HSPA+
Optimization
Process
Summary
Pre-opt start list
RET report
Feeder Loss values audit
Commissioning/SCF/FTM config audits
BFD/AAV audit
Capacity audits and report from capacity team
Cluster Identification process (planning and performance) including
asset analysis, liberty server CQI reports etc
Optimization process using KPI and drivetest.
Explain the flowchart for KPI analysis
New actix process
Using actix for KPI and drivetest analysis
OPTpcs
Change requests
1.
Process for identifying and approving the requests needs to be worked out between the local
market engineers and HSPA+ assigned resources
1.
To be approved by t-mobile employee
Result verification against the reported CQI
1.
Exit criteria
Flow of Optimization work
Flow of Optimization work contd.
Responsibilities
Region
Market
Market to submit & Region to
validate/approve
Detailed Optimization flow
Parameter Audit( If HSPA+ is
enabled)
Audit on all Parameters on RNC and Node B Level.
To make sure they are consistent with the HSPA+ MOP.
Regional tool/script can be used for Parameter Audit
RNC/OSS SME
Verification of all HSPA+ and HSUPA features
To make sure if all the required features in all sites and RNCs
implemented as planned.
List out the Parameter and Feature inconsistencies and
correct them
This step is required first before proceeding with any Optimization
activities.
Basic Capacity Check
Address the following baseband requirement based on the
national capacity guidelines
FSMx
Additional T1 or AAV
HS scheduler
RET Verification
Ensure RET commissioning is complete
RET/Actuator information are in place
Feeder Loss
Feeder Loss is an important design aspect maintain coverage, capacity & performance
Large feeder losses will reduce CPICH and the effective coverage area of the cell
- Other inconsistent values may affect Power Control & RTWP
Identify sites with more than 3dB loss, using the Feeder
Loss Dashboard link given below
On high feeder loss sites, Utilize latest sweep values, if available to
adjust OSS parameters.
If high feeder loss is accurate, identify site for possible mitigation for a
long solution as below
Feeder modification
RRU implementation
Note: These long term solutions will not be implemented during this period of optimization.
Feeder Loss Dashboard
Prioritize clusters
Prioritize clusters = (% of bad CQI) * ( HS traffic)
-1st priority = quality clusters From high to low traffic
-2nd priority = RSCP clusters From high to low traffic
Note: All problem areas in the cluster even with planned
new
sites will still need to be analyzed and optimized to its
full potential
Exceptions
- Bad CQI Sites at coverage edges
For Regions
Only
2nd Priority
1st
Priority
Define Problem Areas
Problem Area Major KPI CQI
EXIT Criteria To meet Baseline KPI
Sl.
No.
1
KPIs
CQI
Exit Threshold (Market wise)
Number of Cells with >5% in CQI 0 to
10, to reduce by 80% AND Market wise
CQI above 10 > 98%
TPIM KPIs
1
Voice Access Failure rate (%)
<= previous round value
2
PS Access Failure rate (%)
<= previous round value
3
Voice Drop Rate (%)
<= previous round value
4
PS Drop Rate (%)
<= previous round value
5
Actual MAC-HS Cell Thpt (Mbps)
>= previous round value
6
Voice Traffic in Erlangs
>= previous round value
7
HSDPA Pay Load-RLC layer (MB)
>= previous round value
8
3G Leakage (% of 3G calls ending on 2G)
<= previous round value
9
% of 3G initiating calls on 2G (CDR Based)
<= previous round value
Note: These KPIs(1-8) are available in TPIM under the content browser by running the report NSN
HSPA KPIs. TPIM (9) can be found: http://dev.reportcenter.internal.t-mobile.com/Reports/Pages/Folder.aspx
under National Performance Reports E2E 3G Leakage Reports
All entry & exit criteria should be measured on a one week average for comparison
The same criteria will be used to monitor the cluster performance during optimization
Above listed TPIM KPIs are expected to be maintained or improved
Justification needed for 8&9 if value degraded by 2% ( to be validated from trial)
Optimization KPI Reporting Only
Sl. No.
