GSM

PS Performance Evaluation and

Analysis Guide

Issue

01

Date

2011-07-20

HUAWEI TECHNOLOGIES CO., LTD.

Copyright Huawei Technologies Co., Ltd. 2012. All rights reserved.

No part of this document may be reproduced or transmitted in any form or by any means without prior
written consent of Huawei Technologies Co., Ltd.

Trademarks and Permissions

and other Huawei trademarks are trademarks of Huawei Technologies Co., Ltd.
All other trademarks and trade names mentioned in this document are the property of their respective
holders.

Notice
The purchased products, services and features are stipulated by the commercial contract made between
Huawei and the customer. All or partial products, services and features described in this document may
not be within the purchased scope or the usage scope. Unless otherwise agreed by the contract, all
statements, information, and recommendations in this document are provided "AS IS" without warranties,
guarantees or representations of any kind, either express or implied.
The information in this document is subject to change without notice. Every effort has been made in the
preparation of this document to ensure accuracy of the contents, but all statements, information, and
recommendations in this document do not constitute a warranty of any kind, express or implied.

Huawei Technologies Co., Ltd.

Address:

Huawei Industrial Base

Bantian, Longgang
Shenzhen 518129
People's Republic of China

Website:

http://www.huawei.com

Email:

support@huawei.com

Customer
service
telephone No.

4008302118

Issue 01 (2011-07-20)

Huawei Proprietary and Confidential

Copyright Huawei Technologies
Co., Ltd

GSM
PS Performance Evaluation and Analysis Guide

Change History

Change History
Date

Issue

Description

Author

2011-04-13

V1.0

Completed the draft.

Shen Huafang,
Zhang Yunfeng

2011-04-20

V1.1

Revised according to review comments.

Shen Huafang

2011-05-06

V1.2

Revised according to review comments.

Shen Huafang

2011-05-14

V1.3

Revised according to review comments.

Shen Huafang

2011-06-10

V1.4

Revised according to training comments.

Shen Huafang

2011-06-15

V1.5

Optimized the document.

Shen Huafang

2011-06-24

V1.6

Revised according to review comments.

Shen Huafang

2012-02-21

V1.7

Updated according to the latest draft rules.

Wang Xiaofen

GSM
PS Performance Evaluation and Analysis Guide

Contents

Contents
Change History................................................................................ ii
1 About This Document....................................................................1
2 Evaluation Items...........................................................................2
3 Analysis.......................................................................................4
3.1 KPI Performance Analysis.................................................................................................................................4
3.1.1 KPI Scoring..............................................................................................................................................4
3.1.2 Accessibility Analysis...............................................................................................................................5
3.1.3 Retainability Analysis.............................................................................................................................10
3.1.4 Transmission Performance Analysis.......................................................................................................14
3.2 User Experience Analysis................................................................................................................................18
3.2.1 LLC Throughput Analysis......................................................................................................................18
1. RLC Single-Timeslot Throughput Analysis................................................................................................19
2. TBF Multiplexing Degree Analysis.............................................................................................................21
3. Channel Satisfaction Degree Analysis.........................................................................................................22
3.2.2 Delay Analysis........................................................................................................................................22
3.2.3 Cell Reselection......................................................................................................................................23

4 Appendix.................................................................................... 25
4.1 Parameters........................................................................................................................................................25
4.2 Features............................................................................................................................................................28
4.2.1 NACC.....................................................................................................................................................28
4.2.2 Packet Si Status......................................................................................................................................28

GSM
PS Performance Evaluation and Analysis Guide

3 Analysis

About This Document

This document describes how to analyze PS performance problems, including PS KPIs and
user experience, based on the evaluation reports generated by the GSM_NetworkAudit tool of
OMStar.
This document is intended for:

R&D personnel for GSM Inventory Solutions Dept.

GNAC personnel

RNAC personnel

Marketing personnel

Field engineers

Before reading this document, familiar with GSM_NetworkAudit tool of OMStar and general
packet radio service (GPRS) basic concepts.

Evaluation Items

PS performance is evaluated from the aspects of KPIs and user experience. The involved
evaluation items can be obtained from the BSC6900 Check Item Summary worksheet in the
Excel report generated by the OMStar. As shown in 1, you can choose Network Information
> General Performance > Network Performance Audit to view all BSC-level KPIs. PS
Network Performance lists the detailed analysis on the KPIs related to PS accessibility,
retainability, transmission performance, and user experience.
1
Category

PS evaluation items

Evaluation
Item

Name

Details

2. General
Performance

Network Performance Audit

Network Performance Audit

PS Accessibility Performance Affect

Factors Audit
BSC Uplink EGPRS TBF Establishment
Succ Rate Analysis
BSC Uplink GPRS TBF Establishment
Succ Rate Analysis
BSC Downlink EGPRS TBF
Establishment Succ Rate Analysis
BSC Downlink GPRS TBF
Establishment Succ Rate Analysis
UL Assignment Success Rate Analysis

PS Accessibility Performance Affect

Factors Audit
BSC Uplink EGPRS TBF
Establishment Succ Rate Analysis
BSC Uplink GPRS TBF
Establishment Succ Rate Analysis
BSC Downlink EGPRS TBF
Establishment Succ Rate Analysis
BSC Downlink GPRS TBF
Establishment Succ Rate Analysis
UL Assignment Success Rate Analysis

DL Assignment Success Rate Analysis

PS Retainability Performance Affect
Factors Audi
EGPRS_TBF_CALL_DROP_RATE

DL Assignment Success Rate Analysis

PS Retainability Performance Affect
Factors Audi
EGPRS_TBF_CALL_DROP_RATE

GPRS_TBF_CALL_DROP_RATE

GPRS_TBF_CALL_DROP_RATE

3.
Packet
Service
1. PS
Accessibility
Performance

2. PS
Retainability
Performance

3. PS
Transmission
Performance

4. PS Resource

Analysis of Rate of BSC LLC

Analysis of Rate of BSC LLC

EGPRS_RLC_BLK_RESEND_RATE
EGPRS Users Uplink and Downlink
Throughput of LLC PDU in Cell
GPRS Users Uplink and Downlink
Throughput of LLC PDU in Cell
RLC SingleSlot Rate Analysis

EGPRS_RLC_BLK_RESEND_RATE
EGPRS Users Uplink and Downlink
Throughput of LLC PDU in Cell
GPRS Users Uplink and Downlink
Throughput of LLC PDU in Cell
RLC SingleSlot Rate Analysis

GPRS_RLC_BLK_RESEND_RATE
PDCH TBF Multiplex Analysis

GPRS_RLC_BLK_RESEND_RATE
PDCH TBF Multiplex Analysis

Utilization

PDCH Resource Utilization

PDCH Resource Utilization

Analysis

PS performance is analyzed by performing the following steps:

1.

Determine whether the overall PS performance meets the standards based on the values
of the KPIs related to overall PS performance generated in the Network Performance
Audit worksheet and the recommended values listed in this chapter.

2.

Analyze the cause for the KPIs that do not meet the standards based on the results of the
corresponding evaluation items in the PS Network Performance worksheet and
determine whether top N cells are available.

3.

Analyze top N cells in a detailed way.

