Figure 15-5 Packet Data MAC Operation States in cdma2000 and TIA IS-95B

Service option connected

Service option not connected Active State

No data exchange
dtch is for more than Tactive;
Suspended State
established release is sent
dmch
Service option established
connected;
dmch_control
Not exchanging
Initialization State established Control Hold State data for more Not exchanging user data
than Thold for more than Tsuspend
PPP terminated;
Release message is sent;
Packet service Packet service release sent
PPP not terminated
request deactivated
Service option Dormant State
connected;
dmch_control established Has data to
Null State send
Reconnect State

PPP terminated;
Release sent dtch: dedicated traffic channel
dmch_control: dedicated MAC channel control
Note: Suspended state, dormant state, and reconnect state are applicable only to packet data services. PPP: point-to-point protocol

Figure 15-6 PLICF Data Services State Diagram

cdma2000 Layering Structure 361

Traffic, power control, and control channels are assigned in the active state. In the control
hold state, a dedicated control channel is maintained between the user and the base station on
which any MAC control command (for example, the command to begin a high-speed data burst)
can be transmitted with virtually no latency. Power control is also maintained so that high-speed
burst operation can begin without any delay due to stabilization of power control.
In the suspended hold state, there are no dedicated channels maintained to or from the
user. However, the state information for RLP is maintained, and the base station and user main-
tain a virtual active set that allows either one to know which base station can best be used
(accessed by the user or paged by the base station) in the event that packet data traffic occurs for
the user. This state also supports a slotted substate that permits the user’s mobile device to pre-
serve power in a highly efficient manner.
A short data burst mode is added to the cdma2000 dormant state to support the delivery of
short messages without incurring the overhead of a transition from dormant to active state. Tran-
sitions between MAC states can be indicated by MAC control signaling or by the expiration of
timers. By properly selecting the values for the timers, the cdma2000 MAC can be adapted to a
wide variety of data services and operating environments.
The states are categorized as either connected or not connected depending on the status of
data service option. The data service option is connected in the control hold, active, and sus-
pended hold states. The data service option is not connected in the null, initialization, dormant,
or reconnect states. Fig. 15-6 shows the state diagram for the PLICF data service option.
PLDCF performs mapping of logical channels from PLICF to logical channels supported
by the specific physical layer. PLDCF performs multiplexing, demultiplexing, and consolidation
of control information with bearer data from the control and traffic channels from multiple
PLICF instances in the same mobile. PLDCF implements QoS capabilities, including resolution
of priorities between competing PLICF instances, and maps QoS requests from PLICF instances
into the appropriate physical layer service requests to deliver the desired QoS. The major func-
tions of this sublayer are to
• Perform any required mapping of the simpler logical channels from the PLICF into the
logical channels supported by physical layer
• Perform any (optional) automatic repeat request (ARQ) protocol functions that are
tightly integrated with physical layer
• Perform some of the physical-layer-specific low-level functions of IS-95B RLP
For cdma2000, four specific PLDCF ARQs are defined.
1. Radio Link Protocol (RLP). This protocol provides a highly efficient streaming ser-
vice that makes a best effort to deliver data between peer PLICF entities. RLP provides
both transparent and nontransparent modes of operation. In the nontransparent mode,
RLP uses ARQ protocol to retransmit data segments that were not delivered properly
by the physical layer. In the nontransparent mode, RLP can introduce some delay. In
the transparent mode, RLP does not retransmit missing data segments. However, RLP
362 Ch. 15 • cdma2000 System

maintains byte synchronization between the sender and receiver and notifies the
receiver of the missing parts of the data stream. Transparent RLP does not introduce
any transmission delay and is useful for implementing voice services over RLP.
2. Radio Burst Protocol (RBP). This protocol provides a mechanism for delivering rela-
tively short data segments with best-effort delivery over a shared access common traffic
channel (ctch). This capability is useful for delivering a small amount of data without
incurring the overhead of establishing a dedicated traffic channel (dtch).
3. Signaling Radio Link Protocol (SRLP). This protocol provides best-effort streaming
service for signaling information analogous to RLP, but optimized for the dedicated
signaling channel (dsch).
4. Signaling Radio Burst Protocol (SRBP). This protocol provides a mechanism to
deliver signaling messages with best-effort delivery analogous to RBP, but optimized
for signaling information and the common signaling channel (csch).

PLDCF includes a Radio Link Access Control (RLAC) function that abstracts the RLP
and RBP from PLICF and coordinates the transmission of data (traffic or signaling) between
RLP and RBP according to the current operational state of MAC (e.g., restricts the use of RBP
to cases in which PLICF is in the packet data dormant state).
The PLDCF MUX and QoS sublayer coordinates multiplexing and demultiplexing of code
channels from multiple PLICF instances. It implements and enforces QoS differences between
instances and maps the data streams and control information on multiple logical channels from
different PLICF instances into requests for logical channels, resources, and control information
from the physical layer.

15.3 cdma2000 Channels

15.3.1 Channel-Naming Convention
A logical channel is denoted by three or four lowercase acronyms followed by “ch” for
channel. The fourth letter applies to common channels used in dormant or suspended states.
Table 15-2 lists the conventions for logical channels.
A physical channel (Table 15-3) is represented by uppercase abbreviations. The first letter
in the name of the channels indicates the direction of the channel, except for the paging and
access channels where the direction is implicitly specified.

15.4 Logical Channels Used by PLICF

The following sections describe the logical channels used by PLICF.

15.4.1 Dedicated Traffic Channel (f/r-dtch)

dtch is the forward or reverse logical channel that is used to carry user data traffic. This
logical channel is a point-to-point channel and is allocated for use throughout the active state of
data service. It carries data dedicated to a single PLICF instance.
Logical Channels Used by PLICF 363

Table 15-2 Logical Channel-Naming Convention

First Letter Second Letter Third Letter
f = forward (BS to MS) d = dedicated t = traffic
r = reverse (MS to BS) c = common m = MAC
s = signaling

Table 15-3 Physical Channel-Naming Convention

Channel Name Physical Channel
F/R-FCH Forward/Reverse Fundamental Channel
F/R-SCH Forward/Reverse Supplementary Channel
F/R-DCCH Forward/Reverse Dedicated Control Channel
F-PCH Forward Paging Channel
R-ACH Reverse Access Channel
F/R-CCCH Forward/Reverse Common Control Channel
F-DAPICH Forward Dedicated Auxiliary Pilot Channel
F-CAPICH Forward Common Auxiliary Pilot Channel
F/R-PICH Forward/Reverse Pilot Channel
F-SYNC Forward Sync Channel

15.4.2 Common Traffic Channel (f/r-ctch)

ctch is the forward or reverse logical channel that is used to carry short data bursts associ-
ated with the data service in the dormant/burst substate of the dormant state. This logical chan-
nel is a point-to-point channel and is allocated for the duration of the short burst. It shares access
among many mobiles and/or PLICF instances.

15.4.3 Dedicated MAC Channel (f/r-dmch_control)

dmch_control is the forward or reverse logical channel that is used to carry MAC mes-
sages. This logical channel is a point-to-point channel and is allocated throughout the active
state and control hold state of data service. It carries control information dedicated to a single
PLICF instance.

15.4.4 Reverse Common MAC Channel (r-cmch_control)

The r-cmch_control is the reverse logical channel used by the mobile while data service is
in the dormant/idle substate of the dormant state or suspended state. This logical channel is used
to carry MAC messages. It is shared by a group of mobiles in the sense that access to this chan-
nel is gained on a contention basis.
364 Ch. 15 • cdma2000 System

15.4.5 Forward Common MAC Channel (f-cmch_control)

The f-cmch_control is the forward logical channel used by the base station while data ser-
vice is in the dormant/idle substate of the dormant state or suspended state. This logical channel
is used to carry MAC messages. It is a point-to-multipoint channel.

15.4.6 Dedicated Signaling Channel (dsch)

dsch carries upper layer signaling data dedicated to a single PLICF instance.

15.4.7 Common Signaling Channel (csch)

csch carries upper layer signaling data with shared access among many mobiles and/or
PLICF instances.

15.5 Physical Layer

The physical layer provides coding and modulation services for a set of logical channels
used by the PLDCF MUX and QoS sublayer. The physical channels (see Fig. 15-7) are clas-
sified as
• Forward/Reverse Dedicated Physical Channels (F/R-DPHCH). The collection of all
physical channels that carry information in a dedicated, point-to-point manner between
the base station and a single mobile (see Fig. 15-8)
• Forward/Reverse Common Physical Channels (F/R-CPHCH). The collection of all
physical channels that carry information in a shared access, point-to-multipoint manner
between the base station and multiple mobile stations (see Fig. 15-9)

15.6 Forward Link Physical Channels

Forward dedicated channels carry information between the base station and a specific mobile;
common channels carry information from the base station to a set of mobiles in a point-to-multi-
point manner. Table 15-4 lists these channels.

Forward Link Reverse Link

Pilot (F-PICH) Pilot (R-PICH)

Paging (F-PCH) Access (R-ACH) or R-CCCH
Sync (F-SYNC) Dedicated Control (R-DCCH)
Mobile
Mobile Base Station
Base Fundamental (F-FCH) Station Station Fundamental (R-FCH)
Station
Supplementary (F-SCH) Supplementary (R-SCH)
Quick Paging (F-QPCH)

Figure 15-7 cdma2000 Physical Channels

Forward Link Physical Channels 365

F-DAPICH Forward Dedicated

Auxiliary Pilot Channel

F-FCH Forward Fundamental

Channel

F-DCCH Forward Dedicated Control

F-SCH1 Forward Supplemental
Channel
Channel-1
F-DPHCH F-SCHT Forward Supplemental F-SCH2 Forward Supplemental
(Forward Dedicated Channel Type Channel-2
Physical Channel) •
•
•

R-PICH Reverse Pilot Channel

R-FCH Reverse Fundamental

Channel

R-DCCH Reverse Dedicated Control

Channel R-SCH1 Reverse Supplemental
Channel-1
R-DPHCH R-SCHT Reverse Supplemental R-SCH2 Reverse Supplemental
(Reverse Dedicated Channel Type Channel-2
Physical Channel) •
•
•

Figure 15-8 cdma2000 Overview of Dedicated Physical Channels

F-PICH Forward Pilot Channel

F-PCH Forward Paging Channel
F-CAPICH Forward Common
Auxiliary Pilot Channel F-CCCH Forward Common Control
Channel
F-CPHCH F-CCHT Forward Common F-SYNC Forward Sync Channel
(Forward Common Channel Type
Physical Channel)

R-ACH Reverse Access Channel

R-CPHCH R-CCHT Reverse Common R-CCCH Reverse Common Control
(Reverse Common Channel Type Channel
Physical Channel)

Figure 15-9 cdma2000 Overview of Common Physical Channels

15.6.1 Forward Pilot Channel (F-PICH)

This channel is continuously broadcast throughout the cell in order to provide timing and
phase information. The common pilot is an all-0s sequence prior to Walsh spreading with Walsh
0. The F-PICH is shared by all traffic channels and is used for
• Estimating channel gain and phase
366 Ch. 15 • cdma2000 System

Table 15-4 Forward Link Channels

Physical Channel Channel Name
Forward Common Forward Pilot Channel F-PICH
Physical Channels
Forward Paging Channel F-PCH
(control and overhead
channels) Forward Sync Channel F-SYNC
Forward Common Control Channel F-CCCH
Forward Common Auxiliary Pilot Channel F-CAPICH
Forward Quick Paging Channel F-QPCH
Forward Broadcast Common Channel F-BCCH
Forward Dedicated Forward Dedicated Auxiliary Pilot Channel F-DAPICH
Physical Channels
Forward Dedicated Common Control Channel F-DCCH
Forward Traffic Channel
—Fundamental F-FCH
—Supplementary F-SCH

• Detecting multipath rays so that RAKE fingers are efficiently assigned to the strongest
multipath
• Cell acquisition and handoff
With a common pilot, it is possible to send the pilot signal without incurring significant
overhead for each user. A system with a common pilot approach can achieve better performance
than a system using a per-user pilot approach. For voice traffic, the common pilot can provide
better channel estimation and lower overhead, resulting in improved receiver performance. It can
also provide improved search and handoff performance.

15.6.2 Forward Sync Channel (F-SYNC)

The sync channel is used by mobiles operating within the coverage area of the base station
to acquire initial time synchronization. There are two types of F-SYNC: shared F-SYNC and
wideband F-SYNC. The shared F-SYNC provides service to both the IS-95B and cdma2000
when using the F-SYNC in an IS-95B underlay channel. This mode is applicable only in overlay
configurations.
The wideband F-SYNC is modulated across the entire wideband channel. The wideband
F-SYNC is modulated as a separate channel within the forward common physical channel (F-
CPHCH). This mode is applicable to both overlay and nonoverlay configurations.

15.6.3 Forward Paging Channel (F-PCH)

A cdma2000 system can have multiple paging channels per base station. A paging channel
is used to send control information and paging messages from the base station to mobiles and
Forward Link Physical Channels 367

operates at a data rate of 9.6 or 4.8 kbps (same as IS-95). The F-PCH carries overhead messages,
pages, acknowledgments, channel assignments, status requests, and SSD updates from the base
station to the mobile.
There are two types of paging channels: shared F-PCH and wideband F-PCH. The
shared F-PCH provides service to both the IS-95B and cdma2000 when using the F-PCH in an
IS-95B underlay channel. This mode is applicable only in overlay configurations. The wide-
band F-PCH is modulated across the entire wideband channel. The wideband F-PCH is modu-
lated as a separate channel within the F-CPHCH. This mode is applicable to both overlay and
nonoverlay configurations.
Figs. 15-10 and 15-11 show F-CPHCH (F-PICH, F-SYNC, and F-PCH) for N = 1 and N ≥ 3.

15.6.4 Forward Common Control Channel (F-CCCH)

The F-CCCH is a common channel used for communication of layer 3 and MAC mes-
sages from the base station to the mobile. Possible frame sizes for F-CCCH are 5 ms, 10 ms, and
20 ms, depending upon the operating environment. It is identical with the F-PCH for a 9.6-kbps
rate (20-ms frame).

15.6.5 Forward Common Auxiliary Pilot Channel (F-CAPICH)

This channel is used with antenna beam-forming applications to generate spot beams. Spot
beams can be used to increase coverage in a particular geographical area or to increase capacity
toward hot spots. The F-CAPICH can be shared among multiple mobiles in the same spot beam.

W0

Pilot Channel (All 0s)

A

Wsync

Sync Channel
k = 9, r = 1/2 Repeat Block
A
Conv. Encoder 2 Times Interleaver
1.2 kbps 2.4 kbps 4.8 ksps

Wpaging

Paging Channel
k = 9, r = 1/2 Repeat Block
A
Conv. Encoder 1 or 2 Times Interleaver
9.6 kbps 19.2 kbps 19.2 ksps
4.8 kbps 9.6 kbps 19.2 ksps

Long-Code Mask Long-Code

for Paging Channel Generator Decimator

Note: Signal Point A feeds into Fig. 15.22.

Figure 15-10 cdma2000 F-CPHCH for N = 1

368 Ch. 15 • cdma2000 System

W0

Pilot Channel (All 0s)

X

Wsync

Sync Channel
k = 9, r = 1/3 Repeat Block
X
Conv. Encoder 2 Times Interleaver
1.2 kbps 3.6 kbps 7.2 ksps

Wpaging

Paging Channel
k = 9, r = 1/3 Repeat Block
X
Conv. Encoder 1 or 2 Times Interleaver
9.6 kbps 28.8 kbps 28.8 ksps
4.8 kbps 14.4 kbps 28.8 ksps

Long-Code Mask Long-Code

for Paging Channel Generator Decimator

Note: Signal Point X feeds into Fig. 15.23.

Figure 15-11 cdma2000 F-CPHCH for N ≥ 3

Auxiliary pilots are code multiplexed with other forward link channels, and they use
orthogonal Walsh codes. Since a common pilot contains no data (all 0s), auxiliary pilots may
use a longer Walsh sequence to lessen the reduction of orthogonal Walsh codes available for
traffic channels. Auxiliary pilots can also be used for orthogonal diversity transmission in the
direct spread forward link. Furthermore, if the CDMA system uses a separate antenna array to
support directional or spot beams, it is necessary to provide a separate forward link pilot for
channel estimation.

