W Access Problem Optimization Guide 20060330 a 3.1

W-Access Problem Optimization Guide

For internal use only

Product name

Confidentiality level

WCDMA RNP


Product version

Total 98 pages

3.1

W-Access Problem Optimization Guide (For internal use only)

Prepared by

Dong Yan

Date

Reviewed by

Wan Liang, Guo Zhengping, Yu Yongxian, and Xu Jianguo

Date

Reviewed by

Qin Yan

Date

Approved by

2005-11-21

Date

Huawei Technologies Co., Ltd. All Rights Reserved

2006-03-06

Huawei Proprietary. No Spreading Without Permission.

Page 1 of 98



Revision Records Date

Version

Description

Author

2004-08-23

1.00

Outline

Guan Shiguo

2004-11-03

1.00

Initial transmittal

Guan Shiguo

2004-12-20

1.20

Revising according to review

Guan Shiguo

2005-11-21

According to V3.0 guide requirements, Dong Yan reorganizing and updating V2.0 guide. The update includes:

3.00



Modifying document structure



Adding optimization objectives



Adding delay optimization



Adding new cases and analysis



Redrawing analysis flow chats





2006-05-19

3.10







2006-03-06

the

Adding analysis of traffic statistics data

V3.10 adds analysis of HSDPA with the following Wang Dekai content updated: 

2006-06-21

Adding appendix which contain background knowledge about access.

3.11

Supplementing admission failure analysis and cases about RRC connection, RAB assignment process in HSDPA service Adding HSDPA-related statistics values

DT

and

traffic

According to traffic statistics indexes of RNC version 1.6C01B064, adding some traffic statistics indexes Updating RRC connection analysis

Adding analysis of traffic statistics in paging Dong Yan problems


Page 2 of 98



Table of Contents Chapter 1 Introduction.................................................................................................................10 Chapter 2 Evaluating Access Performance.................................................................................11 2.1 Accessibility.......................................................................................................................11 2.2 System Availability.............................................................................................................11 2.3 Access Delay....................................................................................................................12 Chapter 3 Analyzing DT/CQT Data...............................................................................................13 3.1 Data Analysis Software.....................................................................................................13 3.2 Definition of Access Failure...............................................................................................13 3.2.1 Definition of Call Failure by GENEX Assistant.........................................................13 3.2.2 Definition of Actix Analyzer......................................................................................14 3.2.3 Definition of TEMS..................................................................................................15 3.3 Flow and Methods for Analyzing Access Failure Problems................................................15 3.3.1 Overall Flow for Analyzing Call Failure Problems....................................................15 3.3.2 Analyzing Paging Problems....................................................................................16 3.3.3 Analyzing RRC Connection Setup Problems...........................................................19 3.3.4 Analyzing Authentication Problems.........................................................................22 3.3.5 Analyzing Security Mode Problems.........................................................................23 3.3.6 Analyzing PDP Activation Failure Problems............................................................24 3.3.7 Analyzing RAB or RB Setup Problems....................................................................24 3.4 Processing Access Delay Problems..................................................................................27 3.4.1 Configuration of Discontinuous Cyclic Period Duration Factor DRX .......................27 3.4.2 Whether to Disable Authentication and Encryption Flow.........................................28 3.4.3 Implementing Early or Late Assignment..................................................................28 3.4.4 Whether the RRC Connection Is Set up on FACH and DCH...................................28 Chapter 4 Analyzing Traffic Statistics Data.................................................................................29 4.1 Tool for Analyzing Data......................................................................................................29 4.2 General Methods for Analyzing Traffic Statistics Data........................................................29 4.2.1 Flow for Analyzing RNC-level Traffic Statistics Data................................................29 4.2.2 Flow for Analyzing Cell-level Traffic Statistics Data.................................................30 4.3 Accessibility Indexes.........................................................................................................31 4.3.1 Paging Traffic Statistics Indexes.............................................................................31 4.3.2 Low Success Rate of RRC Setup...........................................................................35 4.3.3 Low Success Rate of CS RAB Setup......................................................................38 4.3.4 Lower Success Rate of PS RAB Setup...................................................................41 4.3.5 Low Success Rate of RB Setup..............................................................................44 4.4 System Availability Index...................................................................................................46 4.4.1 High Admission Rejection Rate...............................................................................46 4.4.2 High Paging Congestion Rate.................................................................................46 4.4.3 High Rate of Congested Cell...................................................................................46 Chapter 5 Solving Access Problems...........................................................................................47 5.1 Paging Problems...............................................................................................................47 5.1.1 Improper Power Configuration of Paging-related Channels.....................................47 5.1.2 Paging Failure due to UE Location Area Update.....................................................47 5.1.3 Paging Failure due Implicit Detach of UE................................................................49 5.2 Cell Selection and Reselection Problem............................................................................49 5.2.1 Repeating to Send the RRC Connection Request Message due to Cell Reselection49 5.3 RRC Setup Problems........................................................................................................52 5.3.1 Improper Configuration of Parameters of Uplink Access Channel...........................52 5.3.2 Improper Configuration of AICH Power...................................................................55 5.3.3 Improper Configuration of FACH Power..................................................................55 5.3.4 Multiple Times of RRC Connection Request (for Service) and No RAB Assignment Request.............................................................................................................................57 5.3.5 RRC Connection of HSDPA Subscribers Rejected due to Inadequate Code Resource61 5.4 RAB and RB Setup Problems............................................................................................63 2006-03-06


Page 3 of 98



5.4.1 RAB Setup Failure due to Inadequate Resource.....................................................63 5.4.2 Handover Failure before Completion of Signaling Flow...........................................63 5.4.3 Admission Failure due to HSDPA Total Bit Rate Threshold Exceeded by HSDPA Bit Rate of Cell.......................................................................................................................65 5.5 Authentication Problems....................................................................................................66 5.6 Security Mode Problems...................................................................................................66 5.7 Abnormal Equipment Problems.........................................................................................68 5.7.1 Abnormal NodeB....................................................................................................68 5.7.2 Abnormal UE..........................................................................................................70 Chapter 6 Summary......................................................................................................................72 Chapter 7 Appendix 1: Paging Process......................................................................................73 7.1 Paging Origination............................................................................................................73 7.1.1 Paging by CN.........................................................................................................73 7.1.2 Paging by UTRAN..................................................................................................73 7.2 Paging Flow......................................................................................................................73 7.2.1 Paging Type 1.........................................................................................................73 7.2.2 Paging Type 2.........................................................................................................74 7.3 Behaviors of UE after Receiving Paging............................................................................75 7.3.1 UE in Idle Mode......................................................................................................75 7.3.2 UE in Connected Mode...........................................................................................75 7.4 DRX Process of UE...........................................................................................................75 7.4.1 DRX Cyclic Length and Paging Occasion...............................................................75 7.4.2 Relationship of PICH and SCCPCH........................................................................77 7.4.3 PCH Selection........................................................................................................78 7.4.4 DRX Examples of UE.............................................................................................78 Chapter 8 Appendix 2: Access Process Analysis.......................................................................79 8.1 Cell Search.......................................................................................................................79 8.1.1 Timeslot Synchronization........................................................................................79 8.1.2 Frame Synchronization and Scramble Group Identification.....................................79 8.1.3 Identification of Cell Primary Scramble....................................................................80 8.2 Cell Selection and Reselection..........................................................................................80 8.2.1 Cell Selection.........................................................................................................80 8.2.2 Judgment Criterion (Criterion S).............................................................................81 8.2.3 Cell Reselection......................................................................................................82 8.3 Random Access................................................................................................................88 8.3.1 Random Access Channel........................................................................................88 8.3.2 Random Access Process........................................................................................90 Chapter 9 Appendix 3: Authentication Flow...............................................................................93 Chapter 10 Appendix 4: Description of Access-related Parameters.........................................95 10.1 Engineering Parameters..................................................................................................95 10.2 Cell Parameters..............................................................................................................95 10.2.1 Transmit Power of FACH......................................................................................95 10.2.2 Transmit Power of PCH........................................................................................95 10.2.3 Transmit Power of PICH.......................................................................................95 10.2.4 Cell Reselection Parameter: Measurement Hysteresis 2 (Qhyst2s).......................96 10.2.5 Cell Reselection Parameter: Reselection Hysteresis Time (Treselections)............96 10.2.6 Cell Reselection Parameter: Sintrasearch.............................................................96 10.2.7 Cell Reselection Parameter: Qoffset2...................................................................96 10.2.8 Transmit Power of AICH........................................................................................96 10.2.9 PRACH-related Parameters..................................................................................96

2006-03-06


Page 4 of 98



List of Tables Table 1.1Indexes and recommended values for accessibility related to DT...............................11 Table 1.2Indexes and reference values for accessibility related to traffic statistics..................11 Table 1.3Indexes and reference for system availability...............................................................12 Table 1.4Indexes and reference for access delay related to DT..................................................12 Table 11.1Counters related to paging loss for idle UE ................................................................32 Table 12.1Counters related to paging loss for UE in PCH state..................................................33 Table 13.1Counter related to paging loss due to flow control ....................................................34 Table 13.2Counters related to PCH congestion............................................................................34 Table 14.1Counters related to RRC connection request rejection due to lub interface failure .35 Table 14.2Traffic statistics counters related to RRC connection request rejection due to network congestion......................................................................................................................................36 Table 14.3Counter related to RRC connection failure due to no response.................................37 Table 14.4Counters related to RRC connection setup rejection due to redirection...................38 Table 15.1Traffic statistics counters related to CS RAB assignment setup failure due to radio network problems..........................................................................................................................39 Table 15.2Traffic statistics counters related to CS RAB assignment setup failure due to insufficient capability ....................................................................................................................40 Table 15.3Counter related to CS RAB assignment setup failure due to transmission network problems.........................................................................................................................................40 Table 16.1Traffic statistics counters related to PS RAB assignment setup failure due to radio network problems..........................................................................................................................42 Table 16.2Traffic statistics counters related to PS RAB assignment setup failure due to insufficient capability ....................................................................................................................43 Table 16.3Counter related to PS RAB assignment setup failure due to transmission network problems.........................................................................................................................................43 Table 16.4Counter related to PS RAB setup failure due to no resource available......................44 Table 17.1Traffic statistics counters related to RB setup failure.................................................45 Table 17.2Traffic statistics counter related to no response to RB setup....................................46 Table 57.1Mapping relationship between PI and PICH.................................................................77 Table 58.1Parameters and their description in the criterion S.....................................................81 Table 58.2Cell reselection parameters and descriptions.............................................................85 Table 58.3Broadcast parameters and description of cell reselection in system information....86 Table 61.1Relationship among the access subchannel, access timeslot, and SFN ..................92

2006-03-06


Page 5 of 98



List of Figures

2006-03-06


Page 6 of 98



Figure 1.2Overall flow for analyzing call failure problems in DT/CQT.........................................16 Figure 1.3Signaling flow of originated UE in locating paging problems.....................................17 Figure 1.4Flow for analyzing paging problems............................................................................18 Figure 1.5Flow for analyzing RRC connection setup problem ...................................................19 Figure 1.6Authentication failure due to MAC Failure...................................................................22 Figure 1.7Authentication failure due to Synch Failure.................................................................23 Figure 1.8Security mode control...................................................................................................23 Figure 1.9Security mode reject......................................................................................................24 Figure 1.10Flow for analyzing RNC-level traffic statistics data...................................................30 Figure 1.11Flow for analyzing cell-level traffic statistics data.....................................................31 Figure 1.12Position for counting point by counter for paging loss by idle UE ..........................32 Figure 1.13Position for counting point by counter for paging loss by UE in PCH state............33 Figure 1.14Position for counting point by counter for RRC connection rejection.....................35 Figure 1.15Position for counting point by counter for CS RAB assignment failure in RNC traffic statistics starting counting............................................................................................................38 Figure 1.16Position for counting point by counter for PS RAB assignment failure in RNC traffic statistics.........................................................................................................................................41 Figure 1.17Position for counting point by counter for RB setup failure in traffic statistics......45 Figure 1.18Originating signaling flow of paging failure due to UE location area update...........47 Figure 1.19Content of the Disconnect message in paging failure due to UE location area update .........................................................................................................................................................48 Figure 1.20Terminating signaling flow of paging failure due to UE location area update..........48 Figure 1.21Signaling of UE............................................................................................................49 Figure 1.22Signal quality when the UE sends the RRC connection request message..............50 Figure 1.23Signal quality when the UE repeats to send the RRC connection request...............51 Figure 1.24Signaling of UE in a connection process...................................................................52 Figure 1.25Single subscriber tracing signaling on RNC..............................................................52 Figure 1.26Quality of downlink signals.........................................................................................53 Figure 1.27Regular interference in cell 248..................................................................................53 Figure 1.28Part of magnified chart of interference.......................................................................54 Figure 1.29Signaling upon improper configuration of FACH power...........................................55 Figure 1.30Signal strength upon the first sending of RRC connection request.........................56 Figure 1.31Single subscriber tracing signaling by RNC..............................................................56 Figure 1.32Signaling and signal strength upon the second sending of RRC connection request57 Figure 1.33Traced signaling at UE side.........................................................................................58 Figure 1.34Traced signaling at RNC side......................................................................................58 Figure 1.35BLER statistics at UE side...........................................................................................59 Figure 1.36BLER and RRC message at UE side...........................................................................60 Figure 1.37BLER and RRC message at UE side...........................................................................60 Figure 1.38Assignment of HSDPA code tree................................................................................62 Figure 1.39RNC log for HSDPA admission rejection....................................................................62

2006-03-06


Page 7 of 98



Figure 1.40Signaling of Disconnect after completion of RB setup.............................................63 Figure 1.41Signaling of UE upon a connection failure.................................................................64 Figure 1.42Single subscriber tracing signaling by RNC..............................................................64 Figure 1.43Signal strength before release of connection............................................................65 Figure 1.44Rejection messages in security mode........................................................................66 Figure 1.45Content of the RANAP_SECURITY_MODE_REJECT message.................................67 Figure 1.46ciphering mode information configured in previous security mode command.......67 Figure 1.47Security mode message..............................................................................................67 Figure 1.48Content of the RANAP_SECURITY_MODE_COMMAND message............................68 Figure 1.49Signaling of UE upon failure in receiving RRC Connection Setup message...........69 Figure 1.50Single subscriber tracing signaling by RNC..............................................................69 Figure 1.51Signal strength upon occurrence of problems..........................................................69 Figure 1.52Signaling of UE............................................................................................................70 Figure 1.53Downlink signal quality...............................................................................................71 Figure 1.54Flow chart of PAGING TYPE 1 message.....................................................................74 Figure 1.55Flow chart of PAGING TYPE 2 message.....................................................................74 Figure 1.56Schematic drawing of UE paging occasion................................................................76 Figure 1.57Frame structure of PICH..............................................................................................77 Figure 1.58Sequence relationship between PICH and SCCPCH..................................................78 Figure 1.59Number and interval of access timeslots of RACH....................................................88 Figure 1.60Structure of random access transmission.................................................................89 Figure 1.61Timing relation between PRACH and AICH as seen at the UE..................................90 Figure 1.62Definition of access timeslot set (taking the uplink and downlink access timeslot fixed difference τ p-a 7680 chips as example)...............................................................................92 Figure 1.63Successful authentication process............................................................................93 Figure 1.64Authentication Failure (due to MAC Failure)..............................................................94 Figure 1.65Authentication failure (due to Synch failure).............................................................94

2006-03-06


Page 8 of 98



W-Access Problem Optimization GuideKey

words: WCDMA,

radio network optimization, access, DT, and traffic statistics. Abstract: The document describes how to locate and solve access problems in WCDMA network optimization, the definition of access problems, test methods, analysis flows, and solutions. Finally, the appendix provides the fundamental knowledge necessary for analyzing access problems by RNO engineers. Acronyms and abbreviations: Acronyms and abbreviations

2006-03-06

Full spelling

DT

Drive Test

CQT

Call Quality Test

RNC

Radio Network Controller

RNP

Radio Network Planning

RNO

Radio Network Optimization


Page 9 of 98



Chapter 1 Introduction This document aims to meet the requirements on solving access problems by on-site engineers during RNO. It details the methods for evaluating network access performance, test methods, data analysis methods, FAQs, and solutions. The appendix provides the fundamental knowledge about access problems, description of principles, related parameters, and data processing tools. It guides engineers to locate and solve access problems during optimizing network KPI indexes and network O&M. This document excludes the usage of tools. It contains 10 chapters, with the structure as below:          

Chapter 1: Introduction Chapter 2: Evaluating Access Performance Chapter 3: Analyzing DT/CQT Data Chapter 4: Analyzing Traffic Statistics Data Chapter 5: Solving Access Problems Chapter 6: Summary Chapter 7: Appendix 1: Paging Process Chapter 8: Appendix 2: Access Process Analysis Chapter 9: Appendix 3: Authentication Flow Chapter 10: Appendix 4: Description of Access-related Parameters

 Note: On-site engineers must adjust network optimization parameters according to the importance levels of parameters, the impact of adjustment on network services and network equipment in a proper time. The whole adjustment must follow Radio Network Planning Online Data Modification Regulations and data backup and emergency solutions are necessary. Immediate verification must be performed after adjustment so that the adjustment is correct. The access process of HSDPA service is similar to that of original R99 service. There are only some minor differences.

2006-03-06


Page 10 of 98



Chapter 2 Evaluating Access Performance The access performance includes three aspects: accessibility, system availability, and access delay. The specified indexes for measuring access performance are obtainable by DT and traffic statistics. For the definition of indexes, see the UMTS Radio Network KPI Baseline V3.3.

2.1 Accessibility 2.1 lists the indexes and recommended values for accessibility in DT. Table 1.1 Indexes and recommended values for accessibility related to DT Index MOC success rate

MTC success rate PDP context activation success rate

Service

Statistics method

Reference

Voice

DT&CQT

96%

VP

DT&CQT

95%

Voice

DT&CQT

95%

VP

DT&CQT

94%

PS

DT&CQT

96%

HSDPA

DT&CQT

96%

2.1 shows the indexes and reference values for accessibility related to traffic statistics. Table 1.2 Indexes and reference values for accessibility related to traffic statistics Index

Service

Statistics method

Radio paging success rate

N/A

Stat.

85%

RRC setup success rate

N/A

Stat.

97%

Voice

Stat.

98%

VP

Stat.

98%

PS

Stat.

97%

HSDPA

Stat.

97%

RAB setup success rate

Reference

The values previously mentions are just for reference. Determine the actual values according to the detailed requirements of projects or requirements of commercial network contracts.

2.2 System Availability 2.2 lists the indexes and reference for system availability.

2006-06-03


Page 11 of 98



Table 1.3 Indexes and reference for system availability Index

Service

Statistics method

Reference

Admission rejection rate

All services

Stat.

2%

Paging congestion rate

All services

Stat.

0.5%

Congested cell ratio

All services

Stat.

