UMTS Call Drop Analysis. ZTE University

Transcription

1 UMTS Call Drop Analysis ZTE University

2 Content Definition of Call Drop Reasons of Call Drop Analysis of Call Drop Parameters of Call Drop Case of Call Drop

3 Type of Call Drop Definition Call termination not through the normal hang up process, that is dropped calls. Tow types of call drop definition: Definition of drive test indicators Definition of network management indicators

4 Definition of Drive Test Indicators The definition of drive test: Call drop rate= Number of call drop times/number of call setup success times Number of call setup success times+1:after the Alerting message is received

5 Definition of Drive Test Indicators Number of call drop times+1:(air interface signaling at the UE side) The Connect ACK message is not received but the System Information message is received. After the Connect ACK message is received, RRC Release message received and the reasons is Not Normal. After the Connect ACK message is received, any of the CC Disconnect, CC Release, CC Release Complete message received and the reasons is Not Normal.

6 Definition of Drive Test Indicators number of call setup success times+1

7 UE Voluntarily Initiated Signaling Release

8 Definitions of Network Management Indicators The definition of the network management: CSRabrelTriggedByRNC CS _ CDR * 100% CSRABSetupSuccess CSRabrelTriggedByRNC+1:(Iu interface signaling at the RNC side) When the RNC send the Iu Release Request or RAB Release Request message to the CN.

9 Definitions of Network Management Indicators RNC RAB RELEASE REQUEST CN RNC CN IU RELEASE REQUEST

10 Counters of Call Drop Reasons Release Type Iu Release Counter C C C C C C C C C Reason Iu connection release request by UTRAN for CS domain in cell, Repeated Integrity Checking Failure Iu connection release request by UTRAN for CS domain in cell, Release due to UE generated signalling connection release Iu connection release request by UTRAN for CS domain in cell, Radio Connection With UE Lost Iu connection release request by UTRAN for CS domain in cell, timer TRELOCoverall expiry Iu connection release request by UTRAN for CS domain in cell, Failure in the Radio Interface Procedure Iu connection release request by UTRAN for CS domain in cell, O&M Intervention Iu connection release request by UTRAN for CS domain in cell, Release due to Overload Control Iu connection release request by UTRAN for CS domain in cell, Unspecified Failure Iu connection release request by UTRAN for CS domain in cell, UTRAN Generated Reason

11 Counters of Call Drop Reasons Release Type RAB release Counter C C C C C Reason RAB release number request by UTRAN in cell for CS domain,rab pre-empted RAB release number request by UTRAN in cell for CS domain,release due to UTRAN in cell Generated Reason RAB release number request by UTRAN in cell for CS domain,iu UP Failure RAB release number request by UTRAN in cell for CS domain,release due to Overload Control RAB release number request by UTRAN in cell for CS domain,unspecified Failure

12 Content Definition of Call Drop Reasons of Call Drop Analysis of Call Drop Parameters of Call Drop Case of Call Drop

13 Call Drop Reasons Poor Coverage Unconfigured Neighbor Cell Handover Interference PSC Confliction Engineering Error

14 Poor Coverage Service Type Requirement of RSCP (dbm) Requirement of Ec/Io (db) AMR12.2K CS64K PS384K HSDPA RSCP and Ec/Io threshold for different services

15 Poor Coverage The decision whether it is problem of uplink or downlink poor coverage is based on the power of dedicated channel before call drop. UL Poor Coverage: TX power reaches the maximum UL BLER is poor NodeB report RL failure DL Poor Coverage: TX power reaches the maximum DL BLER is poor Scanner: If the RSCP and Ec/Io of the cell with the best coverage are poor, you can infer that the coverage is poor.

16 Unconfigured Neighboring Cell Missed neighbor cell. Removal of key neighbor cells caused by combination of macro diversity. Untimely update of the external cell information.

17 Unconfigured Neighboring Cell

18 Handover The handover process is incomplete Radio Environment becoming bad, UE can not receive the Active Set Update UE RNC Measurement Report (1A event) Active Set Update Active Set Update Complete

19 Handover Ping-pong Handover In a short time, UE send different Report for delete or add cell A UE RNC Measurement Report (1B event, Delete cell A) Active Set Update (Delete cella) Measurement Report (1A event, Add cell A) Active Set Update (Add cell A)

20 Interference Reasons for Pilot Pollution: Cross-cell coverage of high Node-B Node-B in ring layout Signal distortion caused by street effect or strong reflection

21 Interference Judgment of DL interference: CPICH RSCP of the active set is large than -85dBm Ec/Io is lower than -13dB Reason of DL interference: Pilot pollution Unconfigured neighboring cell

22 Interference Judgment of UL interference: The average RTWP of an idle cell exceeds -100dBm The max RTWP is around -90dBm Reason of UL interference: Intra-RAT interferences Inter-RAT interferences.