15.6.6 Forward Broadcast Common Channel (F-BCCH)

This is a paging channel dedicated to carrying only overhead messages and possible SMS
broadcast messages. It removes the overhead messages from the paging channel to a separate
broadcast channel. This improves the mobile initialization time and system access perfor-
mance. At the same time, by reducing the number of messages on the F-PCH, the paging capac-
ity is improved. The F-BCCH has a fixed Walsh code that is communicated to the mobile on the
F-SYNC.

15.6.7 Forward Quick Paging Channel (F-QPCH)

The F-QPCH is a new type of paging channel used by a base station when it needs to con-
tact the mobile in the slotted mode. Its use reduces the time the mobile needs to be “awake”
resulting in increased battery life for the mobile.
Forward Link Physical Channels 369

The F-QPCH will contain a single bit message, the Quick Page message, to direct a slotted-
mode mobile to monitor its assigned slot on the paging channel that immediately follows. The
Quick Page message is sent up to 80 ms before the page message to alert the mobile to listen to
the paging channel. The F-QPCH uses a different modulation, so it will appear as a different
physical channel.

15.6.8 Forward Dedicated Auxiliary Pilot Channel (F-DAPICH)

An optional auxiliary pilot can be generated for a particular mobile. The F-DAPICH is
used with beam-forming applications and beam-steering techniques to increase the coverage or
data rate toward a particular mobile.

15.6.9 Forward Fundamental Channel (F-FCH)

This channel is transmitted at a variable rate as in IS-95B and consequently requires rate
detection at the receiver. Each F-FCH is transmitted on a different orthogonal code channel and
uses frame sizes corresponding to 20 ms and 5 ms. The 20-ms frame structure supports the data
rate corresponding to Rate Set 1 (RS1) and RS2, where the rates are 9.6, 4.8, 2.7, and 1.5 kbps
for RS1 and 14.4, 7.2, 3.6, and 1.8 kbps for RS2. The N = 1 RS1 F-FCH is shown in Fig. 15-12
and N = 1 RS2 F-FCH in Fig. 15-13. For N = 1 and RS1, a rate 1/2 convolutional encoder is
used. For N = 1 and RS2, a rate 1/3 convolutional code followed by puncturing every ninth bit
effectively provides a 3/8 code rate.

Add Add 8-Bit k=9

Full Rate Block
12-Bit Encoder r = 1/2 A
172 Bits CRC Tail 192 Encoder
Interleaver
384
Bits

Add Add 8-Bit k=9

1/2 Rate Repeat Block
8-Bit Encoder r = 1/2 A
80 Bits CRC Tail 96 Encoder 2 TImes
384
Interleaver
384
Bits Bits

Add Add 8-Bit k=9 Delete

1/4 Rate Repeat Block
6-Bit Encoder r = 1/2 Every 9th A
40 Bits CRC Tail 54 Encoder 4 TImes
432 Symbol 384
Interleaver
384
Bits Bits Bits

Add Add 8-Bit k=9 Delete

1/8 Rate Repeat Block
6-Bit Encoder r = 1/2 Every 5th A
16 Bits CRC Tail 30 Encoder 8 TImes
480 Symbol 384
Interleaver
384
Bits Bits Bits
20-ms Frame
5-ms Frame
Add Add 8-Bit k=9
16-Bit Encoder r = 1/2 Block
24 Bits A
CRC Tail Encoder Interleaver
40 48 96
Bits Bits

Note: Signal Point A feeds into Fig. 15.22.

Figure 15-12 F-FCH for N = 1 RS1

370 Ch. 15 • cdma2000 System

Add 1 Add Add 8-Bit k=9

Full Rate Block
Reserved 12-Bit Encoder r = 3/8* A
267 Bits Bit CRC Tail 288 Encoder
Interleaver
768
Bits

Add 1 Add Add 8-Bit k=9

1/2 Rate Repeat Block
Reserved 10-Bit Encoder r = 3/8 A
125 Bits Bit CRC Tail 144 Encoder 384
2 TImes Interleaver
768
Bits Bits

Add 1 Add Add 8-Bit k=9

1/4 Rate Repeat Block
Reserved 8-Bit Encoder r = 3/8 A
55 Bits Bit CRC Tail 72 Encoder 192
4 TImes Interleaver
768
Bits Bits

Add 1 Add Add 8-Bit k=9

1/8 Rate Repeat Block
Reserved 6-Bit Encoder r = 3/8 A
21 Bits Bit CRC Tail 36 Encoder 96
8 TImes Interleaver
768
Bits Bits
20-ms Frame
5-ms Frame
Add Add 8-Bit k=9
16-Bit Encoder r = 1/2 Block
24 Bits A
CRC Tail Encoder Interleaver
48 96
Bits

Note: Signal Point A feeds into Fig. 15.22. *Effective rate after puncturing

Figure 15-13 F-FCH for N = 1 RS2

For N ≥ 3, the RS1 and RS2 F-FCHs are shown in Figs. 15-14 and 15-15, respectively. For
N = 3, 6, 9, and 12 and RS1, a 1/3 code rate is used. For N = 3, 6, 9, and 12 and RS2, code rates
of 1/2 and 1/4 are supported.

15.6.10 Forward Supplemental Channel (F-SCH)

The F-SCH can be operated in two distinct modes. The first mode is used for a data rate
not exceeding 14.4 kbps and uses blind rate detection (no scheduling or rate information pro-
vided). In the second mode the rate information is explicitly provided to the base station. In the
first mode, the variable rates provided are those derived from IS-95B RS1 and RS2. The struc-
tures for the variable-rate modes are identical to the 20-ms F-FCH. In the second mode, the high
data rate modes can have k = 9 convolutional coding or turbo coding with k = 4 component
encoders. For the case of convolutional codes, there are 8 tail bits. For the case of turbo codes, 6
tail bits and 2 reserve bits are used.
There may be more than one F-SCH in use at a given time. The individual F-SCH target
FERs may be set independently with respect to the F-FCH and other F-SCHs, since optimal FER
for data is different than for voice. For classes of data services that have less stringent delay
requirements, the FER may also be managed by retransmissions.
The F-SCH supports 20-ms frames. For data rates derived from RS1, the F-SCH supports
data rates from 9.6 to 307.2 kbps (see Figs. 15-16 and 15-17).
Forward Link Physical Channels 371

Add Add 8-Bit k=9

Full Rate Block
12-Bit Encoder r = 1/3 X
172 Bits CRC Tail 192 Encoder
Interleaver
576
Bits

Add Add 8-Bit k=9

1/2 Rate Repeat Block
8-Bit Encoder r = 1/3 X
80 Bits CRC Tail 96 Encoder 288
2 TImes Interleaver
576
Bits Bits

Add Add 8-Bit k=9 Delete

1/4 Rate Repeat Block
6-Bit Encoder r = 1/3 Every 9th X
40 Bits CRC Tail 54 Encoder 162
4 TImes
648 Symbol
Interleaver
576
Bits Bits Bits

Add Add 8-Bit k=9 Delete

1/8 Rate Repeat Block
6-Bit Encoder r = 1/3 Every 5th X
16 Bits CRC Tail 30 Encoder 90
8 TImes
720 Symbol Interleaver
576
Bits Bits Bits
20-ms Frame
5-ms Frame
Add Add 8-Bit k=9
16-Bit Encoder r = 1/3 Block
24 Bits X
CRC Tail Encoder Interleaver
40 48 144
Bits Bits

Note: Signal Point X feeds into Fig. 15.23.

Figure 15-14 F-FCH for N ≥ 3 and RS1

Add 1 Add Add 8-Bit k=9

Full Rate Block
Reserved 12-Bit Encoder r = 1/4 or 1/2 X
267 Bits Bit CRC Tail 288 Encoder
Interleaver
1152
Bits or 576

Add 1 Add Add 8-Bit k=9

1/2 Rate Repeat Block
Reserved 10-Bit Encoder r = 1/4 or 1/2 X
125 Bits Bit CRC Tail 144 Encoder
2 TImes Interleaver
1152
Bits or 576

Add 1 Add Add 8-Bit k=9

1/4 Rate Repeat Block
Reserved 8-Bit Encoder r = 1/4 or 1/2 X
55 Bits Bit CRC Tail 72 Encoder
4 TImes Interleaver
1152
Bits or 576

Add 1 Add Add 8-Bit k=9

1/8 Rate Repeat Block
Reserved 6-Bit Encoder r = 1/4 or 1/2 X
21 Bits Bit CRC Tail 36 Encoder
8 TImes Interleaver
1152
Bits or 576
20-ms Frame
5-ms Frame
Add Add 8-Bit k=9
16-Bit Encoder r = 1/3 Block
24 Bits X
CRC Tail Encoder Interleaver
40 48 144
Bits Bits

Note: Signal Point X feeds into Fig. 15.23.

Figure 15-15 F-FCH for N ≥ 3 and RS2

372 Ch. 15 • cdma2000 System

Add Encoder Tail k=9

Block
RS1 16-Bit & Reserve r = 1/2 A
Interleaver
CRC Bits Encoder

21 octets 9.6 kbps 384 bits

45 octets 19.2 kbps 768 bits
93 octets 38.4 kbps 1536 bits
189 octets 76.8 kbps 3072 bits
381 octets 153.6 kbps 6144 bits
765 octets 307.2 kbps 12,288 bits

Add Encoder Tail k=9

Puncture Block
RS2 16-Bit & Reserve r = 1/2 A
1 of 9 Interleaver
CRC Bits Encoder

33 octets 14.4 kbps 768 bits

69 octets 28.8 kbps 1536 bits
141 octets 57.6 kbps 3072 bits
286 octets 115.2 kbps 6144 bits
573 octets 230.4 kbps 12,288 bits

Note: Signal Point A feeds into Fig. 15.22.

Figure 15-16 F-SCH for N = 1 and Channel Bandwidth of 1.25 MHz

Add Encoder Tail k=9

Block
RS1 16-Bit & Reserve r = 1/3 X
CRC Bits Interleaver
Encoder

21 octets 9.6 kbps 576 bits

45 octets 19.2 kbps 1152 bits
93 octets 38.4 kbps 2304 bits
189 octets 76.8 kbps 4608 bits
381 octets 153.6 kbps 9216 bits
765 octets 307.2 kbps 18,432 bits
1533 octets 614.4 kbps 36,864 bits

Add Encoder Tail k=9

Block
RS2 16-Bit & Reserve r = 1/4 X
CRC Bits Interleaver
Encoder

33 octets 14.4 kbps 1152 bits

69 octets 28.8 kbps 2304 bits
141 octets 57.6 kbps 4608 bits
285 octets 115.2 kbps 9216 bits
573 octets 230.4 kbps 18,432 bits
1149 octets 460.8 kbps 36,864 bits
2301 octets 921.6 kbps (R = 1/2) 36,864 bits

Note: Signal Point X feeds into Fig. 15.23.

Figure 15-17 F-SCH for N ≥ 3 and Channel Bandwidth of N × 1.25 MHz, N ≥ 3

15.6.11 Forward Dedicated Control Channel (F-DCCH)

The F-DCCH supports 5-ms and 20-ms frames at the 9.6-kbps encoder input rate. Sixteen
CRC bits are added to the information bits for 5-ms frames or 12 CRC bits are added for 20-ms
frames, followed by the addition of 8 tail bits, convolutional encoding, interleaving, and scram-
bling (see Figs. 15-18 and 15.19).
Forward Link Features 373

Add Add 8-Bit k=9

5-ms Block
16-Bit Encoder r = 1/2 A
Frame 24 CRC 40 Tail 48 Encoder 96
Interleaver
Bits Bits Bits Bits

Add Add 8-Bit k=9

20-ms Block
12-Bit Encoder r = 1/2 A
Frame 172 CRC 184 Tail 192 Encoder 384
Interleaver
Bits Bits Bits Bits

Note: Signal Point A feeds into Fig. 15.22.

Figure 15-18 Forward Dedicated Control Channel (F-DCCH) for N = 1

Add Add 8-Bit k=9

5-ms Block
16-Bit Encoder r = 1/3 X
Frame 24 CRC 40 Tail 48 Encoder 144
Interleaver
Bits Bits Bits Bits

Add Add 8-Bit k=9

20-ms Block
12-Bit Encoder r = 1/3 X
Frame 172 CRC 184 Tail 192 Encoder 576
Interleaver
Bits Bits Bits Bits

Note: Signal Point X feeds into Fig. 15.23.

Figure 15-19 Forward Dedicated Control Channel (F-DCCH) for N ≥ 3

15.7 Forward Link Features

The forward link supports chip rates of N × 1.2288 Mcps (where N = 1, 3, 6, 9, 12). For N = 1,
the spreading is similar to IS-95B (see chapter 7); however QPSK modulation and fast closed-
loop power control are used. There are two options for chip rates for N > 1: multicarrier and
direct spread (see Fig. 15-20). The multicarrier approach demultiplexes modulation symbols
onto N separate 1.25-MHz carriers (N = 3, 6, 9, 12). Each carrier is spread with a 1.2288-Mcps
rate. The N > 1 direct-spread approach transmits modulation symbols on a single carrier which is
spread with a chip rate of N × 1.2288 Mcps (N = 3, 6, 9, 12).

1.25 1.25 1.25 1.25 MHz

Multicarrier f1 f2 f3 1.2288 Mcps

N=3
0 1 2 3 4 5 MHz

3.6864 Mcps
Direct Spread
N=3
0 1 2 3 4 5 MHz

Figure 15-20 Multicarrier and Direct Spread Approach on Forward Link in cdma2000
374 Ch. 15 • cdma2000 System

15.7.1 Transmit Diversity

Transmit diversity can reduce the required Eb /It or (required transmit power per channel) and
thus enhance system capacity. Transmit diversity can be implemented in the following ways:
• Multicarrier Transmit Diversity. Antenna diversity can be implemented in a multi-
carrier forward link with no impact on the subscriber terminal, where a subset of carri-
ers is transmitted on each antenna. The main characteristics of the multicarrier
approach are
◆ Coded information symbols are demultiplexed among multiple 1.25-MHz carriers.
◆ Frequency diversity is equivalent to spreading the signal over the entire band-
width.
◆ Both time and frequency diversity are captured by convolutional coder/symbol
repetition and interleaver.
◆ RAKE receiver captures signal energy from all bands.
◆ Each forward link channel may be allocated an identical Walsh code on all carriers.
◆ Fast power control.
In 3 × 1.25-MHz multicarrier transmitter, the serial coded information symbols are
divided into three parallel data streams, and each data stream is spread with a Walsh code and a
long PN sequence at a rate of 1.2288 Mcps. At the output of the transmitter, there are three carri-
ers—A, B, and C (see Fig. 15-21).
After processing the serial coded information symbols with parallel carriers, the multicar-
rier will be transmitted by multiantenna, which is called Multicarrier Transmit Diversity (MCTD).
In the MCTD, the total carriers are divided into subsets; then each subset of the carriers is trans-
mitted on each antenna, where frequency filtering provides near-perfect orthogonality between
antennas. This provides improved frequency diversity and hence increases forward link capacity.
• Direct-Spread Transmit Diversity. Orthogonal Transmit Diversity (OTD) may be
used to provide transmit diversity for direct spread. Coded bits are split into two data
streams and are transmitted via separate antennas. A different orthogonal code is used
per antenna for spreading. This maintains the orthogonality between the two output

Two Antennas Three Antennas

A A
f1 f1
f2 B B
f3 f2
C
f3

Figure 15-21 3 × 1.25-MHz Multicarrier Transmitter

Forward Link Features 375

streams, and hence self-interference is eliminated in flat fading. Note that, by splitting
the coded bits into two separate streams, the effective number of spreading codes per
user is the same as the case without OTD. An auxiliary pilot is introduced for the addi-
tional antenna.

15.7.2 Orthogonal Modulation

To reduce or eliminate intracell interference, each forward link physical channel is modu-
lated by a Walsh code. To increase the number of usable Walsh codes, QPSK modulation is
used before spreading. Every 2 information bits are mapped to a QPSK symbol. As a result, the
available number of Walsh codes is increased by a factor of 2 relative to BPSK (prespreading)
symbols. Walsh code length varies to achieve different information bit rates. The forward link
may be interference limited or Walsh code limited depending on the specific deployment and
operating environment. When a Walsh code limit occurs, additional codes may be generated by
multiplying Walsh codes by the masking functions. The codes generated in this way are called
quasiorthogonal functions.

