UMTS Full Parameter Check Check Rules
UMTS Full Parameter Check Check Rules
To support global UMTS office delivery and reduce the number of online problems caused by
incorrect parameter settings, the UMTS Maintenance Dept issues the UMTS Full Parameter
Check so that field engineers can check parameters in a routine manner to prevent problems
caused by incorrect parameter settings and field engineers, product support engineers, and R&D
personnel can improve the parameter check efficiency when handling online problems.
At present, 698 parameters are collected and sorted out, which cover 16 categories. The parameters
can be checked for their reasonableness, dependency, and consistency. The parameter check
tool can check most parameters. For details, see the UMTS Full Parameter Check Rule Table.
Most of the cases involved in this presentation slide do not go through parameter check first, and
parameter problems are found through analysis based on symptoms. We hope parameter check
to be performed first subsequently before performing other operations to improve the
troubleshooting efficiency. It takes much less time and energy to solve problems after parameter
check is performed first.
From now on, so long as problems are found, a few hours should be spent checking parameters first.
Swapped network
Some parameters use the original network values mapped, and some use
baseline values. Check parameters before equipment are online to prevent low-
level problems caused by incorrect parameter settings.
Legacy network
You are advised to record parameter optimization and regularly check
parameters to prevent low-level problems caused by incorrect parameter settings.
Plan:
Compare parameter settings (MML files)
of the live network with the baseline
scripts and check the reasonable
ranges, dependency, and consistency.
Analyze the parameters that are not in
the reasonable ranges or do not meet
the dependency or consistency
requirement and find out causes:
incorrect settings, improper rules, or
others. That is, explain why the
parameters fail parameter check and
modify improper configurations.
Overview
The customer of O office in country A complains that iPad users are often offline and cannot use PS services.
Problem analysis
Step 1: Determine the scope of access. The customer uses other terminals to perform a test. The problem also
exists. Analysis on the collected problems shows that the problems are not in a centralized area.
Step 2: Analyze the cause of access failure. No traffic statistics are provided for disconnection. Therefore,
analyze the core network tracing logs provided by the customer. After the core network sends an RAU Reject
message to the UE, the UE becomes offline.
Step 3: Check parameters. If parameters are not checked, it takes a long time to analyze the problem.
Step 4: Check equipment and transmission: N/A
Step 5: Check the version. No known problems cause disconnection.
Step 6: Check RF channels: N/A
It takes two days and lots of manpower to
Step 7: Check coverage: N/A analyze and locate the problem. If
Step 8: Analyze the long-term traffic trend: N/A parameters are checked first for their
reasonableness, the problem can be
Step 9: Analyze emergent factors: N/A solved in a minute.
Step 10: Analyze exceptions on the CN side: N/A
1. If the
parameter is set
to a value greater
You are advised to than 15000, state
set the parameter to transition may
10000 in a poor result in long-term
RBRECFGRS Access core Timer
coverage scenario unstable state
PTMR parameter parameter
and retain the and the UE fails
baseline value in to transmit
other scenarios messages. As a
result,
disconnection
may occur.
1 CN coordination parameter 1 0 0 0 0 0 0 0 1
3 Timer parameters 22 22 20 2 0 0 0 0 0
8 Algorithm parameters 15 12 5 7 3 2 1 0 0
Synchronization/loss of synchronization
9 5 2 1 1 3 3 0 0 0
parameters
10 Cell reselection parameters 24 0 0 0 24 8 16 0 0
Whether call drop is measured in abnormal
11 4 4 4 0 0 0 0 0 0
scenarios
12 User plane parameters 5 5 3 2 0 0 0 0 0
Parameter
Baseline Value Parameter Check Reasonableness Handling
Device Parameter Name Type Subtype (Provided in Rule Impact
R13) (Reasonableness) Check Suggestion
Requirement
When the
parameter must
be set to ON,
DL_INNER_LO the inner loop
OP_PC_ACTIV Call drop core Power control This parameter downlink power
E_SWITCH parameter parameter must be set to ON. control state is
set to Active.
Otherwise, it is
set to Inactive.
1 Equipment parameters 1 1 1 0 0 0 0 0 0 0
3 Channel parameters 36 32 21 11 3 1 2 1 1 0
Resource control
4 20 2 0 2 11 9 2 7 6 1
parameters
It takes a long time to analyze the problem because parameters are not checked.