23 Case-Interference UL interference = -93(dBm)

24 Case-Interference

25 PSC Confliction When analyzing such call drop, check Cell ID in the call drop signaling besides PSCs because the neighbor relation is identified by Cell ID.

26 PSC Confliction Cell A and Cell B are configured as neighbor cell for each other. Cell C and Cell B are not configured as neighbor cells for each other. Cell A and Cell C have the same PSC.

27 PSC Confliction Cell A is the neighbor cell of Cell B. Cell D is the neighbor cell of Cell C. Cell A and Cell D have the same PSC.

28 PSC Confliction Cell B and Cell D are not configured as neighbor cell for each other. Cell A is the neighbor cell of Cell B. Cell E is the neighbor cell of Cell D. Cell A and Cell E have the same PSC.

29 Engineering Error Call drops caused by engineering error: Reversely-connected antenna An excessive VSWR Multi-band antenna problem Leakage of signals from indoor distribution system Call drop caused by unsteady transmission

30 Case-Engineering Error

31 Content Definition of Call Drop Reasons of Call Drop Analysis of Call Drop Parameters of Call Drop Case of Call Drop

32 Common Analysis Methods for Call Drop Analyzing Call Drops by DT Analyzing Call Drops by Traffic Statistics

33 Analyzing Call Drops by DT Call drop data Call drop spots Stability of the primary serving cell RSCP and Ec/Io of the primary serving cell Reproducing of problems with DT

34 Analyzing Call Drops by Traffic Statistics The commonly used KPI analysis method is the TOP cell method, which means the top cells will be screened out according to certain index, then these top cells are optimized and then the top cells are selected again. After several repetitions, the related KPI can be speedily converged.

35 Multi-Dimension Analysis Multi-dimension analysis is carried out from different perspectives. For the call drop problem, not only the call drop itself, but also related factors such as access handover traffic statistics time RTWP

36 Multi-Dimension Analysis Cell ID Cell Name CS Call Drop Rate [%] Number of Successful CS RAB establishment Total Abnormal Release Abnormal Iu Released Number, by Cause Abnormal RAB Released Number, by Cause CS Traffic [Erl] PS Traffic [Kbyte] HSDPA RLC Throughput [Mbps] Max Cell Freq RTWP [dbm] Average Cell Freq RTWP [dbm] Max Cell Freq Tcp [dbm] Average Cell Freq Tcp [dbm] Max HSDPA users in cell Average HSDPA users in cell Max HSUPA users in cell Average HSUPA users in cell Repeated Integrity Checking Failure Radio Connection With UE Lost TRELOCoverall expiry Failure in the Radio Interface Procedure Unspecified Failure UTRAN Generated Reason Release due to UTRAN in cell Generated Reason Iu UP Failure Unspecified Failure

37 Trend Analysis

38 Accident Analysis Check the equipment alarm and system log of this period to find out hardware problems; Check the transmission of this period; Check whether the upgrade or cell blocking is performed during this period; Check whether there is an occasion with abrupt high traffic requirements such as a concert, game, or exhibition.

39 Ranking Analysis Ranking analysis Ranking analysis is carried out through classifying data into top N and bottom N data from a large amount of data. Index Reason of failure for Handover Failure times Percent(%) 1 Failure when getting the decision of handover from database Timeout for setup of service channel in handover Failure for decision of NodeB hard handover Failure when build the service channel Timeout for waiting for the handover of UE complete Other errors Basic channel switch, cannot find appointed frequency Basic channel switch, failed to allocate the resource Basic channel switch, failed to setup channel board Not enough resource for channel OVSF code is not engough Timeout for handover of UE The frequency is configured in the neighbouring cell list Overload for power allocation of the frequency for the service cell There is no idle channel in service cell There is not suitable frequency in service cell Different reason of failure for several service cells Can not find UE Abnormal release in CN

40 Cause-and-Effect Analysis For a certain effect, the cause-and-effect analysis is performed to locate the causes that may result in the effect and to determine the influence of the causes. For example, the call congestion of a cell may be caused by insufficient capacity of the hardware, of the downlink, or of the uplink.