PS KPIs are analyzed by KPI performance and user experience. KPI performance involves
accessibility, retainability, and transmission performance. User experience involves
throughput, service interaction delay, and cell reselection.

3.1 KPI Performance Analysis

3.1.1 KPI Scoring
KPI scoring helps you determine the status of network KPIs and identify the key KPIs
affecting scoring.
BSC-level KPIs and KPI scoring are displayed in the Network Performance Audit
worksheet of the OMStar report.
The baseline value for PS KPI scoring is 85.
Table 3.1.1.I.1.1.1.1 KPI scoring
BSC Name

PS
Accessibility

PS
Retainability

PS
Transmission
Performance

PS LLC
Throughput

PS KPI
Scoring
85

3.1.2 Accessibility Analysis

Accessibility measures the capability of an MS to access the network, which affects the
network access success rate and access delay. Accessibility KPIs are displayed in the
Network Performance Audit worksheet and are defined from the aspects of EGPRS/GPRS
TBF establishment success rate and TBF congestion rate. In normal cases, the TBF
establishment success rate is higher than 95% and the TBF congestion rate is lower than 3%.
The actual values depend on operator's requirements. The TBF congestion rate is analyzed for
resource capacity evaluation. Therefore, only the TBF establishment success rate is analyzed
in this section.
2

Accessibility KPIs

KPI

Accessibilit
y

Uplink GPRS TBF

Establishment Succ Rate

> 95%

Downlink GPRS TBF

Establishment Succ Rate

> 95%

Uplink EGPRS TBF

Establishment Succ Rate

> 95%

Downlink EGPRS TBF

Establishment Succ Rate

> 95%

Recommen
ded Value

If the TBF establishment success rate is lower than the baseline value, perform the following
detailed analysis.

TBF Establishment Success Rate Impact Analysis

Table 3.1.2.I.1.1.1.1 lists the KPIs that can be queried by choosing Packet Service > PS
Accessibility Performance > PS Accessibility Performance Affect Factors Audit.
Table 3.1.2.I.1.1.1.1 BSC-level accessibility KPIs
BSC
Name

KPI

TBF Establishment Succ Rate

Rate of Failed TBF
Establishments due to No
Channel
Rate of Failed TBF
Establishments due to MS No
Response

Baseli
ne
Value

Actual
Value

Deviati
on

Affectin
g
Percent
age

Rate of Failed TBF

Establishments due to Others

Baseline value: The baseline value for TBF Establishment Succ Rate is 95%.
Actual value: The actual value is obtained according to the statistical value of the live network. The
value of Rate of Failed TBF Establishments due to MS No Response on the live network is excessively
higher than that of Rate of Failed TBF Establishments due to Others. Therefore, the transmission quality
on the Um interface needs to be optimized if the value of Rate of Failed TBF Establishments due to MS
No Response exceeds 3%. The value of Rate of Failed TBF Establishments due to No Channel is 2%,
and the value of Rate of Failed TBF Establishments due to Others is close to 0%.
Deviation: It specifies the difference between the baseline value and the actual value.
Affecting percentage: It specifies the percentage of deviation for different KPIs. Based on the affecting
percentage, you can identify the main cause decreasing the TBF success rate and take measures
accordingly.

Analysis on Rate of TBF Establishment Failures Due to Various

Reasons on BSC for Top N Cells
The KPIs noted by customers are usually BSC-level KPIs. This section describes how to
optimize top N cells with poor accessibility based on the affecting percentage of BSC-level
KPIs for each cell. This further improves BSC-level KPIs.

Rate of Cell TBF

Establishment Failure

Succ Rate after Cell TBF Establishment

Succ Rate after Cell TBF Establishment

Succ Rate after Cell TBF Establishment

Succ Rate after Cell TBF Establishment

times(Other) to BSCRate of Cell TBF Failure

times(MS No Response) Rate of Cell TBF Failure

Sort data
in
descendin
g order
based on
this
column.

times(No Resource) to Rate of Cell TBF Failure

times to BSC

Establishment Attempt Rate of Cell TBF

Establishments due to Rate of Failed TBF

Cell TBF Establishment

Succ Rate

Establishments due to Rate of Failed TBF

Establishments due to Rate of Failed TBF

Cell Name

BSC Name

Table 3.1.2.I.1.1.1.1 KPIs for measuring top N cells with poor accessibility

BSC Name
Cell Name
Cell TBF Establishment Succ Rate
Rate of Failed TBF Establishments due to No
Channel
Rate of Failed TBF Establishments due to MS No
Response
Rate of Failed TBF Establishments due to Others
Rate of Cell TBF Establishment Attempt times to
BSC
Rate of Cell TBF Establishment Failure times to BSC
Rate of Cell TBF Failure times(No Resource) to BSC
Rate of Cell TBF Failure times(MS No Response) to
BSC
Rate of Cell TBF Failure times(Other) to BSC
Cell TBF Establishment Succ Rate after Optimize
Cell TBF Establishment Succ Rate after Optimize
PDCH resource
Cell TBF Establishment Succ Rate after Optimize MS
No Rsponse
Cell TBF Establishment Succ Rate after Optimize
Other reason

Quantified Analysis on Improvements in Accessibility KPIs

Quantified analysis on improvements in accessibility KPIs quantifies how much BSC KPIs
can improve when KPIs of top N cells reach their baseline values.
Table 3.1.2.I.1.1.1.1 Quantified analysis on accessibility KPIs of top N cells

BSC
Nam
e

Top

Contribution
Degree of
TOP Cells

BSC TBF
Establishmen
t Succ Rate
after
Optimize

BSC TBF
Establishmen
t Succ Rate
after
Optimize
PDCH
resource

BSC TBF
Establishmen
t Succ Rate
after
Optimize MS
No Rsponse

BSC TBF
Establishmen
t Succ Rate
after
Optimize
Other reason

Performance
after optimizing
top 5 cells
Performance
after optimizing
top 10 cells
Performance
after optimizing
top 20 cells

Accessibility KPI Performance Analysis

The TBF establishment failure causes of accessibility KPIs can be analyzed based on the data
in the following worksheets:

BSC Uplink EGPRS TBF Establishment Succ Rate Analysis

BSC Uplink GPRS TBF Establishment Succ Rate Analysis

BSC Downlink EGPRS TBF Establishment Succ Rate Analysis

BSC Downlink GPRS TBF Establishment Succ Rate Analysis

As shown in Table 3.1.2.I.1.1.1.1 and Table 3.1.2.I.1.1.1.2 (using uplink EGPRS KPIs as an
example), the TBFestablishment failure causes are as follows:

No channel resources (resource congestion)

No MS response (poor transmission quality on the Um interface, abnormal G-Abis

transmission, and CCCH congestion)

Other causes (abnormal procedures and flow control, which account for a small
percentage)

Table 3.1.2.I.1.1.1.1 Analysis on BSC-level uplink EGPRS TBF establishment success rate
BSC
Name

Uplink
EGPRS TBF
Establishm
ent Succ
Rate

Rate of
Failed
Uplink
EGPRS TBF
Establishme
nts due to
No Channel

Rate of
Failed
Uplink
EGPRS TBF
Establishme
nts due to
MS No
Response

Rate of
Failed
Uplink
EGPRS TBF
Establishme
nts due to
Others

Table 3.1.2.I.1.1.1.2 Analysis on cell-level uplink EGPRS TBF establishment success rate

BS
C
Na
me

Cel
l
Na
me

Uplink
EGPRS
TBF
Establis
hment
Succ
Rate

Rate of
Failed
Uplink
EGPRS
TBF
Establis
hments
due to
No
Channel

Number
of
Failed
Uplink
EGPRS
TBF
Establis
hments
due to
No
Channel

Rate of
Failed
Uplink
EGPRS
TBF
Establis
hments
due to
MS No
Respons
e

Number
of
Failed
Uplink
EGPRS
TBF
Establis
hments
due to
MS No
Respons
e

Rate of
Failed
Uplink
EGPRS
TBF
Establis
hments
due to
Others

Number
of
Failed
Uplink
EGPRS
TBF
Establis
hments
due to
Others

(1) TBF Establishment Failures Due to No Channel Resources

TBF establishment failures indicate that no PDCH is available when a TBF attempts to
establish. It corresponds to the PDCH congestion rate.