15.7.3 Power Control

A new Fast-Forward Power Control (FFPC) algorithm for the forward link and a power
control for the F-FCH and F-SCH is used in cdma2000. The standards specify a fast closed-loop
power control at 800 Hz. Two schemes of power control have been proposed for the F-FCH and
F-SCH.
• Single-Channel Power Control. This is based on the performance of the higher-rate
channel between the F-FCH and F-SCH. The gain setting for the lower-rate channel is
determined based on its relationship to the higher-rate channel.
• Independent Power Control. In this case, gains for the F-FCH and F-SCH are deter-
mined separately. The mobile runs two separate outer loop algorithms (with different
Eb /Nt targets) and sends two forward Eb /It error bits to the base station.

15.7.4 Walsh Code Administration

IS-95A/B uses fixed-length 64-chip Walsh codes. The new rate sets in cdma2000 require
variable length Walsh codes for traffic channels. The Walsh codes used are from 128 chips to 4
chips in length. The F-FCH Walsh code is fixed (128 chips for RS3 and RS5, and 64 chips for
RS4 and RS6), whereas the length of the Walsh codes for F-SCH decreases as the information
rate increases to maintain a constant bandwidth of the modulated signal. In addition to different
Walsh code lengths, the coordination of the allocation of Walsh codes across the 2G and the 3G
systems is necessary for the overlay systems.
The algorithm must ensure that Walsh codes assigned for different-rate supplemental
channels are always orthogonal to each other as well as to the fundamental traffic channels, pag-
ing channels, sync channel, and pilot channel. For example, if an all-0s 4-chip Walsh code (0 0 0
0) is assigned, then there are two 8-chip Walsh codes that are not to be assigned at the same time
376 Ch. 15 • cdma2000 System

(0 0 0 0 0 0 0 0, 0 0 0 0 1 1 1 1); the remaining six 8-chip codes can be used since they are all
orthogonal to it. By induction, four 16-chip, eight 32-chip, sixteen 64-chip, and thirty-two 128-
chip codes must also be set aside to maintain orthogonality.
The 3G and 2G Walsh code assignments must be coordinated to insure that assigning the
longer-length codes does not block out all of the shorter codes.

15.7.5 Modulation and Spreading

The N = 1 system can be deployed in a new spectrum or as a backward-compatible
upgrade anywhere an IS-95B forward link is deployed in the same RF channel. The new
cdma2000 channels can coexist in an orthogonal manner with the code channels of existing IS-
95B system. The N = 1 spreading is shown in Fig. 15-22. First, the user data is scrambled by the
user long PN code followed by I and Q mapping, channel gain, power control puncturing, and
Walsh spreading. The power control bits may or may not be punctured onto the forward link
depending on the specific logical-to-physical channel mapping. Next, as shown in Fig. 15-22,
the signal is complex PN spread, followed by baseband filtering and frequency modulation.
The multicarrier system can be deployed in new spectrum or as a backward-compatible
upgrade anywhere an IS-95B forward link is deployed in the same N RF channels. The new
cdma2000 channels can coexist in an orthogonal manner with the code channels of existing
IS-95B system.

I Data Puncture
A Channel PC Sym. YI
Gain (optional)

Long-Code
MUX and PC Bits
Mask for User m Bit PC
Long-Code Signal Point 16 bits/frame
Generator Selector Mapping Channel Walsh
±1 values
0 to +1 Gain
1 to –1

Q Data Puncture
I and Q Mapping, and Channel
Walsh Modulation PC Sym. YQ
Gain (optional)

MUX Control 1.2288 Mbps

PNI

+ Baseband
YI
Filter
PNQ – cos (2πfct ) s(t )
PN Q +
Baseband
YQ Filter
+
PN I sin (2πfct )
PN I = I-Channel PN Sequence 1.2288 Mcps
PNQ = Q-Channel PN Sequence 1.2288 Mcps
PC = Power Control

Figure 15-22 I and Q Mapping, Walsh Modulation, PN Spreading, and Frequency

Modulation for N = 1
Forward Link Features 377

The overall structure of the multicarrier CDMA channel is shown in Fig. 15-23. After
scrambling with the long PN code corresponding to user m, the user data is demultiplexed into N
carriers, where N = 3, 6, 9, or 12. On each carrier, the demultiplexed bits are mapped onto I and
Q followed by Walsh spreading. When applicable, power control bits, for reverse closed-loop
power control, may be punctured onto the forward link channel at a rate of 800 Hz. The signal
on each carrier is orthogonally spread by the appropriate Walsh code function in such a manner
as to maintain a fixed chip rate of 1.2288 Mcps per carrier, where the Walsh code may differ on
each carrier. The signal on each carrier is then complex PN spread, as shown in Fig. 15-24, fol-
lowed by baseband filtering and frequency modulation.
N = 1, 3, 6, 9, and 12 direct spreading is shown in Fig. 15-25. The user data is first scram-
bled by the user long PN code followed by I and Q mapping, channel gain, power control

cos(2πfAt )
IA IA(t )
Long-Code Baseband
Mask for Filter
MUX and Complex Carrier A
User m WnA
IQ MAP
0 → +1
PN +
Spreading
1 → –1 QA Baseband QA(t )
Filter
Long-Code
Generator sin(2πfAt )
Z

cos(2πfBt )
IB IB (t )
Bit Baseband
Selector Filter
MUX and Complex Carrier B
WnB
IQ MAP
Z 0 → +1
PN +
Demultiplexer

Spreading
1 → –1 QB Baseband QB (t )
Filter

sin(2πfBt )
X + Z

. . .
. . .
. . .
cos(2πfLt )
IL IL(t )
Baseband
Filter
MUX and Complex Carrier L
WnL
IQ MAP
0 → +1
PN +
Spreading
1 → –1 QL Baseband QL(t )
Filter

sin(2πfLt )
Z

PC Bits • N Carriers at 1.2288 Mcps per Carrier

16 bits/20 ms • QPSK Data
±1 Values • Complex PN Spreading
(optional)

Figure 15-23 Multicarrier CDMA Forward Link Structure

378 Ch. 15 • cdma2000 System

PN I
XQ +

PNQ PN Q

+
XI

+
PN I

Figure 15-24 Complex PN Spreading

MUX Control
Z
cos(2πfCt )
I Data Puncture
Baseband
Channel 1 PC Sym.
Filter
Gain in 1.25 ms

SIgnal Sum with

Point PC Bits PC Complex
Other PN
A Mapping 16 bits/frame Channel Walsh n Code Spreading
0 → +1 ±1 Values Gain
Channels
1 → –1

Data Puncture Baseband

Channel 1 PC Sym. Filter
Gain in 1.25 ms
Q
N × 1.2288 Mcps sin(2πfCt )
Z
MUX Control

• One N × 1.2288-Mcps Carrier

• QPSK Data
• Complex PN Spreading

Figure 15-25 N = 1, 3, 6, 9, and 12, I and Q Mapping, and Walsh Modulation

puncturing, and Walsh spreading. The power control bits may not be punctured onto the for-
ward link channel depending on the specific logical-to-physical channel mapping. Next, the
signal is complex PN spread, followed by baseband filtering and frequency modulation.
Fig. 15-26 provides a comparison between the forward physical channels used in IS-95A/B
and those used in cdma2000.

15.7.6 Key Characteristics of Forward Link

The key characteristics of the forward link are
Reverse Physical Channels 379

Forward Link Channels New Channels

Common to IS-95
& cdma2000

Traffic Traffic
Pilot Sync Paging
RS1 & RS2 RS3 to RS6

F-FCH

F-PCH / F-FCH F-SCH

F-QPCH F-CCCH F-BCCH

Figure 15-26 A Comparison between Forward Physical Channels for IS-95 and cdma2000

• Channels are orthogonal and use Walsh codes. Different-length Walsh codes are used to
achieve the same chip rate for different information bit rates.
• QPSK modulation is used before spreading to increase the number of usable Walsh
codes.
• Forward Error Correction (FEC) is used.
◆ Convolutional codes (k = 9) are used for voice and data.
◆ Turbo codes (k = 4) are used for high data rates on SCHs.
• Supports nonorthogonal forward link channelization.
◆ These are used when running out of orthogonal space (insufficient number of
Walsh codes).
◆ Quasiorthogonal functions are generated by masking existing Walsh functions.
• Synchronous forward link.
• Forward link transmit diversity.
• Fast-forward power control (closed loop) 800 times per second.
• Supplemental channel active set, subset of fundamental channel active set. The maxi-
mum data rate supported for RS3 and RS5 for supplemental channel is 153.6 kbps (raw
data rate). RS4 and RS6 will be supported only for voice calls with the fundamental
channel rates of up to 14.4 kbps (raw data rate).
• Frame lengths:
◆ 20-ms frames are used for signaling and user information
◆ 5-ms frames are used for control information.

15.8 Reverse Physical Channels

Reverse physical channels (see Fig. 15-7) include dedicated channels to carry information from
a single mobile to the base station and common channels to carry information from multiple
mobiles to the base station. Table 15-5 lists the reverse physical channels.
380 Ch. 15 • cdma2000 System

Table 15-5 Reverse Physical Channels

Physical Channels Channel Name
Reverse Common Reverse Access Channel R-ACH
Physical Channel
Reverse Common Control R-CCCH
Channel (9.6 kbps only)
Reverse Dedicated Reverse Pilot Channel R-PICH
Physical Channel
Reverse Dedicated Control R-DCCH
Channel
Reverse Traffic Channel:
—Fundamental R-FCH
—Supplemental R-SCH

15.8.1 The Reverse Access Channel (R-ACH) and the Reverse Common
Control Channel (R-CCCH)
These are common channels used for communication of layer 3 and MAC messages from
the mobile to the base station. The R-CCCH differs from the R-ACH in that the R-CCCH offers
extended capabilities beyond the R-ACH. For example, the R-CCCH supports lower latency
access procedures required for efficient operation of the packet data in suspended state.
The R-ACH and R-CCCH are multiple access channels as mobile stations transmit with-
out explicit authorization by the base station. The R-ACH and R-CCCH use a slotted ALOHA
type of mechanism with higher capture probabilities due to the CDMA properties of the channel
(simultaneous transmission of multiple users). There can be one or more access channels per fre-
quency assignment. Different access channels are distinguished by different long PN codes. The
R-CCCH is identical to the R-ACH for 9.6-kbps, 20-ms frames. Additional rates of 19.2 and
38.4 kbps and frames of 5 ms and 10 ms will be supported. Fig. 15-27 shows the structure of R-
ACH and R-CCCH.

15.8.2 Reverse Pilot Channel (R-PICH)

The pilot channel for the reverse dedicated channels consists of a fixed reference value and
multiplexed forward power control information. The time-multiplexed forward power control
information is referred to as the power control subchannel. This subchannel provides informa-
tion on the quality of the forward link at the rate of 1 bit per 1.25-ms Power Control Group
(PCG) and is used by the forward link channels to adjust their power. The power control symbol
repetition means that the 1-bit value is constant for that repeated symbol’s duration. The power
control bit uses the last portion of each PCG. The +1 pilot symbols and multiplexed power-con-
trol symbols are all sent with the same power level. The binary power control symbols are repre-
sented with ±1 values.
The R-PICH is used for initial acquisition, time tracking, RAKE-receiver coherent refer-
ence recovery, and power control measurements (see Fig. 15-28 for R-PICH structure).
Reverse Physical Channels 381

Modulation
Enhanced Add 12-Bit Add 8 Block Symbols
Convolutional Symbol
Access Frame Encoder Interleaver
Encoder Repetition C
Channel Quality Tail (768
r = 1/4, k = 9 (4 × Factor)
Bits Indicator Bits Symbols)

32 Bits per 9.6 ksps 153.6 ksps

5-ms Frame
• Channel Structure for the Header on the Enhanced Access Channel for Spreading Rate 1

Modulation
Reverse Add Add 8 Symbols
Convolutional
Common Frame Encoder Symbol Block
Encoder Repetition Interleaver
C
Control Quality Tail
r = 1/4, k = 9
Channel Bits Indicator Bits

Bits/Frame Bits Rate (kbps) Factor Symbols Rate (ksps)

172 (5 ms) 12 38.4 1× 768 153.6
360 (10 ms) 16 38.4 1× 1536 153.6
172 (10 ms) 12 19.2 2× 1536 153.6
744 (20 ms) 16 38.4 1× 3072 153.6
360 (20 ms) 16 19.2 2× 3072 153.6
172 (20 ms) 12 9.6 4× 3072 153.6

• Channel Structure for the Reverse Common Control Channel for Spreading Rate 1

Note: Signal Point C feeds into Fig. 15.30

Figure 15-27 Reverse Access and Common Control Channel Structure

Pilot
(+1 Value)
MUX A
1 Power Control Bit per Symbol
PCG; 16 Power Control Repetition Pilot Channel +
Bits per Frame 384 N Copies Power Control Bits

N = 1 for 1.2288 Mcps

N = 3 for 3.6864 Mcps
N = 6 for 7.3728 Mcps
N = 9 for 11.0592 Mcps
N = 12 for 14.7456 Mcps

Pilot Power Control

384 × N PN Chips
1 Power Control Group (PCG)
4 Groups of 384N PN Chips
Note: Signal Point A feeds into Fig. 15.30

Figure 15-28 Reverse Pilot Channel (R-PICH) Structure for Reverse Dedicated Channels

15.8.3 Reverse Dedicated Control Channel (R-DCCH)

The R-DCCH, R-FCH, and R-SCH may or may not be used depending on the service sce-
nario. Each physical channel is spread with a Walsh code sequence to provide orthogonal chan-
nelization among these physical channels. The spread pilot and R-DCCH are mapped to the in-
phase (I) data channel. The spread R-FCH and R-SCH are mapped to the quadrature (Q) data
382 Ch. 15 • cdma2000 System

channel. Then, the I and Q data channels are spread using a complex-multiply-type PN spread-
ing approach (see Fig. 15-29 for reverse dedicated channel structure and Fig. 15-30 for reverse
link I and Q mapping for 1× and 3×).

15.8.4 Reverse Fundamental Channel (R-FCH)

The R-FCH supports 5- and 20-ms frames. The 20-ms frame structures provide rates
derived from the IS-95B RS1 or RS2. The 5-ms frames provide 24 information bits per frame
with 16-bit CRC. Within each 20-ms frame interval, either one 20-ms R-FCH structure, up to
four 5-ms R-FCH structure(s), or nothing can be transmitted. In addition, when a 5-ms R-FCH
structure is used, it can be on or off in each of the four 5-ms segments of a 20-ms frame interval.
The R-FCH is transmitted at different rates. The rates supported for the R-FCH are 1.5, 2.7, 4.8,
and 9.6 kbps for RS3 and RS5, and 1.8, 3.6, 7.2, and 14.4 kbps for RS4 and RS6. (see Fig. 15-31
for R-FCH/R-SCH for 1×, Radio Configuration (RC) 3 and 3×, RC 5)

15.8.5 Reverse Supplementary Channel (R-SCH)

The R-SCH can be operated in two distinct modes. The first mode is used for data rates not
exceeding 14.4 kbps and uses blind rate detection (no scheduling or rate information). In the sec-
ond mode, the rate information is explicitly known by the base station. The R-SCH is used for
data calls and can operate at different prenegotiated rates. Only RS3 and RS5 are supported for
the R-SCH. Since only RS3 is supported on the R-FCH for high-speed packet data calls, the
rates supported for the R-SCH are 9.6, 19.2, 38.4, 76.8, and 153.6 kbps (see Fig. 15-32 for R-
FCH/R-SCH for 1×, RC 4 and 3×, RC 6).

