0

Introduction to UMTS
Radio KPI
3G network planning introduce
www.huawei.com

Huawei Northwest Africa Region RNP

Solution Manage: William.chen
William.chen@huawei.com

HUAWEI TECHNOLOGIES CO., LTD.

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Agenda

 UMTS Radio KPI overview  3G

Network Optimization  Case Study of
KPI Optimization

HUAWEI TECHNOLOGIES CO., LTD. Page 2

PDF created with pdfFactory Pro trial version www.pdffactory.com

 3G KPIs Principle: S.M.A.R.T

the KPI has clear and complete definition and

Specific
formula, include check point and procedure. KPIs are measurable by
given condition, Measurable

measurement equipment and procedure. all the KPIs should be reachable, it

is refer to the Attainable

commercial network.

Relevant the KPI can show the subscriber experience

Time-bound the KPI target should be on the base of time-bound

HUAWEI TECHNOLOGIES CO., LTD. Page 3

PDF created with pdfFactory Pro trial version www.pdffactory.com

 3G KPIs : Specific

Complete KPI definition include the formula combine with the check point or procedure.

The example: RRC Connection Successful Rate

 formula
RRCConnect ionSuccess
RRCS_SR Service = RRCConnect ionAttemptservice ×100%
service

 Check point

The number of “RRC Connection Attempt” is A1 or

A2?

A2 represent the effective number of “RRC

Connection Attempt”.

HUAWEI TECHNOLOGIES CO., LTD. Page 4

PDF created with pdfFactory Pro trial version www.pdffactory.com

 3G KPIs : Measurable - Drive Test

 Field Test -> Drive Test Driver test route is based on

cluster, it is involved about 20
Evaluate the overall network sites in one cluster usually.
performance, usually on the network
pre-launch phase and initial stage of
commercial launch.

Test platform

Multiple test model

UEs can perform
different tests.

Data analysis

Probe & Assistant is used

for analyzing test data

HUAWEI TECHNOLOGIES CO., LTD. Page 5

PDF created with pdfFactory Pro trial version www.pdffactory.com

 3G KPIs : Measurable - Stationary test

System KPI is suggested to be

 Field Test ->stationary Test
test by stationary:
Evaluate the system / equipment performance.  Throughput KPI
 Delay KPI
 Quality KPI
Core Network
Stationary test condition is fixed,
NodeB and usually used in indoor test
scenario.
RNC
NodeB
Radio Network KPI test is suggested
to be test by drive test
System performance Radio Network
performance  Call drop rate
Equipment ability
Radio network  Handover successful rate
Transmission
planning
Network construct
Special environment
…. condition

HUAWEI TECHNOLOGIES CO., LTD. Page 6

PDF created with pdfFactory Pro trial version www.pdffactory.com

 3G KPIs : Measurable - Statistic test

 Statistic Test
Nastar
server

Monitoring & analysis

center

Recommended for security

Essential
Optimization team
OM system

Pre-Condition for Statistic Test

Subscriber care center

There must be enough statistic sample, that

Terminal: ordinary PC or laptop, no distance
mean network traffic should be reach
limitation high
Local networks
level.
Any corresponding dept.

HUAWEI TECHNOLOGIES CO., LTD. Page 7

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Huawei KPI test platform
comprehensive analysis of the iManagerTM M2000
performance, configuration,
alarm obtaining test data
GENEX Nastar 3G

sharing of
analysis results Node B
RNC
 GENEX Probe
GENEX Assistant
 AGILENT 3G Node B
 QUALCOMM HLR/AC
multiple test data
 HUAWEI RNC
Node B

RNS

GENEX Probe PSTN

MSC/SSP/ PLMN
wireless test data WCDMA ISDN
VLR
Node B

RNC
GMSC/SSP

CN
GENEX U- Node B
Net
Node B RNS

network pre-planning/planning

HUAWEI TECHNOLOGIES CO., LTD. Page 8

PDF created with pdfFactory Pro trial version www.pdffactory.com

 3G KPIs: Relevant
Examples of User Experience (QoE)

completion of a End-to-end response The voice and video

download file is fast is Clare

Voice, Video Phone, MMS, WAP, Email, FTP, HTTP…

Service Performance

Radio Network KPI

AMR12.2, CS64k,PS64k/128k/384k, HSDPA/HSUPA

Accessibility Retainability Integrity Mobility Utilization Availability

Statistic / Measurement data

HUAWEI TECHNOLOGIES CO., LTD. Page 9

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Huawei 3G KPIs Architecture

3G KPIs

Accessability Retainability Service Integrity Mobility Delay/Latency Availability

