Professional Documents
Culture Documents
Approved by Date
Revision Records
2006-05-19 V3.10 adds analysis of HSDPA with the following Wang Dekai
content updated:
Supplementing admission failure analysis
and cases about RRC connection, RAB
assignment process in HSDPA service
3.10 Adding HSDPA-related DT and traffic
statistics values
According to traffic statistics indexes of RNC
version 1.6C01B064, adding some traffic
statistics indexes
Updating RRC connection analysis
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.......................................................................................15
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
Figure 3-2 Signaling flow of originated UE in locating paging problems.................................17
3.3.3 Analyzing RRC Connection Setup Problems...........................................................19
3.3.4 Analyzing Authentication Problems..........................................................................22
3.3.5 Analyzing Security Mode Problems.........................................................................24
3.3.6 Analyzing PDP Activation Failure Problems.............................................................25
3.3.7 Analyzing RAB or RB Setup Problems....................................................................25
3.4 Processing Access Delay Problems...................................................................................28
3.4.1 Configuration of Discontinuous Cyclic Period Duration Factor DRX........................28
3.4.2 Whether to Disable Authentication and Encryption Flow.........................................28
3.4.3 Implementing Early or Late Assignment..................................................................29
3.4.4 Whether the RRC Connection Is Set up on FACH and DCH...................................29
Chapter 4 Analyzing Traffic Statistics Data.................................................................................30
4.1 Tool for Analyzing Data....................................................................................................... 30
4.2 General Methods for Analyzing Traffic Statistics Data........................................................30
4.2.1 Flow for Analyzing RNC-level Traffic Statistics Data................................................30
4.2.2 Flow for Analyzing Cell-level Traffic Statistics Data.................................................31
4.3 Accessibility Indexes.......................................................................................................... 32
4.3.1 Paging Traffic Statistics Indexes..............................................................................32
4.3.2 Low Success Rate of RRC Setup............................................................................36
4.3.3 Low Success Rate of CS RAB Setup......................................................................39
4.3.4 Lower Success Rate of PS RAB Setup...................................................................42
4.3.5 Low Success Rate of RB Setup...............................................................................46
4.4 System Availability Index.................................................................................................... 48
4.4.1 High Admission Rejection Rate................................................................................48
4.4.2 High Paging Congestion Rate..................................................................................48
4.4.3 High Rate of Congested Cell...................................................................................48
Chapter 5 Solving Access Problems............................................................................................ 49
5.1 Paging Problems................................................................................................................ 49
5.1.1 Improper Power Configuration of Paging-related Channels.....................................49
5.1.2 Paging Failure due to UE Location Area Update.....................................................49
5.1.3 Paging Failure due Implicit Detach of UE................................................................51
5.2 Cell Selection and Reselection Problem............................................................................51
5.2.1 Repeating to Send the RRC Connection Request Message due to Cell Reselection51
5.3 RRC Setup Problems......................................................................................................... 54
5.3.1 Improper Configuration of Parameters of Uplink Access Channel...........................54
5.3.2 Improper Configuration of AICH Power....................................................................57
5.3.3 Improper Configuration of FACH Power..................................................................57
5.3.4 Multiple Times of RRC Connection Request (for Service) and No RAB Assignment
Request............................................................................................................................ 59
5.3.5 RRC Connection of HSDPA Subscribers Rejected due to Inadequate Code Resource
.......................................................................................................................................... 63
5.4 RAB and RB Setup Problems............................................................................................ 65
5.4.1 RAB Setup Failure due to Inadequate Resource.....................................................65
5.4.2 Handover Failure before Completion of Signaling Flow...........................................66
5.4.3 Admission Failure due to HSDPA Total Bit Rate Threshold Exceeded by HSDPA Bit
Rate of Cell....................................................................................................................... 68