1%


2.3 Access Delay 2.3 lists the indexes and reference for access delay related to DT. Table 1.4 Indexes and reference for access delay related to DT Index

Call setup delay

Service

Reference

Voice

CQT

7s

VP ring

CQT

7s

CQT

13s

PS

CQT

4s

HSDPA

CQT

VP see figure

PDP context activation delay

Statistics method

the

2s (average) < 4s (95%)


2006-06-03


Page 12 of 98



Chapter 3 Analyzing DT/CQT Data 3.1 Data Analysis Software The software for analyzing DT/CQT data includes the following items:   

GENEX Assistant: the post processing software for DT data Rollback tool for Single subscriber tracing on RNC Nastar CHR

3.2 Definition of Access Failure If a service fails to be set up, this is an access failure. In DT, the common access failure problems include the following types:   

Voice call failure VP call failure PDP activation failure

According to the preset judgment conditions, the DT data analyzers can usually judge the access failure problems during test. The analyzers include GENEX Assistant and Actix Analyzer.

3.2.1 Definition of Call Failure by GENEX Assistant I. Originated Call Failure in CS Domain Event definition: the UE sends an RRC REQUEST message. Wherein, the IE establish cause is Originating Conversational Call without reception of the direct transfer message alerting. The following events are defined according to the stages of failure. 1)

2) 3)

4)

2006-06-03

RRC connection setup failure: out of the consideration of retransmission times and waiting time, the UE fails to receive the response from RNC or receives the RRC CONNECTION REJECT message after sending the RRC CONNECTION REQUEST message. Initial direct transfer and security mode setup failure: after the UE sends the RRC CONNECTION SETUP COMPLETE message, it fails to send the NAS SETUP message. RAB assignment failure: after the UE receives the CALL PROCEEDING message, it fails receive the RB SETUP message from RNC. Or the UE responds RB SETUP FAIL message after receiving the RB SETUP message. Or the UE receives the DISCONNECTION message with the cause normal release after receiving the RB SETUP message, when the UE does not send the RB SETUP CMP message. The failure after RAB assignment: after the UE sends the RB SETUP COMPLETE message:  The originated UE receives the DISCONNECT/RELEASE message from CN  The originated UE waits for the CONNECT or ALERTING message until expiration, so the call clearing process is originated. According to the protocols, after the UE sends the CM SERVICE REQUEST message, the timer T303 starts. If the UE fails to receive the CALLPROCEEDING, ALERTING, CONNECT, OR RELEASE COMPLETE message before expiration of T303, the clearing process starts. Huawei Proprietary. No Spreading Without Permission.

Page 13 of 98



Before receiving alerting message, the UE enters the idle state and starts to receive system information. 

 Note: As strictly defined, after the MS enters the CELL_DCH state and before it receives the alerting message, it must send the cell update message with the cause RLC unrecoverable error/ Radio link failure. Take the greater value of the maximum waiting time configured at RLC layer as default and the synchronization time as the judgment time. It is unclear that the UE can report the RLC layer message, so the maximum waiting time is neglected.

II. Terminated Call Failure in CS Domain The terminated UE receives the paging of type 1, but it does not send the RRC CONNECTION REQUEST message with the cause Terminating Conversational Call or does not send the direct transfer message alerting to CN. Detailed failure stages include: 1) 2)

3)

4)

RRC connection setup failure After the UE sends the RRC CONNECTION REQUEST message, it fails to receive reply from RNC or receives the RRC CONNECTION REJECT message from RNC. Initial direct transfer and security mode setup failure After the UE sends the RRC CONNECTION SETUP COMPLETE message, it fails to receive the SETUP direct transfer message, it sends the RELEASE COMPLETE message, or the UE receives the DISCONNECT message from CN. RAB assignment failure After the UE sends the CALL CONFIRM message, it fails to receive the RB SETUP message from RNC. Or after it receives the RB SETUP message, it replies RB SETUP FAIL message. Or after it receives the RB SETUP message, it receives the DISCONNECT message not due to normal release cause, when the UE has not sends the RB SETUP CMP message. Failure after completion of RAB assignment After the UE sends the RB SETUP COMPLETE message, the called UE receives the DISCONNECT/RELEASE message from CN

 Note: As strictly defined, after the UE enters the CELL_DCH state, it sends the cell update message with the cause RLC unrecoverable error/ Radio link failure before receiving the alerting message. The judgment time should be the greater of maximum waiting time and asynchronization time configure by default at the RLC layer. It is unknown that the test UE can report RLC layer messages, so neglect the strict definition.

3.2.2 Definition of Actix Analyzer The Actix Analyzer defines access failure according to the following principles: after the 2006-06-03


Page 14 of 98



originated UE sends the RRC Connection Request message, access failure occurs upon any of the following conditions:    

The UE receives the RRC Connection Reject message It sends the RRC Connection Release message It receives any message from BCCH during Call setup message The timer expires. Namely, a period (T300) after the UE sends the RRC Connection Request message, it fails to receive the RRC Connection setup message

3.2.3 Definition of TEMS The TEMS defines access failure (for originated voice services) according to the following principles: 1)

Random access failure: after dialing, the UE fails to send the RRC Connection Request message. 2) The UE fails to receive the RRC Connection Setup message: after the UE send the RRC Connection Request message, it fails to receive the RRC Connection Setup message. 3) The UE fails to receive the RRC Connection Complete message: after the UE send the RRC Connection Request message, it fails to receive the RRC Connection Setup message. 4) The UE receives the RRC Connection Reject message: the UE receives the RRC Connection Reject message, and does not resend the RRC Connection Request message for try. 5) The UE fails to receive the measurement control message: after the UE sends the RRC Connection Complete message, it fails to receive the measurement control message. 6) The UE fails to send the CM Service Request message: after the UE receives the measurement control message, it fails to send the CM Service Request message. 7) The UE receives the Service Request Reject message. 8) The UE fails to receive the Call Proceeding message: after the UE sends the CC SETUP message, it fails to receive the Call Proceeding message. 9) The UE fails to receive the RB Setup message: after the UE receives the Call Proceeding message, it fails to receive RB Setup message. 10) The UE fails to send the RB Setup Complete message: after the UE receives the RB Setup message, it fails to send the RB Setup Complete message. 11) The UE fails to receive the Alerting or Connect message: after the UE receives the RB Setup Complete message, it fails to receive the RB Setup Complete message. 12) The UE fails to receive the Connect Acknowledge message: after the UE receives the Alerting or Connect message, it fails to send the Connect Acknowledge message.

3.3 Flow and Methods for Analyzing Access Failure Problems 3.3.1 Overall Flow for Analyzing Call Failure Problems 3.3.1 shows the overall flow for analyzing call failure problems.

2006-06-03


Page 15 of 98



Figure 1.2 Overall flow for analyzing call failure problems in DT/CQT By DT data analyzing tool, such as Actix Analyzer and GENEX Assistant, determine the time for Call Fail and obtain the following information:  

Pilot information collected by scanner before and after Call Fail Information about active set, monitor set, and signaling flow collected by UE

Match the signaling collected by UE and the time of single subscriber tracing by messages. Meanwhile locate the points when problems occur in single subscriber tracing on RNC. Based on signaling of single subscriber tracing on RNC and UE's signaling flow, determine the point where call fails according to 3.3.1. Analyze and solve problems according to following sub-flows. The problems include:     

Paging problems RRC setup problems RAB and RB setup problems Authentication and encryption problems Abnormal equipment problems

3.3.2 Analyzing Paging Problems Paging problems usually are: the originated UE completes RAB assignment and CC Setup and waits for Alerting message, when it receives the Disconnect direct transfer message from CN. 3.3.2 shows the signaling flow of originated UE in locating paging problems. According to the signaling flow of called UE, the signaling flow is normal. It occurred that after receiving Paging message it does not originate RRC connection setup request. According to the single subscriber tracing of called UE on RNC, the RNC receives the Paging message from CN without following messages. 2006-06-03


Page 16 of 98



Figure 1.3 Signaling flow of originated UE in locating paging problems1 The causes of paging problems include:   

The RNC does not send the Paging message. The power of paging channel and paging indicator channel is low. The UE reselects a cell.

3.3.2 shows the flow for analyzing paging problems.

1

2006-06-03

The CC CALL PROCEDING message should be after CC SETUP. Huawei Proprietary. No Spreading Without Permission.

Page 17 of 98



Figure 1.4 Flow for analyzing paging problems The following sections described detailed analysis.

I. Failure in Sending Paging Message by RNC After the RNC receives the paging message from CN, the UU interface does not send the message. Probably the capacity of paging channel is inadequate (currently the network load is low, so it is less probable that the paging channel is congested at UU interface. When the network load is heavy, it is probable) or the equipment is abnormal.

II. Over Low Power of PCH or PICH After the RNC sends a paging message, the UE fails to receive it. For this problem, check the Ec/Io of the cell where the UE camps and the cell which it monitors. If the Ec/Io of both cells is lower than –12 dB, the power of PCH or PICH is over low or the coverage is weak.

III. Cell Reselection by UE If the signals of the cell where the UE camps are weak while the signals of monitored cell are strong, the problem might be due to cell reselection. When the UE has its location area (LA) or route area (RA) updated upon paging, the paging message is sent to the original LA or RA, so the UE fails to receive paging message.

2006-06-03


Page 18 of 98



3.3.3 Analyzing RRC Connection Setup Problems The RRC connection setup failure can be analyzed through UE signaling flow and single subscriber tracing on RNC. The RRC connection setup process includes the following steps: 1) 2) 3)

The UE sends RRC Connection Request message on RACH. The RNC sends the RRC Connection Setup message on FACH. After the UE sets up downlink DCH and synchronizes with it, it sends the RRC Connection Setup CMP message on uplink DCH. RRC setup fails due to the following causes:  Uplink RACH problems  Downlink FACH power allocation ratio  Cell reselection parameter problems  Over low transmit power of downlink DCH  Uplink initial power control problems  Congestion  Abnormal equipment problems Among previous problems, the uplink RACH problem, downlink FACH power allocation ratio problem, cell reselection parameter problem, and abnormal equipment problem are probable. shows the flow for analyzing RRC connection setup problem.

Figure 1.5 Flow for analyzing RRC connection setup problem 2006-06-03


Page 19 of 98



The details analysis is as below:

I. After UE Send RRC Connection Request Message, the RNC Fails to Receive It If the Ec/Io of downlink CPICH is over low, the problem is about coverage. If the Ec/Io of downlink CPICH is not over low (such as higher than –14 dB), the problem is about RACH, with the following causes:    

The power step of preamble is small The output power of UE is lower than required. NodeB is problematic with standing wave. The parameter of cell radium is improperly configured.

If the power ramp of preamble is small, you can add the preamble ramp times. For example, increase it from 8 to 20. If the output power of UE is lower than required, there are no specific methods to solve it due to the limitation of UE performance. For NodeB problems, check whether there is standing wave alarm on NodeB. If the parameter of cell radium is set over small, the NodeB cannot synchronize with the UEs beyond cell radium. This causes access failure. This usually occurs in wide coverage scenarios like rural and suburban areas.

II. After the RNC Receives the RRC Setup Request Message from UE, It Sends the RRC Connection Setup Message Which Is Not Received by UE The causes of this problem include:  

Weak coverage Improper parameters of cell selection and reselection

Check the CPICH Ec/Io. If it is lower than –12 dB (the default value is configured based on Ec/Io as –12 dB) and 



If there is no more qualified cell listed in the monitor set, the problem is about coverage. If there is more qualified cell listed in the monitor set, the problem is about cell reselection.

Solutions are as below: 



2006-06-03

When the coverage is weak: If conditions permit, solve coverage problems by enhancing coverage, such as adding sites to cover blind areas and adjusting engineering parameters. If you cannot enhance the coverage, you can increase FACH power. Adjust FACH power according to the coverage conditions of PCPICH Ec/Io. For example, the pilot Ec/Io in all coverage areas after network optimization is larger than –12 dB, the success rate for UE to access from 3G idle mode can be guaranteed if the common channel power is allocated on the condition of Ec/Io equal to –12 dB. When Ec/Io is smaller than –14 dB, the UE reselects a GSM cell. The success rate of RRC setup in weak coverage areas after inter-RAT reselection by UE can be guaranteed if the common power is allocated on the condition of Ec/Io larger than –14 dB. Cell selection and reselection: Adjusting cell selection and reselection parameters accelerates cell selection and reselection and helps solve RRC connection setup failure problems caused by improper parameters of cell selection and reselection.


Page 20 of 98



 Note: The RRC CONNECTION SETUP message is carried by FACH. After the UTRAN side receives the PRACH preamble, the UE sends the RRC CONNECTION REQUEST message on RACH based on current preamble power. The preamble transmit power keeps increasing until response is received (restricted by maximum retransmission times of preamble). Therefore, in poor coverage areas, unbalanced coverage by RACH and FACH is probable. Consequently the UTRAN side can receive the RRC CONNECTION REQUEST message while the UE fails to receive the RRC CONNECTION SETUP message.

III. After the RNC Receives the RRC Setup Request Message from UE, It sends the RRC Connection Reject Message When the RRC Connection Reject message is present, check the cause values, which include congestion and unspecified. If the cause value is congestion, the network is congested. Check the network load conditions, including utilization of power, code, and CE resources. Determine the type of resource that causes congestion and provide ways of network expansion. For details, see W-Network Expansion Guide. The admission of RRC connection for HSDPA subscribers is consistent with that for T99 subscribers, including power, code, and CE resources. Pay special attention to code admission. If the code word of HSDPA subscribers is statically assigned, and excessive code words are assigned to HSDPA subscribers, the RRC connection of HSDAP or R99 subscribers fails probably. This is due to that the code words of downlink signaling channel for HSDPA or R99 subscribers are inadequate. If the cause value is unspecified, check the logs to determine causes of failure.

IV. After Receiving RRC Connection Setup Message, the UE Does Not Send Setup Complete Message If the downlink signals are normal, the UE might be abnormal. Otherwise initial power of downlink DCH is over low so the downlink cannot synchronize. You can solve the problem by adjusting uplink Eb/No of the service.

V. After the UE Sends the RRC Setup Complete Message, the RNC Fails to Receive It It seldom occurs that uplink initial power control leads to increment of UE transmit power. Upon presence of the problems, you can properly raise the Constant Value of DCH so that the initial transmit power of uplink DPCCH of UE increases. This problem is related to whether the initial target value of uplink SIR is rational and has great impact on uplink initial synchronization at the beginning of link setup. 



If it is set over large, the uplink interference from initial link setup of subscriber becomes over large. If it is set over small, the uplink synchronization time increases, and consequently the initial synchronization fails.

This parameter is an RNC-level parameter. It has great impact on network performance, so engineers must be cautious upon adjustment.

2006-06-03


Page 21 of 98



 Note: The RRC Connection Setup Complete message is sent on uplink DPCH. The UE calculates the initial power of DPCCH according to received IE DPCCH_Power_offest and measured CPICH_RSCP. DPCCH_Initial_power = DPCCH_Power_offset - CPICH_RSCP Wherein, DPCCH_Initial_power = Primary CPICH DL TX Power + UL Interference + Constant Value Constant Value can be configured at OMC. If it is set over small, the UE has lower power to send the RRC CONNECTION SETUP COMPLETE message than required. Current default configuration of Constant Value (the default value of it in version V13C03B151 is – 20) usually prevents this problem from happening.

3.3.4 Analyzing Authentication Problems When authentication fails, analyze the problem according to the cause value provided in the authentication failure message replied from UE to the network. Two common cause values include MAC Failure and Sysch Failure.

I. MAC Failure Check the AUTN parameter in the authentication request message send by network side upon the authentication of network by UE. If the MAC information is incorrect, the UE send the authentication failure message with the cause value MAC failure, shown as in I.

Figure 1.6 Authentication failure due to MAC Failure The major causes of the problem include:  

2006-06-03

Unauthorized subscriber USIM and HLR set different Ki and OP (OPc) for the subscriber This problem occurs frequently when a subscriber uses a new USIM. To locate this problem, check whether the Ki and OP (OPc) value of the IMSI are the same. The USIM has default Ki and OP (OPc), but the USIM reader fails to obtain the value. Therefore, the Ki and OP (OPc) of the USIM must be known upon defining a Huawei Proprietary. No Spreading Without Permission.

Page 22 of 98



subscriber or the Ki and OP (OPc) of USIM must be made the same value as in HLR.

II. Sync Failure When the UE detects that the SQN of AUTN message is incorrect, so the authentication fails. The cause value is Synch failure (synchronization failure), as shown in II.

Figure 1.7 Authentication failure due to Synch Failure The causes of the problem include:  

Authorized subscribers Equipment problems

3.3.5 Analyzing Security Mode Problems During the security mode control process, the network side sends encrypted information to radio access network (RAN). During the process, the CN side and RAN negotiate to perform encryption algorithm on UE so that the UE uses the encryption algorithm in the subsequent transfers. After the UE performs handover, it can use the encryption algorithm as possible. Namely, the encryption-related parameters are sent to the target RNC. 3.3.5 shows the security mode control.

Figure 1.8 Security mode control 3.3.5 shows the Security Mode Reject.

2006-06-03


Page 23 of 98



Figure 1.9 Security mode reject The common causes of security mode reject include:  

The UE is unable to support configured encryption algorithm. The encryption model configuration of RNC does not match that of CN. For example, the MSC configures the encryption algorithm UEA0 only but the RNC configures UEA1 only.

I. UE Capability Problems To check the UE capability, refer to the RRC Connect Setup CMP message. Currently the following UEs fail to support encryption algorithm:   

NEC single-mode UEs NEC C606 NEC C616

The following UEs support encryption algorithm:      

Nokia 7600 Nokia 6650 Moto A835 Qualcomm 6200 Qualcomm 6250 Siemens U15

To solve the security mode reject problems due to UE capability, change the UE.

II. Inconsistent Configuration of RNC and CN Encryption Mode Check the MSC or SGSN and RNC whether they have selected the same encryption mode. Namely they must have the same encryption mode. If the encryption modes are different, set the MSC and SGSN to the security mode selecting all. Set the RNC to select UEA0 or UEA1.

3.3.6 Analyzing PDP Activation Failure Problems For analysis of PDP activation failure problems, see the section 5.1 of W-PS Problem Optimization Guide.

3.3.7 Analyzing RAB or RB Setup Problems When the RAB or RB setup fails, the RNC responds RAB assignment setup failure in the RAB Assignment Response message. Locate the specific failure causes through the failure cause value contained in the related cells. Common RAB/RB setup failure problems include: 

2006-06-03

The RNC directly rejects RAB setup request due to incorrect parameter configuration. Huawei Proprietary. No Spreading Without Permission.

Page 24 of 98

W-Access Problem Optimization Guide   


Admission rejection RAB setup fails due to response of RB setup failure from UE RAB setup fails due to RB setup failure at air interface

I. Direct Rejection of RAB Setup Request by RNC Due to Incorrect Parameter Configuration It seldom occurs that the RNC directly rejects RAB setup request due to incorrect parameter configuration. This occurs due to special operations by special subscribers. It occurs when the RNC directly rejects RAB setup request because subscribing information for PS service in HLR exceeds the UE capability. For example, the traffic for a special subscriber is 384K in uplink and downlink, but the maximum uplink capacity is 64K. The subscriber sets the uplink and downlink maximum rate in QoS of activation PDP to 384K by using the AT command and UE software (SonyEricsson UE software can set QoS of activation request). When the RNC receives the RAB assignment request, it finds that the requested uplink maximum rate exceeds the UE capability, so it directly responds RAB setup failure without originating RB setup. After the RAB setup fails due to incorrect parameter configuration that exceeds UE capability, the SGSN will renegotiate to originate new RAB assignment until the UE can support and the system completes RAB assignment. For subscribers, the PDP activation can still be successful and the maximum rate obtained from indicator is the maximum rate supported by UE. If the minimum guaranteed rate requested in QoS setup in PDP activation request by UE exceeds UE capacity, the network accepts the PDP activation request by UE at a negotiated low rate; however, when the negotiated rate of network in the PDP activation acceptance message, the UE originates deactivation PDP request. Therefore, PDP activation fails finally.