Modulation
Reverse Add Add 8 Symbols
Convolutional Symbol
Dedicated Frame Encoder Block
Encoder Repetition
Interleaver
B
Control Quality Tail
r = 1/4, k = 9 (2 × Factor)
Channel Bits Indicator Bits

Bits/Frame Bits Rate (kbps) Symbols Rate (ksps)

24 (5 ms) 16 9.6 384 76.8
172 (20 ms) 12 9.6 1536 76.8

• Reverse Dedicated Control Channel Structure for Radio Configuration (RC) 3 and 5

Modulation
Reverse Add Add 8 Symbols
Add Convolutional Symbol
Dedicated Frame Encoder Symbol Block
Reserved Encoder Repetition
Puncture Interleaver
B
Control Quality Tail
Bits r = 1/4, k = 9 (2 × Factor)
Channel Bits Indicator Bits

Bits/Frame Bits Bits Rate (kbps) Deletion Symbols Rate (ksps)

24 (5 ms) 0 16 9.6 None 384 76.8
172 (20 ms) 1 12 14.4 8 of 24 1536 76.8

• Reverse Dedicated Control Channel Structure for Radio Configuration (RC) 4 and 6
Note: Signal Point B feeds into Fig. 15.30

Figure 15-29 Reverse Dedicated Channel Structure

Reverse Physical Channels 383

Walsh Cover
(+ + – – ) or (+ + – – – – + +) Complex Multiplier

Relative
Supplemental
C Gain
Channel 2 G S2
Pilot + +
Channel & + Baseband
Power Control
A Σ Σ Filter
Gp
Symbols + –
Dedicated Relative
Control B Gain cos(2πfct )
Channel GC

Walsh Cover
(+ + + + + + + + – – – – – – – – ) Σ Gain s(t )

Relative
Fundamental
C Gain
Channel
GF
+ +
Walsh Cover Baseband
(+ + + + – – – – + + + + – – – – ) Σ Σ Filter
Gp

Supplemental + +
Channel 1, Relative
Common Control C Gain sin(2πfct )
Channel, or Enhanced GS1
Access Channel
Walsh Cover
Walsh Cover
(+ – ), (+ + – – ), or (+ + – – + + – – )
(+ – )
for Supplemental Channel
(+ + + + – – – – )
for Common Control Channel and
Enhanced Access Channel Decimator
by Factor
of 2

I-Channel Q-Channel
PN Sequence PN Sequence

Notes :
1-Chip 1. Binary signals are represented with ±1 values with the
Delay mapping +1 for 0 and –1 for 1.
Unused channels and gated-off symbols are represented
with 0 values.
Long- 2. When the Common Control Channel or Enhanced
Long-Code Access Channel is used, the only additional channel is
Code
Mask the Pilot Channel.
Generator Reverse Link 3. All of the digital operations occur at the chip rate.

Figure 15-30 Reverse Link I and Q Mapping for 1× and 3×

Figure 15-33 provides a comparison between the reverse physical channels used in IS-
95A/B and cdma2000.

15.8.6 FEC on Reverse Link

The reverse link uses a k = 9, r = 1/4 convolutional code for R-FCH. The distance proper-
ties of this code are better catered, providing performance gains vs. higher rate codes in fading
and Additive White Gaussian Noise (AWGN) channel conditions. The constraint length k = 9, r
= 1/4 convolutional code provides a gain of about 0.5 dB over a k = 9, r = 1/2 code in AWGN.
The R-SCH uses convolutional codes for data rates of up to 14.4 kbps. Convolutional codes for
higher data rates on the R-SCH are optional, and the use of turbo code is preferred. A common
constituent code is used for the reverse link. Turbo codes of constraint length 4, rate 1/4, 1/3, and
1/2, are used for all R-SCH.
384 Ch. 15 • cdma2000 System

Modulation
Add Add 8 Symbols
Convolutional Symbol Symbol Block
Channel Frame Reserved/
or Turbo
Repetition Puncture Interleaver
C
Bits Quality Encoder
Encoder
Indicator Tail Bits

Bits/Frame Bits Rate (kbps) r Factor Deletion Symbols Rate (ksps)

24 (5 ms) 16 9.6 1/4 2× None 384 76.8

16 6 1.5 1/4 16 × 1 of 5 1536 76.8

40 6 2.7 1/4 8× 1 of 9 1536 76.8
80 8 4.8 1/4 4× None 1536 76.8
172 12 9.6 1/4 2× None 1536N 76.8

360 16 19.2 1/4 1× None 1536N 76.8

744 16 38.4 1/4 1× None 3072N 153.6
1512 16 76.8 1/4 1× None 6144N 307.2
3048 16 153.6 1/4 1× None 12,288N 614.4
6120 16 307.2 1/2 1× None 12,288N 614.4
Notes:
1. The 5-ms frame is used only for the fundamental channels, and only rates of 9.6 kbps or less are used for the
fundamental channels.
2. Turbo coding may be used for the supplemental channels with rates of 19.2 kbps or more; otherwise k = 9
convolutional coding is used.
3. With convolutional coding, the reserved/encoder tail bits provide an encoder tail. With turbo coding, the first 2
of these bits are reserved bits that are encoded and the last 6 bits are replaced by an internally generated tail.
4. N is the number of consecutive 20-ms frames over which the interleaving is done (N = 1, 2 or 4).

Note: Signal Point C feeds into Fig. 15.30

Figure 15-31 R-FCH/R-SCH for 1×, RC3 and 3×, RC5

Modulation
Add 8 Add 8 Symbols
Add Convolutional
Channel Reserved Encoder Symbol Symbol Block
Reserved or Turbo
Repetition Puncture Interleaver
C
Bits /Encoder Tail
Bits Encoder
Tail Bits Bits

Bits/Frame Bits Bits Rate (kbps) r Factor Deletion Symbols Rate (ksps)
24 (5 ms) 0 16 9.6 1/4 2× None 384 76.8

21 1 6 1.8 1/4 16 × 8 of 24 1536 76.8

55 1 8 3.6 1/4 8× 8 of 24 1536 76.8
125 1 10 7.2 1/4 4× 8 of 24 1536 76.8
267 1 12 14.4 1/4 2× 8 of 24 1536N 76.8

552 0 16 28.8 1/4 2× 8 of 24 3072N 153.6

1128 0 16 57.6 1/4 2× 8 of 24 6144N 307.2
2280 0 16 115.2 1/4 2× 8 of 24 12,288N 614.4
4584 0 16 230.4 1/4 1× 8 of 24 12,288N 614.4

Notes:
1. The 5-ms frame is used only for the fundamental channels, and only rates of 14.4 kbps or less are used for the
fundamental channels.
2. Turbo coding may be used for the supplemental channels with rates of 28.8 kbps or more; otherwise k = 9
convolutional coding is used.
3. With convolutional coding, the reserved/encoder tail bits provide an encoder tail. With turbo coding, the first 2
of these bits are reserved bits that are encoded and the last 6 bits are replaced by an internally generated tail.
4. N is the number of consecutive 20-ms frames over which the interleaving is done (N = 1, 2 or 4).

Note: Signal Point C feeds into Fig. 15.30

Figure 15-32 R-FCH/R-SCH for 1×, RC4 and 3×, RC6

Reverse Physical Channels 385

New Channels

Common to IS-95
Reverse Physical Channels & cdma2000

R-ACH or Traffic Traffic

Pilot R-CCCH RS1 & RS2 RS3 to RS6

R-FCH

R-FCH R-SCH

Figure 15-33 A Comparison of Reverse Physical Channels for IS-95 and cdma2000

15.8.7 Reverse Link Physical Layer Characteristics

• Continuous waveform. A continuous pilot and continuous data-channel waveform are
used for all data rates. This continuous waveform minimizes interference to biomedical
devices such as hearing aids and pacemakers and permits a range increase at lower
transmission rates. The continuous waveform also enables the interleaving to be per-
formed over the entire frame rather than just the portions that are not gated off. This
enables the interleaving to achieve the full benefit of the frame time diversity. The base
station uses the pilot for multipath searches, tracking, and coherent demodulation, as
well as to measure the quality of the link for power control purposes. Separate orthogo-
nal channels for the pilot and each of the data channels are used. Thus, the relative lev-
els of the pilot and physical data channels can easily be adjusted without changing the
frame structure or power levels of some symbols of a frame.
• Orthogonal channels with different-length Walsh sequences. The cdma2000 system
uses orthogonal channels for the pilot and the other physical data channels. These
orthogonal channels are provided with different-length Walsh sequences, with the
higher-rate channels with shorter Walsh sequences. A short Walsh sequence allows high
encoder output rates to be accommodated.
• Rate matching. Several approaches are needed to match the data rates to Walsh
spreader input rates. These include adjusting the code rate using puncturing, symbol
repetition, and sequence repetition. The design approach is to first try to use a low rate
code, but not to reduce the rate below r = 1/4 since gains of smaller rates would be
small and the decoder implementation complexity would increase significantly.
• Low spectral side lobes. The cdma2000 system achieves low spectral side lobes with
nonideal mobile power amplifiers by splitting the physical channels into the I and Q
channels and using a complex-multiply-type PN spreading approach.
386 Ch. 15 • cdma2000 System

• Independent data channels. Two types of physical data channels (R-FCH and R-
SCH) are used on the reverse link that can be adapted to a particular type of service.
The use of R-FCH and R-SCH enables the system to be optimized for multiple simulta-
neous services. These channels are separately coded and interleaved and may have dif-
ferent transmit power level and FER set points.
• Reverse power control. There are three components of reverse power control: open
loop, closed loop, and outer loop. Open-loop power control sets the transmit power
based upon the power that is received at the mobile. Open-loop power control compen-
sates for the path loss from the mobile to the base station and handles very slow fading.
Closed-loop power control consists of an 800-bps feedback loop from the base station
to the mobile to set the transmit power of the mobile. Closed-loop power control com-
pensates for medium to fast fading and inaccuracies in open-loop power control. Outer-
loop power control is implementation specific but typically adjusts the closed-loop
power control threshold in the base station to maintain a desired FER.
• Separate dedicated control channel. The reverse link consists of a separate low-rate,
low-power, continuous, orthogonal, dedicated control channel. This allows for a flexi-
ble dedicated control channel structure that does not impact the other pilot and physical
channel frame structures.
• Frame length. The cdma2000 system uses 5- and 20-ms frames for control informa-
tion on fundamental and dedicated control channels and uses 20-ms frames for other
types of data (including voice). Interleaving and sequence repetition are over the entire
frame interval. This provides improved time diversity over systems that use shorter
frames. The 20-ms frames are used for voice. A shorter frame would reduce the total
voice delay, but it would degrade the demodulation performance due to the shorter
interleaving span.

15.8.8 Reverse Link Modulation and Coding

The reverse link uses direct sequence spreading with an IS-95B chip rate of 1.2288 Mcps
(denoted as 1× chip rate) or chip rates that are 3, 6, 9, or 12 times IS-95B chip rate. Higher-chip-
rate systems are denoted as 3×, 6×, 9×, and 12× and they are respectively operated at 3.6864,
7.3728, 11.0592, and 14.7456 Mcps.
The 1× system can be used anywhere that an IS-95B reverse link is used. An IS-95B
reverse link carrier frequency can also be shared with mobiles transmitting the IS-95B waveform
and those transmitting the 1× cdma2000 waveform. The higher-chip-rate reverse links can be
used in applications where larger bandwidth allocations are available. Mobiles that support a
higher chip rate would typically also support the 1× chip rate. This will allow these mobiles to
access base stations that support only the 1× and higher chip rate systems.
Within an operator’s allocated band, the 1× cdma2000 reverse links would typically
occupy the same bandwidth as the IS-95B reverse link system (i.e., 1.25 MHz), and higher-chip-
rate cdma2000 links would typically occupy a bandwidth that is 1.25 MHz times the higher-
Data Services in cdma2000 387

chip-rate factor. A guard band of 625 kHz would typically be used on both sides of the operator’s
allocated band.

15.8.9 Key Characteristics of Reverse Link

The key characteristics of the reverse link are
• Channels are primarily code multiplexed.
• Separate channels are used for different QoS and physical layer characteristics.
• Transmission is continuous to avoid electromagnetic interference (EMI).
• Channels are orthogonalized by Walsh functions and I/Q split so that performance is
equivalent to BPSK.
• Hybrid combination of QPSK and BPSK.
• Coherent reverse link with continuous pilot.
• Forward power control information is time multiplexed with the pilot.
• By restricting alternate phase changes of the complex scrambling, power peaking is
reduced and side lobes are narrowed.
• Independent fundamental and supplemental channels with different transmit power and
FER target.
• Forward error correction:
◆ Convolutional codes (k = 9) are used for voice and data.
◆ Parallel turbo codes (k = 4) are used for high data rates on supplemental channels.
• Fast-reverse power control: 800 times per second.
• Frame lengths:
◆ 20-ms frames are used for signaling and user information.
◆ 5-ms frames are used for control information.

15.9 Data Services in cdma2000

Two types of data services are being considered for cdma2000—packet and high-speed circuit
data services. The packet service and the MAC layer are designed to support a large number of
mobile stations using packet data services. Many packet data services exhibit highly bursty traf-
fic patterns with relatively long periods of inactivity. Due to limited air-interface capacity, lim-
ited base station equipment, and constraints on mobile station power consumption, dedicated
channels for packet service users are allocated on demand and released immediately after the
end of the activity period.
Releasing the dedicated channels and reestablishing them introduces latency and signaling
overhead due to the renegotiation process that has to take place between the BS and MS before
user data exchange. The overhead of reestablishing the dedicated channels includes the cost of
RLP synchronization and the signaling overhead associated with service negotiation to recon-
nect the packet service. The MAC avoids this latency and overhead by allowing the BS to save a
set of state information after the initialization phase is completed.
388 Ch. 15 • cdma2000 System

Signaling, Data, PC, MAC

Voice Packet
Signaling, Voice, PC
Option BS MS Option BS Data Bursts MS
1 1

Voice, PC Signaling, Data, PC, MAC

Voice Packet
Option BS Signaling MS Option BS Data Bursts MS
2 2

Signaling, Data, PC
Packet
MAC, (PC)
Option BS MS
Voice/ Signaling, Voice, Data, PC, MAC 3 Data Bursts
Packet
BS Data Bursts MS
Option
1
Signaling, Voice, PC
Voice/ Fundamental Channel
Packet MAC, Data (Signaling), (PC)
BS MS Dedicated Control Channel
Option Supplemental Channel
Data Bursts
2 PC = Power Control

Figure 15-34 Logical-to-Physical Channels Mapping for Voice and Packet Services

To further reduce the overhead associated with assignment of dedicated channels, the
packet service allows for exchange of short bursts of user data when no dedicated channels are
present. This mode of operation may be suitable for mobile-IP registration, notification services
(e.g., e-mail notification), and location tracking services where the volume of data to be
exchanged is typically small.
Circuit services can be viewed as a special case of the packet services in the sense that
dedicated traffic and control channels are typically assigned to the MS for extended periods of
time during the circuit service sessions. This will lead to a less efficient use of the air-interface
capacity. However, some delay-sensitive services such as video applications require a dedicated
channel for the duration of the call.

15.10 Mapping of Logical Channels to Physical Channels

The mapping from the logical channel to the physical channel is not one to one—multiple logi-
cal channels from multiple service options may be mapped to a single physical channel. This
multiplexing operation is performed at the PLDCF MUX and QoS sublayer. Fig. 15-34 shows
the channel structure for bearer service profiles that include only packet and voice services.