KPIs
Boomer Index - 3G coverage vs Co-loc 2G
Coverage =>1.5
DL Capacity Index - Power per User, SHO Overhead
, 500mwatts
UL Capacity Index RTWP, =>3dB @ busy hour
Boomer Index: The UMTS PRACH Delay propagation measurements are compared against the
co-located GSM cell radius. A calculation is then performed based on a weighted average
Of HHO distances between the cell and its neighbors. Any instance where the 3G cells radius
exceeds the 2G cells radius >= 1.5 is flagged
DL Capacity Index: The Average Power per user is calculated taking into account Voice
Power/RL, RT SHO & Cable loss. Any cell where the Avg Power/User is > 27dBm is flagged
UL Capacity :uplink noise rise. Busy hour minus least busy hour
Cluster/problem area Drive KPIs
Reporting Only
All KPIs below are only for Optimization purposes
1
2
3
4
5
6
HSDPA Application Throughput
HSUPA Application Throughput
HSDPA Access Failure Rate
HSDPA Drop Rate
HSUPA Access Failure Rate
HSUPA Drop Rate
% of time HS UE was served on R99 data
7 (DCH)
8 64 QAM Modulation Rate
9 UE Reported CQI
10 No. of HS Cell Changes
11 HS-PDSCH Code Usage
12 Scanner Based - Too Many Servers
Data / KPI Collection
TPIM Report / KPI Collection
Propagation Delay reports (from Liberty Server or TPIM
content browser) can be used to identify the proximity of
mobiles and cell
Utilize liberty report for optimization and analysis
SHO
PRACH
Power/user
Boomer Index
RTWP
All other KPIs (including EXIT KPIs) need to be collected and
stored for bench marking purpose.
From the content browser in TPIM use and run the
report NSN HSPA KPIs
Detailed Drive In Problem Areas
Detailed Street level Drive Test is for the optimization
purpose in the Problem Areas
Following is the drive test setup for the Problem Area
Detailed Drive
Scanner Analysis is mandatory
Device
HSPA+
Problem
Area Drive
S14
Mode
Priority
Configuration
Details
Function
Scanner (MS1)
Dual
Mandatory
Dual-mode scanner
Dual-mode scanner
Voice Long
(MS2)
UMTS
Additional
1800 sec duration & 10
sec gap
DCR, Quality
HSPA (Rocket)
(MS3)
UMTS
Additional
HSDPA FTP DL - 30MB &
HSUPA UL - 3MB
AFR, DCR, Throughput,
CST
ACTIX ONE/OptPCS Analysis
Using the drive data from the detailed drive in problem
areas, following analysis can be done.
DL and UL Coverage Analysis using Scanners RSCP, Ec/Io and UEs TX Power and
BLER. Identify weaker coverage areas and provide solutions to improve
Pilot Pollution Analysis using Scanner data Identify areas of where there is no
single dominant pilot. Improve the servers coverage and reduce the interferers.
Identify the boomer cells and control them. Identify areas with high SHO overhead and
provide solutions to reduce it
Pilot pollution analysis for data should be done little more aggressively than for voice. Voice
can manage with multiple servers, but for the best data performance it needs only one dominant
server.
DL and UL Data Throughput Analysis In the areas of weaker throughput, correlate it
with coverage, and cell capacity KPIs. Identify if it is a coverage / quality issue or a
capacity issue. Provide solutions as needed to improve the throughput.
CQI target for areas of higher throughput requirement For hotspot areas to achieve
better data performance a higher CQI should be targeted (say, => 25).
HSPA Cell Change Analysis Correlate the cell change locations with Too Many
Server locations. Try to find solutions to reduce the number of cell changes in the
driven area.
Event Failure Analysis Identify and resolve any drops, blocks, IRAT failures, etc that
may have happened in the drive.