5.5 Authentication Problems..................................................................................................... 69
5.6 Security Mode Problems.................................................................................................... 69
5.7 Abnormal Equipment Problems.......................................................................................... 71
5.7.1 Abnormal NodeB...................................................................................................... 71
5.7.2 Abnormal UE............................................................................................................ 72
Chapter 6 Summary....................................................................................................................... 75
Chapter 7 Appendix 1: Paging Process.......................................................................................76
7.1 Paging Origination.............................................................................................................. 76
7.1.1 Paging by CN........................................................................................................... 76
7.1.2 Paging by UTRAN................................................................................................... 76
7.2 Paging Flow....................................................................................................................... 76
7.2.1 Paging Type 1.......................................................................................................... 76
7.2.2 Paging Type 2.......................................................................................................... 77
7.3 Behaviors of UE after Receiving Paging............................................................................78
7.3.1 UE in Idle Mode....................................................................................................... 78
7.3.2 UE in Connected Mode............................................................................................ 78
7.4 DRX Process of UE............................................................................................................ 79
7.4.1 DRX Cyclic Length and Paging Occasion................................................................79
7.4.2 Relationship of PICH and SCCPCH........................................................................80
7.4.3 PCH Selection......................................................................................................... 81
7.4.4 DRX Examples of UE.............................................................................................. 82
Chapter 8 Appendix 2: Access Process Analysis.......................................................................83
8.1 Cell Search......................................................................................................................... 83
8.1.1 Timeslot Synchronization......................................................................................... 83
8.1.2 Frame Synchronization and Scramble Group Identification.....................................83
8.1.3 Identification of Cell Primary Scramble....................................................................84
8.2 Cell Selection and Reselection........................................................................................... 84
8.2.1 Cell Selection........................................................................................................... 84
8.2.2 Judgment Criterion (Criterion S)..............................................................................85
8.2.3 Cell Reselection....................................................................................................... 87
8.3 Random Access................................................................................................................. 92
8.3.1 Random Access Channel........................................................................................ 92
8.3.2 Random Access Process......................................................................................... 94
Chapter 9 Appendix 3: Authentication Flow................................................................................98
Chapter 10 Appendix 4: Description of Access-related Parameters.......................................100
10.1 Engineering Parameters................................................................................................. 100
10.2 Cell Parameters............................................................................................................. 100
10.2.1 Transmit Power of FACH.....................................................................................100
10.2.2 Transmit Power of PCH.......................................................................................100
10.2.3 Transmit Power of PICH......................................................................................100
10.2.4 Cell Reselection Parameter: Measurement Hysteresis 2 (Qhyst2s)....................101
10.2.5 Cell Reselection Parameter: Reselection Hysteresis Time (Treselections)..........101
10.2.6 Cell Reselection Parameter: Sintrasearch...........................................................101
10.2.7 Cell Reselection Parameter: Qoffset2..................................................................101
10.2.8 Transmit Power of AICH.......................................................................................101
10.2.9 PRACH-related Parameters.................................................................................101
List of Tables
Table 2-1 Indexes and recommended values for accessibility related to DT 11
Table 2-2 Indexes and reference values for accessibility related to traffic statistics 11
Table 2-3 Indexes and reference for system availability 12