In normal cases, the GPRS PDCH congestion rate is almost the same as the EGPRS
PDCH congestion rate. If the GPRS PDCH congestion rate is high but the EGPRS
PDCH congestion rate is low, check whether GPRS PDCHs are insufficient because
EGPRS dedicated PDCHs or EGPRS preferred PDCHs are configured. When this
occurs, change EGPRS dedicated PDCHs or EGPRS preferred PDCHs to common
EGPRS PDCHs and set Allow E Down G Up Switch to Yes.

If traffic in the underlaid subcell of a concentric cell is congested, set Dynamic Channel
Conversion Parameter of Concentric Cell to Only convert dynamic channel at OL.
Otherwise, set Dynamic Channel Conversion Parameter of Concentric Cell to Only
convert dynamic channel at UL.

Perform capacity expansion or resource adjustment by referring to chapter "RF Resource and
Capacity Evaluation" in the Guide to GSM Resource and Capacity Analysis.
Suggestion: expand TRX capacity or deploy a new site when frequency resources are
insufficient.
(2) TBF Establishment Failures Due to No MS Response
Due to poor transmission quality of the Um interface, abnormal G-Abis transmission or
CCCH congestion, TBF establishment fails when the MS fails to receive or parse downlink
messages or the BSC fails to parse the uplink response messages from an MS correctly. These
TBF establishment failures are counted as TBF establishment failures due to no MS response.

Poor transmission quality of the Um interface

Weak coverage, interference, and imbalance between the uplink and the downlink result
in poor transmission quality of the Um interface. For details about how to analyze and
optimize the transmission quality of the Um interface, see the Guide to GSM Network
Evaluation of Coverage and Interference.

Abnormal G-Abis transmission

If bit errors or channel out-of-synchronization occurs, uplink and downlink data fails to
be decoded correctly. As a result, TBF establishment fails. In addition, the link delay

prolongs when the G-Abis transmission is abnormal. When this occurs, the timer at the
receiving end may expire. As a result, TBF establishment fails.
For details about how to analyze the G-Abis transmission, see "G-Abis Transmission."

CCCH congestion
You can determine whether an assignment message is delivered normally based on the
packet immediate assignment success rate and the Success Rate of PACCH assignment
success rate. Table 3.1.2.I.1.1.1.3 and Table 3.1.2.I.1.1.1.4 list the KPIs related to BSClevel and cell-level uplink assignment success rates.
Table 3.1.2.I.1.1.1.3 BSC uplink assignment success rate analysis
BS
C
Na
me

Succes
sful UL
Assign
ment
Rate
(%)

Succes
s Rate
of UL
PS
Immedi
ate
Assign
ment

Succes
s Rate
of UL
Assign
ment
on
PACCH

Table 3.1.2.I.1.1.1.4 Uplink assignment success rate analysis

BS
C
Na
me

Cell
Na
me

Success
ful UL
Assign
ment
Rate
(%)

Number
of UL
Assignm
ents

Success
Rate of
UL PS
Immedi
ate
Assign
ment

Number
of UL PS
Immedia
te
Assignm
ents

Success
Rate of
UL
Assign
ment
on
PACCH

Successf
ul UL
Assignm
ents on
PACCH

If the packet immediate assignment success rate is low, but the packet immediate assignment
success rate on the PACCH is high, check whether CCCH overload occurs. If CCCH overload
occurs, the Immediate Assignment messages sent over the CCCH may be discarded. As a
result, TBF establishments fail. You can check whether CCCH overload occurs based on the
Cell CS Service Paging Deletion Check and Paging Deletions Rate for PS Services
Analysis worksheets.
If CCCH overload occurs, set the CCCH load threshold to a large value to avoid TBF
establishment failures due to flow control. The most effective method is to increase the
number of extended BCCHs.
In addition, increase the value of timer T3168 during two-phase access. This prevents MSs
from frequently sending channel requests, alleviating CCCH overload.
To reduce TBF establishment failures due to no channel available and no MS response,
improve the transmission quality of the Um and G-Abis interfaces. Alternatively, optimize the
settings of the following parameters:

Increase the value of T3168 for a satellite cell or a cell where transmission problems are
serious and difficult to resolve, such as a cell where the transmission delay over the GAbis interface is long and unstable.

Increase the values of T3192, Release Delay of Downlink TBF(ms), Release Delay of
Non-extended Uplink TBF(ms), and Inactive Period of Extended Uplink TBF(ms) to
prolong the TBF release delay, increasing TBF establishment success rate on the
PACCH.

Use a low-rate initial coding scheme to increase the TBF establishment success rate.

Suggestion: NPI services (parameter adjustment and optimization)

(3) TBF Establishment Failures Due to Other Causes
TBF establishment may fail when procedures are abnormal or flow control is enabled. TBF
establishment failures of this type seldom occur.

3.1.3 Retainability Analysis

Retainability measures whether call drops occur during conversation, which affect user
throughput and delay. They are listed in the PS Retainability Performance Affect Factors
Audit worksheet. Usually, the call drop rate for a common network should be lower than 5%,
and the call drop rate for a network with excellent performance should be lower than 3%. The
actual call drop rate depends on requirements of telecom operators.
3

BSC-level retainability KPIs

KPI

Recommen
ded Value

Retainabilit
y

Uplink GPRS TBF Call Drop Rate

< 5%

Downlink GPRS TBF Call Drop

Rate

< 5%

UL EGPRS TBF drop rate

< 5%

DL EGPRS TBF drop rate

< 5%

If the TBF call drop rate is lower than the baseline value, perform the following detailed
analysis.