15.10.1 Forward Link

15.10.1.1 Forward Link Dedicated Channel
The mapping of forward logical channels into forward physical channels is given in Table
15-6. The table shows attributes of each physical channel such as variable rate or fixed rate,
physical channel frame size, and sharability of the channel.
Voice Services. In the V1 mode, the F-SCH and F-DCCH are not used. In this mode the
upper layer signaling (f-dsch), voice frames (f-dtch), and power control information are multi-
Mapping of Logical Channels to Physical Channels 389

Table 15-6 Forward Dedicated Channel Structure for Different Services

Forward Services
Forward Link Link Logical
Physical Channel Channel Description V1 V2 V3 P1 P2 VP1 VP2
Fundamental f-dsch upper layer x x x x x
channel (F-FCH) signaling
(soft handoff) messages
f-dtch RLP frames x x x
voice frames x x x x
f-dmch MAC messages x x
pc power control x x x x x x
frame size 20 20 5/20 20 5/20 20
rates V V V V V V
Supplemental f-dtch RLP frames, x x x x x
channel (F-SCH) voice frames
(with or without (for newly
soft handoff) defined services)
rates H V/H H H V/H
Dedicated control f-dsch upper layer x x x
channel signaling
(F-DCCH) messages
(with or without
f-dtch RLP frames x x x
soft handoff)
f-dmch MAC messages x x x
pc power control x
sharable Yes No Yes Yes
frame size 20 5/20 5 5/20
rates F F F F
Key: F = fixed (9.6 kbps or 0); V = variable (9.6, 4.8, 2.7, 1.5 kbps for RS1 and 14.4, 7.2, 3.6, 1.8 kbps
for RS2); H = high data rates (scheduled rates).

plexed in F-FCH (e.g., by using dim-and-burst or blank-and-burst mechanisms). Forward link

continuity and outer-loop power control in this mode is maintained by the F-FCH.
The V2 mode is used to provide higher voice-quality service by transmitting the upper
layer signaling frames on an F-DCCH (e.g., no blank-and-burst or dim-and-burst signaling).
Forward link continuity and outer-loop power control in this mode is maintained by the F-FCH.
Packet Services. The packet data service, P1, is offered on the forward direction by using
F-FCH and F-SCH. The upper layer signaling message (f-dsch), MAC messages (f-dmch), and
390 Ch. 15 • cdma2000 System

user data frames (f-dtch) are time multiplexed in the F-FCH. In this mode, the control of MAC is
performed in a centralized manner since MAC messages are carried on the F-FCH which is typ-
ically in soft handoff to ensure reliability of delivery for upper layer signaling messages. The 5-
ms frames are used to carry short MAC messages in this mode. The F-SCH carries high-rate
RLP frames containing packet data, and transmission on the F-SCH is always scheduled in this
mode. Lower-rate RLP frames may be carried on the F-FCH. The transmission rate of F-SCH is
predefined using MAC messages. Forward link continuity and outer-loop power control in this
mode are maintained by the F-FCH.
The P2 mode of operation is an alternate basic packet data service and is similar to the P1
mode in the sense that upper layer signaling messages (f-dsch), MAC messages (f-dmch), and
user data frames (f-dtch) are time multiplexed in one channel. However, the physical channel
that is used to carry these logical channels is the F-DCCH, which may or may not be in soft
handoff. Thus, the control of the MAC can be performed in either a distributed or centralized
manner. To support the mixing of MAC signaling with RLP frames or upper layer signaling
information, the F-DCCH supports dual-frame-size operation (5 and 20 ms). The F-SCH carries
high-rate, scheduled RLP frames containing packet data as well as lower-rate RLP frames. The
rate for the lower-rate frames carried on F-SCH may be dynamically determined, but the trans-
mission rate of the high-rate scheduled frames carried on the F-SCH is prespecified using MAC
messages. Forward link continuity and outer-loop power control are maintained by the F-DCCH
and, therefore, the F-DCCH becomes unsharable in this mode.
The P3 mode is used for highly optimized packet data service with the potential support for
distributed control of the MAC layer (i.e., the f-dmch is carried in a physical channel that can be
operated with a reduced active set while upper layer signaling information is carried in a channel
with a full active set). In this mode, the F-FCH is primarily used to carry high-reliability, low-
delay upper layer messages. Power control bits are carried by F-FCH. The F-DCCH may be
shared to make a more efficient use of Walsh code resources. The F-DCCH carries the MAC sig-
naling (f-dmch) and may not be in soft handoff. Control of MAC can be performed in either a dis-
tributed or centralized manner. The F-SCH carries high-rate RLP frames containing packet data,
and transmission on the F-SCH is always scheduled in this mode. The transmission rate of F-SCH
is prespecified using MAC messages. Lower-rate RLP frames may be carried on the F-FCH. For-
ward link continuity and outer-loop power control in this mode are maintained by the F-FCH.
Concurrent Voice and Packet Services. The VP1 mode offers simultaneous basic voice
and packet data service by multiplexing upper layer signaling (f-dsch), MAC messages (f-
dmch), voice frames (f-dtch), and potentially low-rate RLP frames (f-dtch) into the F-FCH. Con-
trol of MAC is performed in a centralized manner. To support the mixing of MAC signaling with
RLP frames or upper layer signaling information, the F-FCH supports dual-frame-size operation
(5 and 20 ms). The F-SCH carries high-rate RLP frames containing packet data, and transmis-
sion on the F-SCH is always scheduled in this mode. Forward link continuity and outer-loop
power control are maintained by the F-FCH.
The VP2 mode also provides simultaneous voice and packet data service. To provide a
higher-quality voice service in conjunction with packet data service, the MAC messages (f-
Mapping of Logical Channels to Physical Channels 391

dmch) and potentially upper layer signaling (f-dsch) are carried on the F-DCCH. Control of
MAC can be performed in either a distributed (if F-DCCH is not in soft handoff) or centralized
(if DCCH is in soft handoff) manner. Power control bits are carried by the F-FCH. The F-DCCH
can be shared to make a more efficient use of Walsh code resources. To support the mixing of
MAC signaling with RLP frames or upper layer signaling information, the F-DCCH supports
dual-frame-size operation (5 and 20 ms). The F-SCH carries high-rate, scheduled RLP frames
containing packet data as well as lower-rate RLP frames. The low-rate RLP frames may be sent
on F-SCH to avoid the potential contention between voice and low-rate RLP frames on the F-
FCH. Forward link continuity and outer-loop power control are maintained by F-FCH.
Consult cdma2000 RTT [7] for circuit services and their combinations with voice and
packet services.

15.10.1.2 Forward Link Common Channels

When neither F-DCCH nor F-FCH is allocated to the MS (e.g., in the suspended or dor-
mant states), the upper layer signaling and MAC messages are carried to the MS using F-PCH or
F-CCCH. Messages sent on these channels may be encrypted and must include mobile station
ID or the packet service identifier because both F-PCH and F-CCCH are point-to-multipoint
channels in the sense that there is no one-to-one mapping between the ID of these channels and
mobile station ID. In addition to the control information (i.e., MAC or upper layer signaling),
short data bursts may be carried by the F-PCH or F-CCCH. Table 15-7 shows the mapping of
forward common channels to forward common physical channels.

15.10.2 Reverse Link

15.10.2.1 Reverse Link Dedicated Channels
The mapping of reverse logical channels into reverse physical channels is given in Table
15-8. The table also shows the attributes of each physical channel such as variable rate or fixed
rate and physical channel frame size.

Table 15-7 Mapping of Forward Common Logical Channels to Forward Common Physical
Channels

Forward Link Physical Channels Forward Logical Channels Description

Common control channel (F-CCCH) or f-csch upper layer signaling
paging channel (F-PCH) messages
f-ctch RBP frames
f-cmch MAC messages
Common pilot — common pilot
Auxiliary pilot — auxiliary pilot
Sync channel (F-SYNC) — sync channel information
392 Ch. 15 • cdma2000 System

Table 15-8 Reverse Dedicated Channel Structure for Different Services

Reverse Services
Reverse Link Link Logical
Physical Channel Channel Description V1 V2 V3 P1 P2 VP1 VP2
Fundamental r-dsch upper layer x x x x x x
channel (R-FCH) signaling
(soft handoff) messages
r-dtch RLP frames x x x
voice frames x x x x
r-dmch MAC messages x x
frame size 20 20 5/20 20 5/20 20
rates V V V V V V
Supplemental r-dtch RLP frames, x x x x x
channel (R-SCH) voice frames
(with or without (for newly
soft handoff) defined services)
rates H V/H H H V/H
Dedicated control r-dsch upper layer x x x
channel (R- signaling
DCCH) (with or messages
without soft
r-dtch RLP frames x x x
handoff)
r-dmch MAC messages x x x
frame size 20 5/20 5 5/20
rates F F F F
Key: F = fixed (9.6 kbps or 0); V = variable (9.6, 4.8, 2.7, 1.5 kbps for RS1 and 14.4, 7.2, 3.6, 1.8 kbps
for RS2); H = high data rates (scheduled rates).

Voice Services. In the V1 mode, the R-SCH and R-DCCH are not used. In this case the
upper layer signaling (r-dsch), voice frames (r-dtch), and power control information are multi-
plexed in R-FCH (e.g., using dim-and-burst or blank-and-burst mechanisms).
The V2 mode provides a higher voice-quality service typically by transmitting the upper
layer signaling frames on an R-DCCH (e.g., no blank-and-burst or dim-and-burst signaling).
However, if the mobile station cannot provide sufficient power to transmit on the R-DCCH, the
upper layer signaling information can be transmitted on the R-FCH.
Packet Services. P1 is offered on the reverse direction by using R-FCH and R-SCH. The
upper layer signaling message (r-dsch), MAC messages (r-dmch), and user data frames (r-dtch)
are time multiplexed in R-FCH. To support the mixing of MAC signaling with RLP frames or
upper layer signaling information, the R-FCH supports dual-frame-size operation (5 and 20 ms).
Evolution of cdmaOne (IS-95) to cdma2000 393

The R-SCH carries high-rate RLP frames containing packet data, and transmission on the R-
SCH is always scheduled. Lower-rate RLP frames may be carried on R-FCH.
In the P2 mode, upper layer signaling messages (r-dsch), MAC messages (r-dmch), and
user data frames (r-dtch) are time multiplexed in R-DCCH. To support the mixing of MAC sig-
naling with RLP frames or upper layer signaling information, the R-DCCH supports dual-frame-
size operation (5 and 20 ms). The R-SCH carries high-rate scheduled RLP frames containing
data as well as lower-rate RLP frames.
R-FCH is primarily used to carry high-reliability, low-delay upper layer messages as well
as power control information. The R-DCCH carries the MAC signaling (r-dmch). The R-SCH
carries high-rate RLP frames containing packet data, and transmission on the R-SCH is always
scheduled. Lower-rate RLP frames may be carried on the R-FCH.
Concurrent Voice and Packet Data. The VP1 mode (which offers simultaneous basic
voice and packet data service) multiplexes upper layer signaling (r-dsch), MAC messages (r-
dmch), voice frames (r-dtch), and, potentially, low-rate RLP frames (r-dtch) into the R-FCH. To
support the mixing of MAC signaling with RLP frames or upper layer signaling information, the
R-FCH supports dual-frame-size operation (5 and 20 ms). The R-SCH carries high-rate RLP
frames containing packet data, and transmission on the R-SCH is always scheduled in this mode.
In the VP2 mode, the MAC messages (r-dmch) and potentially upper layer signaling infor-
mation (r-dsch) are carried on R-DCCH (thereby reducing potential disruption due to dim-and-
burst and blank-and-burst signaling). However, if the mobile cannot provide sufficient power to
transmit on the R-DCCH, the upper layer signaling information can be transmitted on the R-
FCH. To support the mixing of MAC signaling with RLP frames or upper layer signaling infor-
mation, the R-DCCH supports dual-frame-size operation (5 and 20 ms). The R-SCH carries
high-rate, scheduled RLP frames containing packet data as well as lower-rate RLP frames. The
low-rate RLP frames may be sent on the R-SCH to avoid the potential contention between voice
and low-rate RLP frames on the R-FCH.

15.10.2.2 Reverse Link Common Channels

When neither R-DCCH nor R-FCH is allocated to the mobile (e.g., in the suspended or
dormant state), the upper layer signaling and MAC messages are conveyed to the base station
using R-ACH or R-CCCH. Messages sent on these channels may be encrypted and must include
the mobile ID or the packet service identifier since there is no one-to-one mapping between the
identity of the R-ACH and R-CCCH channel and the mobile ID. In addition to the control infor-
mation (i.e., MAC or upper layer signaling), short data bursts may be carried by the R-ACH or
R-CCCH. Table 15-9 shows the mapping of the reverse common logical channels to reverse
common physical channels.

15.11 Evolution of cdmaOne (IS-95) to cdma2000

Data is the word as cdmaOne operators eye a host of new network capabilities enabling them to
offer new value-added services that can exploit present and future generations of technology.
With the Internet and corporate intranet becoming more essential to daily business activities, the
394 Ch. 15 • cdma2000 System

Table 15-9 Mapping of Reverse Common Logical Channels to Reverse Common Physical
Channels

Reverse Link Physical Channels Reverse Logical Channels Description

Common control channel (R-CCCH) r-csch upper layer signaling messages
or access channel (R-ACH) a
r-ctch RBP frames
r-cmch MAC messages
Pilot — pilot
pc power control bits
a
r-ctch is carried on the same physical channel (R-CCCH), but can be physically separated by using a dif-
ferent long code.

rush is on to create the wireless office that can easily tie mobile workers to the enterprise. Fur-
ther, there is great potential for push technologies that deliver news and other information
directly to a wireless device—this could create entirely new revenue streams for operators.
Although cdmaOne networks were not the first to offer data access, these networks are
uniquely designed to accommodate data. To start with, they handle data and voice transmissions
in much the same way. cdmaOne’s inherent variable-rate transmission capability allows data rate
determination to accommodate the amount of information being sent, so system resources are
used only as needed. Because cdmaOne systems employ a packetized backbone for voice,
packet data capabilities are already inherent in the equipment. The cdmaOne packet data trans-
mission technology uses a TCP/IP-compliant Cellular Digital Packet Data (CDPD) protocol
stack to enable seamless connectivity with enterprise networks and to expedite third-party appli-
cation development.
Adding data to the cdmaOne network will allow an operator to continue using its existing
radios, backhaul facilities, infrastructure, and handsets while merely implementing a software
upgrade with an interworking function. Upgrading to IS-95B allows for code or channel aggrega-
tion to provide data rates of 64–115 kbps, as well as offering improvements in soft handoffs and
interfrequency hard handoffs. Equipment manufacturers have already announced IS-707 packet
data, circuit-switched data, and digital fax capabilities on its cdmaOne infrastructure equipment.
Mobile IP, the proposed Internet standard for mobility, is an enhancement to basic packet
data services. Mobile IP lets users maintain a continuous data connection and retain a single IP
address while traveling between base station controllers (BSCs) or roaming on other CDMA
networks.
One of the key objectives of ITU IMT-2000 is the creation of standards that will encourage
a worldwide frequency band to promote a high degree of design commonality and to support
high-speed data services. IMT-2000 will utilize small pocket terminals, an expanded range of
operation environments, and the deployment of an open architecture that allows the graceful
introduction of newly created technology. Furthermore, 3G systems promise to deliver wireless
voice services with wireline quality levels, along with the speed and capacity needed to support
Evolution of cdmaOne (IS-95) to cdma2000 395

multimedia and high-speed data applications. Location-based services, on-board navigation,

emergency assistance, and other advanced services will also be supported.
The evolution of 3G systems will open the door of the wireless local loop (WLL) to PSTN
and public data network access, while providing more convenient control of applications and
network resources. It will also provide global roaming, service portability, zone-based ID and
billing, and global directory access. The 3G technology is even expected to support seamless sat-
ellite interworking.
One of the technical requirements for cdma2000 includes cdmaOne backward compatibil-
ity for voice services, vocoders, and signaling structure, as well as for privacy, authentication,
and encryption capabilities.
Phase one of the cdma2000 effort, also known as 1xRTT, employs 1.25 MHz of bandwidth
and delivers a peak data rate of 144 kbps for stationary or mobile applications. Phase two of
cdma2000, called 3xRTT, will use 5 MHz of bandwidth and is expected to deliver a peak data
rate of 144 kbps for mobile and vehicular applications and up to 2 Mbps for fixed applications.
Industry insiders predict that the 3xRTT phase will eventually yield up to 1 Mbps for each traffic
or Walsh channel. By aggregating or bundling two channels, users can achieve the 2-Mbps peak
data rate targeted for IMT-2000.
The primary difference between phase one and phase two of cdma2000 is bandwidth and
the resulting throughput speed, or peak data rate capability. Phase two will introduce advanced
multimedia capabilities and lay the foundation for popular 3G voice services and vocoders, such
as voice over IP. Since the 1xRTT and 3xRTT standards essentially share the same baseband
radio elements, operators can take a major step toward full 3G capabilities by implementing
1xRTT. cdma2000 phase two will include detailed descriptions of signal protocols, data man-
agement, and expected upscale requirements for moving from 5-MHz radios to 10- and 15-MHz
radios in future interactions.
By migrating from the current IS-95 CDMA air-interface technology to 1xRTT of the
cdma2000 standard, operators can reap a twofold increase in radio capacity and the ability to
handle up to 144 kbps of packet data. Phase one capabilities of cdma2000 include a new physi-
cal layer for 1× and 3× 1.25-MHz channel sizes; support for direct-spread and multicarrier for-
ward link 3× options; and definitions for the 1× and 3× numerology. Operators will also enjoy
voice service enhancements that will produce two times the voice capacity.
In the area of extended battery life, phase one will employ a quick paging channel and
gated transmission of 1/8 rate to produce gains of two times the battery life currently available.
Hard-handoff enhancements between 2G and 3G systems and power control enhancements will
also be key factors in the improvement of voice service.
Data services will also be improved with the advent of cdma2000 phase one. Phase one
will feature a MAC framework and a packet data radio link protocol (RLP) definition to support
packet data rates of at least 144 kbps.
Implementation of cdma2000 phase two will bring a host of new capabilities and service
enhancements. Phase two will support all channel sizes (6×, 9×, and 12×) and associated numer-
ology and a framework for advanced cdma2000 3G voice services and vocoders—including
396 Ch. 15 • cdma2000 System