Table 2-4 Indexes and reference for access delay related to DT 12
Table 4-1 Counters related to paging loss for idle UE 33
Table 4-2 Counters related to paging loss for UE in PCH state 34
Table 4-3 Counter related to paging loss due to flow control 35
Table 4-4 Counters related to PCH congestion 36
Table 4-5 Counters related to RRC connection request rejection due to lub interface failure 37
Table 4-6 Traffic statistics counters related to RRC connection request rejection due to network
congestion 38
Table 4-7 Counter related to RRC connection failure due to no response 39
Table 4-8 Counters related to RRC connection setup rejection due to redirection 39
Table 4-9 Traffic statistics counters related to CS RAB assignment setup failure due to radio
network problems 40
Table 4-10 Traffic statistics counters related to CS RAB assignment setup failure due to
insufficient capability 41
Table 4-11 Counter related to CS RAB assignment setup failure due to transmission network
problems 42
Table 4-12 Traffic statistics counters related to PS RAB assignment setup failure due to radio
network problems. 43
Table 4-13 Traffic statistics counters related to PS RAB assignment setup failure due to
insufficient capability 45
Table 4-14 Counter related to PS RAB assignment setup failure due to transmission network
problems 45
Table 4-15 Counter related to PS RAB setup failure due to no resource available 46
Table 4-16 Traffic statistics counters related to RB setup failure 47
Table 4-17 Traffic statistics counter related to no response to RB setup 48
Table 7-1 Mapping relationship between PI and PICH 81
Table 8-1 Parameters and their description in the criterion S 85
Table 8-2 Cell reselection parameters and descriptions 90
Table 8-3 Broadcast parameters and description of cell reselection in system information 90
Table 8-4 Relationship among the access subchannel, access timeslot, and SFN 96
List of Figures
Figure 3-1 Overall flow for analyzing call failure problems in DT/CQT 15
Figure 3-2 Signaling flow of originated UE in locating paging problems 16
Figure 3-3 Flow for analyzing paging problems 17
Figure 3-4 Flow for analyzing RRC connection setup problem 19
Figure 3-5 Authentication failure due to MAC Failure 22
Figure 3-6 Authentication failure due to Synch Failure 22
Figure 3-7 Security mode control 23
Figure 3-8 Security mode reject 23
Figure 4-1 Flow for analyzing RNC-level traffic statistics data 30
Figure 4-2 Flow for analyzing cell-level traffic statistics data 31
Figure 4-3 Position for counting point by counter for paging loss by idle UE 32
Figure 4-4 Position for counting point by counter for paging loss by UE in PCH state 34
Figure 4-5 Position for counting point by counter for RRC connection rejection 35
Figure 4-6 Position for counting point by counter for CS RAB assignment failure in RNC traffic
statistics starting counting 39
Figure 4-7 Position for counting point by counter for PS RAB assignment failure in RNC traffic
statistics 42
Figure 4-8 Position for counting point by counter for RB setup failure in traffic statistics 46
Figure 5-1 Originating signaling flow of paging failure due to UE location area update 48
Figure 5-2 Content of the Disconnect message in paging failure due to UE location area update
49
Figure 5-3 Terminating signaling flow of paging failure due to UE location area update 49
Figure 5-4 Signaling of UE 50
Figure 5-5 Signal quality when the UE sends the RRC connection request message. 51
Figure 5-6 Signal quality when the UE repeats to send the RRC connection request 52
Figure 5-7 Signaling of UE in a connection process 53
Figure 5-8 Single subscriber tracing signaling on RNC 53
Figure 5-9 Quality of downlink signals 54
Figure 5-10 Regular interference in cell 248 54
Figure 5-11 Part of magnified chart of interference 55
Figure 5-12 Signaling upon improper configuration of FACH power 56
Figure 5-13 Signal strength upon the first sending of RRC connection request 57
Figure 5-14 Single subscriber tracing signaling by RNC 57
Figure 5-15 Signaling and signal strength upon the second sending of RRC connection request
58
Figure 5-16 Traced signaling at UE side 59
Figure 5-17 Traced signaling at RNC side 59
Figure 5-18 BLER statistics at UE side 60
Figure 5-19 BLER and RRC message at UE side 61
Figure 5-20 BLER and RRC message at UE side 61
Figure 5-21 Assignment of HSDPA code tree 63
Figure 5-22 RNC log for HSDPA admission rejection 63
Figure 5-23 Signaling of Disconnect after completion of RB setup 64
Figure 5-24 Signaling of UE upon a connection failure 65
Figure 5-25 Single subscriber tracing signaling by RNC 66
Figure 5-26 Signal strength before release of connection 66
Figure 5-27 Rejection messages in security mode 68
Figure 5-28 Content of the RANAP_SECURITY_MODE_REJECT message 68
Figure 5-29 ciphering mode information configured in previous security mode command 69
Figure 5-30 Security mode message 69
Figure 5-31 Content of the RANAP_SECURITY_MODE_COMMAND message 69
Figure 5-32 Signaling of UE upon failure in receiving RRC Connection Setup message 70
Figure 5-33 Single subscriber tracing signaling by RNC 71
Figure 5-34 Signal strength upon occurrence of problems 71
Figure 5-35 Signaling of UE 72
Figure 5-36 Downlink signal quality 72
Figure 7-1 Flow chart of PAGING TYPE 1 message 76
Figure 7-2 Flow chart of PAGING TYPE 2 message 77
Figure 7-3 Schematic drawing of UE paging occasion 79
Figure 7-4 Frame structure of PICH 79
Figure 7-5 Sequence relationship between PICH and SCCPCH 80
Figure 8-1 Number and interval of access timeslots of RACH 92
Figure 8-2 Structure of random access transmission 92
Figure 8-3 Timing relation between PRACH and AICH as seen at the UE 93
Figure 8-4 Definition of access timeslot set (taking the uplink and downlink access timeslot fixed
difference p-a 7680 chips as example) 96
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.