Analysis on Affecting Percentage of Retainability KPIs

Table 3.1.3.I.1.1.1.1 lists the KPIs that can be queried by choosing Packet Service > PS
Retainability Performance > PS Retainability Performance Affect Factors Audit.
Table 3.1.3.I.1.1.1.1 Analysis on BSC-level retainability KPIs
BSC
Name

KPI

TBF Call Drop Rate

Baseli
ne
Value

Actual
Value

Deviati
on

Affectin
g
Percent
age

Rate of TBF Drop Due to MS

No Response
Rate of TBF Drop Due to
FLUSH

Rate of TBF Drop Due to

SUSPEND
Rate of TBF Drop Due to No
Resource
Rate of TBF Drop Due to
Other Reason

Baseline value: The baseline value for TBF establishment success rate is 95%. The baseline
values are obtained based on network evaluation experience. The difference between these
values depends on different network requirements.
Actual value: The actual value is obtained according to the statistical value of the live
network.
Deviation: It is not considered for the KPIs Rate of TBF Drop Due to SUSPEND and Rate of
TBF Drop Due to Other Reason because the optimization on abnormal TBF releases due to
suspend and other reasons is unclear.
Affecting percentage: It specifies the percentage of deviation for different KPIs. Based on the
affecting percentage, you can identify the main cause decreasing the TBF success rate and
take measures accordingly.

Analysis on Rate of TBF Call Drops Due to Various Reasons on

BSC for Top N Cells
The KPIs noted by customers are usually BSC-level KPIs. This section describes how to
optimize top N cells with poor retainability based on the affecting percentage of BSC-level
KPIs for each cell. This further improves BSC-level KPIs.

BSC Name
Cell Name

FLUSHCell TBF Call Drop Rate after Optimize

No RsponseCell TBF Call Drop Rate after Optimize MS

Cell TBF Call Drop Rate after Optimize

Rate of Cell TBF Drop Times(FLUSH) to BSC

PDCH Resource Cell TBF Call Drop Rate after Optimize

Sort
data in
descend
ing
order
based
on this
column.

Response) to BSCRate of Cell TBF Drop Times(MS No

BSCRate of Cell TBF Drop Times(No Resource) to

Rate of Cell TBF Drop Times to BSC

to BSCRate of Cell TBF Establishment Success times

Rate of TBF Drop Due to Other Reason

Rate of TBF Drop Due to SUSPEND

Rate of TBF Drop Due to FLUSH

Rate of TBF Drop Due to MS No Response

Rate of TBF Drop Due to No Resource

Cell TBF Call Drop Rate

Cell Name

BSC Name

Table 3.1.3.I.1.1.1.1 Top N cells with poor retainability

Cell TBF Call Drop Rate

Rate of TBF Drop Due to No Resource
Rate of TBF Drop Due to MS No Response
Rate of TBF Drop Due to FLUSH
Rate of TBF Drop Due to SUSPEND
Rate of TBF Drop Due to Other Reason
Rate of Cell TBF Establishment Success times to BSC
Rate of Cell TBF Drop Times to BSC
Rate of Cell TBF Drop Times(No Resource) to BSC
Rate of Cell TBF Drop Times(MS No Response) to BSC
Rate of Cell TBF Drop Times(FLUSH) to BSC
Cell TBF Call Drop Rate after Optimize
Cell TBF Call Drop Rate after Optimize PDCH Resource
Cell TBF Call Drop Rate after Optimize MS No Rsponse
Cell TBF Call Drop Rate after Optimize FLUSH

Quantified Analysis on Improvements in Retainability KPIs

Quantified analysis on improvements in retainability KPIs quantifies how much
improvements BSC KPIs can make when KPIs of top N cells reach their baseline values.
Table 3.1.3.I.1.1.1.1 Quantified analysis on improvements in retainability KPIs

BSC Name

Top

Performance after
optimizing top 5
cells
Performance after
optimizing top 10
cells
Performance after
optimizing top 20
cells

Contribution
Degree of TOP
Cells

BSC
TBF
Call
Drop
Rate
after
Optimiz
e

BSC
TBF
Call
Drop
Rate
after
Optimiz
e PDCH
Resourc
e

BSC TBF
Call Drop
Rate after
Optimize
MS No
Rsponse

BSC
TBF
Call
Drop
Rate
after
Optimiz
e
FLUSH

Retainability KPI Analysis and Optimization

For the cells with unsatisfied call drop rate, analyze the cause of abnormal TBF releases based
on the analysis items in the EGPRS_TBF_CALL_DROP_RATE and
GPRS_TBF_CALL_DROP_RATE worksheets. Table 3.1.3.I.1.1.1.1 and Table
3.1.3.I.1.1.1.2 list the KPIs related to BSC-level and cell-level abnormal TBF releases.
Following uses abnormal uplink EGPRS TBF release as an example. Other abnormal TBF
releases are similar to this type of abnormal TBF release.
Table 3.1.3.I.1.1.1.1 Analysis on abnormal uplink EGPRS TBF releases
Uplink
EGPRS
TBF
Abnorma
l Release
No
Channel
Resource

BSC
Name

Uplink
EGPRS
TBF
Abnormal
Release
Suspend

Uplink
EGPRS
TBF
Abnormal
Release
Flush

Uplink
EGPRS
TBF
Abnormal
Release
MS No
Response

Uplink
EGPRS
TBF
Abnormal
Releases
Channel
Preemption

Uplink
EGPRS
TBF
Abnormal
Releases
Other
Cause

Uplink
EGPRS
TBF Call
Drop

Table 3.1.3.I.1.1.1.2 Top N cells with high abnormal uplink EGPRS TBF release rate

BSC
Nam
e

Cell
Nam
e

Uplink
EGPRS
TBF
Abnorma
l Release
No
Channel
Resource

Uplink
EGPRS
TBF
Abnorma
l Release
Suspend

Uplink
EGPRS
TBF
Abnorma
l Release
Flush

Uplink
EGPRS
TBF
Abnorma
l Release
MS No
Response

Uplink
EGPRS
TBF
Abnormal
Releases
Channel
Preemptio
n

Uplink
EGPRS
TBF
Abnorma
l Releases
Other
Cause

Uplink
EGPR
S TBF
Call
Drop

DL
EGPR
S TBF
drop
rate
(%)

Sum of
the
precedin
g
abnorma
l releases
due to
various
reasons

(1) Abnormal TBF Releases Due to No MS Response

N3101, N3103, and N3105 are counters for the BSC to measure uplink and downlink TBF
quality. If abnormal TBF releases occur because these counters overflow, the uplink and
downlink TBF quality is poor. Abnormal releases due to no MS response refer to uplink call
drops due to N3101 and N3103 overflow and downlink call drops due to N3105 overflow. In
normal cases, more than 80% call drops are due to N3101, N3103, and N3105 overflow.
N3101, N3103, or N3105 overflows mainly because of poor transmission quality on the Um
interface, unstable transmission quality on the G-Abis interface, or improper MS operations.
In addition, call drops due to N3101, N3103, and N3105 overflow may occur when the
settings of PS coding parameters are inappropriate, or an inappropriate algorithm is selected
for adjusting the coding scheme.

To reduce TBF abnormal releases due to no MS response, improve the transmission quality on
the Um and G-Abis interfaces. Alternatively, optimize the settings of the following
parameters:

Increase the values of N3101, N3103, and N3105 to improve the link quality tolerance of
the network.

Use a low-rate coding scheme to reduce call drops. For details, see the methods for
selecting a low-rate coding scheme.

Select the BTS reporting BER algorithm for adjusting the uplink coding scheme. Call
drops can be reduced by using the algorithm.