II. Admission Rejection When the system resource (power, code, and CE) is inadequate, the admission is rejected and consequently call setup fails. Now you must check the uplink and downlink load, code resource, and CE resource, determine the type of resource that causes congestion, and provide corresponding expansion methods. When excessive code words are statically assigned to HSDPA subscribers, the admission fails due to inadequate code resource for non-HSDPA subscribers. When the system resource is inadequate and admission fails, the V1.5 or higher RNC conducts different operations according to RAB Downsizing Switch and IU QoS negotiation switch. For details, see the description of solving inadequate lub bandwidth. When the power of HSDPA and R99 subscribers is statically assigned, the power admission of non-HSDPA subscribers is judged by (total power of cell - the power statically assigned t HSDPA) * admission threshold. When the power of HSDPA and R99 subscribers is dynamically assigned, the power admission of non-HSDPA subscribers is consistent with that of original R99 subscribers. For the admission rejection of HSDPA subscribers, consider the following aspects: 



2006-06-03

In the method for statically assigning power of HSDPA and R99 subscribers, consider:  HSDPA subscribers supported by NodeB  HSDPA subscribers supported by cell  Total bit rate of cell  Total guaranteed bit rate  Whether the cell transmit power guaranteed bit rate exceeds the prescribed threshold In the method for dynamically assigning power of HSDPA and R99 subscribers, consider: Huawei Proprietary. No Spreading Without Permission.

Page 25 of 98


  


HSDPA subscribers supported by NodeB HSDPA subscribers supported by cell Whether the guaranteed bit rate exceeds the prescribed threshold

For non-HSDPA subscribers, when the configured bandwidth at lub interface is inadequate and the high-rate services are activated, the lub interface is rejected due to bandwidth rejection. This is common for operators outside China. For example, the transmission resource is limited, so the AAL2 bandwidth for service of lub in multiple cells can support a 384K service only. If you activate a PS 384K service after there is a 12.2K voice service, there is inadequate bandwidth at lub interface. The V1.3 RNC responds SGSN RAB assignment failure. The cause of RAB assignment Response message is that requested rate is unavailable, and then the SGSN will renegotiate to originate RAB assignment. For the RNC of V1.5 or higher, it will lower the rate internally when the RAB Downsizing Switch is on and IU QoS negotiation switch is off. If the UE can succeed in requesting lub resources after lowing rate, the RNC will respond SGSN RAB assignment setup success. If the UE fails to request lub resources after lowering rate to 8K, the RNC will respond SGSN RAB assignment setup failure. The SGSN decide to originate negotiation according to internal parameter configuration. When the RAB Downsizing Switch is on and IU QoS negotiation switch is off, the process is consistent with that of V1.3 RNC. For HSDPA subscribers, when the configured bandwidth at lub interface is inadequate, admission rejection will not occur, but the rate become lower. In addition, the AAL2PATH of HSDPA and R99 is respectively configured, and HSDPA AAL2PATH must be configured to HSDPA RT or HSDPA NRT type. If the HSDPA AAL2PATH is configured to R99 AAL2PATH RT or NRT type, RAB assignment will not fail, but the RNC will directly set up HSDPA service to R99 384Kpbs.

III. The UE's Response of RAB Setup Failure due to RB Setup Failure The UE responds RB setup failure due to subscribers' wrong operations. One case is as blow: When subscribers are using a downlink 128K data service, they receive the RB setup request of VP service (originating or terminating VP). Because most UEs cannot support performing VP and high rate PS service simultaneously, the UE directly responds RB setup failure due to unsupported configuration. The other case is as below: The UE called by 3G UE for VP service camps on GSM network, so it does not support VP service. Therefore, after the RNC receives RAB assignment request, the CN sends the Disconnect command after call proceeding due to Bearer capability not authorized. Now the UE has just received RB_SETUP command, so it has not completed RB setup. After receiving the Disconnect message, it immediately responds RB setup failure, so the RNC responds RAB setup failure due to failure in radio interface procedure.

IV. RAB Setup Failure due to RB Setup Failure at Air Interface Another RB setup failure is as below: No response to RB setup is received, so the RNC judges that RB setup fails. In details, no ACK or RB setup complete message is received for RB setup. This occurs in weak coverage areas, because the UE does not camp on the best server and originate access, or the coverage is weak. The UE does not originate to access the network in the best server, so it wishes that the best server (sharp fluctuation of signals leads to sharp fading of signals in the cell on which the UE camps) can be added to the active set during RB setup. The flows cannot be 2006-06-03


Page 26 of 98



nested (both the network and UE does not support nested flow), so the active set is updated after RB setup is complete. This leads to RB setup in weak coverage cells, so RB setup fails probably. For this case, increase the threshold and speed for starting selection of intra-frequency cells so that the UE can camp on the best server as quickly as possible. If the network load at early stage is low, the UE originates to access the network in the best server, set the threshold for starting selection of intra-frequency to –4 dB and set Treselection to 1. For cells at edge of different LACs, set the threshold lower to decrease signaling traffic of location area update. The RB setup failure due to weak coverage includes unqualified uplink and downlink coverage. The RB setup failure due to downlink weak coverage is as below: The UE fails to receive the RB setup command. Unqualified downlink coverage is partially due to poor demodulation performance of UE. It must be solved by RF optimization. The RB setup failure due to downlink weak coverage is as below: The UE receives the RB setup command, but the RAN fails to receive the ACK message or RB Setup Complete message for RB setup. This is probably due to uplink interference. Check RTWP for this.

3.4 Processing Access Delay Problems The access delay is usually affected greatly by equipment factors, so optimizing it is hard. If it is greatly different from default index, check whether the parameter configuration is consistent with the default. A typical process for calls is from sending the RRC CONNECTION REQUEST by originating UE to receiving Alerting message by UE. In terms of signaling flow, the following aspects affect access delay:     

Configuration of discontinuous cyclic period duration factor DRX Whether to disable authentication and encryption Early assignment or late assignment Whether the RRC connection is set up on FACH or DCH The impact from 13.6K and 3.4 K of signaling connection on delay

3.4.1 Configuration of Discontinuous Cyclic Period Duration Factor DRX During proceeding delay of paging one UE by another, the paging delay takes the majority. On the one hand, if the configured power of paging channel and paging indicator channel is so improper that the paging message is resent, this increase proceeding delay. On the other hand, DRX determines the time for sending paging message. An overhigh DRX leads to long delay. When DRX is 6, 7, and 8, the paging period is respectively 640ms, 1280ms, and 2560ms. In terms of statistics probability, if enough UEs originate enough calls, the traffic is in Poisson distribution and the average access delay keeps increasing. According to on-site test result, when DRX is 8, most paging delays are between 1s and 1.5s. The longest paging delay is even longer than 2.5s. When DRX is 6, the paging delay is evenly distributed between 0.35s and 0.95s. Therefore, setting DRX to 6 can effectively lower proceeding delay. Setting DRX to 6 leads to accelerated power consumption by UE, so you must consider this with the actual network conditions. At the beginning of network operation, the key is to raise the RAN performance. According to partners' signaling, such as Nokia, Ericsson, ZTE, and Lucent, setting DRX to 6 is the majority.

2006-06-03


Page 27 of 98



3.4.2 Whether to Disable Authentication and Encryption Flow According to test result, 



For voice calls, the proceeding delay after enabling authentication and encryption flow is 0.6s longer than proceeding delay after disabling authentication and encryption. For VP calls, the proceeding delay after enabling authentication and encryption flow is 0.74s2 longer than proceeding delay after disabling authentication and encryption.

For network security, combined ways of multiple authentications are used at the beginning of network normal operation. For example, 1/2 authentication is used for location area update authentication; some services, such as voice, VP, and short messages, use the synchronization method based on 1/2 authentication; other services use the Always authentication method.

3.4.3 Implementing Early or Late Assignment The different between early and late assignment lies in the different assignment time for TCH.

Early assignment

Late assignment

Terminating call

Starting assignment before the call is answered

Starting assignment after the call is answered

Originating call

Starting assignment before the Alerting message

Starting assignment after the Alerting message

Early assignment increases call completion rate. Late assignment avoids occupation of TCH resource during ringing, so it increases the utilization of TCH resource. According to test result, the UE receives Alerting message 1.28s earlier in early assignment than in late assignment. Using late assignment helps to receive response signals (ringing) from network more quickly, so it is more rational. However, the UE might fail to put through, so late assignment affects call completion rate. You must balance the advantages and disadvantages before using it.

3.4.4 Whether the RRC Connection Is Set up on FACH and DCH According to test result, when the request signaling of RRC connection originated by UE is set up on FACH, the average setup duration of voice calls is 0.601s shorter than that set up on DCH3.4K and 0.491s shorter than that set up on DCH 13.6K. The signaling of RRC connection set up on DCH 13.6K comparatively occupy more resource, it is recommendable if the resource at early stage of network operation is adequate. Note that after RRC connection setup, the signaling is reconfigured on DCH3.4K upon RB setup.

2

The data is from the research and test result of network optimization in a pilot office. The following data is also from this source. 2006-06-03


Page 28 of 98



Chapter 4 Analyzing Traffic Statistics Data Analyze traffic statistics data in the following two situations: 1) 2)

Evaluating network performance Analyze and locate the indexes with different network performance requirements, locate the problem, and enhance network performance. Network performance precaution Find the factors in advance that might lead to deterioration of network performance. This helps avoid deterioration of network performance.

This chapter addresses how to analyze traffic statistics data in the first situation. It first provides a general flow for analyzing traffic statistics data, and then describers the consideration and methods for analyzing major indexes. The version of RNC BSC6800V16C01B064.

used

here

for

analyzing

traffic

statistics

data

is

4.1 Tool for Analyzing Data The tool for analyzing traffic statistics data is GENEX Nastar.

4.2 General Methods for Analyzing Traffic Statistics Data In terms of statistics targets, analyzing traffic statistics data includes analyzing RNC-level data and cell-level data3. Analyzing RNC-level data addresses assessment and analysis of overall network indexes. Analyzing cell-level data helps locate problematic cells. The flow for analyzing RNC-level data contains the flow for analyzing cell-level data. In actual traffic statistics analysis, the flow is as below: 1) 2)

Assess overall network indexes Locate cell-level problems

Therefore, in RNC traffic statistics, the counters for failure cause type is usually cell-level. In the following sections, priority is given to cell-level indexes.

4.2.1 Flow for Analyzing RNC-level Traffic Statistics Data 4.2.1 shows the flow for analyzing RNC-level traffic statistics data.

3

2006-06-03

Due to problems on traffic statistics, current cluster-level analysis is not concerned. It is the same as RNC-level analysis. Huawei Proprietary. No Spreading Without Permission.

Page 29 of 98



Figure 1.10 Flow for analyzing RNC-level traffic statistics data Analyzing RNC-level traffic statistics data proceeds as below: 1) 2)

Check whether RNC-level traffic statistics indexes meet requirements If yes, the analysis ends. If no, find the first N cells with worst indexes and analyze the cell indexes. After locating cell problems and carrying out solutions, analyze the new data of traffic statistics If the new indexes meet requirements, the analysis ends. If there are still problems, continue the analysis until the indexes meet requirements.

For the methods and flow for analyze problematic cells, see the flow for analyzing cell-level traffic statistics data.

4.2.2 Flow for Analyzing Cell-level Traffic Statistics Data The flow for analyzing cell-level traffic statistics data is as below:

2006-06-03


Page 30 of 98



Figure 1.11 Flow for analyzing cell-level traffic statistics data The flow for analyzing cell-level traffic statistics data proceeds as below: 1) 2) 3)

Check whether there are cells with unsatisfied indexes If no, the analysis ends. If there are cells with unsatisfied indexes, analyze the detailed causes, find major causes of indexes deterioration, and provide proper solutions. After carrying out solutions, analyze the new data of traffic statistics until indexes meet requirements.

4.3 Accessibility Indexes 4.3.1 Paging Traffic Statistics Indexes When the CN pages UE, it sends RNC the PAGING message. After receiving the message, the RNC sends UE the PAGING TYPE 1 message or PAGING TYPE 2 message according to the state of UE. If the paged UE is in idle, CELL_PCH, or URA_PCH state, the RNC sends PAGING TYPE 1 message on PCCH to UEs in the paged area. If the paged UE is in CELL_FACH or CELL_DCH state, the RNC sends PAGING TYPE 2 message on DCCH to UEs in the paged area. In addition, the UTRAN sends the PAGING TYPE 1 message to the UE in idle, CELL_PCH 2006-06-03


Page 31 of 98



or URA_PCH state, which triggers the UE to modify system information. The UTRAN also sends the PAGING TYPE 1 message to the UE in CELL_PCH or URA_PCH state, which triggers state transition of UE to support data transmission. When the UE in idle state receives the PAGING TYPE 1 message, it sends RNC the RRC CONNECTION REQUEST message. After the UE in CELL_PCH or URA_PCH state receives the PAGING TYPE 1 message, it sends RNC the CELL UPDATE message. The cause for cell update is paging response. For details, see 3GPP TS 25.331 and 25.413. The indexes related to analyzing paging traffic statistics performance include:    

The UE in idle state loses paging The UE in PCH state loses paging Flow control PCH congestion

I. The UE in Idle State Loses Paging As previously mentioned, if the paged UE is in idle state, the RNC sends the PAGING TYPE 1 message to the cells in paged area. Therefore, the index is based on RNC statistics. I lists the counters related to paging loss for idle UE. Table 11.1 Counters related to paging loss for idle UE Counter name

Counter description

VS.RANAP.Paging.Att.IdleUE

It counts the times that the CN pages idle UEs

VS.RANAP.Paging.Succ.IdleUE

It counts the times that the CN succeeds in paging idle UEs

VS.RANAP.Paging.Fail.IdleUE

It counts the times that the CN fails to page idle UEs

Wherein, VS.RANAP.Paging.Fail.IdleUE is a calculated index, with the calculation as below: VS.RANAP.Paging.Fail.IdleUE = VS.RANAP.Paging.Att.IdleUE VS.RANAP.Paging.Succ.IdleUE II shows the position for counting point by counter for paging loss by idle UE.

Figure 1.12 Position for counting point by counter for paging loss by idle UE When the RNC receives the paging message from CN, and the paged UE is in idle state, the RNC takes statistics of VS.RANAP.Paging.Att.IdleUE at the point B upon sending PAGING TYPE 1 message to the cells in paged area. 2006-06-03


Page 32 of 98



When the RNC receives the RRC CONNECTION REJECT message from UE, the RNC takes statistics of VS.RANAP.Paging.Succ.IdleUE at the point C if the RRC connection setup request is due to one of the following causes:       

Terminating Conversational Call Terminating Streaming Call Terminating Interactive Call Terminating Background Call Terminating High Priority Signaling Terminating Low Priority Signaling Terminating cause unknown

The causes of paging loss by idle UE usually include: 



Parameter configuration problem. For this problem, check the paging-related parameters whether they are configured as the baseline parameters. Weak coverage. For example, the RNC cannot page a UE in indoor UE without being covered by signals, or a UE in blind elevator area.

II. The UE in PCH State Loses Paging As previously mentioned, if the paged UE is in CEPP_PCH or URA_PCH state, the RNC sends the PAGING TYPE 1 message to the cells in paged area so that the UE transits state to support data transmission. Therefore the index is based on RNC statistics. II shows the counters related to paging loss for UE in PCH state. Table 12.1 Counters related to paging loss for UE in PCH state Counter name

Counter description

VS.UTRAN.Paging1.Att

It counts the times that the RNC sends the PAGING TYPE 1 message.

VS.UTRAN.SuccPage1

It counts the times that the RNC succeed in sending the PAGING TYPE 1 message.

VS.RRC.Paging1.Fail.PchUE

It counts the times that the RNC fails to send the PAGING TYPE 1 message.

Wherein, VS.RRC.Paging1.Fail.PchUE is a calculated index, with the calculation as below: VS.RRC.Paging1.Fail.PchUE = VS.UTRAN.Paging1.Att - VS.UTRAN.SuccPage1 II shows the Position for counting point by counter for paging loss by UE in PCH state.

Figure 1.13 Position for counting point by counter for paging loss by UE in PCH state When the RNC sends the PAGING TYPE 1 message to the UE in CELL_PCH or URA_PCH state, it takes statistics of VS.UTRAN.Paging1.Att at the point A. 2006-06-03


Page 33 of 98



When the UE in CELL_PCH or URA_PCH state receives the PAGING TYPE 1 message, it sends RNC the CELL UPDATE message. The cause for cell update is paging response. When the RNC receives the cell update message from UE, with the cause paging response, it takes statistics of VS.UTRAN.SuccPage1 at the point B.

 Note: The RNC will not count the times of sending PAGING TYPE 1 message due to modification of system information.

The causes of paging loss by UE in PCH state include: 



Parameter configuration problem. For this problem, check the paging-related parameters whether they are configured as the baseline parameters. Weak coverage. For example, the RNC cannot page a UE in indoor UE without being covered by signals, or a UE in blind elevator area.

III. Flow Control When the lu interface is in flow control state, it drops the paging messages from CN. III shows the counter related to paging loss due to flow control. Table 13.1 Counter related to paging loss due to flow control Counter name VS.CN.Page.Loss.IUFC

Counter description It counts the times that IU interface drops the paging messages due to flow control

When the RNC receives the paging message from CN, and the IU interface is in paging flow control state, the IU interface drops the paging message and counts the time. When paging messages are dropped due to flow control at IU interface, the traffic of network must be heavy. Therefore, precaution to network expansion must be performed.

IV. PCH Congestion When the RNC receives paging message from CN, and the current paging flow exceeds the maximum capacity of PCH, PCH is congested and paging messages are dropped. IV lists the counters related to PCH congestion. Table 13.2 Counters related to PCH congestion Counter name

2006-06-03

Counter description

VS.CN.Page.Loss.PCHCong

It counts the times that paging messages are dropped due to PCH congestion (RNC level)

VS.RRC.Paging1.Loss.PCHCong.Cell

It counts the times that paging message are dropped due to PCH congestion (cell level)

VS.CRNC.IUB.PCH.Bandwidth

CRNC Iub PCH bandwidth

VS.MAC.CRNCIubBytesPCH.Tx

IUB PCH transport channel flow


Page 34 of 98



If paging messages are dropped in the cell due to PCH congestion, the paging traffic of the cell must have reached the maximum traffic. Check the parameters related to repeated paging and PCH. If this problem is due to heavy traffic volume, split the location area.