DT Drive Test
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.
2.1 Accessibility
Table 1.1 lists the indexes and recommended values for accessibility in DT.
Table 1.2 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
VP Stat. 98%
RAB setup success rate
PS Stat. 97%
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.
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.
Voice CQT 7s
VP ring CQT 7s
Call setup delay
VP see the
CQT 13s
figure
PS CQT 4s
PDP context activation delay 2s (average)
HSDPA CQT
< 4s (95%)
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.
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) 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.
2) 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.
3) 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.
4) 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.
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.
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) 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.
2) 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.
3) 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.
4) 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.
The Actix Analyzer defines access failure according to the following principles: after the
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
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.
Figure 1.1 shows the overall flow for analyzing call failure problems.
Figure 1.1 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 Figure 1.1. 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
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. Figure 1.1 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
1
The CC CALL PROCEDING message should be after CC SETUP.
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.
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.
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.
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) The UE sends RRC Connection Request message on RACH.
2) The RNC sends the RRC Connection Setup message on FACH.
3) 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.
Figure 3.1 shows the flow for analyzing RRC connection setup problem.
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
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.
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.
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
Figure 1.1.
Therefore, the Ki and OP (OPc) of the USIM must be known upon defining a
subscriber or the Ki and OP (OPc) of USIM must be made the same value as in
HLR.
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 Figure 1.1.
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.
Figure 1.1 shows the security mode control.
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.
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.
For analysis of PDP activation failure problems, see the section 5.1 of W-PS Problem
Optimization Guide.
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:
The RNC directly rejects RAB setup request due to incorrect parameter
configuration.
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
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 (Sony-
Ericsson 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.
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:
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:
HSDPA subscribers supported by NodeB
HSDPA subscribers supported by cell
Whether the guaranteed bit rate exceeds the prescribed threshold
III. The UE's Response of RAB Setup Failure due to RB Setup Failure
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 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.
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 over high ؤleads 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.
The different between early and late assignment lies in the different assignment time for
TCH.
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.
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.
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.
Note that after RRC connection setup, the signaling is reconfigured on DCH3.4K upon
RB setup.
Figure 1.1 shows the flow for analyzing RNC-level traffic statistics data.
3
Due to problems on traffic statistics, current cluster-level analysis is not concerned. It is the
same as RNC-level analysis.
The flow for analyzing cell-level traffic statistics data proceeds as below:
1) Check whether there are cells with unsatisfied indexes
If no, the analysis ends.
2) If there are cells with unsatisfied indexes, analyze the detailed causes, find major
causes of indexes deterioration, and provide proper solutions.
3) After carrying out solutions, analyze the new data of traffic statistics until indexes
meet requirements.
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 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
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.
Table 1.1 lists the counters related to paging loss for idle UE.
Figure 1.2 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.
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.
Figure 1.2 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.