voice over IP. With phase two, true multimedia services will be available to bring additional rev-
enue opportunities for operators. Multimedia services will be made possible through enhanced
packet data MAC, full support for packet data services up to 2 Mbps, RLP support for all data
rates up to 2 Mbps, and the advanced multimedia call model.
In the area of signaling and services, phase two cdma2000 will bring native 3G cdma2000
signaling structure to the link access control (LAC) and upper layer signaling structure. This
structure will provide support for enhanced privacy, authentication, and encryption functionality.
An operator’s existing architecture and network equipment can greatly affect the ease of this
migration. A network built on an open, advanced architecture with a clear upward migration
pathway can attain 1xRTT capabilities with a simple modular upward movement of the H-band
operation of the radio. Networks with a less flexible architecture may be required to take the
more costly steps of replacing the entire base transceiver system (BTS). To achieve the expected
144-kbps peak data rate performance, operators can make software upgrades to networks and
base stations to support 1xRTT data protocol.
Packet Data Service Node (PDSN) will be required to support data connectivity to the
Internet/intranet. Many equipment vendors already offer solutions that incorporate PDSN ele-
ments, thus opening a smooth upward pathway to 3G technologies.
The recent agreement between Qualcomm and Ericsson proposes three optional CDMA
modes and the eventual development of a global standard that is compatible with both ANSI
IS-41 and GSM MAP. This approach envisions the use of multimode handsets and various mar-
ket-driven solutions as the surest pathway to a unified CDMA 3G standard in the next genera-
tion of wireless communications. As subscribers demand greater wireless power and
convenience, the migration to 3G technology will benefit operators by supporting higher capa-
bilities, lowering network costs, and increasing overall profitability. Fig. 15-35 shows the
cdmaOne evolution timeline.
cdmaOne operators will be able to upgrade to 3G system without acquiring additional
spectrum, a key component to minimum time to market without additional, significant invest-
ment. The design of cdma2000 will allow for deployment of the 3G enhancements while main-
taining existing 2G support for cdmaOne in the spectrum being used by an operator today.
Both cdma2000 phase one and phase two can be intermingled with cdmaOne to maximize
the effective use of spectrum according to the needs of an individual operator’s customer base.
For example, an operator with a strong demand for high-speed data service may choose to
deploy a combination of cdma2000 phase one and cdmaOne that uses more channels for
cdma2000 (see Fig. 15-36). In another market, users may not be as quick to adopt high-speed
data services, and more channels will remain dedicated to cdmaOne services. As cdma2000
phase two capabilities become available, an operator has even more choices of ways in which to
use the spectrum to support the new services (see Figs. 15-37 and 15-38).

15.12 Major Technical Differences between cdma2000 and W-CDMA

Table 15-10 lists the major technical differences between the two wideband CDMA 3G proposals.
Major Technical Differences between cdma2000 and W-CDMA 397

2G 3G
cdmaOne
2 Mbps cdma2000
Phase Two

144 kbps cdma2000

Phase One

64 kbps IS-95B

14.4 kbps IS-95A

1999 2001
Figure 15-35 cdmaOne Evolution Timeline

5 MHz

Guard cdmaOne cdmaOne cdma2000 Guard

F1 F2 Phase One

0.625 1.25 MHz 1.25 MHz 1.25 MHz 0.625

MHz MHz

Figure 15-36 Intermixing of cdmaOne and cdma2000 Phase One

7.5 MHz

cdma2000 Phase Two cdma2000

Phase One

Guard 5 MHz 1.25 MHz Guard

Figure 15-37 Intermixing of cdma2000 Phase One and cdma2000 Phase Two

cdma2000 Phase Two cdma2000 cdmaOne cdmaOne

Phase One

5 MHz 1.25 MHz 1.25 MHz 1.25 MHz

10 MHz

Figure 15-38 Intermixing of cdmaOne, cdma2000 Phase One, and cdma2000 Phase Two
398 Ch. 15 • cdma2000 System

Table 15-10 Major Technical Differences between cdma2000 and W-CDMA

cdma2000 W-CDMA
Core network ANSI-41 GSM MAP
Chip rate 3.6864 Mcps 4.096 Mcps (DOCOMO) 3.84 Mcps (UMTS)
Synchronized BS yes no/yes (optional)
Frame length 20 ms 10 ms
Multicarrier spreading option yes no
Voice coder EVRC new
Overhead low (because of shared high (because of nonshared pilot code
pilot code channel) channel)

15.13 Summary
This chapter first presented cdma2000 layering structure and logical and physical channels and
then concentrated on the cdma2000 physical layer, providing details of forward/reverse dedi-
cated and common physical channel structures. We described forward and reverse link features
and pointed out the improvements of cdma2000 over cdmaOne.
We briefly discussed data services in cdma2000 and considered the mapping of logical
channels. Next we presented the evolution plans for cdmaOne to cdma2000. We concluded the
chapter by providing the major technical differences between cdma2000 and W-CDMA.

15.14 References
1. “CDMA for Next Generation Mobile Communications Systems,” IEEE Communications
Magazine 36(9), September 1998.
2. Dahlman, E., Gumudson, B., Nilsson, M., and Skold, J., “UMTS/IMT-2000 Based Wide-
band CDMA,” IEEE Communications Magazine 36(9), September 1998.
3. Garg, V. K., Halpern, S., and Smolik, K. F., “Third Generation (3G) Mobile Communication
Systems,” 1999 IEEE International Conference on Personal Wireless Communications,
Jaipur, India, February 1999.
4. Knisley, D., Quinn, L., and Ramesh, N., “cdma2000: A Third Generation Radio Transmis-
sion Technology,” Bell Labs Technical Journal 3(3), July–September 1998.
5. Rao, Y. S., and Kripalani, A., “cdma2000 Mobile Radio Access for IMT-2000,” 1999 IEEE
International Conference on Personal Wireless Communications, Jaipur, India, February
1999.
6. Shanker, B., McClelland, S., “Mobilising the Third-Generation [Cellular Radio],” Journal of
Telecommunications (International Edition), August 1997.
7. TIA TR 45.5, “The cdma2000 ITU-RTT Candidate Submission,” TR 45-ISD/98.06.02.03,
May 15, 1998.
 A P P E N D I X A

Traffic Tables

This appendix provides traffic tables for a variety of blocking probabilities and channels. The
blocked-calls-cleared (Erlang B) call model is used. In Erlang B, when traffic arrives in the sys-
tem, we assume that it either is served, with probability from the table, or is lost to the system. A
customer attempting to place a call will therefore either see a call completion or will be blocked
and will abandon the call. This assumption is acceptable for low blocking probabilities. In some
cases, the call will be placed again after a short period of time. If too many calls reappear in the
system after a short delay, the Erlang B model will no longer hold.

Table A-1 Offered Load for Given Blocking Probability

Number Blocking Probability
of
Servers 0.005 0.01 0.015 0.02 0.03 0.05 0.07 0.1
1 — 0.01011 0.01524 0.02041 0.03093 0.05264 0.07527 0.1111
2 — 0.1527 0.1904 0.2235 0.2816 0.3814 0.4705 0.5955
3 — 0.4556 0.5352 0.6022 0.7152 0.8994 1.057 1.271
4 — 0.8693 0.9919 1.092 1.259 1.525 1.748 2.045
5 — 1.361 1.524 1.657 1.875 2.219 2.504 2.881
6 — 1.909 2.112 2.276 2.543 2.961 3.305 3.759
7 — 2.501 2.741 2.936 3.25 3.738 4.139 4.666
8 — 3.127 3.405 3.627 3.987 4.543 4.999 5.597
9 — 3.783 4.095 4.345 4.748 5.371 5.88 6.547
10 — 4.462 4.808 5.084 5.53 6.216 6.777 7.511

399
400 Apx. A • Traffic Tables

Table A-1 Offered Load for Given Blocking Probability (Continued)

Number Blocking Probability
of
Servers 0.005 0.01 0.015 0.02 0.03 0.05 0.07 0.1
11 — 5.159 5.54 5.842 6.328 7.076 7.688 8.487
12 — 5.877 6.288 6.615 7.142 7.95 8.61 9.474
13 — 6.607 7.05 7.401 7.967 8.835 9.543 10.47
14 — 7.352 7.825 8.201 8.804 9.73 10.49 11.47
15 — 8.109 8.61 9.01 9.651 10.63 11.43 12.48
16 — 8.876 9.406 9.829 10.51 11.54 12.39 13.5
17 — 9.653 10.21 10.66 11.37 12.46 13.35 14.52
18 — 10.44 11.03 11.49 12.24 13.39 14.32 15.55
19 — 11.23 11.85 12.33 13.12 14.31 15.29 16.58
20 — 12.03 12.67 13.18 14 15.25 16.27 17.61
21 11.86 12.84 13.51 14.04 14.89 16.19 17.25 18.65
22 12.63 13.65 14.35 14.9 15.78 17.13 18.24 19.69
23 13.42 14.47 15.19 15.76 16.68 18.08 19.23 20.74
24 14.2 15.3 16.04 16.63 17.58 19.03 20.22 21.78
25 15 16.13 16.9 17.5 18.48 19.99 21.22 22.83
26 15.8 16.96 17.75 18.38 19.39 20.94 22.21 23.89
27 16.6 17.8 18.62 19.27 20.31 21.9 23.21 24.94
28 17.41 18.64 19.48 20.15 21.22 22.87 24.22 26
29 18.22 19.49 20.35 21.04 22.14 23.83 25.22 27.05
30 19.04 20.34 21.23 21.93 23.06 24.8 26.23 28.11
31 19.85 21.19 22.1 22.83 23.99 25.77 27.24 29.17
32 20.68 22.05 22.98 23.73 24.92 26.75 28.25 30.24
33 21.51 22.91 23.87 24.63 25.85 27.72 29.26 31.3
34 22.34 23.77 24.75 25.53 26.78 28.7 30.28 32.37
35 23.17 24.64 25.64 26.44 27.71 29.68 31.29 33.43
36 24.01 25.51 26.53 27.34 28.65 30.66 32.31 34.5
37 24.85 26.38 27.42 28.26 29.59 31.64 33.33 35.57
38 25.69 27.25 28.32 29.17 30.53 32.62 34.35 36.64
39 26.54 28.13 29.22 30.08 31.47 33.61 35.37 37.72
40 27.38 29.01 30.12 31 32.41 34.6 36.4 38.79
Apx. A • Traffic Tables 401

Table A-1 Offered Load for Given Blocking Probability (Continued)

Number Blocking Probability
of
Servers 0.005 0.01 0.015 0.02 0.03 0.05 0.07 0.1
41 28.23 29.89 31.02 31.92 33.36 35.59 37.42 39.86
42 29.09 30.77 31.92 32.84 34.31 36.57 38.45 40.94
43 29.94 31.66 32.83 33.76 35.25 37.57 39.47 42.01
44 30.8 32.55 33.74 34.68 36.2 38.56 40.5 43.09
45 31.66 33.43 34.65 35.61 37.16 39.55 41.53 44.17
46 32.52 34.32 35.56 36.54 38.11 40.55 42.56 45.24
47 33.38 35.22 36.47 37.46 39.06 41.54 43.59 46.32
48 34.25 36.11 37.38 38.39 40.02 42.54 44.62 47.4
49 35.12 37.01 38.3 39.32 40.98 43.54 45.65 48.48
50 35.99 37.9 39.21 40.25 41.93 44.53 46.69 49.56
51 36.86 38.8 40.13 41.19 42.89 45.53 47.72 50.64
52 37.73 39.7 41.05 42.12 43.85 46.53 48.76 51.73
53 38.6 40.61 41.97 43.06 44.81 47.54 49.79 52.81
54 39.47 41.51 42.9 44 45.78 48.54 50.83 53.89
55 40.36 42.41 43.82 44.94 46.74 49.54 51.86 54.98
56 41.23 43.32 44.74 45.88 47.71 50.54 52.9 56.06
57 42.11 44.23 45.67 46.82 48.67 51.55 53.94 57.15
58 43 45.13 46.6 47.76 49.64 52.55 54.98 58.23
59 43.87 46.04 47.52 48.7 50.6 53.56 56.02 59.32
60 44.75 46.95 48.45 49.65 51.57 54.57 57.06 60.4
61 45.65 47.86 49.38 50.59 52.54 55.57 58.1 61.49
62 46.53 48.78 50.31 51.54 53.51 56.58 59.14 62.58
63 47.42 49.69 51.24 52.48 54.48 57.59 60.18 63.66
64 48.3 50.6 52.18 53.43 55.45 58.6 61.22 64.75
65 49.2 51.52 53.11 54.38 56.42 59.61 62.27 65.84
66 50.09 52.43 54.05 55.33 57.4 60.62 63.31 66.93
67 50.98 53.36 54.98 56.28 58.37 61.63 64.35 68.02
68 51.87 54.27 55.92 57.23 59.34 62.64 65.4 69.11
69 52.76 55.2 56.85 58.18 60.32 63.66 66.44 70.2
70 53.66 56.12 57.79 59.13 61.29 64.67 67.49 71.29
402 Apx. A • Traffic Tables

Table A-1 Offered Load for Given Blocking Probability (Continued)

Number Blocking Probability
of
Servers 0.005 0.01 0.015 0.02 0.03 0.05 0.07 0.1
71 54.55 57.04 58.73 60.08 62.27 65.68 68.53 72.38
72 55.46 57.96 59.67 61.04 63.25 66.69 69.58 73.47
73 56.35 58.88 60.61 61.99 64.22 67.71 70.63 74.56
74 57.26 59.81 61.55 62.95 65.2 68.72 71.67 75.65
75 58.15 60.73 62.5 63.9 66.18 69.74 72.72 76.74
76 59.06 61.66 63.44 64.86 67.16 70.75 73.77 77.83
77 59.96 62.58 64.38 65.82 68.14 71.77 74.81 78.93
78 60.87 63.51 65.32 66.77 69.12 72.79 75.86 80.02
79 61.77 64.44 66.27 67.73 70.1 73.8 76.91 81.11
80 62.67 65.36 67.21 68.69 71.08 74.82 77.96 82.2
81 63.58 66.3 68.16 69.65 72.06 75.84 79.01 83.3
82 64.47 67.23 69.11 70.61 73.04 76.86 80.06 84.39
83 65.38 68.16 70.05 71.57 74.03 77.88 81.11 85.49
84 66.3 69.09 71 72.53 75.01 78.89 82.16 86.58
85 67.2 70.02 71.95 73.49 75.99 79.91 83.21 87.67
86 68.11 70.95 72.9 74.46 76.98 80.93 84.26 88.77
87 69.03 71.89 73.85 75.42 77.96 81.95 85.31 89.86
88 69.93 72.81 74.8 76.38 78.95 82.97 86.36 90.96
89 70.85 73.75 75.74 77.34 79.93 83.99 87.41 92.05
90 71.76 74.69 76.7 78.31 80.92 85.02 88.46 93.15
91 72.67 75.62 77.65 79.27 81.9 86.04 89.52 94.24
92 73.58 76.56 78.6 80.24 82.89 87.06 90.57 95.34
93 74.5 77.5 79.56 81.2 83.88 88.08 91.62 96.43
94 75.41 78.43 80.51 82.17 84.86 89.1 92.67 97.53
95 76.33 79.37 81.46 83.14 85.85 90.13 93.73 98.63
96 77.25 80.3 82.42 84.1 86.84 91.15 94.78 99.72
97 78.16 81.25 83.37 85.07 87.83 92.17 95.83 100.8
98 79.08 82.19 84.33 86.03 88.82 93.19 96.89 101.9
99 80 83.12 85.28 87.01 89.81 94.22 97.94 103
100 80.92 84.07 86.24 87.98 90.8 95.24 99 104.1
 A P P E N D I X B