4.3.2 Low Success Rate of RRC Setup This section analyzes possible causes of low success rate of RRC setup. It also describes phenomena of problems about traffic statistics indexes and corresponding solutions. In traffic statistics, the major causes of RRC connection setup failure include the following types: 1) 2) 3) 4)

The RRC connection request is rejected due to lub interface failure The RRC connection request is rejected due to network congestion The RRC connection fails due to no response The RRC connection fails due to redirection

At the RNC side, the RRC connection setup failure includes two situations: 

After the RNC receives the RRC Connection Request message from UE, it sends UE the RRC Connection Reject message. This corresponds to the first two major causes listed previously. The counter starts counting at the point A shown in 4.

Figure 1.14 Position for counting point by counter for RRC connection rejection 

After the RNC sends the RRC CONNECTION SETUP message, it fails to receive the RRC CONNECTION SETUP COMPLETE or RRC CONNECTION SETUP FAILED message from UE. This corresponds to the third cause listed previously.

I. RRC Connection Request Rejection due to lub Interface Failure The RRC connection request rejection is rejected due to lub interface failure, with the following detailed causes:  

RRC connection setup rejection due to RL setup failure RRC connection setup rejection due to AAL2 setup failure

The corresponding traffic statistics counters I lists the counters related to RRC connection request rejection due to lub interface failure. Table 14.1 Counters related to RRC connection request rejection due to lub interface failure Counter name VS.RRC.Rej.RL.Fail

2006-06-03

Counter description It counts the times that RRC connection setup is rejected due to RL setup failure


Page 35 of 98


VS.RRC.Rej.AAL2.Fail


It counts the times that RRC connection setup is rejected due to AAL2 synchronization failure

The indexes listed in I are cell-level indexes. 1)

2)

RRC connection setup rejection due to RL setup failure RL setup seldom fails. It might be due to:  Hardware problems of NodeB. For example, power amplifiers are overheated (seldom).  Restricted number of CEs on NodeB. When the estimation of NodeB credits are too incorrect to actually reflect the usage conditions of NodeB CEs, the RNC judges that the NodeB CEs are enough, so the RNC sends NodeB the RL setup message. Consequently, the NodeB responds RL setup failure due to restriction of CEs. When the RL setup failure leads to that RRC connection rejected times is unequal to 0, you must check the cell load to confirm that restriction on number of CEs is not present. Check whether there are equipment alarms. Confirm that there is no failure due to air-conditioner and power amplifier problems. For example, in a period, for an operator's network, the RRC connection is rejected hundreds of times due to this cause. According to the following analysis, there are overheating alarms on NodeB and the causes might be load or ambient temperature, such as air-conditioner failure. For these problems, contact NodeB maintenance employees; otherwise, the problems will greatly affect access and handover success rate. RRC connection setup rejection due to AAL2 setup failure AAL2 setup seldom fails. It fails when AAL2 resource is restricted or the cell becomes problematic.

II. RRC Connection Request Rejection due to Network Congestion Find the type of resource that causes RRC connection request rejection due to network congestion. The congestion of radio resources includes the following types:     

Failure in application for power resource Failure in application for uplink CE resource Failure in application for downlink CE resource Failure in application for code resource Others4

II lists the counters related to RRC connection request rejection due to network congestion. Table 14.2 Traffic statistics counters related to RRC connection request rejection due to network congestion Counter name RRC.FailConnEstab.Cong

Counter description It counts the times that RRC connection setup is rejected due to network congestion. It is the total rejection times.

VS.RRC.Rej.Power.Cong

It counts the times that RRC connection setup is rejected due to failure in application for cell power resource.

VS.RRC.Rej.UL.CE.Cong

It counts the times that RRC connection setup is rejected due to failure in application for uplink CE resource.

4

Others are calculated indexes, equal to total failure times (RRC.FailConnEstab.Cong) minus previous failure times. 2006-06-03


Page 36 of 98



VS.RRC.Rej.DL.CE.Cong

It counts the times that RRC connection setup is rejected due to failure in application for downlink CE resource.

VS.RRC.Rej.Code.Cong

It counts the times that RRC connection setup is rejected due to failure in application for code resource.

The previous indexes are cell-level indexes. 1)

2)

3) 4)

Failure in application for power resource When application for power resource fails, you must check whether the configuration of admission parameters is consistent with the default. If the parameters are properly configured, you need check the current network load by traffic counter and the counter for equivalent number of subscriber. If the network load and congestion rate actually meet the expansion requirements, start network expansion. For detailed expansion methods, see W-Network Expansion Guide. Failure in application for uplink/downlink CE resource When the NodeB CE resource is inadequate; you must check the configuration of NodeB CE resource based on the current actual traffic load. For the querying method, see W-Equipment Room Guide. Failure in application for code resource When the code resource is inadequate, you must provide rational expansion methods based on actual traffic load. For details, see W-Network Expansion Guide. Others There are few such cases, and they are usually product problems. Therefore, this part neglects it.

III. RRC Connection Failure due to No Response III lists the counter related to RRC connection failure due to no response. Table 14.3 Counter related to RRC connection failure due to no response Counter name

Counter description

RRC.FailConnEstab.NoReply

It counts the times that RRC connection setup fails due to no response

The RRC.FailConnEstab.NoReply is a cell-level index. The major causes of the problem include the following two types: 



The UE fails to receive the RRC CONNECTION SETUP message from RNC. The cause of the problem is irrational configuration of coverage, cell selection and cell reselection parameters. For details, see 3.3.3. After the UE send the RRC CONNECTION SETUP COMPLETE message, but the RNC does not receive it. Maybe the initial transmit power of uplink DCH is over low. For solutions, see 3.3.3.

IV. RRC Connection Request Rejection due to Redirection After the UE sends the RRC connection setup request message, the redirection algorithm is triggered if the cell is congestion or assigning resources (mainly the admission and code resource assignment) fails, and the entire RRC direct retrial algorithms fail. If the serving cell of originating UE has inter-frequency neighbor cell or GSM cell, the UE is indicated by the IE Redirection info of RRC connection reject message to redirection to the frequency point of inter-frequency neighbor cell or GSM cell. If there is no inter-frequency neighbor cell or GSM cell, the IE Redirection info of RRC connection reject message is not configured. 2006-06-03


Page 37 of 98



IV lists the counters related to RRC connection setup rejection due to redirection Table 14.4 Counters related to RRC connection setup rejection due to redirection Counter name

Counter description

VS.RRC.Rej.Redir.Inter.Att

interfrequency cell

VS.RRC.Reject.Redir.Intrat

interfRAT cell info

These two indexes are cell-level indexes.

4.3.3 Low Success Rate of CS RAB Setup The causes of CS RAB assignment setup failure in traffic statistics include:   

Radio network problems Transmission network problems Other problems5

When CS RAB assignment fails, the counter starts counting at the point B shown in 4.3.3.

Figure 1.15 Position for counting point by counter for CS RAB assignment failure in RNC traffic statistics starting counting At the point B in 4.3.3, when the RNC sends CN the RAB ASSIGNMENT RESPONSE message with the cause failure, the corresponding counter starts working according to specific failure causes. The RB SETUP process is marked in broken line and is optional.

I. Radio Network Problems RAB assignment setup fails due to radio network problems with the following detailed types:   

CS RAB assignment setup failure due to relocation CS RAB assignment setup failure due to air interface failure CS RAB assignment setup failure due to insufficient capability

5

This index is a calculated index, equal to RAB request times minus the sum of RAB success times and other failure times. 2006-06-03


Page 38 of 98

W-Access Problem Optimization Guide 


CS RAB assignment setup failure due to other problems of radio networks6

I shows the traffic statistics counters related to CS RAB assignment setup failure due to radio network problems. Table 15.1 Traffic statistics counters related to CS RAB assignment setup failure due to radio network problems Counter name VS.RAB.FailEstabCS.RNL

Counter description It counts the times that CS RAB assignment setup fails due to radio network problems. It counts the total failure times.

VS.RAB.FailEstCS.Relo

It counts the times that CS RAB assignment setup fails due to relocation.

VS.RAB.FailEstCS.RIPFail

It counts the times that CS RAB assignment setup fails due to air interface failure.

VS.RAB.FailEstCS.Unsp

It counts the times that CS RAB assignment setup fails due to insufficient capability.

The indexes listed in I are cell-level indexes. For detailed VS.RAB.FailEstCS.Unsp, see 3. 1)

2)

3)

CS RAB assignment setup fails due to relocation When the RNC carries out relocation, it receives the RAB ASSISNMENT REQUEST message, and then it will not respond to the message but respond the RAB ASSISNMENT RESPONSE message directly to CN (due to Relocation Triggered). This index is seldom present, so neglect it. CS RAB assignment setup fails due to air interface failure After the RNC receives the RB Setup Failure message from UE, it sends the RAB Assignment Response message to CN due to Failure in the Radio Interface Procedure. To analyze CS RAB assignment setup fails due to air interface failure, you must analyze the causes of RB setup failure. For details, see RB setup failure in 4.3.5. CS RAB assignment setup fails due to insufficient capability The detailed causes of CS RAB assignment setup failure due to insufficient capability include:  Requested Traffic Class not Available (18)  Requested Maximum Bit Rate not Available (20)  Requested Maximum Bit Rate for DL not Available (33)  Requested Maximum Bit Rate for UL not Available (34)  Requested Guaranteed Bit Rate not Available (21)  Requested Guaranteed Bit Rate for DL not Available (35)  Requested Guaranteed Bit Rate for UL not Available (36)  Requested Transfer Delay not Achievable (22) CS RAB assignment setup fails due to insufficient capability when the cell is congested, such as Requested Maximum Bit Rate not Available. Note that the causes of the indexes include the following causes of failure due to radio resource congestion:     

CS RAB is rejected due to inadequate power CS RAB is rejected due to uplink CE resource CS RAB is rejected due to downlink CE resource CS RAB is rejected due to code resource CS RAB is rejected due to inadequate IUB bandwidth

6

This index is a calculated index, equal to total failure times (VS.RAB.FailEstabCS.RNL) minus previous failure times. 2006-06-03


Page 39 of 98



Others7

3 lists the traffic statistics counters related to CS RAB assignment setup failure due to insufficient capability. Table 15.2 Traffic statistics counters related to CS RAB assignment setup failure due to insufficient capability Counter name

Counter description

VS.RAB.FailEstCs.Power.Cong

It counts the times that CS RAB assignment fails due to power resource congestion

VS.RAB.FailEstCs.ULCE.Cong

It counts the times that CS RAB assignment fails due to uplink CE congestion

VS.RAB.FailEstCs.DLCE.Cong

It counts the times that CS RAB assignment fails due to downlink CE congestion

VS.RAB.FailEstCs.Code.Cong

It counts the times that CS RAB assignment fails due to code resource congestion

VS.RAB.FailEstCs.IUB.Band

It counts the times that CS RAB assignment fails due to inadequate IUB bandwidth

The previous indexes are cell-level indexes. By querying related indexes, determine the type of resource that causes failure and perform corresponding expansion solutions. 4)

CS RAB assignment setup fails due to other problems of radio networks Other causes may lead to CS RAB assignment setup failure, such as no response to RB setup. For RB setup failure, see 4.3.5. CS RAB assignment setup fails due to other problems of radio networks, and obtaining the causes directly from traffic statistics is difficult. You might locate problems by DT or other tests.

II. Transmission Network Problems The detailed causes of CS bearer setup failure which causes RAB assignment setup failure include:  

Signaling Transport Resource Failure(65) Iu Transport Connection Failed to Establish(66)

II lists the counter related to CS RAB assignment setup failure due to transmission network problems. Table 15.3 Counter related to CS RAB assignment setup failure due to transmission network problems Counter name VS.RAB.FailEstabCS.TNL

Counter description It counts the times that CS RAB assignment setup fails due to transmission network problems

The VS.RAB.FailEstabCS.TNL is a cell-level index. 7

This index is a calculated index, equal to total failure times (VS.RAB.FailEstCS.Unsp) minus times of other four failures. 2006-06-03


Page 40 of 98



If VS.RAB.FailEstabCS.TNL is present, there are probably transmission problems. You must check whether the transmission at lu interface is normal.

III. Other Causes The index seldom occurs, so this document neglects it.

4.3.4 Lower Success Rate of PS RAB Setup The causes of PS RAB assignment setup failure in traffic statistics include:    

Radio network problems Transmission network problems No resource available Other problems8

4.3.4 shows the counter for PS RAB assignment failure starting counting in RNC traffic statistics.

Figure 1.16 Position for counting point by counter for PS RAB assignment failure in RNC traffic statistics At the point B in 4.3.4, when the RNC sends CN the RAB ASSIGNMENT RESPONSE message with the cause failure, the corresponding counter starts working according to specific failure causes. The RB SETUP process is marked in broken line and is optional.

I. Radio Network Problems RAB assignment setup fails due to radio network problems with the following detailed types:     

PS RAB assignment setup failure due to parameter errors PS RAB assignment setup failure due to relocation PS RAB assignment setup failure due to air interface failure PS RAB assignment setup failure due to insufficient capability PS RAB assignment setup failure due to other problems of radio networks9

I shows the traffic statistics counters related to PS RAB assignment setup failure due to 8

This index is a calculated index, equal to RAB request times minus the sum of RAB success times and other failure times 9

2006-06-03

This index is a calculated index, equal to total failure times minus other failure times. Huawei Proprietary. No Spreading Without Permission.

Page 41 of 98



radio network problems. Table 16.1 Traffic statistics counters related to PS RAB assignment setup failure due to radio network problems. Counter name VS.RAB.FailEstPS.Par

Counter description It counts the times that PS RAB assignment setup fails due to parameter errors. It counts the total failure times.

VS.RAB.FailEstPS.Relo

It counts the times that PS RAB assignment setup fails due to relocation.

VS.RAB.FailEstPS.RIPFail

It counts the times that PS RAB assignment setup fails due to air interface failure.

VS.RAB.FailEstPS.Unsp

It counts the times that PS RAB assignment setup fails due to insufficient capability.

The indexes listed in Iare cell-level indexes. For detailed VS.RAB.FailEstPS.Unsp, see II. 1)

2)

3)

4)

PS RAB assignment setup fails due to parameter errors The detailed causes include:  Invalid RAB Parameters Value  Invalid RAB Parameters Combination  Condition Violation for SDU Parameters  Condition Violation for Traffic Handling Priority  Condition Violation for Guaranteed Bit Rate. The index is seldom present, so this part neglects it. PS RAB assignment setup fails due to relocation When the RNC carries out relocation, it receives the RAB ASSISNMENT REQUEST message, and then it will not respond to the message but respond the RAB ASSISNMENT RESPONSE message directly to CN (due to Relocation Triggered). This index is seldom present, so neglect it. PS RAB assignment setup fails due to air interface failure After the RNC receives the RB Setup Failure message from UE, it sends the RAB Assignment Response message to CN due to Failure in the Radio Interface Procedure. To analyze PS RAB assignment setup fails due to air interface failure, you must analyze the causes of RB setup failure. For details, see RB setup failure in 4.3.5. PS RAB assignment setup fails due to insufficient capability The detailed causes of PS RAB assignment setup failure due to insufficient capability include:  Requested Traffic Class not Available (18)  Requested Maximum Bit Rate not Available (20)  Requested Maximum Bit Rate for DL not Available (33)  Requested Maximum Bit Rate for UL not Available (34)  Requested Guaranteed Bit Rate not Available (21)  Requested Guaranteed Bit Rate for DL not Available (35)  Requested Guaranteed Bit Rate for UL not Available (36)  Requested Transfer Delay not Achievable (22) The index is present when the cell is congested. For detailed causes, such as Requested Maximum Bit Rate not Available, see CDL. Note that the causes of the indexes include the following causes of failure due to radio resource congestion:   

2006-06-03

PS RAB is rejected due to inadequate power PS RAB is rejected due to uplink CE resource PS RAB is rejected due to downlink CE resource Huawei Proprietary. No Spreading Without Permission.

Page 42 of 98

W-Access Problem Optimization Guide  


PS RAB is rejected due to code resource Others10

4 lists the traffic statistics counters related to PS RAB assignment setup failure due to insufficient capability. Table 16.2 Traffic statistics counters related to PS RAB assignment setup failure due to insufficient capability Counter name

Counter description

VS.RAB.FailEstPs.Power.Cong

It counts the times that PS RAB assignment fails due to power resource congestion

VS.RAB.FailEstPs.ULCE.Cong

It counts the times that PS RAB assignment fails due to uplink CE congestion

VS.RAB.FailEstPs.DLCE.Cong

It counts the times that PS RAB assignment fails due to downlink CE congestion

VS.RAB.FailEstPs.Code.Cong

It counts the times that PS RAB assignment fails due to code resource congestion

VS.RAB.FailEstPs.IUB.Band

It counts the times that PS RAB assignment fails due to inadequate IUB bandwidth

The previous indexes are cell-level indexes. By querying related indexes, determine the type of resource that causes failure and carry out corresponding expansion solutions. 5)

PS RAB assignment setup fails due to other problems of radio networks Other causes may lead to PS RAB assignment setup failure, such as no response to RB setup. For RB setup failure, see 4.3.5. PS RAB assignment setup fails due to other problems of radio networks, and obtaining the causes directly from traffic statistics is difficult. You might locate problems by DT or other tests.

II. Transmission Network Problems The detailed causes of PS bearer setup failure which causes RAB assignment setup failure include:  

Signaling Transport Resource Failure(65) Iu Transport Connection Failed to Establish(66)

II lists the counter related to PS RAB assignment setup failure due to transmission network problems. Table 16.3 Counter related to PS RAB assignment setup failure due to transmission network problems Counter name VS.RAB.FailEstabPS.TNL

Counter description It counts the times that PS RAB assignment setup fails due to transmission network problems

The VS.RAB.FailEstabPS.TNL is a cell-level index. 10

This index is a calculated index, equal to total failure times (VS.RAB.FailEstCS.Unsp) minus times of other four failures. 2006-06-03


Page 43 of 98



If VS.RAB.FailEstabPS.TNL is present, there are probably transmission problems. You must check whether the transmission at lu interface is normal.

III. No Resources Available No resource available causes PS RAB setup failure III shows the counter related to PS RAB setup failure due to no resource available. Table 16.4 Counter related to PS RAB setup failure due to no resource available Counter name VS.RAB.FlEstPS.Str.NResAvail

Counter description It counts the times that PS RAB setup fails due to no resource available

VS.RAB.FlEstPS.Str.NResAvail is a cell-level index. The resource referred in III includes the equipment resources excluding radio layer resources (such as power resource, code resource, and CE resource). VS.RAB.FlEstPS.Str.NResAvail is seldom present, so this document neglects it.

IV. Other Causes The index seldom occurs, so this document neglects it. For HSDPA service, the cause of low success rate of RAB assignment is the same as that of R99 PS RAB assignment. The traffic statistics indexes of PS RAB involve R99 PS service and HSDPA service.

