2014 International Conference on Circuits, Systems, Communication and Information Technology Applications (CSCITA)

Adaptive Modulation and Coding for Mobile-WiMAX

Using SDR in GNU Radio
B. Siva Kumar Reddy and B. Lakshmi
Research Scholar and Associate Professor,
Department of Electronics and Communication Engineering,
National Institute of Technology Warangal, Andhra Pradesh-506004, India
E-mail: [email protected], [email protected], www.nitw.ac.in.

Abstract-The demand for high data rates and a single transmission signal is separated into numerous
low bit error rates has prompted the ascent of narrowband signals [5]. The point of interest of OFDM
developing remote innovations, such as WiMAX. over a single carrier systems is its capacity to endure the
Adaptive Modulation and Coding (AMC) plans adverse and harsh channel conditions. OFDM addition-
when joined into WiMAX helps the framework ally reduces fading effect and increments the data rate
to alertly respond to the channel varieties giv- in transmission. The advantage of multiple access in Or-
ing higher information rates and better BER ex- thogonal Frequency Division Multiple Access (OFDMA)
ecution. Switching the modulation and coding is its ability to endure multipath interference.
rates dependent upon channel state information In current communication systems, the utilization of
permits to attain higher throughput and better AMC (Adaptive modulation and Coding) techniques [5]
spectral efficiency. This paper deals with pro- empowers the framework to realize high spectral effi-
posal of efficient adaptive modulation coding sys- ciency. The core idea of AMC is to dynamically change
tem which maximizes the throughput while up- the modulation and coding plans dependent upon the
holding a target bit error rate (BER). Software channel facts in order to enhance overall spectral effi-
Defined Radio (SDR) idea has been utilized to do ciency. The receiver chooses the modulation and coding
most of the signal processing over a software than scheme (MCS) [6] to be utilized on channel state infor-
in hardware. It proposes to select a specific mod- mation and reaction to the transmitter. In this paper,
ulation coding schemes in addition to the relating two AMC systems are examined. The two strategies
coding rates dependent upon the threshold given. point at switching the MCS dependent upon target BER
Hence, a selector block is created in GNU Ra- and most maximum throughput approach [5]. In target
dio companion using Python script. This block BER technique, the modulation and coding schemes are
switches to proper modulation and coding rates selected with an objective to maintain the bit error rate
depends upon the thresholds considered. This below a target BER value, using a set of threshold val-
paper investigations the Mobile-WiMAX frame- ues, the estimated signal to noise ratio is partitioned into
work regarding BER and throughput by consid- a set of regions. Each such region is associated with a
ering SNR and channel attenuation factor as the different MCS such that the bit error rate is maintained
threshold qualities. Obtained results derived a below a target BER. Every such area is connected with
trade-off between BER and throughput. an alternate MCS such that the bit error rate is main-
keywords: AMC, BER, MCS, OFDM, SDR, tained beneath a target BER. An elective AMC strategy
WiMAX. is to switch the MCS keeping in mind the end goal to
maximize the throughput. This is called most Maxi-
mum Throughput (MT) procedure. In most maximum
1 INTRODUCTION throughput approach, less accentuation is given to sup-
port a specified BER [7].
Software Defined Radio (SDR) is a radio communica-
Channel coding is the technique to make the transmit-
tion framework in which the signal handling blocks are
ted signals less vulnerable to errors. Channel coding in-
executed in programming in place of a hardware [1]. Uti-
cludes scrambling, Forward Error Correction (FEC), and
lizing USRP (Universal Software Radio Peripheral) [2] as
interleaving. This paper uses convolution coding system
equipment, we can transmit and receive real time signals.
of forward error correction with code rates of 1/2 and
In between the transmitter and receiver, the signal pro-
3/4.
cessing can be performed using GNU Radio Companion
software [3].
The IEEE 802.16 (for WiMAX) [4] standard employs 2 ORTHOGONAL FREQUENCY DIVISION
Orthogonal Frequency Division Multiplexing (OFDM) MULTIPLEXING
to realize high data rates. Utilization of OFDM in
WiMAX eliminates inter symbol interference (ISI) and OFDM is a blend of modulation and multiplexing. Mul-
inter carries interference impacts in multipath transmis- tiplexing refers to independent signals, those processed
sion. OFDM is a strategy for advanced balance in which by diverse sources. As indicated in Fig. 1, multi-carrier

978-1-4799-2494-3/14/$31.00 ©2014 IEEE 173

 2014 International Conference on Circuits, Systems, Communication and Information Technology Applications (CSCITA)

modulation separates the wideband high data rate R algorithm, intends to admiration a certain target Bler
bps approaching data stream and with a passband data (Block Error Rate) because of particular Qos (i.e., data
transmission B/L, each of which is then transmitted integrity) prerequisites. The target quality of the Bler
over an alternate orthogonal-frequency sub channel. The might be the same for every SNR (i.e., the mean signal-
to-noise ratio averaged over the frame length), or can
change with it.