Suggestion: NPI services (parameter adjustment and optimization)

(2) Abnormal TBF Releases Due to Suspended PS Services
Ongoing PS services are suspended when MSs perform location updates or CS services. In
this situation, abnormal TBF releases due to suspended PS services are measured. This type of
abnormal release reflects the impact of CS services on PS services.
If abnormal releases of this type occur, the location area settings for cells where the call drop
rate is high and the setting of the timer for periodic location updates may be inappropriate.
You are advised to check these settings.
(3) Abnormal TBF Releases Due to FLUSH_LL Messages
When the serving GPRS support node (SGSN) detects cell reselection initiated by an MS
during PS services, it sends a FLUSH_LL message to the source cell. After the source cell
receives the FLUSH_LL message, one abnormal TBF release due to the FLUSH_LL
messages is counted. This type of TBF abnormal release reflects the frequency of cell
reselection initiated by MSs. If a flood of abnormal releases of this type occur, check the
settings of coverage parameters and reselection parameters of a serving cell and its
neighboring cells to avoid ping-pong cell reselection. For details about the optimization
method, see section 3.2.3"Cell Reselection."
(4) Abnormal TBF Releases Due to No Channel
The causes of abnormal TBF releases due to no channel available are as follows:

Unstable channel status

Frequent channel out-of-synchronization

Channel faults

Manual channel blocking

Channel deactivation

PDCH preemption by CS services

When abnormal releases of this type occur, you are advised to whether transmission or the
clock is faulty.
If the number of times the BSC reclaims dynamic PDCHs is equal to the number of times the
BSC reclaims dynamic PDCHs in load state, ongoing CS services preempt channels occupied
by PS services. In this situation, you need to add static PDCHs. In addition, set Level of
Preempting Dynamic Channel to LEVEL1(No preempt of CCHs).
Suggestion: expand TRX capacity or deploy a new site when frequency resources are
insufficient.

3.1.4 Transmission Performance Analysis

Transmission performance measures transmission quality of the network, which affects
accessibility and retainability as well as user experience. 4 lists the transmission performance
KPIs displayed in the Network Performance Audit worksheet. Usually, the downlink
EGPRS RLC data block retransmission rate is lower than 10%, other retransmission rates are
lower than 5%, and the transmission frame error rate is lower than 0.1%. The actual rates
depend on the requirements of telecom operators.
4

BSC-level transmission performance KPIs

KPI

Recommen
ded Value

Transmission
performance

Retransmission rate of UL RLC

data block (%)

< 5%

Retransmission rate of DL RLC

data block (%)

< 5%

Retransmission rate of UL
EGPRS RLC data block (%)

< 5%

Retransmission rate of DL
EGPRS RLC data block (%)

< 10%

Rate of Transmitted Error Frames

< 0.10%

Cell transmission performance can be analyzed based on the analysis items that can be
queried by choosing Packet Service > PS Transmission Performance.
5

Transmission performance evaluation items

Evaluation Item

Details

EGPRS_RLC_BLK_RESEND_RATE

EGPRS_RLC_BLK_RESEND_RATE

GPRS_RLC_BLK_RESEND_RATE

GPRS_RLC_BLK_RESEND_RATE

RLC SingleSlot Rate Analysis

RLC SingleSlot Rate Analysis

RLC Retransmission Rate

If RLC data blocks are lost or incorrect because of poor transmission quality on the Um or GAbis interface, the receiving end requests the sending end to retransmit data. Therefore, the
RLC retransmission rate reflects the transmission quality of the Um and G-Abis interfaces to
some extent.
For details about how to analyze and optimize the transmission quality of the Um interface,
see the Guide to GSM Network Evaluation of Coverage and Interference. You can determine
the transmission quality of G-Abis interface based on the FER over the G-Abis interface,
transmission alarms, and clock alarms.

The coding scheme is dynamically adjusted according to the transmission quality on the Um
interface during GPRS data transmission. When a low-rate coding scheme is used, data blocks
contain more redundant data and the data transmission rate is low, but the reliability of data
transmission is high. When a high-rate coding scheme is used, data blocks contain less
redundant data and the data transmission rate is high, but the reliability of data transmission is
low. Therefore, if the transmission quality of the Um or G-Abis interface is poor, you are
advised to use a low-rate coding scheme to obtain high-quality data transmission, improving
user experience. Using a high-rate coding scheme may not improve user experience as
expected and may even deteriorate user experience. The methods for selecting a low-rate
coding scheme are as follows:

Select a low-rate initial coding scheme.

For GPRS services, increase difficulties in converting a low-rate coding scheme to a

high-rate coding scheme and decrease difficulties in converting a high-rate coding
scheme to a low-rate coding scheme.

For EGPRS services, shorten the BEP period to improve the sensitivity of measurement
reports (MRs) to changes in the transmission quality of the Um interface to ensure that
appropriate coding schemes are dynamically selected and used.

Select the BTS reporting BER algorithm for adjusting the uplink coding scheme to
ensure that the uplink coding scheme is appropriate to the transmission quality of the Um
interface.

6 lists the parameters related to the coding scheme.

6

Parameters related to the coding scheme

Category

Parameter

Default
Value

Recommended
Value

Setting Principle

Coding
scheme
adjustment

ADJUSTULMCSTYPE

The uplink coding scheme is

adjusted according to uplink
quality. The adjusted coding
scheme can better suit the uplink
quality.

BEP Period

When this parameter is set to a

small value, the PS performance
is sensitive to the changes in the
transmission quality of the Um
interface. If the transmission
quality of the Um interface is
poor, decrease the value of this
parameter.

Uplink Fixed MCS Type

UNFIXED

UNFIXED

Uplink Default MCS Type

MCS2

MCS2

If the transmission quality of the

Um interface is unstable and a
high-rate initial uplink coding
scheme is used, the TBF
establishment success rate may be
low. If a low-rate coding scheme
is used, the TBF establishment
success rate increases but the
uplink throughput and proportion
of high-rate coding schemes

decrease.
Downlink Fixed MCS Type

UNFIXED

UNFIXED

Downlink Default MCS

Type

MCS6

MCS6

Use a high-rate initial coding

scheme to increase the proportion
of high-rate coding schemes. This
may increase the retransmission
rate.

Uplink Fixed CS Type

UNFIXED

UNFIXED

Uplink Default CS Type

CS1

CS1

If the transmission quality of the

Um interface is unstable and a
high-rate initial uplink coding
scheme is used, the TBF
establishment success rate may be
low. If a low-rate coding scheme
is used, the TBF establishment
success rate increases but the
uplink throughput and proportion
of high-rate coding schemes
decrease.

Downlink Fixed CS Type

UNFIXED

UNFIXED

Downlink Default CS Type

CS2

CS2

Use a high-rate initial coding

scheme to increase the proportion
of high-rate coding schemes. This
may increase the retransmission
rate.