Analyze distribution of online users through the OMStar network assessment function. The ratio of uplink 2ms users to
uplink 10ms users is about 7:3. Most terminals support the uplink 2ms services. In the live network configuration, the
HSUPA 2ms policy does not comply with the baseline configuration and needs to be optimized.
Scheduling probability before swap: 55% Scheduling probability after swap: 40%
Compare the gain with the baseline value and inform field engineers to change the parameter value to the
baseline value. The gain obtained through a drive test is 3% to 4%, which meets the requirement.
HUAWEI TECHNOLOGIES CO., LTD. Huawei Confidential Page 23
04. Voice Quality Core Parameters
Totally, there are 12 voice quality core parameters, which cover one subcategory.
12 RNC-level parameters, 0 cell-level parameter, and 0 NodeB-level parameter
5 MUST parameters and 7 SHOULD parameters
9 parameters can be checked for their reasonableness by the tool and 3 need to be
checked manually.
Category
RNC
No. Subcategory of voice quality Total
parameters Subtotal MUST SHOULD
Total 12 12 5 7
[INTERNAL]
Appointment of the Problem Tackling Team for Deterioration of 3G MOS after Swap in Office D of Country G
1. Background
Problem description:
The first FSA was completed at the end of July. After the equipment of MSN is swapped, the MOS drops by 0.2 points after the customer
performs a test by dialing a fixed-line number on a mobile phone using the Probe test tool in the quipment room. The test performed by
frontline engineers shows that the MOS also drops by 0.2 points after calls are made between mobile phones. This problem affects the
acceptance of the FSA.
After lots of time and manpower are used, the root cause is found in the end: The dynamic power control
function is disabled: SET UCORRMPARA: ReservedSwitch0=RESERVED_SWITCH_0_BIT24-0; The
baseline BLER is set to -20, which is equivalent to 1%. Because the dynamic power control function is
disabled (the uplink BLER value is not corrected), the MOS deteriorates. After the dynamic power control
function is enabled, the MOS meets the requirement.
It takes lots of time and manpower to locate this problem, and a complaint is even filed to the higher level. If
parameter check is performed, it takes only a few minutes to locate the problem.
Known Unknown Unknown Baseline Parameter Parameter Parameter Parameter
Device Constraint of Constraint of Constraint Parameter Type Subtype Value Check Rule Reasonable Check Rule Check Rule Handling Impact
the Basic the Basic of the Live Name (Provided (Reasonable ness Check (Dependency) (Consistency) Suggestion
Network Network Network in R13) ness) Requirement
Total 84 29 17 12 55 24 31
1 Power control 5 0 0 0 5 1 4
2 Access 11 5 0 5 6 0 6
3 Switch 2 1 1 0 1 0 1
4 Handover 47 23 16 7 24 16 8
5 Cell reselection 19 0 0 0 19 7 12
Note: In this case, drive test analysis rather than parameter check is performed first, which delays problem locating.
RSCP HW ER
EcNo Huawei Ericsson EcNo
(-65 to -30) 4530 (48%) 4529 (46%)
EcNo
EcNo>-10 62.61% 73.47% (-75 to -65) 2698 (29%) 2798 (28%)
-14<EcNo<-10 33.87% 25.13% (-85 to -75) 1721 (18%) 2000 (20%)
-16<EcNo<-14 2.58% 1.17% (-95 to -85) 366 (4%) 476 (5%)
EcNo<-16 0.94% 2.23% (-105 to -95) 26 (0%) 31 (0%)
(-115 to -105) 3 (0%) 2 (0%)
Note: This is partly because of mapping of CS service inter-system 2D/2F threshold.
It can be seen, although the coverage in some areas becomes good (new sites are added
after swap), that the overall coverage is becomes poor.
Root cause: Parameters are set incorrectly (the power does not match).
Locating method: Check parameters and analyze drive test data.