3.1.1 MT algorithm
The thresholds needed for the MCS determination for
the MT methodology are indicated in Table 1. The
MT algorithm seeks to maximize the generally inter-
face throughput by picking for every SNR worth the
MCS that permits the most extreme throughput. Subse-
quently, the the transmission efficiency is improved how-
ever the execution as far as error rate could be worse, no
Figure 1: A basic multicarrier system constraint on the maximum acceptable error probability
is introduced. The goal is to pick, for every threshold and
data rate of WiMAX is given by [4], SNR worth, the balance request that expand the sum
throughput. The last outcome is that the most effective
B.Ld .log2 (M )
R= (1) plan as far as throughput for every α quality is chosen.
L(1 + G) This algorithm has no level of opportunity and thresh-
In OFDM-WiMAX PHY layer, if 16-QAM is used as olds could be considered by looking at the throughput
of the recognized transmission schemes. Towards this
modulation scheme, then the data rate is as follows end, the traditional throughput definition gave in ref. [5]
has been altered here to enhanced match the opening
107 M Hz.768.log2 (16)
R= = 24M bps (2) structure of the proposed framework: the throughput
(1024)(1.125) (η) could be communicated as a capacity of the channel
In expressions, every data convey subcarriers of data constriction calculate α and utilized MCS for a settled
SNR as
transmission B/L conveys bits of information. An ex-
tra overhead punishment of (1 + G) must be paid for the 4.(1 − Pblock (α, SN R, M, Rc ))Nc .Nsd log2 M.Rc )
cyclic prefix, since it comprises of repetitive data and η=
4.Tf
tributes the transmission of actual data symbols. (3)
where M= 4,16,64 represents the modulation order,
3 OFDMA- MODULATION SWITCHING pblock stands for the block error probability related to
each modulationcoding rate couple, nc is the amount of
In this section, the proposal of AMC (Adaptive Modu-
subcarriers for every user,nsd is the amount of OFDM
lation and Coding) frameworks exploiting the OFDMA
symbols forming the Downlink OFDMA subframe, rc
abilities is portrayed considering that the regulation re-
is the received coding rate and tf is the span of the
quest and coding rate of every complex symbol ck asso-
OFDMA outline.
ciated to the comparing k-th subcarrier (0 ≤ k ≤ 1023)
The symbol error probability psym in a Rayleigh chan-
could be changed as per the physical channel state and
nel can be expressed as follows [5]:
the MCS is changed likewise for all the subcarriers of
the slot. The selection of the modulation is dependent
upon the channel parameters. In this paper, the Channel √ s
4( M − 1) 3.SN R.α2
state methodology is acknowledged as Adaptive Modu- PSN R,α,M = √ .Q( )
lation techniques. M M −1
√ s (4)
2 M −1 3.SN R.α2
3.1 Channel state technique −[ √ Q( ]
M M −1
In this section two algorithms named Maximum
Throughput (MT) and Target Block error rate (Tbler) Both pb,M QAM and psym,M QAM are functions of SNR,
dependent upon the channel state methodology are pro- and modulation order (M);
posed. Every adaptation algorithm is essentially char-
acterised by five thresholds. The MCS (Modulation and
3.1.2 TBLER Technique
Coding Scheme) remains same until the measured pa-
rameter (SNR, BER, throughput or channel attenuation The thresholds needed for the MCS determination for
factor) overcomes a threshold [7]. In this paper, channel the Tbler approach are demonstrated in Table 2. The
attenuation factor is recognized as channel state param- fundamental characteristic of the TBLER procedure is
eter. The Maximum Throughput (MT) algorithm plans that of keeping the error rate beneath a target limit,
to maximize the system throughput without demands on subsequently ensuring a fixed level of QoS as far as P-
target Block Error Probability. The Target Bler (Tbler) block [5]. Therefore this algorithm is more suitable to