Suggestion: NPI services (parameter adjustment and optimization)

G-Abis Transmission
FER over the G-Abis interface can be used to identify uplink transmission problems based on
check error frames and out-of-synchronization frames received on the uplink. FER over the
G-Abis interface mainly reflects the uplink transmission problems and reflects the downlink
transmission problems to some extent. However, it does not map downlink KPIs. Therefore,
you must analyze transmission alarms and clock alarms when analyzing downlink
accessibility, retainability, and transmission performance KPIs.
Table 3.1.4.I.1.1.1.1 lists the KPIs that are used for analyzing G-Abis frame error rate. These
KPIs can be queried by choosing Packet Service > PS Transmission Performance > RLC
SingleSlot Rate Analysis.
Table 3.1.4.I.1.1.1.1 Cell-level transmission FER analysis
BSC
Nam
e

Cell
Name

Average
Throughput
of Uplink
GPRS RLC
per
PDCH(kbps
)

Uplink
CS3CS4
Ratio

Rate of
Fail
Applicatio
n Attempts
of Abis
Timeslot
Because of

Rate of
Transmitte
d Error
Frames

no Idle
Timeslot

E432
E432

MD41964
MD42542

8.80

0.00%

0.00%

0.03%

8.85

0.00%

0.00%

0.06%

FER over the G-Abis interface can be used to preliminarily determine the transmission quality
of the G-Abis interface.

In normal cases, the FER is less than 10e-5, that is, an average of one error frame is sent
every four minutes on one channel. In this situation, the link quality is favorable, and
MSs can transmit data properly.

For a transmission link with poor quality, the FER is less than 10e-4, that is, an average
of one to three error frames are sent every minute on one channel. Due to the error
frames, the affected MS may encounter the problems of decreased data rate, long
transmission delay, and even call drops or disconnection from the network.

If the FER is greater than 10e-4, the link is quite unstable and tends to be out of
synchronization. In this situation, the proportion of out-of-synchronization frames
increases. The MS may only be able to perform small-traffic data services, such as the
upper-layer signaling and some WAP services. Mass data transmission, such as FTP
services, becomes difficult.

If the FER is less than 5, the link quality is acceptable. If the FER of a cell is always high, a
transmission problem occurs in this cell. In this situation, check the transmission cables and
improve the transmission quality.

3.2 User Experience Analysis

This section describes how to analyze PS user experience from the aspects of LLC
throughput, delay, and cell reselection.

3.2.1 LLC Throughput Analysis

7 lists the KPIs for reflecting the average LLC throughput per user on the BSC.
7

BSC-level LLC throughput KPIs

KPI

Baseline
Value

Uplink Throughput of GPRS Users

LLC PDU(kbit/s)

> 16

Downlink Throughput of GPRS

Users LLC PDU(kbit/s)

> 36

Uplink Throughput of EGPRS Users

LLC PDU(kbit/s)

> 45

Downlink Throughput of EGPRS

Users LLC PDU(kbit/s)

> 120

The LLC throughput on the live network can be obtained based on the weighted value of the
uplink and downlink GPRS/EGPRS LLC throughputs obtained by choosing Packet Service >
PS Transmission Performance > Analysis of Rate of BSC LLC. Then, check whether the
LLC throughput meets the standards by comparing it with its baseline value.
Table 3.2.1.I.1.1.1.1 BSC-level LLC throughput analysis
BSC
Name

LLC
Throughput
Baseline
Value
(kbit/s)

LLC
Throughput
on the Live
Network
(kbit/s)

Analysis
Result

If the LLC throughput on the live network does not reach its baseline value, check the impact
of each affecting factor on the LLC throughput. As listed in Table 3.2.1.I.1.1.1.2, the factors
affecting the LLC throughput are RLC single-timeslot throughput, TBF multiplexing degree,
and channel satisfaction degree.
Table 3.2.1.I.1.1.1.2 Analysis on factors affecting BSC-level LLC throughput
BSC Name
DoubleHAJ
J
DoubleHAJ
J
DoubleHAJ
J

Affecting Factor

KPI
Baseline
Value

Actual
Value

Affecting
Percentage

27.65

25.16

17.14%

1.91

0.00%

0.85

0.62

82.86%

Throughout
Capacity of RLC
Signal-time(kbps)
Frequence of
TBF
Content of
Channel

Obtain and analyze the top N cells with unsatisfied LLC throughput based on the cell-level
LLC throughput analysis table.
Table 3.2.1.I.1.1.1.3 Cell-level LLC throughput analysis

BSC
Name

Cell
Name

Rate of
LLC(kbp
s)

Signaltime Rate

Frequen
ce

Conte
nt of
Chann
el

Percenta
ge of
Signaltime
Affect
LLC
Rate

DoubleHA
JJ

MD3701
-1

24.91

10.26

2.45

0.78

84.31%

Percenta
ge of
Frequen
ce Affect
LLC
Rate

Percenta
ge of
Content
of
Channel
Affect
LLC
Rate

15.51%

0.18%

Suggestion: If only EGPRS-capable MSs experience low LLC throughput, EGPRS and GPRS
channels can be separated.
Following describes how to analyze each affecting factor.

1. RLC Single-Timeslot Throughput Analysis

The average RLC single-timeslot throughput can be obtained based on the weighted value of
the average uplink and downlink GPRS/EGPRS RLC single-timeslot throughputs obtained by
choosing Packet Service > PS Transmission Performance > RLC SingleSlot Rate
Analysis. Then, check whether the average RLC single-timeslot throughput meets the
standards by comparing it with its baseline value.
Table 3.2.1.I.1.1.1.4 BSC-level RLC single-timeslot throughput

BSC
Name

Average
Through
put of
Uplink
GPRS
RLC per
PDCH(k
bps)

Average
Through
put of
Downlin
k GPRS
RLC per
PDCH(k
bps)

Average
Through
put of
Uplink
EGPRS
RLC per
PDCH(k
bps)

Average
Through
put of
Downlin
k
EGPRS
RLC per
PDCH(k
bps)

Avg-Rate of
RLC Signaltime(kbps)

Baseline of
Avg-Rate of
RLC Signaltime(kbps)

DoubleH
AJJ

9.74

11.91

27.42

35.75

25.16

27.65

If the average RLC single-timeslot throughput does not meet the standards, isolate the main
affecting factor based on the affecting percentages shown in Table 3.2.1.I.1.1.1.5. The two
factors affecting the RLC single-timeslot throughput shown in Table 3.2.1.I.1.1.1.5 are BEP
19-31 ratio and proportion of Abis timeslot application failures due to no idle timeslots.
Table 3.2.1.I.1.1.1.5 Analysis on BSC-level RLC single-timeslot affecting factors
BSC Name

Affecting
Factor

KPI
Baseline
Value

Actual
Value

Affecting
Percentage

DoubleHAJ
J

BEP19~31 Ratio

85.00%

83.89%

5.85%

DoubleHAJ
J

Rate of Fail
Application
Attempts of Abis
Timeslot
Because of no
Idle Timeslot

30.00%

83.49%

94.15%

Obtain and analyze the top N cells with unsatisfied RLC single-timeslot throughput based on
the cell-level single-timeslot throughput analysis table.