Abbreviations

A
AC Authentication Center
ACCLOC Access Overload Class
ACH Access Channel
ACK Acknowledgment
ACM Address Complete Message
ACSE Association Control Service Element
ADPCM Adaptive Differential Pulse Code Modulation
AMPS Advanced Mobile Phone Service
ANSI American National Standard Institute
ARPA Advanced Research Project Agency
ARQ Automatic Repeat Request
AT Prefix for dialing using a modem
AUC Authentication Center
AWGN Additive White Gaussian Noise

B
BB Baseband filter
BCAF Bearer Control Agent Function
BCCH Broadcast Control Channel

403
404 Apx. B • Abbreviations

BCF Bearer Control Function

BCFr Bearer Control Function for radio bearer
BER Bit Error Rate
BHCA Busy Hour Call Attempts
bps bits per second
BPSK Binary Phase Shift Keying
BS Base Station
BSAP Base Station Application Part
BSC Base Station Controller
BSMAP Base Station Management Application Part
BSS Base Station System
BTS Base Transceiver System

C
CASE Common Application Service Element
CAPICH Common Auxiliary Pilot Channel
CBSEED Codebook Seed
CC Connection Confirm
CC Country Code
CCAF Call Control Agent Function
CCCH Common Control Channel
CCF Call Control Function
CCR Commitment Concurrency and Recovery
CCAF’ Call Control Agent Function (enhanced)
CCF’ Call Control Function (enhanced)
CDG CDMA Development Group
CDMA Code Division Multiple Access
CDPD Cellular Digital Packet Data
CELP Code-Excited Linear Predictor
CH Channel
CI Cell Identity
CGI Cell Global Identification
CLIP Connectionless Interworking Protocol
CM Connection Management
COUNT Call history parameter
COST Committee On Standards and Technology
CR Connection Request
Apx. B • Abbreviations 405

CRC Cyclic Redundancy Check

CSMA/CA Carrier-Sense Multiple Access with Collision Avoidance

D
DAPICH Dedicated Auxiliary Pilot Channel
dB Decibels
dBm Decibels with respect to 1 milliwatt
DCCH Dedicated Control Channel
DLCI Data Link Connection Identifier
DMH Data Message Handler
DPC Destination Point Code
DQPSK Differential Quadrature Phase Shift Keying
DS Direct Sequence
DSP Digital Signal Processing
DSSS Direct Sequence Spread Spectrum
DTAP Direct Transfer Application Part
DTMF Dual-Tone Multifrequency

E
EIR Equipment Identity Register
ETSI European Telecommunications Standards Institute
ESN Electronic Serial Number
EVRC Enhanced Variable Rate Codec

F
FAF Floor Attenuation Factor
FCH Fundamental Channel
FCC Federal Communication Commission
FD Full Duplex
FDD Frequency Division Duplex
FDMA Frequency Division Multiple Access
FE Functional Entity
FEC Forward Error Correction
FER Frame Error Rate
FFPC Fast-Forward Power Control
406 Apx. B • Abbreviations

FHSS Frequency Hopping Spread Spectrum

FLPC Forward Link Power Control
FPICH Forward Pilot Channel
FPLMTS Future Public Land Mobile Telephone System

G
G Codebook Gain
GBN Go-Back-N
GFSK Gaussian Frequency Shift Keying
GHz gigaHertz
GSM Global System of Mobile Communications
GoS Grade of Service
GPRS General Packet Radio Service

H
HCM Handoff Completion Message
HD Half Duplex
HDM Handoff Direction Message
HLR Home Location Register
Hz Hertz

I
IAM Initial Address Message
ICMP Internet Control Message Protocol
IF Intermediate Frequency
IMT-2000 International Mobile Telecommunications in year 2000
IMSI International Mobile Subscriber Identifier
IN Intelligent Network
INAP Intelligent Network Application Part
IP Internet Protocol
IPCP Internet Protocol Control Protocol
IPR Intellectual Property Rights
ISDN Integrated Services Digital Network
ISM Industrial Scientific Medical
ISO International Organization for Standardization
Apx. B • Abbreviations 407

ISUP ISDN User Part (of Signaling System 7)

ITU International Telecommunications Union
IWF Interworking Function

J
J Joules

K
K Kelvin
kbps kilobits per second
kHz kilohertz
km kilometers
km/h kilometers per hour

L
LAC Link Access Control
LAC Location Area Code
LBT Listen-Before-Talk
LCP Link Control Protocol
LLC Logical Link Control
LPC Linear Predictive Coding
LPF Low Pass Filter
LOS Line-of-Sight
LS Least Squares
LSP Linear Spectral Pairs

M
MAC Medium Access Control
MAN Metropolitan Area Network
MAP Mobile Application Part
MBS Mobile Broadband Systems
Mbps Megabits per second
Mcps Megachips per second
MC Message Center
408 Apx. B • Abbreviations

MCC Mobile Country Code

MCTD Multicarrier Transmit Diversity
MDN Mobile Directory Number
MDR Medium Data Rate
MHz megahertz
MIN Mobile Identification Number
MM Mobility Management
MMSE Minimum Mean Square Error
MNC Mobile Network Code
MOS Mean Opinion Score
MPL Mean Path Loss
mph miles per hour
MRRC Mobile Radio Resource Control
MRTR Mobile Radio Transmission and Reception
MS Mobile Station
ms milliseconds
MSC Mobile Switching Center
MSIN Mobile Station Identification Number
MSS Mobile Satellite Services
MSS Mobile Switching Subsystem
MT Mobile Termination
MTP Message Transfer Part
MUD Multiuser Detection
MUX Multiplexer
mW milliwatts

N
NID Network Identification
NE Network Element
NLUM Neighbor List Update Message
NMSI National Mobile Subscriber Identity

O
OA&M Operation Administration and Maintenance
OMC Operation Maintenance Center
OQPSK Offset Quadrature Phase Shift Keying
Apx. B • Abbreviations 409

OS Operations System
OSI Open System Interconnection
OSS Operations Support System
OTAF Over-the-Air Function
OTASP Over-the-Air Service Provisioning
OTD Orthogonal Transmit Diversity

P
PACA Priority Access Channel Assignment
PBX Private Branch Exchange
PCG Power Control Group
PCH Paging Channel
PCM Pulse Code Modulation
PCMCIA Personal Computer Memory Card International Association
PCS Personal Communication Services
PDA Personal Digital Assistance
PICH Pilot Channel
PIN Personal Identification Number
PMRM Pilot Strength Measurement Message
PN pseudonoise
PPDN Public Packet Data Network
PPP Point-to-Point Protocol
PSK Phase Shift Keying
psd power spectral density
PSMM Power Measurement Report Message
PSTN Public Switched Telephone Network

Q
QCELP Qualcomm Code-Excited Linear Prediction
QoS Quality of Service
QPSK Quadrature Phase Shift Keying

R
RACE Research in Advanced Communications Equipment
RACF Radio Access Control Function
410 Apx. B • Abbreviations

RAM Random Access Memory

RAND Random Number
RBP Radio Burst Protocol
RCF Radio Control Function
RCLP Relaxed Code-excited Linear Prediction
RF Radio Frequency
RFTR Radio Frequency Transmission and Reception
RLP Radio Link Protocol
rms Root Mean Square
ROLPC Reverse Outer Loop Power Control
ROSE Remote Operation Service Element
RRC Radio Resource Control
RS Radio System
RS1 Rate Set 1
RS 2 Rate Set 2
RT Random Time
RTF Radio Terminal Function
RTT Radio Transmission Technology

S
SACF Service Access Control Function
SASE Specific Application Service Element
SBS Switched Beam System
SCCP Signaling Connection Control Part
SCF Service Control Function
SCH Supplementary Channel
SCI Synchronized Capsule Indicator
SCP Switching Control Point
SDF Service Data Function
SID System Identification
SIR Signal-to-Interference Ratio
SLS Signaling Link Selection
SMF Service Management Function
SMRS Specialized Mobile Radio Services
SMS Short Message Service
SNR Signal-to-Noise Ratio
SNDCF Sub-Network Dependent Convergence Function
Apx. B • Abbreviations 411

SNDCP Sub-Network Dependent Convergence Protocol

SOM Start of Message
SP Signaling Point
SRBP Signaling Radio Burst Protocol
SRF Specialized Resource Function
SRLP Signaling Radio Link Protocol
SRP Selective Repeat Request
SS Spread Spectrum
SSD Shared Secret Data
SSF Service Switching Function
SSP Switching System Platform
STP Signal Transfer Point
SYNC Sync Channel

T
TACAF Terminal Access Control Agent Function
TACF Terminal Access Control Function
TCAP Transaction Capabilities Application Part
TCH Traffic Channel
TCP Transport Control Protocol
TDD Time Division Duplex
TDM Time Division Multiplex
TDMA Time Division Multiple Access
TE Terminal Equipment
THSS Time-Hopped Spread Spectrum
TIA Telecommunication Industry Association
TIMF Terminal Identification Management Function
TMA Telesystems Micro-cellular Architecture
TMSI Temporary Mobile Subscriber Identity
TRAC Telecommunication Research and Action Center

U
UDP User Data Protocol
UIMF User Identification Management Function
UMTS Universal Mobile Telephone Service
412 Apx. B • Abbreviations

V
VLR Visitor Location Register
VLSI Very Large Scale Integration

W
WARC World Administration Radio Conference
W-CDMA Wideband CDMA
WLAN Wireless Local Area Network
WLL Wireless Local Loop
WORM Window control Operation-based Reception Memory
WPBX Wireless Private Branch Exchange

X
XC Transcoder
 A P P E N D I X C

Additional References

Adachi, F., Sawahashi, M., and Suda, H., “Wideband DS-CDMA for Next Generation Mobile
Communications Systems,” IEEE Communication Magazine 36(9), September. 1998.
Applebaum, S. P., “Adaptive Arrays,” IEEE Transactions on Antenna and Propagation, AP-24,
September 1976, pp. 585–98.
Bang S. C., “Performance Analysis of a Wideband CDMA System for FPLMTS,” Proceedings
of IEEE VTC, May 1997.
Benedetto, S., and Montorsi, G., “Design of Parallel Concatenated Convolutional Codes,” IEEE
Transactions on Communications 44(5), May 1996.
Benedetto, S., Garello, R., and Montorsi, G., “A Search for Good Convolutional Codes to be
Used in the Construction of Turbo Codes,” IEEE Transactions on Communications 46(9), Sep-
tember 1998.
Braun, W. R., and Dersch, U., “A Physical Mobile Radio Channel Model, “ IEEE Transactions
on VT, VT-40(2), May 1991, pp. 472–82.
Chan, M., C., and Woo, T. Y. C., “Next Generation Wireless Data Services: Architecture and
Experience,” IEEE Personal Communications 6(1), February 1999, pp. 20–33.
Chang, C. R., Wan, J. Z., and Yee, M. F., “PN Offset Planning Strategies for Non-Uniform
CDMA Networks,” Proceeding of IEEE VTC, Arizona, May 1997, pp. 1543–47.
Chopra, M., Rohani, K., and Reed, J., “Analysis of CDMA Range Extension Due to Soft Hand-
off,” Proceedings of IEEE VTC, Chicago, July 1995, pp. 917–21.
Compton, R. T., Adaptive Antennas (Concepts and Performance), Prentice Hall, Englewood
Cliffs, NJ, 1988.

413
414 Apx. C • Additional References

Compton, R. T., “An Adaptive Antenna in a Spread Communication Systems,” Proceedings of

IEEE 66, March 1978, pp. 289–98.
Feher, K., Wireless Digital Communications, Prentice Hall, Upper Saddle River, NJ, 1995.
Fleming, G., et al., “A Flexible Network Architecture for UMTS,” IEEE Personal Communica-
tions 5(2), April 1998, pp. 8–15.
Gabriel, W. F., “Adaptive Arrays ( an Introduction)”, Proc. IEEE 64, February 1976, pp. 239–72.
Gilhousen, K. S., Jacobs, I. M., Padovani, R., Viterbi, A. J., Weaver, L. A., and Wheatley, C. E.,
“On the Capacity of a Cellular CDMA System,” IEEE Transaction of Vehicular Technology 40,
May 1991, pp. 472–80.
Goodman, D. J., “Trends in Cellular and Cordless Communications,” IEEE Communication
Magazine 29(6), June 1991, pp. 31–40.
Hall C. J., and Foose, W. A., “Practical Planning for CDMA Networks (A Design Process Over-
view),” Proceedings of the 1996 Southcon Conference, Orlando, FL, June 1996, pp. 66–71.
Hong, D., and Rappaport, S. S., “Traffic Model and Performance Analysis for Cellular Radio
Telephone Systems with Prioritized and Non-Prioritized Handoff Procedures,” IEEE Trans. VT,
VT-35(3), August 1986, pp. 77–92.
Hu, L.-R., and Rappaport, S. S., “Personal Communications Systems Using Hierarchical Cellu-
lar Overlays,” Proc. IEEE ICUPC, 1994, pp. 397–401.
“Harmonized Global 3G (G3G),” Technical Framework for ITU IMT-2000 CDMA proposal,
OHG Technical Framework, June, 1998.
“IMT-2000: Standards Efforts of the ITU,” Special Issue, IEEE Personal Communications 4(4),
August 1977.
Manji, S., and Zhuang, W., “Capacity Analysis of an Integrated Voice/Data DS-CDMA Net-
work,” Proceedings of IEEE International Conference of Communications, 1997, pp. 979–83.
Markoulidakis, J. G., et al., “Mobility Modeling in Third-Generation Mobile Telecommunica-
tions Systems,” IEEE Personal Communications, August 1997, pp. 41–56.
McClelland, S., “Europe’s Wireless Future,” Microwave Journal 42(9), September 1999,
pp. 78–108.
Mende, W., “On the Hand-Over Rate in Future Cellular Systems,” IEEE VTC, 1998, pp. 358–61.
Naguib, A., and Paulraj, A., “Performance of CDMA Cellular Networks with Base Station
Antenna Arrays,” Proc. International Zurich Seminar on Digital Communications, Zurich, Swit-
zerland, March 1994, pp. 87–100.
Naguib, A., and Paulraj, A., “Performance of Wireless CDMA with M-ary Orthogonal Modula-
tion and Cell Site Antenna Arrays,” IEEE Journal on Selected Areas in Communications 14(9),
December 1996, pp. 1770–83.
Apx. C • Additional References 415