174
 2014 International Conference on Circuits, Systems, Communication and Information Technology Applications (CSCITA)

way, with a specific end goal to get a code rate of 3/4,

Table 1: MCS selection for the MT approach.
the data must be first encoded utilizing an essential en-
Parameter Value Modulation Scheme coder and after that transmit each first and third bit in
K < 0.09 QPSK 1/2 the first extension and each first and second bit in the
0.09 < K ≤ 0.2 QPSK 3/4 second branch. For decoding purpose, Viterbi decoder
0.2 < K ≤ 0.33 16QAM 1/2 is utilized.
0.33 < K ≤ 0.51 16QAM 3/4
0.51 < K ≤ 0.57 64QAM 1/2
0.57 < K 64QAM 3/4
5 SELECTOR BLOCK

The selector block has been designed in python scripting

language. The same block has additionally been inte-
services asking for a certain maximum error rate value
grated with the essential OFDM framework in Gnu Ra-
as best exertion traffics. This strategy has single level
dio programming that incorporates an encoder-decoder
of flexibility, constituted by the infringed target BLER;
block for choosing the coding rate. It chooses the mod-
diverse qualities of the target BLER compare to distinc-
ulation scheme of different coding rates consistent with
tive framework execution. For every quality of SNR we
the methodologies of Rayleigh channel blurring [8]. The
can determine BLER as a capacity of α. Subsequently,
methodologies, as talked about in the recent past, are
we can definitive target α qualities (i.e., thresholds) in
Maximum throughput approach and target Bler ap-
correspondence of which state moves for the modula-
proach. This block expedites facilitates in choice of the
tion switching process occur. The α values guarantee
modulation schemes of distinctive code rates being uti-
the target BLER, for a certain SNR and a certain MCS.
lized as a part of OFDM.
Thresholds could be inferred by selecting the α worth as
Consequently, there is no further need to physically
per the specified Target BLER.
select the modulation scheme and after that run the
.grc file in GNU radio companion programming. We can
Table 2: MCS selection for the TBLER approach. straight run the project in which the block has recently
Parameter Value Modulation Scheme been incorporated and after that as per the parameter, it
K < 0.46 QPSK 1/2 will immediately select the modulation scheme and give
us the modulated output. The calculation being utilized
0.46 <K ≤ 0.65 QPSK 3/4
is indicated as a part of Fig. 2.
0.65 <K ≤0.95 16QAM 1/2
0.95 <K ≤ 1.38 16QAM 3/4
1.38 <K ≤ 1.95 64QAM 1/2
1.95 <K 64QAM 3/4

4 CONVOLUTION CODING

Convolution codes are utilized as a part of requisitions

that require good performance and low usage cost. They
work on information stream and not static block. It in-
cludes expansion of equality bits to decrease the suscepti-
bility to errors in the transmitted signals. Convolutional
coding is incorporated into the WiMAX framework for
reliable data transfer and WiMAX supports coding rates
of 1/2 and 3/4 for encoding the information. In AMC
system, coding rate is also switched as well as modu-
lation method. Switching the coding rate strategy as
per the channel conditions empowers dependable data Figure 2: MCS selection strategy
transmission. In unreliable channel conditions, coding
rate 1/2 is utilized for encoding the information, while After the execution of the above restrictive stream ex-
in dependable channel conditions coding rate 3/4 could planations, a modulation scheme with suitable coding
be utilized. rate might have been chosen and the same could be uti-
Convolution codes are normally specified by three pa- lized as a part of the OFDM framework.
rameters (n,k,m); n = number of yield bits, k = number
of information bits, m = number of memory registers. 6 SIMULATION RESULTS
The amount k/n is called code rate. The code rate
runs from 1/8 to 7/8. The amount L is known as the The needed OFDM parameters for WiMAX [4] setup
stipulation length of the code and is characterized by are demonstrated in the Table 3. OFDM could be ac-
L = k(m − 1). The code rate of 1/2 is produced with tualized by utilizing an open source programming GNU
the requirement length of 7. The code rate of 3/4 is Radio companion as indicated in the Fig. 3. OFDM pa-
produced by puncturing the information. The punctur- rameters are designed consistent with the Table. 3 and
ing matrix for coding rate of 3/4 is [(101)(110)]. In this executed for 1/2 64QAM waveform in the Figs 4,5,6 and

175
 2014 International Conference on Circuits, Systems, Communication and Information Technology Applications (CSCITA)

7. In this paper, a selector block has been made utilizing

Python scripting for the automatic selection of modula-
tion scheme with different coding rates in OFDM block.

Table 3: Simulation parameters defined.