41 Content Definition of Call Drop Reasons of Call Drop Analysis of Call Drop Parameters of Call Drop Case of Call Drop

42 Optimization Methods for Call Drop Engineer optimization: Directional angle, downtilt, position of Antenna, type of Antenna, Transmit power of BS, position of BS, new BS. Radio parameter optimization: Time to Trigger, CIO, threshold of enabling/disabling Compression Mode, Maximum transmit power of DL RL, Threshold of Interfrequency and Inter-system.

43 Time To Trigger Abbreviated Name Description TrigTime[MAX_INTRA_MEAS_EVENT] This parameter indicates the time difference between having detected the event generation and reporting the event. Only when the event generation is detected and still meets all requirements of event triggering after Time to trigger, the event can be triggered and reported. The larger the value is, the stricter the judgment is for the event to be triggered. The parameter should be set according to the actual requirements. Sometimes, if it is set too large, the quality of calls may decrease. Range and Step (0, 10, 20, 40, 60, 80, 100, 120, 160, 200, 240,320, 640, 1280, 2560, 5000)ms Default Value UE Event Report Parameters for CPICH Ec/No: [200,640,320,320,200,200, 320]ms UE Event Report Parameters for CPICH RSCP: [200,200,200,200,200,200,200]ms Detected Set Measurement Parameters for CPICH Ec/No: [200]ms Detected Set Measurement Parameters for CPICH RSCP: [200]ms Time To Trigger (TTT) refers to the interval between the detection and reporting of events (1A, 1B, 1C, and 1D). The setting of TTT has an impact on the promptness of handovers.

44 Cell Individual Offset Abbreviated Name Description CellIndivOffset An offset is allocated for each cell being monitored. The offset can be positive or negative. Before the UE judges whether an event occurs, adds the offset to the measurement result. If the PCPICH uses a positive offset, the UE sends the measurement report just like PCPICH is x db better than the actual case. Or, if the PCPICH uses a negative offset, the PCPICH report is limited. When the cell individual offset is used, the corresponding cell is possibly (at least temporarily) the target cell of the handover or removed from the active set. Range and Step OMCR: [-10, 10] db, step 0.5 RNC: D=(P+10)*2, [0, 40] Default Value 0dB A higher value of this parameter results in easier soft handovers, more UEs in the soft handover state, and more used resources. A lower value results in more difficult handovers. The CIO has an impact on the non-best cell. In detail, the CIO is effective for 1a events in neighboring cells and effective for 1b events in cells to be deleted.

45 Threshold of Enabling/Disabling Compression Mode Abbreviated Name Description Range and Step Default Value ThreshUsedFreq[MAX_INTER_MEAS_EVENT] This parameter indicates the absolute threshold that is required to be configured for event 2b/2d/2f (used when judging the quality of the currently used frequency. CPICH RSCP: [-115, -25] dbm, step 1dBm CPICH Ec/No: [-24, 0] db, step 1dB UE Inter-frequency Event Report Parameters for CPICH Ec/No: [-24,-13, -24,-13,-24,-8]dB UE Inter-frequency Event Report Parameters for CPICH RSCP: [-115,-95,-115,-95,-115,-80]dBm The compression mode is used in inter-frequency and inter-system handovers. The compression mode is enabled before the handover. Currently, the compression mode is enabled by the 2D event and disabled by the 2F event. The measurement can be RSCP or Ec/Io. By default, the RSCP is currently used.

46 Maximum Downlink Transmit Power of Radio Link Abbreviated Name Description Range and Step MaxDlDpchPwr This parameter indicates the maximum downlink DPCH transmission power. When performing the downlink inner loop power control, the new transmission power must be smaller than or equal to the configured DPCH Maximum DL Power. If the newly computed transmission power is larger than the configured DPCH Maximum DL Power, make it equal to the configured DPCH Maximum DL Power. [-35, 15] db, step 0.1 db If call drop occurs frequently in a cell due to coverage problem, increase the maximum downlink transmit power of services. However, a user in the edge area may consume great transmit power, which affects other users and reduces the downlink capacity of the system. If users fail to access a cell due to heavy traffic, consider changing the value of this parameter to a smaller value.