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.
When the lu interface is in flow control state, it drops the paging messages from CN.
Table 1.1 shows the counter related to paging loss 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.
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.
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.
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) The RRC connection request is rejected due to lub interface failure
2) The RRC connection request is rejected due to network congestion
3) The RRC connection fails due to no response
4) 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 Figure
4.1.
Figure 4.1 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.
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
Table 1.1 lists the counters related to RRC connection request rejection due to lub
interface failure.
Table 1.1 Counters related to RRC connection request rejection due to lub interface
failure
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
Table 1.1 Traffic statistics counters related to RRC connection request rejection due to
network congestion
Table 1.1 lists the counter related to RRC connection failure due to no response.
4
Others are calculated indexes, equal to total failure times (RRC.FailConnEstab.Cong) minus
previous failure times.
2006-06-03 Huawei Proprietary. No Spreading Without Permission. Page 38 of 102
W-Access Problem Optimization Guide For internal use only
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.
Table 1.1 lists the counters related to RRC connection setup rejection due to redirection
Table 1.1 Counters related to RRC connection setup rejection due to redirection
5
This index is a calculated index, equal to RAB request times minus the sum of RAB success
times and other failure times.
Figure 1.1 Position for counting point by counter for CS RAB assignment failure in RNC
traffic statistics starting counting
At the point B in Figure 1.1, 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.
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
CS RAB assignment setup failure due to other problems of radio networks 6
Table 1.1 shows the traffic statistics counters related to CS RAB assignment setup failure
due to radio network problems.
Table 1.1 Traffic statistics counters related to CS RAB assignment setup failure due to
radio network problems
6
This index is a calculated index, equal to total failure times (VS.RAB.FailEstabCS.RNL) minus
previous failure times.
The indexes listed in Table 1.1 are cell-level indexes. For detailed
VS.RAB.FailEstCS.Unsp, see 3Table 1.1.
1) 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.
2) 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.
3) 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
Others7
Table 1.1 lists the traffic statistics counters related to CS RAB assignment setup failure
due to insufficient capability.
Table 1.1 Traffic statistics counters related to CS RAB assignment setup failure due to
insufficient capability
7
This index is a calculated index, equal to total failure times (VS.RAB.FailEstCS.Unsp) minus
times of other four failures.
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)
Table 1.1 lists the counter related to CS RAB assignment setup failure due to
transmission network problems.
Table 1.1 Counter related to CS RAB assignment setup failure due to transmission
network problems
8
This index is a calculated index, equal to RAB request times minus the sum of RAB success
times and other failure times
Figure 1.1 Position for counting point by counter for PS RAB assignment failure in RNC
traffic statistics
At the point B in Figure 1.1, 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.
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 networks 9
Table 1.1 shows the traffic statistics counters related to PS RAB assignment setup failure
due to radio network problems.
Table 1.1 Traffic statistics counters related to PS RAB assignment setup failure due to
radio network problems.
9
This index is a calculated index, equal to total failure times minus other failure times.
Table 1.1 Traffic statistics counters related to PS RAB assignment setup failure due to
insufficient capability
10
This index is a calculated index, equal to total failure times (VS.RAB.FailEstCS.Unsp) minus
times of other four failures.
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)
Table 1.1 lists the counter related to PS RAB assignment setup failure due to
transmission network problems.
Table 1.1 Counter related to PS RAB assignment setup failure due to transmission
network problems
Table 1.1 Counter related to PS RAB setup failure due to no resource available
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 Figure 1.1.
Figure 1.1 Position for counting point by counter for RB setup failure in traffic statistics
At the point A shown in Figure 1.1, 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.
Table 1.1 lists the traffic statistics counters related to RB setup failure.
1) 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.
2) Physical channel failure
It seldom occurs.
3) Cell update occurrence
Cell update occurs during RB setup. It seldom occurs, so this document neglects it.
4) 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
Table 1.1 lists the traffic statistics counter related to no response to RB setup.