Naguib, A., Paulraj, A., and T. Kailath, “Capacity Improvement with Base Station Antenna
Arrays in Cellular CDMA,” IEEE, Trans. on VT, VT-43(3), August 1994, pp. 691–4.
Nanda, S., “Teletraffic Models for Urban and Suburban Microcells: Cell Sizes and Handoff
Rates,” IEEE Trans. Vehicular Tech. 42(4), November 1993, pp 673–82.
Nikula, E., et al., “FRAMES Multiple Access for UMTS and IMT-2000,” IEEE Personal Com-
munications 5(2), April 1998, pp. 16–24.
Nobelen, R. V., et al., “An Adaptive Radio Link Protocol with Enhanced Data Rates for GSM
Evolution,” IEEE Personal Communications 6(1), February 1999, pp. 54–63.
Ojanpera, Tero, and Prasad, R., “An Overview of Third-Generation Wireless Personal Commu-
nications: A European Perspective,” IEEE Personal Communications 5 (6), December 1998,
pp. 59–65.
Ojanpera, Tero, et al., “Comparison of Multiple Access Schemes for UMTS,” Proceedings of
IEEE VTC, May 1997.
Ojanpera, Tero, et al., “Analysis of CDMA and TDMA for Third Generation Mobile Radio Sys-
tems,” Proceedings of IEEE VTC, May 1997.
Padgett, J. E., Gunther, C. G., and Hattori, T., “Overview of Wireless Personal Communica-
tions,” IEEE Communications Magazine 33(1), January 1995, pp. 28–41.
Pollini, G. P., “Trends in Handover Design,” IEEE Communications Magazine, March 1996, pp.
82–90.
Prasad, N. R., “GSM Evolution Towards Third Generation UMTS/IMT-2000,” 1999 IEEE Inter-
national Conference on Personal Wireless Communications, Jaipur, India, February 1999.
Proakis, J. G., Digital Communications, McGraw-Hill, New York, 1989.
Skold, J., et al., “Cellular Evolution into Wideband Services,” Proceedings of IEEE VTC, May
1997.
Soroushnejad, M., and Geraniotis E., “Multi-Access Strategies for an Integrated Voice/Data
CDMA Packet Radio Network,” IEEE Transaction of Communications 43, February 1995,
pp. 934–45.
Suard, B., Naguib, A., and Paulraj, A., Performance Analysis of CDMA Mobile Communication
Systems Using Antenna Arrays, Proceeding IC-ASSP’93, Volume VI, Minneapolis, MN, April
1994, pp. 153–56.
Tripathi, N. D., Reed, J. H., and Vanlandingham, H. F., “Handoff in Cellular Systems,” IEEE
Personal Communications 5(6), December 1998, pp. 26–37.
Viterbi, A. J., and Padovani, R., “Implications of Mobile Cellular CDMA,” IEEE Communica-
tion Magazine 30(12), December 1992, pp. 38–41.
Viterbi, A. J., et al., “Soft Handoff Extends CDMA Cell Coverage and Increases Reverse Link
Capacity,” IEEE JSAC 12(8), October 1994, pp. 1281–87.
416 Apx. C • Additional References

Winters, J. H., “Signal Acquisition and Tracking with Adaptive Arrays in Digital Mobile Radio
System IS-54 with Flat Fading, IEEE Trans. VT 2(4), July 1993, pp. 377–84.
Wong, D., “Soft Handoff in CD,” IEEE Personal Communications,” December 1997, pp. 6–17.
Wu, J. T.-H., and Geraniotis, E., “Power Control in Multi-Media CDMA Networks,” Proceeding
IEEE Vehicular Technology Conference, 1995, pp. 789–93.
Zou, J., and Bhargava, V. K., “On Soft Handoff, Erlang Capacity and Service Quality of a
CDMA Cellular System: Reverse Link Analysis,” IEEE International Symp. Personal, Indoor,
Mobile Radio Communications, Toronto, Canada, September 1995, pp. 603–7.
Index

Analysis by Synthesis, 46
A
Antenna Gain, 73–76
A-interface, 82, 156–167 Asynchronous Data and Group 3 FAX, 334–338
A-key, 218 Authentication, 149–150
Access Channel, 92, 96, 111, 118–122 Center, 82
Modulation Parameters, 100 Challenge Response Message, 118–119, 127,
Access Methods, 325–329 131
Fixed-Assignment, 325–326 MS Data Bursts, 222
Random, 326–329 MS Origination, 220–221
Access Parameters Message, 117 MS Registration, 219
Access Technologies, 5–7 MS Terminations, 221
Acknowledgement procedure, 140 TMSI Assignment, 222–223
Adaptive Antenna Array, 77 Unique Challenge Response Procedure, 219–
Adaptive Differential Pulse Code Modulation, 43– 220
44 Autocorrelation, 29–30
Adaptive Predictive Coding, 45 Automatic Repeat Request, 154
Additive White Gaussian Noise, 11
Advanced Mobile Phone System, 103, 105–106
Advanced Research Project Agency, 325 B
Alliance for Telecommunications Industry
Association, 83 Bandpass Filter, 13, 16
Analog System to Digital System Transition, 255– Base Station, 6, 81, 159
268 Application Part, 157
Integrated Design, 256–257 Controller, 81
Overlay Design, 255–256 Management Application Part, 83, 157, 163–
Partial CDMA Coverage, 257–261 166
Advanced Mobile Phone System, 103, 105–106 Base Transceiver System, 81

417
418 Index

Bearer Control Agent Function, 87 Logical Channels, 362–364

Bearer Control Function For Radio Bearer, 85 Logical to Physical Channels Mapping, 388–
Binary Phase Shift Keying, 14–15 393
Bit Error Rate Physical Layer, 364
Binary Phase Shift Keying, 15 Reverse Link Characteristics, 385–386
Quadrature Phase Shift Keying, 17 Reverse Link Modulation, 387–388
Bit Repetition, 109 Reverse Physical Channels, 379–385
Bit Scrambling, 18–20 cdma2000 and W-CDMA Differences,
Blank and Burst, 111–112 396–398
Block Interleaving, 110, 113, 116, 119, 123, 128 Cell Identity, 160
Channel Coding, 50, 111
Code Division Multiple Access, 1–4, 8, 91–93,
C 100, 103–111
Code-Excited Linear Predictor, 41, 46–48
Call Control Agent Function, 89 Coherence Bandwidth, 245–247
Enhanced, 87 Combining Methods, 63
Call Control Function (Enhanced), 85 Equal Gain, 67–70
Call Processing, 160–161 Maximal Ratio, 65–67
CDMA Call Processing State, 133–145 Selection, 63–65
Idle, 137–142 Comparison Threshold, 187
Mobile Control on Traffic Channel, 143–145 Connectionless Interworking Protocol, 155
System Access, 142–143 Convolutional Codes, 52
System Initialization, 135–137 Encoder, 113, 116, 119, 123–124, 128–129
CDMA Cell Size, 300–302
CDMA Channel Number, 108
CDMA Development Group, 1, 50 D
CDMA Registration, 145–149
Distance-based, 147 Data Burst Randomizer, 128–129
Implicit, 149 Data Communications Services, 314–315
Parameter Change, 148–149 Data Services in IS-95, 333–334
Power Down, 147 Delay Spread, 243–247
Power Up, 146 Design Consideration, CDMA System Boundary,
Timer-based, 147 280
Traffic Channel, 149 Dim and Burst, 111–112
Zone-based, 147–148 Direct Application Part, 83
CDMA System Direct Sequence Spread Spectrum, 9–10
Performance, 20–23 Performance, 14–17
cdmaOne to cdma2000 Evolution, 393–396 Direct Transfer Application Part, 157–158, 162–
cdma2000 System, 353–356 163, 165
Channels, 362 Diversity
Data Services, 387–388 BPSK Modulation, 71–72
Forward Link Features, 373–379 Macro, 58–59
Forward Physical Channels, 364–373 Micro, 59–60
Layering Structure, 356–362 Reception, 57–58
Index 419

Doppler Spread, 247–249

Drop Timer Threshold, 188 G
Dual Mode CDMA Mobiles, 253–255 Global Positioning System, 94
Global System of Mobile communications (GSM),
1–2
E

Effective Interference Power, 18

Energy Per Bit, 13, 15, 16–18 H
Energy Per Bit to Noise Ratio, 13, 15, 17–18,
21–24 Handoff, 171–176
Enhanced Variable Rate Codec, 49–51, 55 Messages, 188–190
Equipment Identity Register, 82 Parameters, 187
Erlang Capacity of a Single Cell, 230–231 Procedures, 190–192
Error Control Schemes, 330–333 Home Location Register, 81, 167–170
Go-Back-N, 331
Selective Repeat Protocol, 331
Stop and Wait, 330–331
I
Variable Window and Frame, 332
WORM, 331–332 Intellectual Property Rights, 4
Estimation of Cell Count, 230–231 Intercell Interference, 291–292
European Telecommunications Standards Interfrequency Handoff, 280–281
Institution, 4 Intermediate Frequency, 11
External Networks, 82 International Mobile Telecommunications-2000,
3–4
International Telecommunication Union, 79
F Internet Protocol, 155
Intersymbol Interference, 249
Facilities Engineering, 268–279
Interworking Function, 82
Forward Link, 104
Budgets, 305–310
Capacity, 294–300
Forward Traffic Channels, 93–94, 122–128 L
Modulation Parameters, 95–96
Frequency Band, 305–310 Layered Structure, 153–155
Band Class 0, 106 Line Spectral Pairs, 46–48
Band Class 1, 107 Linear Pass Filter, 109
Frequency Division Duplex, 7, 91 Linear Predictive Coding, 45, 48
Frequency Division Multiple Access, 4–5, 7 Link Access Control, 154–155
Frequency Hopped Systems, 9–10 Link Balance
Fundamental Code Channels, 92 Forward and Reverse, 302–305
Forward, 93 Link Budget and Cell Coverage, 249–253
Reverse, 96 Location Area Code, 160
Future Land Public Mobile Telephone Systems, 3 Logical Channels, 111–132
420 Index

M O

Mac Sublayer, 154–155 Offset Quadrature Phase-Shift Keying, 99, 109

Maximal Length Pseudorandom Sequences, Operations System, 82
27–29 Orthogonal Functions, 31–36
Message Transfer Part, 156 OSI Upper Layers, 315–316
Mobile Antennas, 72–73 Over-the-Air Service Provisioning, 176–178
Corner Reflector, 76
Quarter Wave Vertical, 73–74
Smart, 76–77 P
Stacked Dipoles, 75
Packet Data Service for CDMA Cellular/PCS
Mobile Broadband Systems, 3
Systems, 342–351
Mobile Control Function, 87
Applicable Mobile Type, 345
Mobile Country Code, 160
Features, 348–351
Mobile Identification Parameters, 213–217
Network Elements, 342–343
Access Overload Class, 216
Network Reference Model, 342
Directory Number, 214
Network Reference Points, 343
Discontinuous Reception, 217
Protocol Options, 343–344
Electronic Serial Number, 214
Protocol States, 345–348
Home System and Network Identification,
Paging Channel, 92, 94, 111, 115–118
216
Modulation Parameters, 95
Identification Number, 213–214
Parameter Update, 225
Local Control Plan, 216
Personal Communications Services, 1, 91
Preferred Operation Selection, 217
Physical Channel, 103–108
Registration Memory, 215
Pilot Channel, 92, 94, 96, 111–113
Station Class Mark, 215
Pilot Detection Threshold, 187
Mobile Network Code, 160 Pilot Drop Threshold, 188
Mobile Radio Resource Control, 88 Pilot Sets, 195–196
Mobile Radio Transmission and Reception, Power Control, 111, 123–124
88 Forward Link, 206–210
Mobile Station, 6, 81 Need, 196–197
Mobile Switching Center, 81, 159 Reverse Link, 198–206
Mobility Management, 162 Preferred Set of Frequency, 108
Modes of Operation in Wireless Communications, Primary CDMA Carriers, 105
7 Propagation Models, 232–243
Modulation, 109 Correction Factor for Attenuation Due to
Multicell Network, 290–291 Trees, 237–238
Hata-Okumura, 233–235
IMT-2000 Models, 242–243
N Indoor Environment, 238–242
Outside Environment, 232–243
Network Identification, 114 Walfisch-Ikegami, 235–237
Index 421

Pseudorandom Noise Sequences, 24–31 Service Data Function, 87

Pulse Code Modulation, 41–44, 48, 159 Service Management Function, 89
Service Switching Function, 87
Service Switching Function/Call Control Function,
Q 84
Shared Secret Data, 223–225
QCELP, 159 Short Message Service, 155, 338–342
Quadrature Phase-Shift Keying, 14–17, 94 Signal-to-Noise Ratio, 5, 12
Quality of Service, 154 Signaling Connection Control Part, 156–157, 164–
167
Soft Handoff, 181–184
R Benefit, 184
Forward Link, 184
Radio Access Control Function, 88 Reverse Link, 184
Radio Control Function and Radio Access Control Set Up and End, 192–195
Function, 83, 89 Softer Handoff, 184
Radio Coverage Planning, 231–232 Specialized Resource Function, 87
Radio Design for a Cellular/PCS Network, 229 Specialized Resource Functional Element and
Radio Frequency, 11 Data Interworking, 85
Radio Frequency Transmission and Reception, 87 Spread Spectrum, 7
Radio Link Design, 230 Concept, 11
Radio Link Protocol, 154–155 Supplemental Code Channel
Radio Network Planning, 229–230 Forward, 93
Radio Resource Control, 87 Reverse, 96
Radio Resource Management, 162 Supplementary Services, 160–161
Radio Terminal Function, 83, 88 Symbol Repetition, 113, 116, 119, 123–124, 128
RAKE Receiver, 60–62, 93 Sync Channel, 93, 111, 113–115
Rate Set 1, 109–111, 123, 126, 130 Modulation Parameters, 95
Rate Set 2, 109, 124, 126 System Identification, 114
Relaxed Code-Excited Linear Prediction, 50 System Processing Gain, 12
Residual-Excited Linear Prediction, 45–46
Reverse Link, 96–100, 104
Capacity, 284–290 T
Reverse Traffic Channels, 96–97, 126–132
Modulation Parameters, 96–97 Telecommunication Research and Action Center, 1
Terminal Access Control Agent Function, 87
Terminal Access Control Function, 87
S Terminal Identification Management Function, 87
Third Generation Systems, 3–4
Search Windows, 185–187 Time Division Duplex, 7
Secondary CDMA Carriers, 105 Time Hopped System, 10–11
Service Access Control Function, 86 Time Division Multiplex, 6
Service Control Function, 86 Time Division Multiplex Access, 1–4, 6–7, 103
422 Index

Traffic Channel, 92
Forward, 93, 95–96 W
Reverse, 97–98 Walsh Codes, 112–113, 116, 123–124
Transmission Control Protocol, 155 Walsh Functions, 34–37, 91
Transmission Facilities Management, 162 Wideband CDMA, 4
TR-45/TR-46 Reference Model, 79–83 Wireless Data Systems, 316
Functional Model, 83–85 Local Area Networks, 317–321
Wide Area, 3, 316–317
Wireless Local Area Standard, 321–325
U High Performance Radio
IEEE 802.11, 321–324
User Data Protocol, 155
Local Area Network, 324–325
User Identification Management Function, 87
Wireless Information Network Forum, 324

Very Large Scale Integrated Digital Circuits, 5

Visited Location Register, 82
Voice Privacy, 225–227
About the Author

Dr. Vijay Garg is an adjunct professor in the Electrical

Engineering and Computer Science department at the
University of Illinois at Chicago, where he teaches
courses in wireless communications and wireless net-
working. Dr. Garg is also a Distinguished Member of
Technical Staff at the Bell Labs of Lucent Technologies
in Naperville, Illinois. Dr. Garg received his Ph.D. from
the Illinois Institute of Technology, Chicago, in 1973
and his MS from the University of California at Berke-
ley, California, in 1966. Dr. Garg has co-authored sev-
eral technical books, including four in
Telecommunications; one in Rigid Body Dynamics;
and one in Railway Vehicle Dynamics, which has been
translated into Russian and Chinese. He is a Fellow of ASCE and ASME and a Senior Member
of IEEE. Dr. Garg is a Registered Professional Engineer in the states of Maine and Illinois. He is
an Academic Member of the Russian Academy of Transport. He has served on various technical
committees of IEEE and ASME, and he is a feature editor for a PCS series in IEEE Communica-
tion Magazine.

423
 P R E N T I C E H A L L

Professional Technical Reference

Tomorrow’s Solutions for Today’s Professionals.

Keep Up-to-Date with

www.phptr.c

PH PTR Online!
We strive to stay on the cutting-edge of what’s happening in
professional computer science and engineering. Here’s a bit of what
you’ll find when you stop by www.phptr.com:

Special interest areas offering our latest books, book series, software,
features of the month, related links and other useful information to
help you get the job done.

Deals, deals, deals! Come to our promotions section for the latest
bargains offered to you exclusively from our retailers.

Need to find a bookstore? Chances are, there’s a bookseller near you

that carries a broad selection of PTR titles. Locate a Magnet bookstore
near you at www.phptr.com.

What’s New at PH PTR? We don’t just publish books for the professional
community, we’re a part of it. Check out our convention schedule, join
an author chat, get the latest reviews and press releases on topics of
interest to you.

Subscribe Today! Join PH PTR’s monthly email newsletter!

Want to be kept up-to-date on your area of interest? Choose a targeted

category on our website, and we’ll keep you informed of the latest PH PTR
products, author events, reviews and conferences in your interest area.
Visit our mailroom to subscribe today! http://www.phptr.com/mail_lists