47 Maximum Downlink Transmit Power of Radio Link Conversational DL 3.4kbps Signaling(AM): 0 db Conversational DL 13.6kbps Signaling:0dB Conversational DL WAMR 6.60~23.85kbps: 0dB Conversational DL 64kbps (PS Conversational Video): 3dB Conversational DL NAMR 4.75~12.2kbps: 0dB Conversational DL CS 28.8kbps: -6dB Conversational DL CS 32kbps: -4dB Conversational DL CS 64kbps: 3dB Streaming DL CS 14.4kbps: -7dB Streaming DL CS 28.8kbps: -6dB Streaming DL CS 57.6kbps: -4dB Streaming DL CS 64kbps: 3dB Streaming DL PS 16kbps:-4dB Streaming DL PS64kbps: 1dB Streaming DL PS384kbps: 4dB Streaming DL PS128kbps: 2dB Interactive DL PS 8kbps:-8dB Interactive DL PS 16kbps:-4dB Interactive DL PS64kbps: 1dB Interactive DL PS384kbps: 4dB Interactive DL PS128kbps: 2dB Background DL PS 8kbps:-8dB Background DL PS 16kbps:-4dB Background DL PS64kbps: 1dB Background DL PS384kbps: 4dB Background DL PS128kbps: 2dB

48 Maximum Downlink Transmit Power of Radio Link Streaming DL CS64kbps: 3dB Interactive Interactive DL PS 1.2Mbps: 4dB DL PS 3.65Mbps: 4dB Interactive DL PS 7.2Mbps: 4dB Interactive Interactive DL PS 10.1Mbps: 4dB DL PS 14.0Mbps: 4dB Background DL PS 1.2Mbps: 4dB Background Background DL PS 3.65Mbps: 4dB DL PS 7.2Mbps: 4dB Background DL PS 10.1Mbps: 4dB Background Streaming DL PS 14.0Mbps: 4dB DL PS 1.2Mbps: 4dB Streaming DL PS 3.65Mbps: 4dB Streaming DL PS 7.2Mbps: 4dB Streaming DL PS 10.1Mbps: 4dB Streaming DL PS 14.0Mbps: 4dB Interactive DL PS 17.6Mbps: 4dB Interactive DL PS Mbps: 4dB Background DL PS 17.6Mbps: 4dB Background DL PS Mbps: 4dB Streaming DL PS 17.6Mbps: 4dB Streaming DL PS Mbps: 4dB

49 Threshold of Inter-system handover Abbreviated Name Description Range and Step Default Value Thresh[MAX_RAT_MEAS_EVENT] This parameter indicates the absolute threshold of the UTRAN cell quality that UE uses to judge event 3a. The range and unit of the parameter relate to the measurement quantity of UTRAN system. CPICH RSCP: [-115, -25] dbm, step 1dBm CPICH Ec/No: [-24, 0] db, step 1dB UE Event Report Parameters for Own System CPICH Ec/No: [-6,-24,-24,-24] UE Event Report Parameters for Own System CPICH RSCP: [-95,-115,-115,-115] When the measured value of inter-system neighboring cell signal exceeds the specified threshold, handover is triggered. If you set this parameter to a small value, handover is triggered ahead of time. If you set this parameter to a great value, handover is delayed.

50 Timer and Counter Related to Call Drop Name Description Value Range T312 Connected N312 Connected T313 N313 T314 T315 N315 T309 T312 of connection mode, that is, the time when the UE waits for L1 synchronization indicator when the special physical channel is set up N312 of connection mode, that is, the number of synchronization indicators that the UE should receive continuously from L1 before the special channel is set up successfully Waiting time after the DPCCH channel set up in CELL_DCH mode loses synchronization Maximum number of lost synchronization indicators that the UE receives continuously from the L1 Time of cell update, existing in T314-related radio bearer when wireless connection fails Time of cell update, existing in T315-related radio bearer when wireless connection fails Maximum number of synchronization indicators that the UE receives continuously from L1 in T313 activated state Waiting time after initiating requests to access other RATs, such as GSM (1..15)s (1, 2, 4, 10, 20, 50, 100, 200, 400, 600, 800, 1000) (0..15)s (1, 2, 4, 10, 20, 50, 100, 200) (0, 2, 4, 6, 8, 12, 16, 20)s (0,10, 30, 60, 180, 600, 1200, 1800)s (1, 2, 4, 10, 20, 50, 100, 200, 400, 600, 800, 1000) (1..8)s Default Value 1s 1 3s 20 4s 30s 1 3s

51 Content Definition of Call Drop Reasons of Call Drop Analysis of Call Drop Parameters of Call Drop Case of Call Drop

52 Neighboring Relation Adjustment Reason High call drop rate caused by improper neighbor list configuration Description The call drop rate of the TRI135W-1 cell corresponding to RNC1 in Libya is always around 3%, and no hardware alarm is generated. The cell coverage is mainly on the sea.

53 Neighboring Relation Adjustment

54 Neighboring Relation Adjustment After the neighbor relation is adjusted, the CS call drop rate of TRI135W-1 decreases from 3% to 1.3%.