Table 3.2.1.I.1.1.1.6 Analysis on cell-level RLC single-timeslot throughput

BSC Name

Cell
Name

Signaltime Rate
of
RLC(kbp
s)

DoubleHAJ
J2

MD37031

16.53

BEP19~3
1 Ratio

Rate of
Fail
Applicatio
n
Attempts
of Abis
Timeslot
Because
of no Idle
Timeslot

94.45%

91.00%

Effect on
Rate of
RLC
Caused by
LinkQuali
ty

Effect
on Rate
of RLC
Caused
by Abistime

0.00%

100.00%

BEP19-31 Ratio
BEP means the bit error probability, which indicates the proportion of bit errors measured on
the receiving end. The BEP is classified into 32 classes, ranging from 0 to 31. The higher the
BEP, the more favorable the transmission quality of the Um and G-Abis interfaces. The BSC
determines the coding scheme based on the BEP. The higher the BEP, the higher-rate the
coding scheme.
For the cells with BEP 19-31 ratio lower than the baseline value, locate the problems on the
cells based on the transmission quality of the G-Abis and Um interfaces.
For details about the analysis on the transmission quality of the G-Abis interface, see "G-Abis
Transmission." For details about the transmission quality of the Um interface, see the Guide
to GSM Network Evaluation of Coverage and Interference.
Table 3.2.1.I.1.1.1.1 describes the receive level and carrier-to-interference ratio supported by
each coding scheme. When the carrier-to-interference ratio is higher than 23.5 dB, the coding
scheme higher than MCS7 can be used, and the corresponding BEP class must be 19 or
higher.
Table 3.2.1.I.1.1.1.1 Receive level and carrier-to-interference ratio supported by each coding
scheme
Coding
Scheme
MCS1
MCS2
MCS3
MCS4
MCS5
MCS6
MCS7
MCS8
MCS9

Receive
Level
(dBm)
102
101
99
97
98
96
93
90.5
86

TU3 Carrier-toInterference Ratio (dB)

13
15
16.5
19
18
20
23.5
28.5
30

Proportion of Abis Timeslot Application Failures Due to No Idle

Timeslots
Abis timeslot sufficiency is a prerequisite for cells using high-rate coding schemes. If idle
timeslots are insufficient, channels cannot use high-rate coding schemes. Table 3.2.1.I.1.1.1.1
describes the mapping between the number of required idle timeslots and coding schemes.
Table 3.2.1.I.1.1.1.1 Mapping between the number of idle timeslots and coding schemes
Coding Scheme
CS1-CS2
CS3-CS4
MCS1-MCS2
MCS3~MCS6
MCS7
MCS8-MCS9

Number of Required
Idle Timeslots
0
1
0
1
2
3

In the TDM_Abis Interface Resource Evaluation worksheet of the OMStar report, check
whether the number of idle timeslots is sufficient, the number of required idle timeslots, and
whether to expand E1 timeslots in the site idle timeslot capacity analysis table. Field
engineers can determine to add idle timeslots or expand E1 timeslots based on the number of
required idle timeslots.
Suggestion: transmission capacity expansion, and GBFD-117301 Flex Abis.

2. TBF Multiplexing Degree Analysis

PDCHs carry RLC control signaling and RLC data. The rate for a single PDCH is definite.
When multiple MSs are multiplexed onto the same channel, the single-user throughput will
decrease because block resources on the channel are shared by MSs. In this situation, the
PDCH multiplexing degree needs to decrease to improve the single-user throughput.
For the cells whose uplink and downlink PDCH TBF multiplexing degrees are higher than 2,
perform the following optimization operations:

For the cells where channels are insufficient and TCHs are not congested, more PDCHs
can be obtained by improving the maximum PDCH rate threshold, decrease the uplink or
downlink multiplexing dynamic channel conversion threshold, and set the PDCH
downlink multiplexing threshold to 80.
Suggestion: NPI services (parameter optimization)

For the cells where TCHs are congested, the recommended optimization operation is
capacity expansion.

Suggestion: expand TRX capacity or deploy a new site when frequency resources are
insufficient.

3. Channel Satisfaction Degree Analysis

Channel satisfaction degree is the ratio of the actual number of allocated channels to the MS
multi-timeslot capability. The higher the channel satisfaction degree, the more the number of
channels allocated to cells, the higher the probability that the number of allocated channels
meets the requirements of MS multi-timeslot capability.

If the channel satisfaction degree is lower than 85%, the number of PDCHs in the cell is
insufficient. The optimization methods are as follows:

If the number of times the BSC reclaims dynamic PDCHs is equal to the number of
times the BSC reclaims dynamic PDCHs in load state, ongoing CS services preempt
channels occupied by PS services. In this situation, you need to add static PDCHs. In
addition, set Level of Preempting Dynamic Channel to LEVEL1(No preempt of
CCHs).

Check the analysis items in the DSP CPU Occuption Rate of DPUP worksheet. When
the value of Average Number of PDCHs Activated on DSP is greater than 40, dynamic
PDCHs in some cells cannot be converted because a single DSP reaches the channel
activation upper limit. As a result, channels are insufficient. When this occurs, the
number of activated channels on certain DSP is great. It is recommended that cell
redistribution be performed to balance cell distribution on DSPs. If most DSPs
experience this type of problem, it is recommended to add DPUPs.
Suggestion: board expansion

For the cells where channels are sufficient and TCHs are not congested, improve the
maximum PDCH rate threshold in the cell to obtain more PDCHs.

For the cells where TCHs are congested, the recommended optimization operation is
capacity expansion.
Suggestion: expand TRX capacity or deploy a new site when frequency resources are
insufficient.

3.2.2 Delay Analysis

Delay can be analyzed based on PCHR data instead of traditional traffic statistics. PCHRs are
improved since GBSS 13.0, and therefore there is no tool for batch analyzing PCHRs. In this
situation, you need to analyze PCHRs one by one by using the InsightSharp tool.
Delay can be analyzed based on the data in the PCHR delay-sensitive service information
area. The key KPI to be analyzed is average delay in service interaction response. The
formula for calculating the average delay in service interaction response is as follows:
Average delay in service interaction response = Average delay in sending PDUs + Average
delay in receiving PDUs + Average delay in PDU round trip over the Gb interface
Table 3.2.2.I.1.1.1.1 describes the formulas for calculating each KPIs.
Table 3.2.2.I.1.1.1.1 Delay KPIs
KPI
Average delay in sending
PDUs
Average delay in receiving
PDUs
Average delay in PDU
round trip over the Gb
interface

Formula
Total duration for sending PDUs/Total number of
sent PDUs
Total duration for receiving PDUs/Total number of
received PDUs
Total interval between uplink and downlink
PDUs/Statistical times for measuring the interval
between uplink and downlink PDUs

The average delay in sending PDUs and average delay in receiving PDUs reflect the downlink
and uplink delays on the BSS side, respectively. The average delay in PDU round trip over the
Gb interface reflects the delay above the Gb interface, including the uplink and downlink
processing duration on the Gb interface and core network.

For the cells with long delay on the BSS side, check whether accessibility, retainability, and
transmission performance KPIs are normal. If they are abnormal, optimize them by referring
to associated methods.
For the cells with long delay above the Gb interface, check the links over the Gb interface or
contact core network engineers to locate problems on the core network.