To be supplemented.
To be supplemented.
To be supplemented.
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.
Figure 1.1 shows the originating signaling flow of paging failure due to UE location area
update.
Figure 1.1 Originating signaling flow of paging failure due to UE location area update
According to Figure 1.1, the RNC receives the Disconnect message from CN.
Figure 1.2 shows the content of the Disconnect message.
Figure 1.2 Content of the Disconnect message in paging failure due to UE location area
update
According to Figure 1.2, the cause value for the Disconnect message is no route to
destination. Therefore, the connection is released because the destination UE cannot
be paged.
Figure 1.3 shows the terminating signaling flow of paging failure due to UE location area
update.
Figure 1.3 Terminating signaling flow of paging failure due to UE location area update
According to Figure 1.3, 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.
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.
5.2.1 Repeating to Send the RRC Connection Request Message due to Cell
Reselection
Figure 1.1 shows the signaling of UE when cell reselection causes repeating to send
RRC Connection Request message.
The interval between two times of repeating to send request by UE is about 1.2s.
Figure 1.2 Signal quality when the UE sends the RRC connection request message.
Figure 1.2 Signal quality when the UE sends the RRC connection request message.
Figure 1.3 shows the Signal quality when the UE resends the RRC connection request.
Figure 1.3 Signal quality when the UE repeats to send the RRC connection request
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:
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.
After the table shown in Figure 1.3 is sorted by time order, you can see that the RNC
responds to the second RRC setup connection message from UE.
Figure 1.3 shows the quality of downlink signals.
According to Figure 1.3, 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 Figure 1.4.
Note:
Interference lasts for 40s each time. The last peak shown in Figure 1.4 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) 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.
2) 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.
3) 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.
4) 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.
5) 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
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.
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.
Figure 1.1 shows the signaling upon improper configuration of FACH power.
Figure 1.2 shows the signal strength upon the first sending of RRC connection request.
Figure 1.2 Signal strength upon the first sending of RRC connection request
In Figure 1.2,
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.
Figure 1.3 shows the 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.
Figure 1.4 shows the signaling and signal strength upon the second sending of RRC
connection request after 2s.
Figure 1.4 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
Figure 1.4, 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.
Figure 1.1 shows the traced signaling at UE side and RNC side.
Figure 1.2 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.
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.
Figure 1.3 shows the 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.
Figure 1.4 shows the BLER and RRC message.
According to Figure 1.4, 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 Figure 1.5.
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 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.
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.
Figure 1.1 shows the assignment of HSDPA code tree.
According to the assignment of HSDPA code tree in Figure 1.1, the number of code
words assigned for HSDPA subscribers is clear, as well as the rest code words and the
occupation by R99 subscribers. Figure 1.1 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.
Figure 1.2 shows RNC log for HSDPA admission rejection.
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.
According to Figure 1.1, 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
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.
Figure 1.1 shows the signaling of UE upon a connection failure according to Analyzer
software.
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
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:
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 .
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.
Figure 1.1 shows the rejection messages in security mode during tracing single
subscriber.
According to Figure 1.2, 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.
Figure 1.3 shows the ciphering mode information configured in previous security mode
command.
Figure 1.3 ciphering mode information configured in previous security mode command
According to Figure 1.3, 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
Figure 1.4.
According to Figure 1.5, 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 Figure 1.5, it selects the
UEA1 as the encryption algorithm. It then receives the security mode command shown in
Figure 1.3 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.
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.
Figure 1.1 shows the signaling of UE upon failure in receiving RRC Connection Setup
message.
Figure 1.1 Signaling of UE upon failure in receiving RRC Connection Setup message
Figure 1.3 shows the normal signal strength upon occurrence of problems.
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.
The UE cannot connect to the network for a period. Figure 1.1 shows the signaling of UE.
In Figure 1.2,
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.
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:
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
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.
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.
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.
Besides previous situations, the UTRAN sends the PAGING TYPE 2 message on DCCH,
which is called the cooperation paging.