55 Unconfigured Neighboring Cell 9 74

56 Unconfigured Neighboring Cell Main parameters: Cells in the system Troubleshooting process: The cell with SC 9 is in the detection set and cannot be added to the active set when the quality of the serving cell with SC 74 is extremely poor. This is a typical unconfigured neighboring cell. Solution: Adjust the neighbor relation: Add the cell with SC 9 to the neighbor list of the cell with SC 74. Result: In the same test, the cell with SC 9 is in the active set of the serving cell with SC 74. When the signal cell with SC 9 is strong enough, the UE hands over from the serving cell to the cell with SC 9.

57 Unconfigured Neighboring Cell

58 Soft Handover Parameter Optimization Reason: Low handover success rate because of improper soft handover parameter configuration Description: The success rate of the handover from sector 1 (SC 436) of the Shuqian Lu site to sector 2 (SC 434) of the Meihuacun hotel is low. This area is within the Shuqian Lu section. Main parameters: Soft handover 1a/1b event handover threshold, trigger time

59 Soft Handover Parameter Optimization The signal quality of the Shuqian Lu section is poor and unstable because there are overpasses in this section, Because the comparative threshold decision algorithm is used, a cell with poor signal quality may be added to the active set if the 1a threshold is excessively high. If the RNC sends the ActiveSet Update Command message to instruct the UE to enter this cell, the soft handover may fail because the radio link cannot be set up due to poor and unstable signal quality of this cell.

60 Soft Handover Parameter Optimization Adjust the 1a/1b event handover threshold and trigger time of cell 436. Lower the 1a event handover threshold and shorten the trigger time to ensure that the cells with good signal quality can firstly enter the active set. Raise the 1b event handover threshold and extend the trigger time so that the cells are not deleted too early due to drastic signal deterioration.

61 Soft Handover Parameter Optimization Result After the parameter optimization, cell 434 of the BS-1 (Meihuacun hotel site) can be added to the active set quickly and is not deleted too early. According to the drive test result from more than 100 times of handover tests, the success rate of the handover between the BS-2 (Shuqian Lu site) and the BS-1 (Meihuacun hotel site) increases greatly.

62 Data Configuration Optimization in 2G/3G Handover Reason Call drop because of incorrect data configuration Description When the 2G/3G handover tests are performed at the boundary of the 3G network, the handover from the 3G network to the 2G network succeeds in the west to east direction, but the handover fails in the east to west direction. Main parameters BSC ID, LAC, NCC, and BCCH of the 2G neighboring cell fail 3G 2G

63 Data Configuration Optimization in 2G/3G Handover Troubleshooting process Because the handover from 3G network to 2G network fails, you should firstly check whether 2G neighboring cells are configured. If all 2G neighboring cells are configured, go to the next step. Record whether the 2G Sagem UE starts the compression mode when the signal quality of a 3G UE is lower than the threshold for starting the compression mode, and record the CI of the 2G cell where the compression mode is started. You can find that the UE starts the compression mode after the preceding step is performed. The signaling is as follows:

64 Data Configuration Optimization in 2G/3G Handover However, after the UE starts the compression mode, the repositioning fails. The signaling is as follows: And the cause of the repositioning failure is as follows: According to the preceding signaling analysis, you can infer that the UE does not recognize the BSC of the 2G cell during the 3G-to-2G handover. In this case, the failure may be caused by unconfigured BSC ID or LAC.

65 Data Configuration Optimization in 2G/3G Handover Solution: Check the BSC and LAC of the target 2G cell on the CN. You can find that the LAC is not configured. Then, reconfigure the LACs of all 2G neighboring cells on the CN. Result: After the data is configured again, all 3G-to-2G handovers succeed. Suggestion for similar problems: In the 3G-to-2G handover, the BSC ID, LAC, NCC, and BCCH of 2G neighboring cells must be configured in the 2G neighboring cell database of the OMCR and on the CN.

66 Call Drop Case-Handover

67 Call Drop Case-Handover Delete Cell 51 and Cell 53

68 Call Drop Case-Handover Report e1a to add cell 51 and Cell 53 into Active Set after 0.4 second

69 Call Drop Case-Handover Report e1a to add cell 51,64,53,52 to Active Set

70 Call Drop Case-Handover

71 Call Drop Case-Handover Continue report e1a to add stronger cells to Active Set But UE cannot receive the AcitiveUpdate message because the poor signal of the cells in ActiveSet.

72 Call Drop Case-Handover

73 Call Drop Case-Handover Call Drop Happened

74