3.2.3 Cell Reselection

The key KPI for analyzing cell reselection on the BSS side is TBF abnormal release rate due
to flush. If the value of this KPI is greater than 1%, cell reselection greatly affects services. In
this situation, you need to set the cell reselection parameters and adjust coverage.
The KPIs listed in Table 3.2.3.I.1.1.1.1 can be obtained by choosing Packet Service > PS
Retainability Performance > PS Retainability Performance Affect Factors Audit. The
baseline values for these KPIs are 1%.
Table 3.2.3.I.1.1.1.1 BSC-level cell reselection analysis
BSC
Nam
e

UL GPRS TBF
drop rate due
to Flush

DL GPRS TBF
drop rate due to
Flush

UL EGPRS
TBF drop rate
due to Flush

UL EGPRS
TBF drop rate
due to Flush

Obtain top N cells with high abnormal TBF release rate due to flush based on the cell-level
cell reselection analysis table, and analyze whether short-interval cell reselection or ping-pong
cell reselection occurs.
Table 3.2.3.I.1.1.1.2 Cell-level cell reselection analysis

BSC
Nam
e

Cell
Name

TBF drop rate

due to Flush

Numbe
r of
Times a
Cell Is
Selecte
d by
MS

Ratio of
Cell
Reselection
s by MS at
Short
Interval

Ratio of
Cell PingPong
Reselection
s by MS

CR
H

For the cells with high short-interval cell reselection rate, check whether they are under poor
coverage due to incorrect power settings, hardware faults, or quick attenuation.
For the cells with high ping-pong cell reselection rate, stabilize the serving cell by controlling
coverage or adjusting CRH parameters.
Suggestion: NPI services (cell reselection optimization), GBFD-116301 Network Assisted
Cell Change (N/ACC), and GBFD-119801 Packet SI Status (PSI).

Appendix

4.1 Parameters
8

PS parameters

Category

Accessibilit
y

Resource

Parameter
Name

Default
Value

Recommended
Value

T3168

500

T3192

500

BS_CV_MAX

10

10

NO

YES

MODE4_1

MODE4_1

Support 11BIT
EGPRS Access
Multiplexing
Mode

Maximum
Ratio
Threshold of
PDCHs in a
Cell(%)
Uplink
Multiplex

Setting Principle
After eliminating the effect of
link quality and radio resource
congestion, increasing the value
of T3168 can improve the uplink
TBF assignment success rate.
However, this also prolongs the
delay.
When network resources are
sufficient, that is, the GPRS
service congestion rate is low,
set T3192 to a large value,
accelerating TBF establishments
and improve data transmission
performance.
If this parameter is set to a small
value, the MS retransmits data
before receiving an
acknowledgement message. This
increases retransmission rate and
wastes resources.

30

When the number of MSs

carried on a channel reaches the

Threshold of
Dynamic
Channel
Conversion
Downlink
Multiplex
Threshold of
Dynamic
Channel
Conversion
PDCH Uplink
Multiplex
Threshold
PDCH
Downlink
Multiplex
Threshold
Timer of
Releasing Idle
Dynamic
Channel
Level of
Preempting
Dynamic
Channel

threshold divided by 10,

dynamic channel conversion is
triggered. A high multiplexing
degree may affect service rates
for MSs.
2

20

20

LEVEL0

LEVEL0

Timer of
Releasing Abis
Timeslot

15

15

Dynamic
Channel
Conversion
Parameter of
Concentric
Cell

Only
convert
dynamic
channel at
UL

Allow E Down
G Up Switch

Yes

Yes

Resource
Reallocation to
Different Trx

ALLOW

ALLOW

Once

Multiple

Applying for
Multiple Abis
Timeslots

Decrease the value of this

parameter when channels are
insufficient.
If TCHs are sufficient but the
reclaimed dynamic PDCH
proportion is high, set this
parameter to LEVEL1.
Set this parameter to prevent
frequent Abis resource requests,
shorten timeslot binding
duration, and increase the rate.
When the traffic is heavy and
idle timeslots are insufficient,
decrease the value of this
parameter.
If traffic in the underlaid subcell
is congested, dynamic channels
are converted in the overlaid
cell.
If traffic in the overlaid subcell
is congested, dynamic channels
are converted in the underlaid
subcell.
When this parameter is set to
No, the number of available
GPRS resources decreases. Set
this parameter to Yes only if
demonstrating the EGPRS
downlink rate or the operator
has a demanding requirement for
EGPRS downlink rate.

Set this parameter to accelerate

channel assignment or coding
scheme adjustment to improve

the rate.

Once

Coding
scheme
adjustment

EGPRS UL
Coding
Scheme
Dynamic
Conversion

BTS
reporting
BER

BTS reporting
BER

BEP Period

Uplink Fixed
MCS Type

UNFIXED

UNFIXED

Uplink Default
MCS Type

MCS2

MCS2

Downlink
Fixed MCS
Type

UNFIXED

UNFIXED

Downlink
Default MCS
Type
Uplink Fixed
CS Type
Uplink Default
CS Type
Downlink
Fixed CS Type
Downlink
Default CS
Type
PS timer and
counter

Release Delay
of Nonextended
Uplink TBF
(ms)

MCS6

MCS6

UNFIXED

UNFIXED

CS1

CS1

UNFIXED

UNFIXED

CS2

CS2

120

120

The uplink coding scheme is

adjusted according to uplink
signal quality. The adjusted
coding scheme can better suit
the uplink signal quality.
If this parameter is set to a small
value, the PS performance is
sensitive to changes in the
transmission quality of the Um
interface. If the transmission
quality of the Um interface is
poor, decrease the value of this
parameter.

If the transmission quality of the

Um interface is unstable and the
uplink coding scheme is high,
the TBF establishment success
rate may be low. If a low-rate
coding scheme is used, the TBF
establishment success rate
increases but the uplink
throughput and proportion of
high-rate coding schemes
decrease.

Use a high-rate initial coding

scheme to increase the
proportion of high-rate coding
schemes. This increases the
retransmission rate.

Same as the setting principle of

Uplink Default MCS Type.

Use a high-rate initial coding

scheme to increase the
proportion of high-rate coding
schemes. This increases the
retransmission rate.
Increase the value of this
parameter to increase the
probability of TBF
establishment on the PACCH for

Inactive Period
of Extended
Uplink TBF
(ms)
Release Delay
of Downlink
TBF (ms)
Maximum
Value of N3101
Maximum
Value of N3103
Maximum
Value of N3105
Support NACC

PS cell
reselection

2000

2000

2400

2400

20

20

10

10

NO

YES

PACKET SI

NO

YES

FLUSH-LL
PDU Transfer
Switch

OPEN

OPEN

MSs and increase the TBF

establishment success rate. This
wastes TBF resources.

When the Um interface quality

fluctuates significantly, increase
the value of this parameter to
reduce call drops. However,
resources may be occupied for a
long time.
Adjust the value of this
parameter to reduce the
interruption duration caused by
cell reselection.

4.2 Features
4.2.1 NACC
Before an MS reselects a cell, it reports the target cell to the BSC. The BSC sends the system
information (SI) message about the target cell to the MS in advance through RLC data blocks.
In this way, the MS hands over to the target cell without receiving the SI message about the
target cell. This facilitates the MS to process PS services and reduces data transmission
interruption caused by cell reselection.

4.2.2 Packet Si Status

The MS can initial PS resource request in the target cell before receiving all SI messages. The
MS sends a Packet Si Status message to the BSS to notify the network of system information
messages that are not received. The BSS sends the requested SI messages through RLC data
blocks. In this way, the MS does not need to exit from the PS state and enters the idle state to
receive these messages.