4.3.5 Low Success Rate of RB Setup RB setup failure does not serve as one cause of RAB assignment failure in current version, so no special counter is used for RB setup failure. In traffic statistics of current version, CS RB setup failure and PS RB setup failure are not distinguished, so matching the traffic statistics indexes of RB setup failure and the specific causes of RAB setup failure one by one is temporarily impossible. The major causes of lower success rate of RB setup include the following two types:  

RB setup failure No response to RB setup

I. RB Setup failure RB setup failure: after the RNC sends the RB Setup message, it receives the RB Setup Failure message from UE. The detailed causes of RB setup failure include:    

Unsupported configuration Physical channel failure Cell update occurrence Invalid configuration

In traffic statistics of RNC, the counter for RB setup failure starts counting at the point A shown in I.

2006-06-03


Page 44 of 98



Figure 1.17 Position for counting point by counter for RB setup failure in traffic statistics At the point A shown in I, when the RNC receives the RADIO BEARER SETUP FAILURE message from UE, it takes statistics according to various causes of RB setup failure in the cell where UE camps. I lists the traffic statistics counters related to RB setup failure. Table 17.1 Traffic statistics counters related to RB setup failure Counter name

Counter description

VS.FailRBSetup.CfgUnsup

RB setup fails due to unsupported cell configuration

VS.FailRBSetup.PhyChFail

RB setup fails due to cell physical channel failure

VS.FailRBSetup.CellUpd

RB setup fails due to cell update

VS.FailRBSetup.IncCfg

RB setup fails due to invalid cell configuration

The previous indexes are cell-level indexes. 1)

2) 3) 4)

Unsupported configuration This is due to maloperations by subscribers. For example, when using downlink 128K data service, a subscriber receives the RB setup request (originating or terminating VP) of VP service, the UE directly responds RB setup failure due to unsupported configuration, namely, most UEs fails to support using VP and high speed (>= 64K) PS service simultaneously. Physical channel failure It seldom occurs. Cell update occurrence Cell update occurs during RB setup. It seldom occurs, so this document neglects it. Invalid configuration This is a common causes of RB setup failure. It is possible that: the 3G UE originates VP service to an terminating MS that camps on GSM cells and that does not support VP service, so after the RNC receives the RAB assignment request, the CN immediately sends the Disconnect command with the cause Bearer capability not authorized after call proceeding. Consequently, the UE receives the RB_SETUP message and has not completed RB setup, so it responds RB setup failure upon receiving the Disconnect message, and then the RNC responds RAB setup failure.

II. No response to RB Setup II lists the traffic statistics counter related to no response to RB setup.

2006-06-03


Page 45 of 98



Table 17.2 Traffic statistics counter related to no response to RB setup Counter name VS.FailRBSetup.NoReply

Counter description RB setup fails due to no response to cell RB setup

The previous index is a cell-level index. This is a common cause of RB setup failure. The RB setup fails because that the UE fails to receive RB SETUP message or the RNC fails to receive response from UE. This occurs in weak coverage areas due to two causes:  

The UE does not originates a call in the best server The coverage is weak

For locating and solving no response to RB setup, see 3.3.7.

4.4 System Availability Index 4.4.1 High Admission Rejection Rate To be supplemented.

4.4.2 High Paging Congestion Rate To be supplemented.

4.4.3 High Rate of Congested Cell To be supplemented.

2006-06-03


Page 46 of 98



Chapter 5 Solving Access Problems 5.1 Paging Problems 5.1.1 Improper Power Configuration of Paging-related Channels Two channels are related to paging: PICH and PCH. When the power of these two channels is too low to meet the requirements for UE demodulation, the UE fails to receive paging messages correctly. By default, the power of PCH is –2 dB and that of PICH is –7 dB. According to the result of parameter optimization, the power configuration guarantees the paging request that Ec/Io is larger than –12 dB. If the network coverage is even worse than –12 dB, consider raising PCH power. If the paging indexes are bad, without DT data of UE and single subscriber tracing data of RNC, you need analyze the distribution chart of network coverage conditions and check whether raising the power allocation ratio of these two channels is necessary.

5.1.2 Paging Failure due to UE Location Area Update I. Description and Analysis I shows the originating signaling flow of paging failure due to UE location area update.

Figure 1.18 Originating signaling flow of paging failure due to UE location area update According to I, the RNC receives the Disconnect message from CN. I shows the content of the Disconnect message.

2006-06-03


Page 47 of 98



Figure 1.19 Content of the Disconnect message in paging failure due to UE location area update According to I, the cause value for the Disconnect message is no route to destination. Therefore, the connection is released because the destination UE cannot be paged. I shows the terminating signaling flow of paging failure due to UE location area update.

Figure 1.20 Terminating signaling flow of paging failure due to UE location area update According to I, the terminating UE has location area and route areas updated. During the process, the UTRAN fails to page UE, so the call fails.

II. Solution No general solution is for this problem. You can rationally configure the location area and route area to avoid frequent location area update in hot spot areas.

2006-06-03


Page 48 of 98



5.1.3 Paging Failure due Implicit Detach of UE The UE has location area updated in a period of usually 2 hours. The CN has a timer, with a longer period than the timer for location area update. If the CN does not receive location area update message of the UE in the preset time by the timer, the CN originates implicit detach and sets the permitted call flag of this UE to be false; therefore, paging the UE fails. This is due to the following causes: The UE stays in the blind area permanently. The network is usually full coverage by GSM network, so the UE will reselect a GSM cell if coverage by UMTS network is unavailable in some area. This seldom occurs. The causes also include mal-operations, such as directly removing UE battery or USIM card.

5.2 Cell Selection and Reselection Problem 5.2.1 Repeating to Send the RRC Connection Request Message due to Cell Reselection I. Description and Analysis I shows the signaling of UE when cell reselection causes repeating to send RRC Connection Request message.

Figure 1.21 Signaling of UE The interval between two times of repeating to send request by UE is about 1.2s. I Signal quality when the UE sends the RRC connection request message. 2006-06-03


Page 49 of 98



Figure 1.22 Signal quality when the UE sends the RRC connection request message. I shows the Signal quality when the UE resends the RRC connection request.

2006-06-03


Page 50 of 98



Figure 1.23 Signal quality when the UE repeats to send the RRC connection request According to default system parameter configuration,    

Treselection = 1 Qhyst2 = 2 dB Qoffset2 = 0 dB Sintrasearch = 5

When the signals of target cell are stronger than the serving cell, completing reselection takes 1s. Therefore, the signal fluctuation of target cell and serving cell is similar to previous description. There is little space to optimize the parameters of cell reselection. The minimum Treselection is 1. If Treselection is set to 0, the reselection time will be 8*DRX, much longer than 1s, because the minimum DRX is set to 0.64s. If Treselection is set to 0, the Ec/Io of target cell must be 3 dB higher than that of serving cell according to protocols. After multiple statistics, the time for cell reselection is between 1.2s and 1.4s.

II. Solution To reduce the time for cell reselection as possible, modify Qhyst2 to 0, SintraSearch to 7. During walking test, ping-pong cell reselection occurs without decrement of reselection time. It is recommended that:  

2006-06-03

Qhyst2 remains 2 dB SintraSearch is set to enable UE to start intra-frequency measurement early. If the modification does not have great impact on UE power consumption, SintraSearch is set to 7.


Page 51 of 98



5.3 RRC Setup Problems 5.3.1 Improper Configuration of Parameters of Uplink Access Channel I. Description and Analysis I shows the signaling of UE in a connection process.

Figure 1.24 Signaling of UE in a connection process I shows the single subscriber tracing signaling on RNC

Figure 1.25 Single subscriber tracing signaling on RNC After the table shown in I is sorted by time order, you can see that the RNC responds to the second RRC setup connection message from UE. I shows the quality of downlink signals.

2006-06-03


Page 52 of 98



Figure 1.26 Quality of downlink signals According to I, the downlink signals are strong, so the uplink signals should be strong. Why does the first connection fail? After static test, the problem reoccurs. The problem occurs every half hour. Sometimes, call fails because repeating to send RRC connection request fails four times. According to analysis of signaling for tracing TMSI by RNC, the RNC fails to receive the RRC Connection Request message. The signal strength of the cell: RSCP ranges from –60 dBm to –70 dBm; Ec/No ranges from –2 dB to –4 dB. According to previous interference analysis, regular interference is present in the cell, as shown in I.

Figure 1.27 Regular interference in cell 248 I shows the part of magnified chart of interference.

2006-06-03


Page 53 of 98



Figure 1.28 Part of magnified chart of interference

 Note: Interference lasts for 40s each time. The last peak shown in I is not the periodic interference as previous ones. It lasts for a short time (a sampling point). The interference is present from 8:00 to 21:00. There is no interference in other time.

At first, engineers guess that the interference causes the problem. After sampling data, based on RNC messages, UE messages, and recorded RTWP, there is no interference one minute before and after call fails. Therefore the problem is irrelevant to interference. The following tests are to locate the causes of problems: 1) 2) 3) 4)

5)

2006-06-03

Test with Qualcomm handset (6200). During one-hour call, no similar phenomenon is present, but the interference is still present. This proves that Qualcomm test handset is normal. To check that the problem is not due to AICH, raise the AICH power to 0 dB. Test with Moto handset and the problem is still present. Namely, the problem is irrelevant to AICH power. Restore the AICH power to –7 dB, restore the retransmission times of preamble from 8 to 20. During the test more than one hour, the problem is not present. Signals are stable and strong during indoor static test: Ec/Io is about –3 dB and RSCP is about –50 dBm. So engineers doubt that the power of Moto handset is problematic in areas with strong signals. After engineers lower the Ec/Io to –7 dB by increasing downlink load, the RRC setup problem remains. To further confirm that the problem is not caused by interference, test after 22:00. According to the RTWP, there is no interference, but repeating to send request message occurs fours times in the test longer than one hour, with two times of Call Fail. According to comparative analysis of interference record by NodeB and system information by UE, the interference value is changed. During repeating to send request message, the interference before and after system information remain the same (–105 dB), which proves that the repeating to send request problem of Moto Huawei Proprietary. No Spreading Without Permission.

Page 54 of 98



handset is irrelevant to external interference. According to previous test, a conclusion is drawn that the problem is irrelevant to uplink interference and power configuration of AICH. Since the Qualcomm handset (6200) is normal in the test, the problem must be with uplink RACH of Moto handset.

II. Solution After engineers change the retransmission times of preamble from 8 to 20, the problem never occurs again.

5.3.2 Improper Configuration of AICH Power The power allocation of AICH directly affects the demodulation of AI by UE. If the power is over low, the UE cannot demodulate AI correctly; therefore connection to network fails. In earlier times, the AICH power is set to –12 dB, so the RRC setup problems usually occur. Now the default AICH power is –6 dB and it is enough to meet the demodulation by Moto, Qualcomm, and NEC handsets when Ec/Io is –12 dB. Different UEs varies in demodulation capability. If a UE has not performed for IOT test, focus on the power allocation of AICH if PRACH problems are with it.

5.3.3 Improper Configuration of FACH Power I. Description and Analysis I shows the signaling upon improper configuration of FACH power.

Figure 1.29 Signaling upon improper configuration of FACH power I shows the signal strength upon the first sending of RRC connection request.

2006-06-03


Page 55 of 98



Figure 1.30 Signal strength upon the first sending of RRC connection request In I,   



The second column is the signal strength of serving cell The third column is the scramble of the serving cell The fourth and fifth column is the signal strength and scramble of the best monitored cell. The signals of these two cells keep fluctuating.

I shows the single subscriber tracing signaling by RNC

Figure 1.31 Single subscriber tracing signaling by RNC Because the downlink coverage is weak, so the UE originates the RRC connection request message. Consequently, the RNC receives the RRC connection request message and sends the RRC connection setup message. However, the downlink signals are weak, so the UE fails to receive the RRC connection setup message. I shows the signaling and signal strength upon the second sending of RRC connection request after 2s.

2006-06-03


Page 56 of 98



Figure 1.32 Signaling and signal strength upon the second sending of RRC connection request When the UE sends the RRC connection request message the second time, the downlink signal strength is about –13, so the connection succeeds. According to the I, when the Ec/Io of downlink signals is lower than –12 dB, it is not guaranteed that the UE can correctly demodulate data from downlink FACH. The current FACH power is –1 dB, which is provided based on the relationship curve of FACH Ec/No and power allocation rate tested on field on the assumption that Ec/Io at cell edge is –12 dB. To raise the receiving success rate when Ec/Io power is –14 dB, raising the FACH power by 2 dB is recommended out of the consideration for the threshold for starting inter-RAT measurement.

II. Solution After the FACH power is set to 1 dB, the problem no longer exists that RRC connection fails because the UE cannot receive RRC Setup message in downlink.

5.3.4 Multiple Times of RRC Connection Request (for Service) and No RAB Assignment Request There are no alarms at RNC and NodeB side. According to test, originated out from the cell is unavailable, but (soft handover) SHO is normal. 5.3.4 shows the traced signaling at UE side and RNC side.

2006-06-03


Page 57 of 98



Figure 1.33 Traced signaling at UE side 5.3.4 shows the traced signaling at RNC side. According to UE side, after the UE sends the initial direct transfer message, it does not receive any message. Therefore, it repeats to resend the RRC connection request message after 5s.

Figure 1.34 Traced signaling at RNC side According to UU interface of RNC, after the RNC receives the initial direct transfer message, it sends the authentication request message (probably there is no authentication, so the RNC directly sends the Security mode command). But there is no response, so the RNC release RRC connection after expiration. 5.3.4 shows the BLER statistics at UE side.

2006-06-03


Page 58 of 98



Figure 1.35 BLER statistics at UE side By comparison of messages at UE side and RNC side, the UE fails to receive the measurement control and authentication request message (sometimes Security mode command). According to the BLER statistics at UE side, after the UE sends the initial direct transfer message, the BLER with 32 as the Trch ID is 100%. Therefore decoding signaling RB on downlink transport channel is all wrong, so the UE fails to receive any message from downlink DCH. The RRC setup message is sent on CCH, so the UE can receive it. From previous analysis, the downlink DCH might be problematic. 5.3.4 shows the BLER and RRC message.

2006-06-03


Page 59 of 98



Figure 1.36 BLER and RRC message at UE side According to 5.3.4, when the cell of PSC 205 is listed in active set and multiple braches of SHO are combined, the BLER increases but call drop does not occur. When signals of other cell are weak, call drops easily, as shown in 5.3.4.

Figure 1.37 BLER and RRC message at UE side The signals from the cell of SC 205 are strong, but the downlink BLER is 100%, so the call drops. Therefore, the cell of SC 205 is problematic in downlink. Meanwhile, another cell under the same NodeB is normal. Therefore engineers doubt that 2006-06-03


Page 60 of 98



a DSP of downlink NDLP board on NodeB is problematic. After connect a normal cell to the DSP, the problem is still present. Therefore, the DSP must be problematic. After NDLP reset, the DSP becomes normal. The DSP is fixed not after activation and deactivation, but after reset.

5.3.5 RRC Connection of HSDPA Subscribers Rejected due to Inadequate Code Resource I. Description and Analysis In the test of maximum throughput rate of HSDPA cells, 15 code words (SF = 16) are statically assigned to HS-PDSCH. The HS-SCCH configures 4 code words (SF = 128). The RRC connection rejection for first HSDPA subscriber to access the network fails. The cause of RRC connection rejection is congestion, namely, code resource congestion, power resource congestion, and CE resource congestion. The first UE connects to the network without other subscribers, so the cause of RRC connection rejection cannot be power and CE restriction. After tracing performance of cells under RNC and tracing assignment of code tree, engineers find: 

 

CCH uses a code word (SF = 32) (PCPICH SF=256, PCCPCH SF=256, AICH SF=256, PICH SF=256, and SCCPCH SF=64) Four HS-SCCHs use a code word (SF = 32) 15 HS-PDSCHs use 30 code words (SF=32)

When an HSPDA subscriber accesses the network, a code word (SF = 128) is necessary for 13.6K signaling. However, no code word is available now, so the RRC connection is rejected. I shows the assignment of HSDPA code tree.

2006-06-03


Page 61 of 98



Figure 1.38 Assignment of HSDPA code tree According to the assignment of HSDPA code tree in I, the number of code words assigned for HSDPA subscribers is clear, as well as the rest code words and the occupation by R99 subscribers. I shows a sample. Besides tracing code tree, engineers can obtain the cause of admission failure from RNC logs based on the time for admission rejection, and IMSI. I shows RNC log for HSDPA admission rejection.

Figure 1.39 RNC log for HSDPA admission rejection 2006-06-03


Page 62 of 98



II. Solution After engineers assign 14 code words for HS-PDSCH, the HSDPA subscriber succeeds in access the network. When code words are statically assigned to HSDPA subscribers, the admission is usually rejected due to code word restriction. The PS 384Kbps R99 subscribers and other R99 subscribers in the cell use most code words, so the admission of HSDPA subscribers in downlink is rejected because the DCH obtains no code word when HSDPA subscribers connect to the network.

5.4 RAB and RB Setup Problems 5.4.1 RAB Setup Failure due to Inadequate Resource I. Description and Analysis A UE originates a call in a cell with strong signals, but fails. I shows the signaling of Disconnect after completion of RB setup.

Figure 1.40 Signaling of Disconnect after completion of RB setup According to I, the UE receives the Disconnect message after completion of RB setup. The CN releases the connection, so the connection fails. The cause value of Disconnect is requested circuit channel not available. The terminating UE is in the equipment room of RNC. There is indoor coverage system in the equipment room, so the coverage is good. But excessive subscribers are using the network, so the network is congested and the connection fails.

II. Solution Solve access failure problems due to inadequate capacity by network expansion. This failure occurs under a special background, because excessive subscribers use the network in the equipment room. Therefore, the test becomes problematic. To guarantee normal test, engineers must restrict the number of subscribers using the network in equipment room. This problem must be noticed during optimization.

5.4.2 Handover Failure before Completion of Signaling Flow I. Description and Analysis The UE might hand over from RRC setup completion to RAB assignment or after RB setup completion. If handover fails during this period, the subscriber might feel that connection to the network fails. I shows the signaling of UE upon a connection failure according to Analyzer software.

2006-06-03


Page 63 of 98



Figure 1.41 Signaling of UE upon a connection failure I shows the single subscriber tracing signaling by RNC.

Figure 1.42 Single subscriber tracing signaling by RNC I shows the signal strength before release of connection.

2006-06-03


Page 64 of 98



Figure 1.43 Signal strength before release of connection According to RNC signaling, the signals from the cell of SC 121 attenuate sharply. The cell of SC 56 needs adding to the active set, the UE cannot receive the ActiveSet Update message from RNC.

II. Solution To solve the problem, adjust the SHO parameters to enable the target cell to be added to the active set earlier as possible. For details, see guidebooks related to call drop analysis.

5.4.3 Admission Failure due to HSDPA Total Bit Rate Threshold Exceeded by HSDPA Bit Rate of Cell I. Description and Analysis After the RNC enables the admission switch of HSDPA subscribers, it sets the number of admission subscribers to 16. The maximum rate upon registration and for RAB assignment is 2048 kpbs. The CELLCAC command involves two parameters as below: 



Average HSDPA throughput per HS-PDSCH code It is 200 kpbs by default. Multiplier factor HS-PDSCH transport channel It is 10.