1 Timer parameters 1 1 0 1 0 0 0
3 Access parameters 1 1 0 1 0 0 0
4 Switch parameters 3 3 2 1 0 0 0
5 Threshold parameters 6 6 6 0 0 0 0
7 Channel parameters 1 1 0 1 0 0 0
8 Paging parameters 5 5 3 2 0 0 0
Category NodeB
No. Total
RF channel core parameters Subtotal MUST
Total 5 5 5
1 RF channel parameters 5 5 5
Action 6: Analyze main and diversity RTWP. The traced values show that the
main RTWP values of the cells are around –106 dBm, which is in the normal
range, and the diversity RTWP values are below –114 dBm and does not
fluctuate, which indicates that the alarm is not a false alarm. Incorrect
configuration (cross-connect mode) or RF module failure may cause this
problem.
Field problem solving result: Field engineers determine that the problem is caused by the CME
template and do not execute subsequent actions.
1 Timer parameters 4 4 4 0 0 0 0
2 Function parameters 27 21 21 4 4 2 2
3 Threshold parameters 21 14 14 7 7 0 0
Problem analysis:
1.After channels are checked, the RTWP is still abnormal. Field engineers perform many interference tests and still cannot
find the root cause several days later.
2.During the check on power control parameters, comparison of different network logs shows that power control
BETAC/BETAD is set to an incorrect value. If parameter check is performed in advance, the problem can be located within
several minutes.
Total 8 7 7 1 1
1 Process parameters 8 7 7 1 1
Process
parameter
1 Delay parameters 5 2 2 3 2 1
2 Switch parameters 4 4 4 0 0 0
3 Threshold parameters 30 15 15 15 15 0
Network planning
4 4 0 0 4 4 0
parameters
Problem analysis
Step 1: Confirm the scope and causes. After the 3G network is swapped, the number of handovers on the entire network
increases sharply, which is unrelated to top cells. Handover failures are caused by other reasons, which account for more
than 55%.
Step 2: Check parameters. (Parameters are not checked.)
Step 3…
Required Action Data Analysis and Conclusion Handling Result
Analysis Symptom
Parameter check The inter-RAT The setting of the inter-RAT handover Enable the The number of inter-
handover attribute attribute parameter does not match that of handover RAT handovers
parameter is set tothe original network. More than 384,000 attribute. restores to the number
off. users are allowed to hand over to the 2G before the swap.
network.
Caution: After the 3G network is swapped, field engineers do not perform parameter mapping or check
parameters. The field engineers responsible for locating this problem do not check parameters in time, which
extend the problem locating duration.
Problem analysis
Step 1: Determine the scope of call drop. The analysis on traffic measurement data shows
that the call drop rate is normal.
Step 2: Determine the specific causes of call drop. Because this problem always occurs in
3G calls, field engineers directly trace VIP signaling. The tracing results show the following law:
The interval from call setup to call release initiated by the core network ranges from 5s to 10s;
the RB/RAB is established successfully on the RNC side; the UE interworks with the core
network at the NAS layer through direct messages normally; after calls are connected
successfully, the core network sends a disconnect message (an NAS layer message
transparently transmitted by the RNC) within 5s to 10s to release the calls.
This problem is actually caused by incorrect parameter settings and incompatibility with the core
network of a peer vendor and can be located through parameter check. In problem locating, however,
parameters are not checked. Instead, signaling is analyzed to locate the problem. Because the core
network sends a disconnect message, the problem lies in the core network. Then ask the peer vendor
to check the core network. Pushed by field engineers, NSN explains why calls are released. Later, after
NSN disables strict check on RTCP packets, the problem is solved.
Clarification
for RTCP Issue
After the problem occurs, parameter check is not performed. Instead, lots of alarms, traffic
measurement data, and tracing data are analyzed. It takes field engineers one week to locate the
problem: a parameter is not set to the recommended value.
The analysis on SCTP tracing data captured by field engineers shows that (assume the packet length is
52): The RNC sends a piece of data with TSN 74370857 at 13:25:24(38) and another piece of data with
TSN 74370858 140 ms ((52 – 38) x 10) later. Because no SACK message is received, the RNC
retransmits the two pieces of data 160 ms later.
The RNC receives the data 2831696010 at 16:54:16(30) and 16:54:16(46). The interval is 160 ms.
The duration of the peer retransmission timer is 150 ms. Therefore, the MGW retransmits the data
after the retransmission timer expires.
Category RNC
No. Equipment core Total
Subtotal MUST SHOULD
parameters
Total 2 2 1 1
1 Equipment parameters 2 2 1 1
Problem overview
An emergent problem occurs in an office in country S. The service cannot be established when a
subrack is added.