Figure 1.1 shows the flow chart of PAGING TYPE 1 message.
PAGING
RANAP RANAP
According to Figure 1.1, 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.
CN SRNC UE
PAGING
RANAP RANAP
According to Figure 1.1, 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.
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.
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.
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).
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.
Figure 1.1 shows the frame structure of PICH.
12 bits (transmission
288 bits for paging indication off)
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.
Table 1.1 describes the mapping relationship between {PI0, .., PIN-1} and PICH bits
{b0,..,b287}.
Number of PI per
PIp = 1 PIp = 0
frame (NP)
The UE determines by calculating its paging indicator suffix p that p is relevant to the qth
bits of PICH frame.
q PI 18 SFN SFN / 8 SFN / 64 SFN / 512 mod 144 Np mod Np
144
(3)
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.
Figure 1.2 shows the time sequence relationship between PICH and SCCPCH.
PICH
The end of PICH radio frame is 7680 chips earlier than associated S-CCPCH frame.
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.
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.
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.
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.
The following parameters describe the trigger time, process, and principles for judging
suitable cell for cell selection.
I. Trigger Time
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.
If yes, the UE searches for other SIBs according to the scheduling information of
other SIBs contained in MIB, and obtain their content.
If no, the UE starts cell search from the next frequency.
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
Table 1.1 lists the parameters and their description in the criterion S.
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.
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. Trigger Time
Hs = Qmeas_LEV,s - Qhcss
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
Where:
Ln = 0 if HCS_PRIOn = HCS_PRIOs
Ln = 1 if HCS_PRIOn <> HCS_PRIOs
The timer Tn is implemented for each neighbor cell. T n shall be started from zero when
one of the following conditions becomes true:
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 TO n is valid only if the
associated timer Tn is still running else TOn shall be set to zero.
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 1.2 Broadcast parameters and description of cell reselection in system information
This specifies the HCS priority level (0-7) for serving cell
HCS_PRIOs, HCS_PRIO
and neighbor cells.
5120 chips
The subscriber can originate random access transmission at the beginning of each
access timeslot. Figure 1.2 shows the structure of random access transmission. The
structure includes message part of 10ms or 20ms.
Preamble Preamble Preamble Message part
4096 chips
10 ms (one radio frame)
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*2 k 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
Figure 1.3 Timing relation between PRACH and AICH as seen at the UE
The preamble-to-preamble distance p-p shall be larger than or equal to the minimum
preamble-to-preamble distance
p-p,min, i.e. p-p p-p,min.
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.
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
Table 1.1 Relationship among the access subchannel, access timeslot, and SFN
0 0 1 2 3 4 5 6 7
1 12 13 14 8 9 10 11
2 0 1 2 3 4 5 6 7
3 9 10 11 12 13 14 8
4 6 7 0 1 2 3 4 5
5 8 9 10 11 12 13 14
6 3 4 5 6 7 0 1 2
7 8 9 10 11 12 13 14
Figure 8.2 shows the definition of access timeslot set (taking the uplink and downlink
access timeslot fixed difference p-a = 7680 chips as example).
AICH access
slots SFN mod 2 = 0 SFN mod 2 = 1
PRACH
access slots Access slot set 1 Access slot set 2
10 ms 10 ms
Figure 8.2 Definition of access timeslot set (taking the uplink and downlink access timeslot fixed
difference p-a 7680 chips as example)
Sometimes, after the UE receives the authentication request message, it reports that the
authentication fails. Typical causes of authentication failure include:
1) When the UE authenticates the network, it checks the AUTN in authentication
request 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 Figure 1.2.
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.
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.
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.
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.
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.
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.
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.
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 analysis and Problem
Location Guidance-20050926-A-1.0.doc. 2005-09
9. RAN Radio Performance Dept.. UMTS Radio Network KPI baseline (V3.3). 2006-01
10. RRNP. WCDMA RNO Network Event Definition Baseline 1.0. 2005-11