Configure 13 codes to HS-PDSCH when multiple subscribers connect to the network. When the fifteenth HSDPA subscriber accesses the network, the RNC admission is rejected. The first 14 subscribers is connecting to the network and downloading data. After engineers change average HSDPA throughput per HS-PDSCH code to 300 kpbs, more than 16 subscribers can access the network. The following information is from analysis of tracing log of RNC: 









2006-06-03

BM_CraCacHsDschMaxReqRateAdm: DL Hsdpa Admission Fail! ulPreHsDschMaxReqRate=30720000 ulHsDschMaxReqRateThd=26000000 NBM_CraCacHsDschMaxReqRateAdm: DL Hsdpa Admission Fail! CellHsDschMaxReqRate=28672000 HsPdschNum=13 RatioTotalRate=10 usKave=200 NBM_CraCacHsdpaAdm: Hsdpa Admission Failure Because Hsdpa Max Req Rate over threshold NBM_CraProcPreCac: Cell ulUCId= 656374 Reject RncapInst= 44224 Request Because Hsdpa Adimisson Failed . NBM_CraPreAdmUserRsrc: UCId 656374 Alloc Power Resource Failure. Huawei Proprietary. No Spreading Without Permission.

Page 65 of 98




    


NBM_CraExecDchUserReqFlow: UCId 656374 RncapInst 44224 Alloc Cell Resource Failure. NBM_CraSendCellRrFailRsp: SF128 mapping: 4294967295 4294967295 4294967295 4278321151. File Name: D:\boardproject\V16C01B061\epu/app/src/rr/rncap/rab/RncapRabSetup.c Line No: 1709 SRC WARNING->Err In PS RAB Setup: RB Setup Fail. Cause = 168658899 ulErrorID is 184944608 CORRM_FrcRabSetupRslt: RAB setup result is failure, error code = 168658899

Obviously, the admission for the 15th subscriber is rejected because the HSPDA bit rate after the access of 15th subscriber exceeded the HSPDA total bit rate, so the admission fails. The analysis is as below: Configure 13 codes to HS-PDSCH. The rate of each code is 200 kpbs. The multiplier factor of HS-PDSCH transport channel is 10. Therefore, the maximum requested rate of a cell is 13*200Kbps*10=26000000bps. However, during test, after the 14th subscriber accesses the network, the HSDPA admission is accepted. Therefore, CellHsDschMaxReqRate=14*2048Kbps = 28672000bps, which exceeds the threshold. Therefore, HSDPA admission of fifth subscriber fails.

II. Solution After engineers change the average HSDPA throughput per HS-PDSCH code to 300 kbps, 16 subscribers can access the network. The problem is solved.

5.5 Authentication Problems To be supplemented.

5.6 Security Mode Problems I. Description and Analysis I shows the rejection messages in security mode during tracing single subscriber.

Figure 1.44 Rejection messages in security mode I shows the 15:33:28(28).

2006-06-03

content

of

the

RANAP_SECURITY_MODE_REJECT message


Page 66 of 98

at



Figure 1.45 Content of the RANAP_SECURITY_MODE_REJECT message According to I, the cause of security mode rejection is conflict with already existing integrity protection and or ciphering information. This cause means that the latest integrity protection or ciphering information is inconsistent with the configuration. I shows the ciphering mode information configured in previous security mode command.

Figure 1.46 ciphering mode information configured in previous security mode command According to I, the encryption algorithm configured in the security mode command is no encryption, namely, no encryption is conducted on the message. According to previous command, another security mode command is found, as shown in I.

Figure 1.47 Security mode message Setting the security mode succeeds. Check the RANAP_SECURITY_MODE_COMMAND message, as shown in I.

2006-06-03


Page 67 of 98



Figure 1.48 Content of the RANAP_SECURITY_MODE_COMMAND message According to I, there are two encryption methods: UEA1 and no encryption. According to protocols, there are two encryption methods: encryption and no encryption, engineers need select the algorithm to be encrypted. Namely, UEA1 is needed here. By comparison of these two encryption mode commands, they are from different CN domain: The CN domain No. of the successful command is 4669, while that of the failed one is 4666. Namely, the CS and PS domain configures different encryption methods. After the RNC receives the security mode command shown in I, it selects the UEA1 as the encryption algorithm. It then receives the security mode command shown in I and this command requires no encryption. Therefore, the RNC rejects this security mode command.

II. Solution After setting the CS and PS encryption algorithm to the same, engineers solve the problem successfully.

5.7 Abnormal Equipment Problems There are various problems about abnormal equipment. Some are about network equipment. Others are about UEs. Engineers must analyze the problem in details. The following sections provide a typical example.

5.7.1 Abnormal NodeB I. Description and Analysis The UE cannot connect to the network from a cell all the time during DT. The UE keeps sending the RRC Connection Request message. According to single subscriber tracing by RNC, the RNC receives the RRC Connection Request message and responds the RRC Connection Setup message which the UE fails to receive. I shows the signaling of UE upon failure in receiving RRC Connection Setup message.

2006-06-03


Page 68 of 98



Figure 1.49 Signaling of UE upon failure in receiving RRC Connection Setup message I shows the single subscriber tracing signaling by RNC.

Figure 1.50 Single subscriber tracing signaling by RNC I shows the normal signal strength upon occurrence of problems.

Figure 1.51 Signal strength upon occurrence of problems

2006-06-03


Page 69 of 98



The messages at IUB interface of NodeB and internal message are normal and without alarms according to tracing. To further locate problem in the NodeB equipment room, by test, the UE can sometimes access to the network by antenna where the Ec/Io is about –3 dB, but in rest time it fails with the same phenomena. The RSCP is about –70 dBm. If the UE moves farther, the Ec/Io is about –5 dB and the UE cannot connect to the network in a probability of 80%. The phenomenon is that the UE fails to receive setup message in downlink. According to detection of NodeB console, the output power of NodeB is 24 dBm, but the normal output power is 36 dBm. Therefore the power amplifier is problematic.

II. Solution After changing the power amplifier, engineers solve the problem successfully.

5.7.2 Abnormal UE There are abundant phenomena about abnormal UE. The following paragraphs provide an example.

I. Description and Analysis The UE cannot connect to the network for a period. I shows the signaling of UE.

Figure 1.52 Signaling of UE I shows the downlink signal quality.

2006-06-03


Page 70 of 98



Figure 1.53 Downlink signal quality In I,    

The second column is the downlink scramble measured by UE. The third column is the CPICH Ec/Io measured by UE of the cell. The fourth and fifth column are the Ec/Io and scramble measured by scanner. The quality of the signals measured by UE and the signals measured by scanner is much different from each other during the unconnected time.

II. Solution This problem is related to UE performance. There are no more solutions except changing UE.

2006-06-03


Page 71 of 98



Chapter 6 Summary Compared with V2.0, V3.0 addresses operability. It is closer to on-site engineers and can guide on-site engineers to solve actual problems during network optimization. Therefore this guidebook has more updated parts compared with the V2.0 guidebook and addresses the flow for analyzing problems. It guides engineers to solve problem step by step. The fundamental knowledge serves as appendix for reference by on-site engineers. Compared with V3.0, the V3.1 guidebook adds the following content:    

RRC connection of HSPDA service Analysis and cases of admission failure in RAB assignment process HSDPA-related DT and traffic statistics indexes Some traffic statistics indexes according to that of RNC 1.6C01B064.

In the following versions, the following content needs adding or updating:  



2006-06-03

Analyzing of HSUPA access problems Updating the methods for analyzing traffic statistics and detailed indexes according to RNC version Updating the analysis of common problems


Page 72 of 98



Chapter 7 Appendix 1: Paging Process 7.1 Paging Origination During paging, paging messages are sent on paging control channel (PCCH) to UE in idle mode CELL_PCH or URA_PCH state. The CN might request paging, such as setting up a signaling connection. The UTRAN originates paging the UE in CELL_PCH or URA_PCH state to trigger cell update process. In addition, the UTRAN originates paging the UE in idle mode, CELL_PCH or URA_PCH state to enable UE to read updated system information. In details, the network side originates paging in the following conditions.

7.1.1 Paging by CN The CN originates paging so that the CN can request UTRAN to connect to UE. The paging process is the signaling process without connection at IU interface. The CN triggers paging by sending paging messages. The UTRAN sends the paging message from CN to UE in the paging process at UU interface so that the paged UE is connected to CN.

7.1.2 Paging by UTRAN When the system information changes, the UTRAN triggers paging process to inform UE in idle mode, CELL_PCH and URA_PCH state of updating system information. After this, the UE reads the updated system information. To trigger the UE in idle mode, CELL_PCH and URA_PCH state for state transition (such as transiting to CELL_FACH state), the UTRAN originates a paging process. As a response to the paging, the UE originates cell update or URA update.

7.2 Paging Flow 7.2.1 Paging Type 1 To setup a call, the CN sends paging message to UTRAN through lu interface. The UTRAN sends the paging message from CN to UE in the paging process at UU interface so that the paged UE is connected to CN. Paging messages are sent in non-connection message at IU interface. The RNC sends PAGING TYPE 1 message on PCCH in the following two situations: 



After the RNC receives the paging message from CN, the cell Non Searching Indication is specified to non-searching (the RNC does not search whether the UE is in the connected state) in the paging message. After the RNC receives the paging message from CN, the cell Non Searching Indication is specified to searching; but the UTRAN cannot find SRNTI (UE in idle state) by IMSI.

If the paging message at IU interface contains LAI or RAI, the RNC will send the PAGING TYPE 1 message to all cells in the specified location area or routing area. If the paging message at IU interface contains no LAI or RAI, the RNC will send the PAGING TYPE 1 message to all cells under the RNC. 2006-06-03


Page 73 of 98



Besides previous situations, the UTRAN sends the PAGING TYPE 2 message on DCCH, which is called the cooperation paging. 7.2.1 shows the flow chart of PAGING TYPE 1 message. CN

RNC1

RNC2

NODEB1. 1

NODEB2. 1

UE

PAGI NG RANAP

RANAP

RANAP

PAGI NG

RANAP

PCCH: PAGI NG TYPE 1

PCCH: PAGI NG TYPE 1

Figure 1.54 Flow chart of PAGING TYPE 1 message According to 7.2.1, the CN originates paging in a location area which is distributed under two RNCs. After the RNC receives the paging message, it searches for the matching cells and calculates the paging occasion. It sends the PAGING TYPE 1 message to the cell on PCCH at the paging occasion.

7.2.2 Paging Type 2 7.2.2 shows the flow chart of PAGING TYPE 2 message. CN

SRNC

UE

PAGI NG RANAP

RANAP

DCCH:

PAGI NG TYPE 2

RRC

RRC

Figure 1.55 Flow chart of PAGING TYPE 2 message

2006-06-03


Page 74 of 98



According to 7.2.2, if the UTRAN judges that the paging is cooperation paging, the UE must be in CELL_DCH or CELL_FACH state, so the UTRAN immediately sends the paged UE the PAGING TYPE 2 message on DCCH. If the UE is in CELL_PC or URA PCH state, the UTRAN sends the PAGING TYPE 1 to UE. After the UE receives the PAGING TYPE 1 message, it originates the cell update process to transit to CELL_FACH state. Conclusion: If the UE is in CELL_DCH or CELL_FACH state, the network side sends the PAGING TYPE 2 message. If the UE is in other state, the network side sends the PAGING TYPE 1 message.

7.3 Behaviors of UE after Receiving Paging The UTRAN can page multiple UEs in the same paging occasion. The information about the paged UE is contained in the Paging record of PAGING TYPE 1 message. The UE must check every event in Paging record upon receiving PAGING TYPE 1 message. The UE needs to compare the identity of each event and the identity of UE.

7.3.1 UE in Idle Mode If the UE is in idle mode, for the presence of each Paging record IE in the message, the UE will:  If the IE Used paging identity paging originator is a CN identity, 1) Compare the IE type of UE identity type in CN IE and the identity of all assigned UE. 2) If one pair matches, this means that UE accepts the paging and forwards the IE CN domain identity, UE identity, and Paging cause to upper layer.  Otherwise, the UE neglects the paging record.

7.3.2 UE in Connected Mode If the UE is in connected mode, for the presence of each Paging record IE in the message, the UE will: 



If the IE Used paging identity is UTRAN, and the U-RNTI is the same as assigned U-RNTI of UE:  If the Paging record contains the optional IE CN originated page to connected mode UE, the UE responds that paging is received, and it sends the upper-layer the IE CN domain identity, Paging cause, and Paging record type identifier.  If the Paging record does not contain the IE CN originated page to connected mode UE, the UE starts cell update process with the cause paging response. If the IE is Used paging identity is not UTRAN, the UE will neglects the paging record. If the Paging record contains the IE BCCH modification info, the UE in idle mode, CELL_PCH or URA_PCH state must read the system information again, without reading content of Paging record.

7.4 DRX Process of UE 7.4.1 DRX Cyclic Length and Paging Occasion UE monitors paging in idle mode in two ways:  

2006-06-03

Decode data on SCCPCH every 10ms directly Decode data on PICH periodically. Only when there is paging indicator, the UE will Huawei Proprietary. No Spreading Without Permission.

Page 75 of 98



decode associated data on SCCPCH, namely, DRX. The UE reduces power consumption by using DRX. In idle mode, the DRX paging length is calculated as below:

DRX _ Cycle _ Length = 2 K × PBP Wherein, 



K is the IE CN domain specific DRX cycle length coefficient, which is broadcast in system information. Now the K in CS and PS domain is 8. PBP is the paging block period. In FDD mode, the PBP = 1.

Therefore, the previous formula is simplified to: DRX_Cycle_Length = 28 (1) The UE paging occasion is calculated as below: Paging Occasion (CELL SFN) = {(IMSI mod M) mod (DRX cycle length div PBP)} * PBP + n * DRX cycle length + Frame Offset Wherein, 

 

n = 0, 1, 2…. . It is necessary to prove that Paging Occasion is smaller than the maximum value of SFN: 4096. Frame Offset = 0 (in FDD mode) M is the number of SCCPCHs that bears PCH, usually equal to 1

Therefore the previous formula is simplified to: SFN = IMSI mod 2K + n * 2K (2) The UE only need to monitor paging indicator on PICH frames. 7.4.1 shows the schematic drawing of UE paging occasion.

Figure 1.56 Schematic drawing of UE paging occasion The UE must monitors the frames (paging occasions) indicated by red dots in each paging period, and then decode the qth PI. For the calculation of q, see the formula (3).

2006-06-03


Page 76 of 98



7.4.2 Relationship of PICH and SCCPCH The paging indicator channel (PICH) is a physical channel of fixed rate (the spreading factor is 256). It carries paging indicator. The PICH is relevant to SCCPCH mapped by PCH. 7.4.2 shows the frame structure of PICH. 12 bits (transmission off)

288 bits for paging indication b0 b1

b287 b288

b299

One radio frame (10 ms) Figure 1.57 Frame structure of PICH A 10ms (length) PICH consists of 300 bits (b0, b1… b299). Wherein, the first 288 bits (b0, b1…b288) are to carry paging indicator. The rest 12 bits is for following use. Each PICH frame carries NP paging. NP is the number of paging indications per frame. It defines the maximum paging indicators supported by each frame on PICH. The UE obtains the value of NP in cell system information. The NP is 18, 36, 72, and 144; namely, the 288 bits are divided by NP, so each division has 288/NP bits. Each division is a paging indicator. 7.4.2 describes the mapping relationship between {PI0, .., PIN-1} and PICH bits {b0,..,b287}. Table 57.1 Mapping relationship between PI and PICH Number of PI per frame (NP)

PIp = 1

PIp = 0

NP=18

{b16p, .. b16p+15} ={1,1,..,1}

{b16p, .. b16p+15} = {0,0,..,0}

NP=36

{b8p, .. b8p+7} = {1,1,..,1}

{b8p, .. b8p+7} = {0,0,..,0}

NP=72

{b4p, .. b4p+3} = {1, 1,1,1}

{b4p, .. b4p+3} = {0, 0,0,0}

NP=144

{b2p, b2p+1} = {1,1}

{b2p, b2p+1} = {0,0}

The UE determines by calculating its paging indicator suffix p that p is relevant to the qth bits of PICH frame.

 Np    q =  PI + ( (18 × ( SFN +  SFN / 8 +  SFN / 64 +  SFN / 512  ) ) mod 144 ) ×  mod Np (3) 144     Wherein,  

PI = DRX index mod NP = (IMSI div 8192) mod NP. SFN is the paging occasion for UE. It is the PCCPCH SFN when PICH is present.

From the formula (3), the UE can know the suffix of PI so that the UE can monitor relevant bits on PICH only. Once the UE detects that the bits are set to 1, it knows that it is paged. It starts receiving and decoding paging messages from 7680 chips after completion of PICH radio frame. 2006-06-03


Page 77 of 98



7.4.2 shows the time sequence relationship between PICH and SCCPCH.

PICH frame containing paging indicator Associated S-CCPCH frame τ PICH

Figure 1.58 Sequence relationship between PICH and SCCPCH The end of PICH radio frame is 7680 chips earlier than associated S-CCPCH frame.

7.4.3 PCH Selection The system information block 5 (SIB5) defines the PCH used in idle mode. In a cell, one or more PCHs are built. In system information, each SCCPCH bears a PCH, so each prescribed PCH corresponds to a unique PICH. If the SIB5 defines more than one PCH and relevant PICH, the UE selects SCCPCH listed in SIB5 based on IMSI as below: Index of selected SCCPCH = (IMSI div ((“DRX cycle length”div PBP)*Np*NPICH)) mod K, Wherein, the K equals to the number of SCCPCHs bearing PCH (for example, those SCCPCHs which bear only one FACH does not counter). These SCCPCHs are marked by 0 to K-1 in the order contained in SIB5. The K is usually 1. Namely, only one SCCPCH bears PCH. The Index of selected SCCPCH shows the SCCPCH selected by UE-used PCH and the unique corresponding PICH identity. Now Huawei configures only one PICH and a SCCPCH for a cell. The SCCPCH bears two FACHs and a PCH.

7.4.4 DRX Examples of UE After the cell is set up, the parameter configuration about paging in the broadcast system information is:  

CN domain specific DRX cycle length coefficient = 8 Number of PI per frame = 36

After the UE receives the information, it calculates the paging occation, PI, and p. The IMSI of a UE is 448835805669362, so the related parameters are calculated as below:    

DRX cycle length = 28 = 256 Cell SFN = 448835805669362 mod 28+ n * 28 = 242 + 256 * n (n = 0, 1, 2...) PI = (448835805669362 div 8192) mod 36 = 14 q = (14 + [((18 * (242 + [242 / 8] + [242 / 64] + [242 / 512])) mod 144) * 0.25]) mod 36 = 27

From previous data, each frame of the cell PICH carries 36 PIs. Each PI consists of 8 bits (288/36). The UE must monitor the bit216 (27x 8) to bit223 of each PICH radio frame. If these 8 bits changes to 1, the UE knows that it might be paged, so it receives paging message on SCCPCH. 2006-06-03


Page 78 of 98



Chapter 8 Appendix 2: Access Process Analysis The UE has two basic operation modes: idle mode and connected mode. When the power is on, the UE is in idle mode. It is identified by non-access layer identity, such as IMSI, TMSI, or P-TMSI. The UTRAN does not save the information of UE in idle mode. It pages respectively the UE that powers on and camps on a cell, or pages all UE in idle mode under an RNC simultaneously. After UE completes RRC connection setup, UE transits from idle mode to the CELL_FACH or CELL_DCH state of connected mode. After the RRC connection is released, the UE transits from connected mode to idle mode. According to access layer, the access process is the process of UE transiting from idle mode to connected mode. It includes:    

Cell search Receiving system information broadcast in the cell Cell selection and reselection Random access

Once the UE is in connected mode, it can carry out the following non-access layer activities:    

PLMN selection and reselection Location registration Service application Authentication

8.1 Cell Search When the UE is searching for a cell, it selects the corresponding process according to whether there is information about RF channel of UTRA carrier. 