• Call Setup • Call Drop • PS Throughput • SHO • Call Setup Delay • Worst Cell
Success Rate Rate Success (Voice/VP) Rate
(Voice/VP) (Voice/VP) Rate
• PDP Context
• PDP Context • PDP • Hard Activation Time
Activation Context Handover
• RTT
Success Rate Drop Rate Success
Rate
• SMS/MMS
Success Rate • Inter-RAT
Handover
Success
Rate

HUAWEI TECHNOLOGIES CO., LTD. Page 10

PDF created with pdfFactory Pro trial version www.pdffactory.com

 3G KPIs: It is not only a value
3G KPI is not only a value, but a system, for example of CDR (Call Drop Rate)

Network Planning Network Construct Transmission optimization

Root cause drop-call

cause
59.45% 10.02% 7.29% 4.56% 0.68% 0.68% 3% 2.28% 11.94%

HUAWEI TECHNOLOGIES CO., LTD. Page 11

PDF created with pdfFactory Pro trial version www.pdffactory.com

 3G KPIs: Attainable

3G KPI is output of combination with investment include technology, HR, cost,

competition strategy etc. Cost

Huawei KPI comply principle:

 Huawei will be responsibility for every complied KPI;
le
 Cooperate with Mobilis to work out a Competitive KPI; t
 Use the experience from other commercial network;

Technical
AttainableCommon target for Mobilis and Huawei

HUAWEI TECHNOLOGIES CO., LTD. Page 12

PDF created with pdfFactory Pro trial version www.pdffactory.com

 3G KPIs: Time-bound
Mature Network
 Drive Test KPIs
10 90

Ongoing Optimization
80

70 60

 Stationary test KPIs

50

40 30

20 10

 Network Statistics Number of Sites Number of Sites

 Network Statistics
Commercial Launch  Benchmark test
Nastar
Probe & Assistant
Probe & Assistant

 Pre-Launch
Probe & Assistant
KPI Acceptance

 Drive Test KPIs 

Stationary test KPIs

Network construct
 Site verification

HUAWEI TECHNOLOGIES CO., LTD. Page 13

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Recommend test methodology
Test Methodology Pre-launch Acceptance On-going Optimization
RRC connection success rate Drive Test -

Call setup success rate Drive Test Statistic Test

PDP activation success rate Drive Test Statistic Test

CS Call drop rate Drive Test Statistic Test

PS Call drop rate Drive Test Statistic Test

CS Call complete rate - Statistic Test

PS Call Complete rate - Statistic Test

HSPA Call drop rate Stationary Stationary

R99 PS Throughput Stationary Stationary

HSPA Throughput Stationary Stationary

Soft handover success rate Drive Test Statistic Test

Inter-RAT handover success Drive Test Statistic Test

rate

HUAWEI TECHNOLOGIES CO., LTD. Page 14

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Contents

• UMTS Radio KPI overview •3G

Radio Network Optimization • Case
Study of KPI Optimization

HUAWEI TECHNOLOGIES CO., LTD. Page 15

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Optimization throughout Life Cycle

Drop call •RF adjustment

•Para. adjustment
Congestion
•Maintenance and upgrade
Poor quality
•Contingency planning
Access failure
Pre-launch optimization •Network expansion
……
•…….

Customized and
differentiated
services
Single site verification
Billing strategies

Investment Return

RF plan review DT & CQT KPI Complaints

analysis
Construction
Commercia Booming phase Mature phase Time
phase
l launch Optimizing Balancing
Troubleshooting

HUAWEI TECHNOLOGIES CO., LTD. Page 16

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Network Optimization Phases

Optimization
preparation
RF optimization

New site integrated Service test and

parameter optimization

Routine drive test and

Single site verification stats. analysis

Acceptance criteria
N satisfied?
80 ％ sites in
Y
the Y
N cluster ready?