Parameters Values
FFT size (NFFT) 1024
Occupied Tones 840
Sampling rate 10.66667 M
Centre Frequency 2.48 GHz
Convolutional Code 1/2 Figure 6: 64QAM 3/4 output by manual selection using
Cyclic Prefix length 184 TBLER approach.
Useful symbol duration 91.43 µs
Carrier spacing (1/Tu ) 10.94 KHz
Guard time (Tg=(1/4)* Tu) 11.43 µs
OFDM symbol duration 102.86 µ s
Digital Modulations BPSK, QPSK,
16QAM, 64QAM

The modulation coding plans are chosen consistent

with the MT methodology are accounted for in Table 1.
With respect to the MT Technique, at whatever point
a dissimilar propagation environment is conceived, is di-
rect to note that, because of different performance in
terms of BLER, the thresholds have different values. The
coming about MCSs are identified switching thresholds Figure 7: 64QAM 3/4 output by automatic switching
as per the TBLER approach as indicated in Table 2. using TBLER approach.
The particular modulation scheme has been chosen de-
pendent upon two methodologies.
The effects were likewise confirmed for their BER [7]
qualities. The BER got for the coding rate of 1/2 is
0.6138883809 units and the BER acquired for the coding
rate of 3/4 is 0.7769923210 units. The above outcomes
for the bit error rate qualities likewise demonstrates that
as the coding rate is increasing the BER value is addi-
tionally expanding. Along these lines we can say that,
we enhance the error susceptibility of the signal with the
trade-off of the Bit Error Rate.
The Fig. 8 is a plot of BER Vs. SNR for differ-
ent modulation and coding procedures. Balance strate-
gies QPSK, 16QAM, 64QAM are utilized for reproduc-
Figure 4: 64 QAM 1/2 output obtained by manual se- tion. The information is initially encoded utilizing cod-
lection using MT approach. ing rates 1/2 and 3/4 and afterwards transmitted uti-
lizing a suitable modulation technique. From the Fig.
8, it could be gathered that the BER rate diminishes
with increase in SNR. Likewise the BER execution of
QPSK is superior to what that of 64QAM. As the order
of modulation expands the BER execution diminishes.
Subsequently in an extremely boisterous channel QPSK
regulation must be utilized. In a channel with low noise
a regulation with higher request might be utilized. From
the Fig.8 it is observed that the coding rate 1/2 has an
improved BER execution than the coding rate 3/4. So
in a channel prone to high interference and noise coding
rate 1/2 must be utilized within request to recoup the
transmitted indicator at the collector with least mistake.
In a channel with interference effects and low noise the
Figure 5: 64 QAM 1/2 output obtained by automatic coding rate 3/4 is suitable for encoding the information
switching using MT approach. before transmission.
So in AMC we combine different modulation and cod-

176
 2014 International Conference on Circuits, Systems, Communication and Information Technology Applications (CSCITA)

Figure 3: GNU schematic for 1/2 64QAM implementation with WiMAX physical layer

In the Fig. 9, the MCS is switched as per the divert

conditions with a specific end goal to maintain a BER
threshold of 0.001. For different SNR qualities the MCS
is exchanged to get a BER beneath the threshold level.
This method of restricting the BER to a certain thresh-
old is called Target BER system. Here much stress is
laid on restricting the BER and there is a slight com-
promise in the throughput of the framework. This is an
outline exchange off and AMC tries to accomplish an
equalization between throughput and BER execution of
the framework. At low SNR qualities QPSK 1/2 is uti-
lized for transmission. At high SNR where there is low
noise 64QAM with coding rate 3/4 is utilized. For SNR
Figure 8: BER Vs. SNR characteristics of different MCS going from 10 dB to 11 dB, QPSK 1/2 is utilized and
schemes. for SNR running from 12 to 15dB 16QAM with coding
rate is utilized for transmission. 16QAM with coding
rate is utilized to transmit the signal for SNR qualities
ing schemes to realize better BER execution [7]. Ex-
going from 15 db to 20 db [9]. For SNR qualities more
changing the MCS as per the channel conditions ex-
greater than 20 db 64-QAM with coding rate of 3/4 is
pands the BER execution and guarantees dependable
utilized. The partitioning of the SNR qualities into nu-
data transmission with high data rates. Joining to-
merous districts hinges on the limit BER level. For ev-
gether the modulation and coding strategy it is observed
ery district an alternate MCS is utilized to utmost the
that QPSK 1/2 has the best BER execution around the
BER. To acquire an alternate limit of BER partitioning
MCS utilized. In traditional communication systems the
of the SNR values must be done again in order to limit
framework was intended for worst case channel condi-
the overall BER to the desired threshold. In Fig. 10,
tions. Designing the system as per the most exceedingly
the throughput qualities as capacity of SNR for target
awful case channel conditions break down framework
BER of 0.01 is indicated. Assuming that include infor-
execution. Subsequently AMC is acquainted in frame-
mation streams are transmitted utilizing 64QAM with
work plan with increment framework execution. So in
coding rate of 3/4, the throughput will be augmented
AMC the MCS is exchanged as per the channel condi-
yet there will be lower BER execution. The through-
tions consequently enhancing the framework execution
put attributes of AMC system for target BER of 0.001
and prompts ideal use of communication resources.
are demonstrated in Fig. 11. Throughput execution has
been thought about for Target BER 0.01 and 0.001, in-
dicated in Fig. 13.