Actix One Analysis Flow
OptPCS Analysis Flow
Drive Test naming convention
S14 mobility name
Sample:
For Dallas= SODA_C01a_S14_K01_D01_05252010_001.log
Sample Process & Asset Set-up
Asset set-up &
procedue
Asset Set-up & procedure
Sample Proble
area procedure
Sample problem area
procedure
HSPA+ Tracker
Problem area log tracker
Performance Improvements via
Capacity Enhancements
Performance Improvements via Capacity Enhancements
document link: http://docs.eng.tmobile.com/InfoRouter/docs/~D2271915
Progress Reporting
Optimization Status Update
Week
HSPA+ Project Progress
Ending
July 1st 2010
CQI Optimization Progress Tracking
Confidential and Proprietary Information of T-Mobile USA
24
Optimization Status Update
Week
Data Source: CQI Dashboard from Liberty Report
HSPA+ Project Progress
Ending
Center
July 1st 2010
re CQI samples (Range 0-10) Total More Than 5% Of A
Optimization Status Update
Week
Data Source: CQI Dashboard from Liberty Report
HSPA+ Project Progress
Ending
Center
July 1st 2010
Trending Cells Where CQI (Range Of 0-10) Total >5%
T-PIM KPI Reporting
Reports now available in T-PIM for KPIs to be
monitored
CQI distribution reporting per
cluster
Cluster
DA00
Cluster
<empty>
CQI distribution
reporting based
upon Asset
clusterdefinition.
Begin vs End of
project comparison
in terms of CQI
distribution (graph
2)
Modulation Distribution
Similar
reporting based
upon the used
modulation
technique in
the cluster.
FSC reported
WoW report for
CQI
Program Tracker
Regional Level
Market Level
Activities List
Preliminary Justification Package
Carolina
8CO0517C
Cluster: CR_01
HSPA+
Before
High Cost Site Mod Package
Solution Type
RSCP >= -92dBm
RSCP >= -97dBm
RSCP >= -104dBm
Traffic Loading at: 50 %
Fiber on Tower Conversion
Antenna Relocation
Antenna Change
Antenna Height Change
3G Macro Site Addition
3G Micro/Pico Site Addition
Priority
Critical
Important
Needed
Expected Cost
Other
$
Optimization Efforts Completed to Date
After
Antenna Tilts
< 30 Days
30-60 Days
Parameters
Other
Implementation Time
60 90 Days
90 180 Days
Before Stats
POPS:
R99 Traffic:
HS Traffic:
CQI AVG:
Ec/Io AVG:
RSCP AVG:
Neighbors
6-12 months
12-24 months
After Stats
POPS:
R99 Traffic:
HS Traffic:
CQI AVG:
Ec/Io AVG:
RSCP AVG:
Drive Test
Other notes of interest:
______________________________________________________
______________________________________________________
______________________________________________________
______________________________________________________
______________________________________________________
______________
HSPA+
High Cost Site Mod Package
Before
After Maps
Three
RSCP and
levels compound
array at Ec/Io =12db
-92 dBm indoor RSCP
-97 dBm indoor RSCP
-105 dBm indoor RSCP
50% loading/or preferably TPIM traffic
Other information on Map
Store Locations
BANS
POPS
Solution Type
This box lists the most common
solutions that would fall into the
high cost site mod category.
Other Solutions can be written in
under the other category
Cluster ID can be alternated
with sub-market
Priority
Priority is based on customer
impact against other sites that
need modifications.
Expected Cost
On Air Sites
Include all costs associated with
planning and implementing this
solution
Planned or Proposed Sites or
sector
GSM only on-air site
Carolina
8CO0517C
Cluster: CR_01
Optimization Efforts Completed to Date
Answers the question of what has been completed to date to improve the
performance of the area in question
Drive Test Maps
At a minimum this map needs to contain RSCP and
Ec/No drive test of the area question
Use offset feature to display both features on the
same map
Display 3G and 2G sites. Each colored differently
preferably using the above legend
Overlay RSCP and Ec/No legend
Implementation Time
The forecasted amount of time it will take to implement this solution from the
date of approval
Before Stats
Lists basic KPIs before any
optimization had been performed
on the area in question
After Stats
Lists basic KPIs after the low
cost optimization changes had
been completed.
This is NOT a prediction of
performance after this package
is implemented