If the UE has no information about RF channel of UTRA carrier, it scans all frequency bands within UTRA band to find the proper cell to camp on in the selected PLMN. In each carrier, the UE searches for the cell with strongest signals. If the UE has the saved information about UTRA carrier and cell parameters (such as primary scramble of cell) obtained from previously received measurement control information, the UE tries to camp on the saved cell. If it fails to camp, it scans all frequency bands within UTRA band to find the proper cell to camp on in the selected PLMN.

After the UE locks a frequency, it completes cell search by timeslot synchronization, frame synchronization, and scramble synchronization.

8.1.1 Timeslot Synchronization All synchronization codes of primary SCH in UTRAN are the same, so they are sent in the first 256 chips of each timeslot. The UE can easily synchronize by using a matched filter or similar technologies.

8.1.2 Frame Synchronization and Scramble Group Identification Frame synchronization is fulfilled by using synchronization code of secondary SCH. The secondary SCH has 16 synchronization codes, and they are various in each timeslot. They form 64 groups of code sequence according to code word of each timeslot. The 64 groups of code sequence features that the result after their cyclic shift is unique. The cell scramble group and frame synchronization is determined by performing SSC corelation, FWHT and 2006-06-03


Page 79 of 98



RS decoding on secondary synchronization.

8.1.3 Identification of Cell Primary Scramble In 8.1.2, the UE obtains the scramble group of the serving cell. Each scramble group has 8 primary scrambles. The UE keeps searching for the most relevant scramble according to symbol corelation until it determines the primary scramble. After obtaining the code word, the UE can read data from broadcast channel because both the CPICH and PCCPCH use the same scramble and their channel codes are fixed.

8.2 Cell Selection and Reselection After the UE powers on, it searches for cells. After this , it judges whether the current PLMN is suitable according to system information. If the current PLMN is suitable, it performs cell measurement. It judges according to the criterion S whether camping on the cell is suitable. This is the cell selection process. If the current cell fails to meet the criterion S, the UE starts PLMN selection and cell reselection (First, the UE performs cell reselection in the current PLMN. If no cell meets the conditions, the UE searches for PLMN, and performs cell selection and reselection in other PLMNs). It measures neighbor cells, and then it sorts the measured cells by the criterion R or the criterion H. If a cell meets the criterion S, the UE can camp on it. Cell selection and resection does not only start upon power on, but due to other causes.

8.2.1 Cell Selection The following parameters describe the trigger time, process, and principles for judging suitable cell for cell selection.

I. Trigger Time The UE starts cell selection in the following situations:    

The UE powers on The UE transits from connected mode to idle mode The UE lost cell information in connected mode When the UE fails to find a cell for normal camp for cell reselection according to the cell list provided by measurement control system information (TS25.133)

II. PLMN Selection The UE obtains the PCCPCH scramble according to 8.1.3. The channel code (SF (=256,1)) of PCCPCH is known, and it is unique in the whole UTRAN. Therefore the UE can read the information on the broadcast channel. First, the UE obtains SFN from system information sent on BCH (PCCPCH). The first domain of the message is SFNprime. Its value is the initial SFN of the transport block, with its range (0, 2, 4, 6…4094). The rage of SFNprime is (0…2047) after PER coding. The BCH TTI is 20ms. It includes two radio frames, so the step of SFNprime is 2. The scheduling information is known; namely, SIB_POS = 0 and SIB_REP = 8. After the UE obtains SFN, it can read MIB in the radio frame (SFN = 0, 8, 16…). After reading MIB, the UE judges according to PLMN identity in MIB whether the current PLMN is the needed PLMN. 

2006-06-03

If yes, the UE searches for other SIBs according to the scheduling information of other SIBs contained in MIB, and obtain their content. Huawei Proprietary. No Spreading Without Permission.

Page 80 of 98



If no, the UE starts cell search from the next frequency.

8.2.2 Judgment Criterion (Criterion S) If the current PLMN is the PLMN needed by UE, the UE reads SIB3 for information about cell selection and reselection. In the IE (Cell selection and re-selection info for SIB3/4),, the UE obtains the following parameters:    

Qqualmin Qrxlevmin Maximum allowed UL TX power (UE_TXPWR_MAX_RACH) Other parameters

After obtaining previous parameters, the UE judge with the criterion S whether the current cell is suitable to camp on. The criterion S is: If Srxlev > 0 and Squal > 0, the cell is suitable for UE to camp on. Wherein,

Squal = Qqualmeas − Qqual min Srxlev = Qrxlevmeas − Qrxlev min − Pcompensation 8.2.2 lists the parameters and their description in the criterion S. Table 58.1 Parameters and their description in the criterion S Parameter name

Parameter description Cell Selection quality value, (dB)

Squal

Not applicable for TDD cells or GSM cells.

Srxlev

Cell Selection RX level value (dB)

Qqualmeas

Measured cell quality value. The quality of the received signal expressed in CPICH Ec/N0 (dB) for FDD cells. Not applicable for TDD cells or GSM cells.

Qrxlevmeas

Measured cell RX level value. This is received signal, CPICH RSCP for FDD cells (dBm), P-CCPCH RSCP for TDD cells (dBm) and RXLEV for GSM cells (dBm).

Qqualmin

Minimum required quality level in the cell (dB). Not applicable for TDD cells or GSM cells.

Qrxlevmin

Minimum required RX level in the cell. (dBm)

Pcompensation

max(UE_TXPWR_MAX_RACH – P_MAX, 0) (dB)

UE_TXPWR_MAX_RACH

Maximum TX power level an UE may use when accessing the cell on RACH (read in system information), (dBm)

P_MAX

Maximum RF output power of the UE, (dBm)

If the cell meets the criterion S, the UE judges the cell as a suitable cell. Therefore it camps on the cell, reads other needed system information, and originates location registration. 2006-06-03


Page 81 of 98



If the cell does not meet the criterion S, the UE searches for the cell meeting the criterion S in the neighbor cells of the cell in the following procedures.

I. Intra-frequency Neighbor Cell The UE reads the following parameters from SIB11:        

Measurement control system information Intra-frequency measurement system information Intra-frequency cell info list Cell info Primary CPICH info Reference time difference to cell Cell Selection and Re-selection info for SIB11/12 Others

In the CPICH info, the UE obtains primary scrambling code. The channel code of CPICH is unique in the entire UTRAN. The UE can measure Qqualmeas and Qrxlevmeas (timeslot synchronization and frame synchronization are needed) according to primary scrambling code and reference time difference to cell. In the Cell Selection and Re-selection info for SIB11/12, the UE obtains the following parameters of neighbor cell:   

Maximum allowed UL TX power Qqualmin Qrxlevmin

After this, the UE can calculate the Squal and Srxlev of neighbor cells, and judges whether the neighbor cell meets the previous criterion.

II. Inter-frequency Neighbor Cell The UE reads the following information from SIB11:     

Inter-frequency measurement system information Inter-frequency cell info list Frequency info and cell info Cell info Others

The Frequency info contains:  

UARFCN uplink (Nu) UARFCN downlink (Nd)

By previous information, the UE can calculate the Squal and Srxlev of neighbor cells, and judges whether the neighbor cell meets the previous criterion. If the UE finds that no cell meets the criterion S, it judges that there is no coverage. Therefore it continues PLMN selection and reselection. In addition, the UE in idle mode randomly monitors the signal quality of the serving cell and neighbor cells to select a best cell for service. This is cell reselection. If the UE finds a neighbor cell that meets the criterion S, the UE camps on the cell and reads other needed system information. After this, the UE starts random access and originates location registration.

8.2.3 Cell Reselection The UE completes the following tasks in normally camped state in UTRAN:  

2006-06-03

Monitor PCH and PICH as indicated by system information Monitor related system information Huawei Proprietary. No Spreading Without Permission.

Page 82 of 98



Carry out cell measurement and provide data for evaluating cell reselection

The following paragraphs introduce the trigger time and measurement rule for cell reselection, as well as the principle for evaluating cell reselection.

I. Trigger Time UE reselects a cell in the following conditions: 





Idle mode time trigger (measured quality value of the serving cell is lower than that of intra-frequency measurement threshold). In idle mode, the serving cell in continuous Nserv DRX cannot meet the criterion S (however the system information is configured). When the UE detects itself in non-service area.

II. Measurement Rules The measurement rules when HCS is not used: If the cell broadcast system information indicates not to use HCS, the UE decides to start the corresponding measurement. In the CPICH Ec/Io measurement state, the Squal corresponds to Sx. In CPICH RSCP measurement state, the Srxlev corresponds to Sx. 1)

2)

3)

Intra-frequency measurement  If Sx > Sintrasearch, the UE need not to start intra-frequency measurement.  If Sx <= Sintrasearch, the UE starts intra-frequency measurement.  If the system information does not provide Sintrasearch, the UE always starts intra-frequency. Inter-frequency measurement  If Sx > Sintersearch, the UE need not to start inter-frequency measurement.  If Sx <= Sintrasearch, the UE starts inter-frequency measurement.  If the system information does not provide Sintrasearch, the UE always starts inter-frequency. Inter-RAT measurement  If Sx > SsearchRATm, the UE needs not to measurement the system m.  If Sx <= SsearchRATm, the UE starts inter-RAT measurement on the system m.  If the system information does provide SsearchRATm, the UE always starts inter-RAT measurement on the system m.

The measurement rules when HCS is used: If the cell broadcast system information indicates not to use HCS, the UE decides to start the corresponding measurement. 1)

2)





2006-06-03

For intra-frequency and inter-frequency threshold-based measurement rules  If Srxlevs <= SsearchHCS or if FDD and Sx <= Sintersearch, the UE measures all inter-frequency and intra-frequency cells.  Else the UE measures on all intra-frequency and inter-frequency cells, which have higher HCS priority level than the serving cell unless measurement rules for fast-moving UEs are triggered.  Else, the UE measure on all intra-frequency and inter-frequency cells, which have equal or higher HCS priority level than the serving cell unless measurement rules for fast-moving UEs are triggered. For intra-frequency and inter-frequency measurement rules for fast-moving UEs If the number of cell reselections during time period TCRmax exceeds NCR, high-mobility has been detected. In this high-mobility state, UE shall Measure intra-frequency and inter-frequency neighbor cells, which have equal or lower HCS priority than serving cell. Prioritize re-selection of intra-frequency and inter-frequency neighbor cells on lower HCS priority level before neighbor cells on same HCS priority level. When the number of cell reselections during time period TCRmax no longer exceeds NCR, UE shall Huawei Proprietary. No Spreading Without Permission.

Page 83 of 98


– –


Continue these measurements during time period TCrmaxHyst, Revert to measurements according to the threshold based measurement rules.

The inter-RAT measurement rules with HCS: 1)

2)

Inter-RAT threshold-based measurement rules  If Srxlevs <= SHCS,RATm or if FDD and Squal <= SSearchRATm, then the UE shall measure on all inter-RATm cells.  Else if Sx > Slimit,SearchRATm, the UE need not measure neighbor cells in RAT "m"  Else the UE shall measure on all neighbor cells in RAT "m", which have equal or higher HCS priority level than the serving cell unless measurement rules for fast-moving UEs are triggered. Inter-RAT measurement rules for fast-moving UEs If the number of cell reselections during time period TCRmax exceeds NCR, high-mobility has been detected. In this high-mobility state, UE shall  Measure the neighbor cells in RAT "m", which have an equal or lower HCS priority than the serving cell  Prioritize re-selection of neighbor cells in RAT "m" on lower HCS priority level before neighbor cells in RAT "m" on same HCS priority level. When the number of cell reselections during time interval TCRmax no longer exceeds NCR, UE shall  Continue these measurements during time period TCrmaxHyst,  Revert to measure according to the threshold-based measurement rules.

III. Judgment Criterion (Criterion H and Criterion R) Evaluating cell reselection occurs in the following situations:  

Internal trigger of UE. For details, see TS 25.133. The information for evaluating cell reselection on BCCH changes.

The following cell re-selection criteria are used for intra-frequency cells, inter-frequency cells and inter-RAT cells: The quality level threshold criterion H for hierarchical cell structures is used to determine whether prioritized ranking according to hierarchical cell re-selection rules shall apply, and is defined by:

Hs = Qmeas_LEV,s - Qhcss Hn = Qmeas_LEV,n - Qhcsn – TOn * Ln

If it is indicated in system information that HCS is not used, the quality level threshold criterion H is not applied. The cell-ranking criterion R is defined by:

Rs = Qmap,s + Qhysts Rn = Qmap,n - Qoffsets,n - TOn * (1 – Ln) Where:

2006-06-03


Page 84 of 98



TOn = TEMP_OFFSETn * W(PENALTY_TIMEn – Tn) Ln = 0 Ln = 1

if HCS_PRIOn = HCS_PRIOs if HCS_PRIOn <> HCS_PRIOs

W(x) = 0 W(x) = 1

for x < 0 for x >= 0

TEMP_OFFSETn applies an offset to the H and R criteria for the duration of PENALTY_TIMEn after a timer Tn has started for that neighbor cell. TEMP_OFFSETn and PENALTY_TIMEn are only applicable if the usage of HCS is indicated in system information. The timer Tn is implemented for each neighbor cell. Tn shall be started from zero when one of the following conditions becomes true: If HCS_PRIOn <> HCS_PRIOs and Qmeas_LEV,n > Qhcsn, Or if HCS_PRIOn = HCS_PRIOs and for serving FDD and neighbor FDD cells if the quality measure for cell selection and reselection is set to CPICH RSCP in the serving cell, and: Qmap,n > Qmap,s + Qoffset1s,n for serving FDD and neighbour FDD cells if the quality measure for cell selection and reselection is set to CPICH Ec/No in the serving cell, and: Qmeas_LEV,n > Qmeas_LEV,s + Qoffset2s,n for all other serving and neighbour cells: Qmap,n > Qmap,s + Qoffset1s,n Tn for the associated neighbour cell shall be stopped as soon as any of the above conditions are no longer fulfilled. Any value calculated for TOn is valid only if the associated timer Tn is still running else TOn shall be set to zero. At cell-reselection, a timer Tn is stopped only if the corresponding cell is not a neighbour cell of the new serving cell, or if the criteria given above for starting timer T n for the corresponding cell is no longer fulfilled with the parameters of the new serving cell. On cell re-selection, timer Tn shall be continued to be run for the corresponding cells but the criteria given above shall be evaluated with parameters broadcast in the new serving cell if the corresponding cells are neighbors of the new serving cell. Table 58.2 Cell reselection parameters and descriptions Parameter name Sn

2006-06-03

Parameter description Cell Selection value of the neighbor cell, (dB)


Page 85 of 98



Qmap,n

Quality of the neighbor cell, after mapping function is applied, derived from CPICH Ec/N0 or CPICH RSCP for FDD cells, from P-CCPCH RSCP for TDD cells and from RXLEV for GSM cells. For FDD cells, the measurement that is used to derive the quality value is set by the Cell_selection_and_reselection_quality_measure information element.

Qmap,s

Quality of the serving cell, after mapping function is applied, derived from CPICH Ec/N0 or CPICH RSCP for FDD cells and from P-CCPCH RSCP for TDD cells. For FDD cells, the measurement that is used to derive the quality value is set by the Cell_selection_and_reselection_quality_measure information element.

Qmeas_LEV

Quality value. The quality value of the received signal expressed in CPICH_Ec/No or CPICH_RSCP_LEV for FDD cells as set by the Cell_selection_and_reselection_quality_measure information element, P-CCPCH_RSCP_LEV for TDD cells and RXLEV for GSM cells.

The UE shall perform ranking of all cells that fulfil the S criterion among: 



All cells with highest HCS_PRIO meets criterion H, namely, the cells with H > = 0, Note that this rule is not valid when UE high-mobility is detected. All cells, not considering HCS priority levels, if no cell fulfil the criterion H >= 0. This case is also valid when it is indicated in system information that HCS is not used, that is when serving cell does not belong to a hierarchical cell structure.

In all cases, the UE shall reselect the new cell, only if the following conditions are met:  

The new cell is better ranked than the serving cell during a time interval Treselection. M than 1 second has elapsed since the UE camped on the current serving cell.

Table 58.3 Broadcast parameters and description of cell reselection in system information Parameter name

2006-06-03

Parameter description

Qoffset1s,n

This specifies the offset between the two cells. It is used for TDD and GSM cells and for FDD cells in case the quality measure for cell selection and re-selection is set to CPICH RSCP.

Qoffset2s,n

This specifies the offset between the two cells. It is used for FDD cells in case the quality measure for cell selection and re-selection is set to CPICH Ec/No.

Qhyst1s

This specifies the hysteresis value (Qhyst). It is used for TDD and GSM cells and for FDD cells in case the quality measure for cell selection and re-selection is set to CPICH RSCP.

Qhyst2

This specifies the hysteresis value (Qhyst). It is used for FDD cells if the quality measure for cell selection and reselection is set to CPICH Ec/No.

HCS_PRIOs, HCS_PRIO

This specifies the HCS priority level (0-7) for serving cell and neighbor cells.


Page 86 of 98


2006-06-03


Qhcss, Qhcsn

This specifies the quality threshold levels for applying prioritized hierarchical cell re-selection.

Qqualmin

This specifies the minimum required quality level in the cell in dB. It is not applicable for TDD cells or GSM cells.

Qrxlevmin

This specifies the minimum required RX level in the cell in dBm.

PENALTY_TIMEn

This specifies the time duration for which the TEMPORARY_OFFSETn is applied for a neighbor cell.

TEMPORARY_OFFSET1n

This specifies the offset applied to the H and R criteria for a neighbor cell for the duration of PENALTY_TIMEn. It is used for TDD and GSM cells and for FDD cells in case the quality measure for cell selection and re-selection is set to CPICH RSCP.

TEMPORARY_OFFSET2n

This specifies the offset applied to the H and R criteria for a neighbor cell for the duration of PENALTY_TIMEn. It is used for FDD cells in case the quality measure for cell selection and re-selection is set to CPICH Ec/No.

TCRmax

This specifies the duration for evaluating allowed amount of cell reselection(s).

NCR

This specifies the maximum number of cell reselections.

TCRmaxHyst

This specifies the additional time period before the UE can revert to low-mobility measurements.

Treselections

This specifies the cell reselection timer value.

SsearchHCS

This threshold is used in the measurement rules for cell reselection when HCS is used. It specifies the limit for Srxlev in the serving cell below which the UE shall initiate measurements of all neighbor cells of the serving cell.