Problem analysis
An emergency plan is available for this type of problem. First check related configurations. The
check on configuration scripts shows that the SCU port between the new subrack and subrack 0
is not enabled. As a result, the new subrack is disconnected from subrack 0. Heartbeat messages
cannot be received from the SCU port, the new subrack resets repeatedly, and the service cannot
be established. After the SCU port is enabled, the service recovers.
Because parameter check is performed first, the problem is solved within 5 minutes with R&D
engineer support by telephone.
Category RNC
No. Total
Subcategory of terminal
Subtotal MUST
compatibility parameters
Total 8 8 8
1 Switch parameters 8 8 8
Set up the test environment. After MacD PDU size is set to 1296, the problem reoccurs. After field
engineers modify MacD PDU size to 656, the problem is solved. If parameter check is performed
first, the problem can be solved within several minutes.
Parameter Parameter Check Parameter Check
Device Known Constraint of Unknown Constraint Parameter Name Type Subtype Baseline Value Parameter Check Rule Reasonableness Rule Rule Handling Impact
the Basic Network of the Basic Network (Provided in R13) (Reasonableness) Check (Dependency) (Consistency) Suggestion
Requirement
If the parameter
is set to a large
HS-DSCH value, the
MAC-D PDU size You are advised performance of
PS data to use the weak coverage
1 transmission Channel baseline value. areas is affected
core parameter parameter and terminals
may be
incompatible.
Category RNC
No. Subcategory of license Total
Subtotal SHOULD
parameters
Total 5 5 5
1 License parameter 5 5 5
Query the configuration scripts of RNC. The RNC IUPS interface is configured with two IP paths, the
IP path ping function is enabled, and alarms about IP path are not reported on the RNC side, which
indicates that the IP paths are normal. In addition, field engineers ping the IP path and find the peer
address of the IP path can be pinged.
Analyze the peer IP address carried in the RANAP_RAB_ASSIGMENT_REQ message sent by the CN
to the RNC and convert the IP address into a decimal IP address. The decimal IP address is
consistent with the peer IP address 172.15.3.3, which indicates that the peer IP address is sent
normally.
It takes a long time to locate the problem. Later, R&D engineers check tracing logs and, based
on codes, find that the license is not activated. Then field engineers confirm that the PS services
fail to be established because the IP license is not activated on the Iu interface.
Problem overview: An alarm indicating that a public transmission channel fails to be established on the NodeB. Trace messages on
the Iub interface. The NBAP_COMM_TRANSP_CH_SETUP_FAIL message is found, with the cause value power-level-not-
supported(2).
Problem analysis: The cause value power-level-not-supported occurs because the remaining
power is not enough to establish a transmission channel. The power of channels is based on the
power offset of pilots. Therefore, when the pilot configuration is similar to the power configuration of
logical cells, the remaining power is not enough to establish a transmission channel.
After confirming the problem by tracing data, the analysis personnel first query the pilot configuration
and the power configuration of logical cells. The difference between the two is smaller than 53
(0.1dbm). After the difference between the two is increased, the problem is solved.
After the problem occurs, first check parameters, which helps quickly solve the problem.
D. Tool
Tool guide
Manual check method
E. Appendixes
Maturity of function commercialization
Summary of deliverables
UMTSÈ«²ÎÊý¼ì²é
¹¤¾ß²Ù×÷Ö¸µ¼.docx
Step 2: In the result obtained in step 1, search in the live network scripts based on
column B MML Command. If a unique value can be found, fill the value in column Y
Live Network Configuration. If you cannot determine the unique value, go to step 3.
Step 3: In the result obtained in step 2, search in live network scripts based on column
D Known Constraint of Basic Network and fill the searched value in column Y Live
Network Configuration.
If this parameter is set to ON, more code reshuffling procedures will be initiated. The failure of the procedures You are advised to use the baseline
results in a higher call drop rate. value.
D. Tool
Tool guide
Manual check method
E. Appendixes
Maturity of function commercialization
Summary of deliverables
UMTSÈ«²ÎÊý¼ì²é CFGMML-RNCxxx
20120820-È˹¤¼ì²é½»¸¶¼þ .rar
Output results of manual check Output results of the check using the tool