7 CONCLUSIONS

This paper Adaptive Modulation and Coding technique

has been integrated with OFDMA frameworks. The se-
lector block has been produced in python scripting lan-
guage. The designed selector block immediately picks
the modulation schemes with the suitable coding rate
dependent upon the two methodologies, such as Max-
imum Throughput (MT) and TBLER (Target BLER)
channel estimation strategies. The outcomes acquired
Figure 9: Performance of Target BER of 0.001 from the selector block have been compared and ana-

177
 2014 International Conference on Circuits, Systems, Communication and Information Technology Applications (CSCITA)

lyzed. Whichever, condition gets fulfilled, the relating

Modulation scheme gets chose and the OFDM frame-
work is executed.
The BER execution and throughput attributes are uti-
lized as a measure for AMC execution and investigated
the procedure of Target BER for AMC conspires by plot-
ting BER vs. SNR attributes. From simulations, one can
see that, at more level SNR value the QPSK modulation
with 1/2 convolutional coding modulation is preferred
over others as far as BER in light of the fact that it
gives the most minimal BER however the throughput or
spectral efficiency is very low. Additionally, the through-
put of 64QAM with 3/4 convolutional coding is higher
than other MCS but experiencing highest BER. We can
infer that spectral efficiency of AMC scheme with target
BER 0.01 is higher than that of AMC scheme with target
BER 0.001. So we can conclude that there is a trade-off
Figure 10: Throughput performance of AMC system un-
between BER performance and spectral efficiency.
der Target-BER 0.01.

References
[1] Chen, Chi-Yuan and Tseng, Fan-Hsun and Chang,
Kai-Di and Chao, Han-Chieh and Chen, Jiann-
Liang and others, ”Reconfigurable software defined
radio and its applications” Tamkang Journal of
Science and Engineering vol 13, 2010.

[2] Ettus, Matt, USRP Users and Developers Guide,

Ettus Research LLC, 2005.
[3] GNU Radio Trac, http://gnuradio.org/trac/wiki.
[4] Andrews, Jeffrey G and Ghosh, Arunabha and
Muhamed, Rias, ”Fundamentals of WiMAX:
understanding broadband wireless networking,”
Prentice Hall PTR,2007.
0
[5] Fantacci, Romano and Marabissi, Dania and
Figure 11: Throughput performance of AMC system un- Tarchi, Daniele and Habib, Ibrahim, ”Adaptive
der Target-BER 0.001 modulation and coding techniques for OFDMA
systems,” Communications Letters, IEEE, pages
4876–4883,2009.
[6] Reddy, B Siva Kumar and Lakshmi, B ”Adaptive
Modulation and Coding in COFDM for WiMAX
Using LMS Channel Estimator,” Conference on
Advances in Communication and Control Sys-
tems,Pages 23-29, 2013.
[7] Goldfeld, Lev and Lyandres, Vladimir and Wulich,
Dov, ”Minimum BER power loading for OFDM in
fading channel,” Communications, IEEE Transac-
tions on, vol 50, pages 1729–1733, 2002.
[8] TProakis, John G and Manolakis, Dimitris
G,” Digital communications,” McGraw-hill New
York,vol 3,1995.
[9] Takeda, Daisuke and Yuk, C CHOW and Strauch,
Paul and Tsurumi, Hiroshi,” Threshold controlling
scheme for adaptive modulation and coding sys-
Figure 12: Throughput comparison of Target-BER 0.001
tem,” IEICE transactions on communications,vol
and BER 0.01.
89, pages - 1598 - 1604 ,2006.

178