Acceptance

HUAWEI TECHNOLOGIES CO., LTD. Page 17

PDF created with pdfFactory Pro trial version www.pdffactory.com

 RF Optimization Target

 To optimize coverage  To

minimize pilot pollution  To

optimize cell dominance  To

optimize neighbor cell list  To

resolve RF-related drop calls

HUAWEI TECHNOLOGIES CO., LTD. Page 18

PDF created with pdfFactory Pro trial version www.pdffactory.com

 RF Parameters Optimization

 Engineering parameters adjustment

To adjust antenna down tilt

To adjust antenna azimuth

To adjust antenna location
To adjust antenna height To
replace antenna To replace
site To add new cell

 Radio part parameters adjustment

To optimize neighbor cell list

HUAWEI TECHNOLOGIES CO., LTD. Page 19

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Pilot Pollution Minimization
Pilot pollution definition:
 SHO candidates (A) Pilot pollution will result in:
(1) low signal quality
 Active set size (B)
(2) decreased system
 If A>B, pilot pollution exists capacity
(3) Call drops easier

-60

-65 Margin

-70
Ways to optimize:
-75 (1) Antenna adjustment
-62
-64 -66 (e.g. azimuth or down tilt)
-68 -69
-80
(2) Pilot power optimization
-85 -81

-90
SC1 SC2 SC3 SC4 SC5 SC6

Not
Active Set Pilot Pollution
Pilot Pollution

HUAWEI TECHNOLOGIES CO., LTD. Page 20

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Pilot Pollution: Example

HUAWEI TECHNOLOGIES CO., LTD. Page 21

PDF created with pdfFactory Pro trial version www.pdffactory.com

 RF-related Drop Calls
Problems Appearance

 Poor coverage Poor RSCP & Ec/Io

 High interference Poor Ec/Io or RTWP

 Poor UL coverage UE max Tx power

 Poor dominance many SHO events

 Pilot pollution many cells present

 Missing neighbors

Fast change of RF conditions usually causes

drop calls, e.g. turning a corner.

HUAWEI TECHNOLOGIES CO., LTD. Page 22

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Para. Optimization Target

 To reduce access failures Para. optimization

is an important step
after RF optimization
 To reduce drop calls

 To enhance service quality

Service quality and network resources

utilization will be improved after para.
optimization

HUAWEI TECHNOLOGIES CO., LTD. Page 23

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Parameters to be Optimized

 CCCH power allocation para.

 RL maximum power para.

Seldom
 Cell re-selection para. adjusted

 Intra-freq. handover para.

An effective way to
 Inter-RAT handover para. reduce drop calls

 Power control para.

 Access power para. The influence caused by the

cell parameters adjustment
 Other related para. should be analyzed carefully.

HUAWEI TECHNOLOGIES CO., LTD. Page 24

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Intra-freq. Re-selection Para.
Optimize Treselection, Qhysts and Sintrasearch so as to have
minimum ping-pong and minimum camped onto wrong cell

The black dotted line

Qhyst2 CPICH_Ec/Io is the serving cells.
CPICH_Ec/Io
Qhyst2
Qhyst2
Qhyst2
Qhyst2

The black dotted line

UE doesn’t camp is the serving cells.
onto the best cells.
Sintrasearch
UE doesn’t camp
onto the best cells.
Qqualmin

Qqualmin
Treselection Treselection
Treselection Treselection
Treselection Treselection Time

Treselection, Qhyst2, Sintrasearch = 0 Treselection, Qhyst2, Sintrasearch != 0

HUAWEI TECHNOLOGIES CO., LTD. Page 25

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Contents

• UMTS Radio KPI overview •3G

Radio Network Optimization • Case
Study of KPI Optimization

HUAWEI TECHNOLOGIES CO., LTD. Page 26

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Case study from commercial projects

• Example of Accessibility KPI analysis • Example of

Call Drop KPI issue & solution • Example of
Throughput KPI issue analysis

HUAWEI TECHNOLOGIES CO., LTD. Page 27

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Example of accessibility KPIs issue

Trouble-shooting of RRC
setup and RAB assignment Breakdown of CS RAB failure (Counters)
problems:
VS.RAB.FailEstCS.I VS.RAB.FailEstabCS. VS.RAB.FailEstPS.
• STEP1: Identify the counter with the UB.Band Other.Cell RIPFail
highest number
• STEP2: Identify the Top 10 cells with
the same problem Breakdown of CS RRC failure (Counters)

VS.RRC.Rej.R VS.RRC.Rej.Cod VS.RRC.Rej.UL.C RRC.FailConnEstab.