SsearchRAT 1 - SsearchRAT k

This specifies the RAT specific threshold in the serving cell used in the inter-RAT measurement rules.

SHCS,RATm

This threshold is used in the measurement rules for cell reselection when HCS is used. It specifies the RAT specific threshold in the serving cell used in the inter-RAT measurement rules.

Sintrasearch

This specifies the threshold (in dB) for intra frequency measurements and for the HCS measurement rules.

Sintersearch

This specifies the threshold (in dB) for intra frequency measurements and for the HCS measurement rules.

Slimit,SearchRATm

This threshold is used in the measurement rules for cell reselection when HCS is used. It specifies the RAT specific threshold (in dB) in the serving UTRA cell above which the UE need not perform any inter-RAT measurements in RAT "m".


Page 87 of 98



8.3 Random Access Random access is: after the MS request the system for connection, it receives the response from the system and is probably assigned with DCH. This process occurs in the signaling connection setup process of the following situations:     

Attach upon power on Detach upon power off Location area update Routing area update Carrying out any service

3GPP 25.211 defines that the timing relationship of frame structure and physical layer of RACh, PRACH, and access channel. 3GPP 25.213 defines the modulation of preamble on access channel and the spreading modulation of message part (data and control). It also defines the preamble, scramble, and spreading code. 3GPP 25.213 defines the access process.

8.3.1 Random Access Channel RACH is an uplink common transport channel, corresponding to the uplink common physical channel PRACH. The data from RACH is received by NodeB in the whole cell. The feature is with collision and using open loop power control. The transmission RACH uses the timeslot ALOHA method with fast acquisition indicator. The UE starts transmission in a preset time offset, namely, the access timeslot. Every two 10ms radio frames forms a 20ms access frame, which is divided by 15 access frames into intervals of 5120 chips (1.332ms). 8.3.1 shows the timing information and acquisition indicator of access timeslot, and the interval between access timeslots and timeslot number. Whether the information of an access timeslot in the serving cell is available is decided by upper-layer signaling. radio frame: 10 ms

radio frame: 10 ms

5120 chips Access slot

#0

#1

#2

#3

#4

#5

#6

#7

#8

#9

#10

#11

#12

#13

#14

Random Access Transmission Random Access Transmission

Random Access Transmission Random Access Transmission

Figure 1.59 Number and interval of access timeslots of RACH The subscriber can originate random access transmission at the beginning of each access 2006-06-03


Page 88 of 98



timeslot. 8.3.1 shows the structure of random access transmission. The structure includes message part of 10ms or 20ms.

Preamble

Preamble

Preamble

4096 chips

Preamble

Message part 10 ms (one radio frame)

Preamble

Preamble

4096 chips

Message part 20 ms (two radio frames)

Figure 1.60 Structure of random access transmission The preamble length of random access is 4096 chips. It includes a SIGNATURE. The SIGNATURE is 16 chips and is repeated 256 times. In total there are 16 different SIGNATURE. The 10 ms message part radio frame is split into 15 slots, each of length T slot = 2560 chips. Each slot consists of two parts, a data part to which the RACH transport channel is mapped and a control part that carries Layer 1 control information. The data and control parts are transmitted in parallel. A 10 ms message part consists of one message part radio frame, while a 20 ms message part consists of two consecutive 10 ms message part radio frames. The message part length is equal to the Transmission Time Interval of the RACH Transport channel in use. This TTI length is configured by higher layers. The data part consists of 10*2k bits, where k=0,1,2,3. This corresponds to a spreading factor of 256, 128, 64, and 32 respectively for the message data part. The control part consists of 8 known pilot bits to support channel estimation for coherent detection and 2 TFCI bits. This corresponds to a spreading factor of 256 for the message control part. The pilot bit pattern is described in 3GPP TS 25.211 table 8. The total number of TFCI bits in the random-access message is 15*2 = 30. The TFCI of a radio frame indicates the transport format of the RACH transport channel mapped to the simultaneously transmitted message part radio frame. In case of a 20 ms PRACH message part, the TFCI is repeated in the second radio frame. The downlink AICH is divided into downlink access slots, each access slot is of length 5120 chips. The downlink access slots are time aligned with the P-CCPCH. The uplink PRACH is divided into uplink access slots, each access slot is of length 5120 chips. Uplink access slot number n is transmitted from the UE τp-a chips prior to the reception of downlink access slot number n, n = 0, 1, …, 14. Transmission of downlink acquisition indicators may only start at the beginning of a downlink access slot. Similarly, transmission of uplink RACH preambles and RACH message parts may only start at the beginning of an uplink access slot. 8.3.1 shows the PRACH/AICH timing relation.

2006-06-03


Page 89 of 98



Figure 1.61 Timing relation between PRACH and AICH as seen at the UE The preamble-to-preamble distance τp-p shall be larger than or equal to the minimum preamble-to-preamble distance τ p-p,min, i.e. τp-p ≥ τp-p,min. In addition to τp-p,min, the preamble-to-AI distance τp-a and preamble-to-message distance τ pm are defined as follows: 1)

2)

When AICH_Transmission_Timing is set to 0, then  τ p-p,min = 15360 chips (3 access slots)  τ p-a = 7680 chips  τ p-m = 15360 chips (3 access slots) When AICH_Transmission_Timing is set to 1, then  τ p-p,min = 20480 chips (4 access slots)  τ p-a = 12800 chips  τ p-m = 20480 chips (4 access slots)

The parameter AICH_Transmission_Timing is signaled by higher layers.

8.3.2 Random Access Process I. Related Information When the physical layer of UE receives the PHY-DATA-REQ primitives, it starts physical random access process. For details, see 3GPP TRAFFIC STATISTICS 25.321. Before the UE starts physical random access process, the UE receives the following system information at layer 1 (physical layer) from upper-layer:    

    

Scramble of preamble Length of message part, 10ms or 20ms Value of AICH_Transmission_Timing (0 or 1) The signature set and RACH subchannel set assigned for each ASC (access subchannel) number Power_Ramp_Step (integer > 0) Preamble_Retrans_Max (integer > 0) Preamble_Initial_Power P p-m = Pmessage-control – Ppreamble (in dB) TFS parameter. It includes the power offset corresponding to each transmission format, data part and control part of random access message.

Note that the upper-layer might update previous parameters before the UE starts physical random access process. 2006-06-03


Page 90 of 98



In addition, before the UE starts physical random access process, the layer 1 shall receive the following information from MAC layer:   

Transmission format for PRACH message part ASC transmitted on PRACH The data (TBS) to be sent

II. Process for Starting Physical Random Access The Process for starting physical random access proceeds as below: 1)

2) 3) 4) 5)

6)

7) 8)

Decide available RACH access subchannel set and available uplink access timeslot set of next complete access timeslot set (SFN mod 2 = 0 and SFN mod 2 = 1, the former is called the access timeslot set 1 while the latter is called the access timeslot set 2). Select an uplink access timeslot randomly. The rule for random selection is to select by equivalent probability. If there is no available uplink access timeslot in the current access timeslot set, select one from next access timeslot set. According to provided ASC, select randomly the signature used by access in the signature set. Set the initial value of PRACH preamble retransmission counters to Preamble_Retrans_Max. Set the parameter Commanded Preamble Power to Preamble_Initial_Power. If the parameter Commanded Preamble Power exceeds the maximum allowed value, set the transmit power of preamble to maximum allowed transmit power. If it is lower than the needed minimum value (prescribed by 3GPP TS 25.101), set the transmit power of preamble to the current value to be calculated. This value might be larger or smaller than, equal to Commanded Preamble Power. Otherwise, set the transmit power of preamble to Commanded Preamble Power. Send the preamble with the selected uplink access timeslot, signature, and preamble transmit power. The UE waits for confirmation signals corresponding to the used signature from NodeB. If the UE fails not detect the +1 or –1 acquisition indicator in the downlink access timeslot which has the same number of uplink access timeslot used by transit preamble code, it randomly select the next available access timeslot. According to power ramp step, it increases the Command Preamble Power, deduct the preamble code reset counter by 1. If the Command Preamble Power is 6 dB larger than the maximum allowed power, the UE report layer 1 state ("No ack on AICH") to MAC layer, and then quits the physical random access process. If the retransmission counter value is larger than 0, repeat the sixth step; otherwise, the UE reports layer 1 state ("No ack on AICH") to MAC layer, and then quits physical random access process. If the UE receives an –1 acquisition indicator, it reports layer 1 state ("Nack on AICH received ") to MAC layer, and then quits physical random access process. If the UE receives a +1 acquisition indicator, it sends the random access message part after 3 or 4 uplink access timeslots before last transmission according to the value of AICH_Transmission_Timing. The power for sending control part of random access message is Pp-m higher than the power for sending preamble the last time. For the transmit power of data part, see protocols.

According to previous operation flow for random access, when the UE accesses the network, it first sends preamble, and then waits for the confirmation signals from NodeB in the fixed downlink access timeslot. If the NodeB detects a preamble signal transmitted by UE, the NodeB responds an acquisition indicator signal on downlink AICH. After sending preamble, it detects acquisition indicator (AI) signal in the fixed downlink access timeslot. If the UE is permitted, it keeps sending message part and completes a physical random access. If the UE fails to receive AI, it keeps repeating the handshaking process of "sending preamble to detecting AI" for preset times until permitted. Then it sends the message part and completes a physical random access process. If the UE receives the signal indicating that access is prohibited, it quits this random access process and reports the state. The message part of random access carries the sign information of UE, the type of applicated service, and so on.

2006-06-03


Page 91 of 98



8 describes the relationship among the access subchannel, access timeslot, and SFN. Table 61.1 Relationship among the access subchannel, access timeslot, and SFN SFN modulo 8 of corresponding PCCPCH frame

Sub-channel number 0

1

2

3

4

5

6

7

0

0

1

2

3

4

5

6

7

1

12

13

14

2 3

9

10

4

6

7

11

5 6

3

4

0

1

2

12

13

14

8

9

10

5

6

7

7

3

4

8

9

10

11

8

9

10

11

5

6

7 8

11

8

0

1

2

12

13

14

9

10

11

3

4

5

0

1

2

12

13

14

8 shows the definition of access timeslot set (taking the uplink and downlink access timeslot fixed difference τ p-a = 7680 chips as example). AICH access slots

SFN mod 2 = 0 τp-a

#0 PRACH access slots

#0

#1

#2

#1

#3

#2

#4

#3

#5

Access slot set 1

10 ms

#4

#6

SFN mod 2 = 1 #5

#7

#6

#8

#7

#9

#8

#9

#10

#10

#11

#11

#12

#12

#13

#13

#14

#14

Access slot set 2

10 ms

Figure 1.62 Definition of access timeslot set (taking the uplink and downlink access timeslot fixed difference τ p-a 7680 chips as example)

2006-06-03


Page 92 of 98



Chapter 9 Appendix 3: Authentication Flow 9 shows a successful authentication process.

Figure 1.63 Successful authentication process The authentication flow starts from network side. It aims as below:     

The network checks the UE whether the UE is allowed to access the network. The authentication flow provides random array of authentication quintuple parameters The UE calculates cipher key (CK). The UE can calculate the integrity key (IK) related to network side. The UE can authenticate the network.

Compared with the authentication flow of GSM networks, the authentication flow of 3G networks adds consistency check and the authentication of network by UE. These features further enhance the security of 3G networks. Before the network side originates authentication, if the VLR has not authenticated the authentication quintuple parameters, the network side first originates the process to HLR for obtaining authentication set, and then waits for response of authentication quintuple parameters. The authentication quintuple parameters include:     

RAND XRES AUTN CK IK

When detecting that the authentication quintuple parameters are present, the network side sends the authentication request message. This message contains the RAND and AUTN of authentication quintuple parameters. After the UE receives the message, the USIM authenticates AUTN. Namely, the UE authenticates the network. If the authentication passes, the USIM calculates CK, IK, and signature XRES. If the USIM judges that the authentication succeeds, the UE responds XRES in authentication response message. After the network side receives the authentication response message, it compares the XRES of the message with the XRES of authentication quintuple parameters saved in VLR database. This confirms whether the authentication succeeds. If yes, the flow proceeds. If no, an abnormal processing flow starts. The flow releases the connection between network side and UEs, and releases the occupied network resources and radio resources. After successful authentication, the UE saves CK and IK to USIM card. Sometimes, after the UE receives the authentication request message, it reports that the authentication fails. Typical causes of authentication failure include: 1) 2006-06-03

When the UE authenticates the network, it checks the AUTN in authentication request Huawei Proprietary. No Spreading Without Permission.

Page 93 of 98



message sent by network side. If the MAC is faulty, the UE sends the authentication failure message with the cause MAC Failure, as shown in 1.

Figure 1.64 Authentication Failure (due to MAC Failure) The network side decides according to subscriber identity reported by UE whether to originate identification process. If the current identity is TMSI (or P-TMSI), the network side originates identification process which requests UE of reporting IMSI. Then the network side restarts authentication flow. 2)

When UE detects that the SQN of AUTN message is faulty, the authentication fails with the cause Synch failure.

Figure 1.65 Authentication failure (due to Synch failure) Now the VLR removes all authentication quintuple parameters and starts the process of synchronization with HLR. This process requires HLR to reuse authentication quintuple parameters and to start authentication process.

2006-06-03


Page 94 of 98



Chapter 10 Appendix 4: Description of Accessrelated Parameters 10.1 Engineering Parameters There are limited adjustable engineering parameters:    

Antenna azimuth Antenna down tilt Antenna beamwidth Antenna gain

To solve access problems caused by coverage, engineers consider adjusting these engineering parameters. For example, constructing new sites in blind areas, increase the antenna gain of serving cells, or decrease the down tilt of antennas in neighbor cells. Avoid the impact on original coverage area by adjustment.

10.2 Cell Parameters The following sections list multiple parameters closely related to access problems. Engineers can adjust these parameters according to the causes when locating the problems.

10.2.1 Transmit Power of FACH This defines the transmit power of FACH. 



If it is set over small, the UE at the cell edge cannot correctly receive the signaling carried by FACH. This impacts the downlink CCH coverage, and consequently impacts the cell coverage. If it is set over large, other channels will be impacted, downlink transmit power will be used, and the cell capacity is impacted.

The default power of FACH is –1 dB, and it is based on that the CPICH Ec/Io of coverage cell edge is –12 dB. If the field coverage is worse, raise the FACH power according to CPICH Ec/Io.

10.2.2 Transmit Power of PCH This parameter defines the transmit power of PCH. 



If it is set over small, the UE at the cell edge cannot correctly receive paging messages. Therefore the paging delay increases, paging success rate drops, and access success rate drops. If it is set over large, power consumption increases and downlink interference increases.

10.2.3 Transmit Power of PICH It defines the transmit power of PICH. 

2006-06-03

If it is set over small, the UE at the cell edge cannot correctly receive paging indicator information. Therefore, the paging delay increases, or mal-operation of reading PCH data is probably performed, the UE consumes more power, the downlink CCH coverage is affected, and finally the cell coverage is affected. Huawei Proprietary. No Spreading Without Permission.

Page 95 of 98



Because the PICH keeps sending paging indicator information, if the transmit power of PICH is set over large, the PICH will interfere with other channels, consume downlink transmit power, and affects cell capacity. So increasing transmit power of PICH is not recommended. To increase the coverage by PICH signals, decrease NP to 18. Decreasing NP will lead to decrement of paging capacity at UU interface. At the early stage of network construction, setting NP to 18 keeps an adequate paging capacity and it is a typical configuration in the industry.

10.2.4 Cell Reselection Parameter: Measurement Hysteresis 2 (Qhyst2s) According to criterion R, for cell reselection, the cell performs ranking by the sum of measured value of serving cell and hysteresis. The value of Qhyst2s is closely related to slow fading feature of the area where the cell is. Qhyst2s avoids cell ping-pong reselection due to slow fading when the UE is at the cell edge, because ping-pong reselection leads to frequency location area update (idle mode), URA update (URA_PCH), cell update (CELL_FACH, CELL_PCH), and consequently, network signaling load increases and power consumption by UE increases.

10.2.5 Cell Reselection Parameter: Reselection Hysteresis Time (Treselections) If the quality of another cell signals (CPICH Ec/No measured by UE) is better than that of serving cell in the time specified by Treselection, the UE reselects the cell to camp on. Treselection avoids ping-pong reselection between cells by UE.

10.2.6 Cell Reselection Parameter: Sintrasearch It is the threshold for starting intra-frequency cell measurement: when the Ec/Io of serving cell is lower than QRelxmin + 2 * Sintrasearch, the intra-frequency cell measurement starts. Sintrasearch affects the speed of cell reselection, and consequently affects the first access success rate of UE and the first paging success rate at IU interface. Set Sintrasearch as large as possible based on little impact on power consumption by UE.

10.2.7 Cell Reselection Parameter: Qoffset2 Before evaluation by criterion R, a neighbor cell has its signal strength deducted by a offset, namely, Qoffset2. For a single-layer cell, set Qoffset2 to 0. Engineers can reach the same goal by Qhyst. Adjusting it is not recommended.

10.2.8 Transmit Power of AICH If it is set over small, the UE at the cell edge cannot correctly receive acquisition indicator. Therefore, the downlink CCH coverage is affected. The default transmit power of AICH is – 6 dB. According to optimization result, the default configuration meets the downlink coverage. Since the channel sends data continuously, increasing power lead to increment of needed downlink capacity.

10.2.9 PRACH-related Parameters To solve uplink PRACH problems, adjust PRACH-related parameters, including:   

2006-06-03

Retransmission times of preamble Power step of preamble Power offset of preamble Huawei Proprietary. No Spreading Without Permission.

Page 96 of 98



Power offset between preamble and message

The previous parameters affect one another. Upon occurrence of PRACH problems, adjust the retransmission times of preamble. The default configuration is 8. It is recommended to set retransmission times of preamble to 20 to avoid PRACH problems.

2006-06-03


Page 97 of 98



List of Reference 1.

3GPP R99 TS 24.008 V3.7.0. 2001-03

2.

3GPP R99 25_series. 2002-09

3.

URNP-SANA. W-RNO Access Procedure Analysis Guidance 20041101-A-1.0.doc. 2003-05

4.

URNP-SANA. W-Paging Procedure Analysis Guidance 20041101-A-1.0.doc. 2003-12

5.

URNP-SANA. W-Paging Problem Analysis Guidance 20041101-A-1.0.doc. 2003-12

6.

Joint-research team on RNO project. Call Delay Test Report. 2005-12

7.

RAN Radio Performance Dept.. WCDMA RAN Radio Performance Call Access Delay Optimization Test Report. 2004-02

8.

RAN Radio Performance Dept.. W-RAN Traffic Statistics Location Guidance-20050926-A-1.0.doc. 2005-09

9.

RAN Radio Performance Dept.. UMTS Radio Network KPI baseline (V3.3). 2006-01

analysis and Problem

10. RRNP. WCDMA RNO Network Event Definition Baseline 1.0. 2005-11

2006-06-03


Page 98 of 98

W Access Problem Optimization Guide 20060330 a 3.1

Recommend Documents