Definition of Counters L.Fail e.Cong E.Cong NoReply
 VS.RAB.FailEstPs.IUB.Band: The number of
RABs unsuccessfully established in a cell due
to IUB resource congestion causes. • VS.RAB.FailEstabCS.Other.Cell: • VS.RRC.Rej.UL.CE.Cong: RRC
Pre-admission control module reports setup failure due to lack of UL CE
a resource congestion. resources

HUAWEI TECHNOLOGIES CO., LTD. Page 28

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Analysis

The “worst” sites (cells) that have been found to have both IUB and CE congestion
issues:

0109TO (cell id 578/579/580)

0111TO (3882/3883/3884)
0070TO (18641/18642/18643)

NodeB UL CE Occupation Rate

Name
2007-02-19 2007-02-20 2007-02-21 2007-02-22

0070TO 120.83% 132.29% 132.29% 171.88%

0109TO 142.71% 175.00% 175.00% 180.21%

0111TO 158.33% 132.29% 132.29% 148.96%

HUAWEI TECHNOLOGIES CO., LTD. Page 29

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Suggested solution
A comparative analysis shows that, if excluding the problematic sites, the KPI would
improve substantially, say, AMR RAB assignment success rate from 97.79% to
98.09%:
KPIs (including IUB KPIs (without IUB
Congestion sites) Congestion sites)

RRC Connection Setup Success

99.17%(313992/316615) 99.35%(256449/258126)
Rate (CS)(>98%)

AMR RAB Assignment Success Suggested action plan:

97.79%(294499/301169) 98.09%(241127/245815)
Rate(>98%)
• CE/IUB expansion
Video Call RAB Assignment
98.31%(3314/3371) 98.67%(2891/2930) • Perform regular neighbour list
Success Rate(>98%)
check to further tune the
PS64 RAB Assignment Success performance
84.93%(2462/2899) 83.10%(1981/2384)
Rate(>97%)

PS128 RAB Assignment

98.48%(16674/16932) 98.80%(13638/13803)
Success Rate(>97%)
PS384 RAB Assignment
97.53%(58114/59584) 97.81%(48854/49946)
Success Rate(>97%)

HUAWEI TECHNOLOGIES CO., LTD. Page 30

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Case study from commercial projects

• Example of Accessibility KPI analysis • Example of

Call Drop KPI issue & solution • Example of
Throughput KPI issue analysis

HUAWEI TECHNOLOGIES CO., LTD. Page 31

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Example of Call drop rate issues
Drop Call Rate Voice(3G)
AMR12.2 Call Drop Rate Drop Call Rate Video(3G)
Drop Call Rate (2G)

Voice call drop rate

3.5%

3.0%

2.5%
changes over a three-
2.0% month optimization
1.5% period
1.0%

0.5%

0.0%

Zone1 Zone2
Solving Voice call drop Zone3
rate problems in Top 3
zones

HUAWEI TECHNOLOGIES CO., LTD. Page 32

PDF created with pdfFactory Pro trial version www.pdffactory.com

 What measures were taken to solve the
call drop problems?

Zone Action time Call drop rate Optimization measures

reduction
Zone 1 Feb. 14-17 1.57%->1.27%  Scrambling code planning
adjustment
 System/handover
Zone 2 Feb. 23-25 1.38%->1.21%
parameter tuning
 Neighbour list optimization
 RF optimization, etc.
Zone 3 Mar.2-6 1.34%->1.10%

HUAWEI TECHNOLOGIES CO., LTD.

Page 33

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Case study from commercial projects

• Example of Accessibility KPI analysis • Example of

Call Drop KPI issue & solution • Example of
Throughput issue analysis

HUAWEI TECHNOLOGIES CO., LTD. Page 34

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Analysis: coverage problems

RNC 123

RNC 122

Areas that lack

dominant pilot

Drive test route

involves 3 Huawei
RNCs and 1 Nortel
RNC

RNC 121

HUAWEI TECHNOLOGIES CO., LTD. Page 36

PDF created with pdfFactory Pro trial version www.pdffactory.com

 The impact of missing neighbours
• DL R99 throughput has been impacted by the interference generated by missing
neighbors, mostly concerning the new sites that had been recently put on air
• Areas with low throughput corresponds to bad Ec/Io figures and HHO to 2G

Low throughput

UE handover to GSM

HUAWEI TECHNOLOGIES CO., LTD. Page 37

PDF created with pdfFactory Pro trial version www.pdffactory.com

 Other causes
In addition to coverage and missing neighbour causes, the HSDPA
minimum throughput problem is due to:

• DT route crosses the IUR area where HSDPA calls are reconfigured to DCH (H2D)

• UE moves from a non-hotspot to a hotspot site in HSDPA connected mode

RNC 122
RNC 123 RNC 122
RNC 123

RNC 121 RNC 121

HUAWEI TECHNOLOGIES CO., LTD. Page 38

PDF created with pdfFactory Pro trial version www.pdffactory.com

 0

Thank You
www.huawei.com

PDF created with pdfFactory Pro trial version www.pdffactory.com