How to do DT road test in GSM event analysis

1. 1. 1 The terrain environment leads to weak coverage.

The problem phenomenon road test shows that the RSRP of the current cell and all neighboring cells of the calling and called phones is less than-1 10dbm, and the SINR is caused by the lack of the primary cell.

The problem analysis combines the site environment and the base station topology diagram to investigate and verify the coverage of all the surrounding base stations one by one. For the road sections that cannot be completely solved due to terrain reasons, submit the follow-up project construction plan.

Solution:

1: Check the station planning scheme of this bureau. If there is a planned station, improve the priority of construction and opening; if there is no construction scheme, submit the station construction scheme;

2. Test the coverage of the 2G network, and turn on the SRVCC function when the coverage is good; At the same time, ensure that the FR function on the 2G side is turned on, and ensure that the 4G is dropped in time.

3. In the case of testing and evaluation requirements, take evasive route to reduce the probability of abnormal events;

1. 1.2 neighborhood mismatch leads to weak coverage.

Problem phenomenon

In the DT test, the main symptoms of such problems are as follows:

1. Frequent A3 events make it impossible to switch;

2. Generally speaking, RSRP varies greatly with the cell occupied by the calling party and the called party;

3. The switch chain is out of order;

4. It is easy to cause dropped calls, reconstruction and handover failure.

As shown in the figure below:

problem analysis

1: Test and analyze the foreground, and check whether the cell information reported by A3 event is included in RRCConnectionReconfiguration. If not, it is determined as a lost neighbor. Check the topology diagram of the base station to determine whether it is necessary to increase the neighboring cells. Eliminate over-coverage and needle-shaped coverage areas (such as missing areas and off-road coverage areas) and add them carefully to prevent the problem that they cannot be cut out in time after switching.

As shown in the figure below, two base stations are adjacent to each other with a distance of 200m, so it is suggested to supplement the neighbor relationship.

2. The background IMSI tracking signaling analysis can filter the UnknowPciNotify through the UDA tool. For the phenomenon of continuously reporting undefined PCI, it is necessary to further determine whether to add neighbor cells by combining the base station topology map.

As shown in the figure below, the background signaling analysis also found that the candidate neighbor cell for reporting multiple measurement reports is 486. Combined with the topology diagram, it is finally suggested to increase the relationship between neighboring cells.

Solution:

1: Neighboring base stations usually adopt the scheme of adding missing neighbor cells;

2. Overcoverage cell gives priority to control coverage;

3: Needle-shaped covering scene is not recommended, which generally affects short roads, so the covering should be controlled first;

1. 1.2. 1 instance 1

Problem phenomenon

After the calling and called party occupied the newly-opened base station 983529 133(PCI 149), the wireless environment deteriorated due to the lack of neighbor relationship.

problem analysis

The following figure shows that after the calling party and the called party occupy the cell, the RSRP becomes weak from strong, and the wireless environment gradually deteriorates. A3 events are frequently reported but handover is not initiated. Looking at the configuration of adjacent cells, it is found that the station only configures its own two cells as adjacent cells. Through understanding, this station is a new base station, and there is no single inspection and network access optimization work, so there are frequent incidents such as dropped calls and re-establishment in this section.

solution

1: optimize the single station and supplement the neighbor relationship in time to ensure that the service performance of the single station passes the acceptance;

2. Inform the optimization team to conduct parameter verification, neighbor verification and performance test as soon as possible after the site is opened to ensure the normal operation after network access.

1. 1.3 site failure leads to weak coverage.

Problem phenomenon

In the test, this problem is as follows: (1) can't occupy nearby base stations, which will be accompanied by the situation of missing allocation and over-coverage in neighboring cells, and it is prone to disconnection, network disconnection and chaotic switching chain;

As shown below, the base station 462682 is located at the intersection of Lufeng Road and Xianjiahu West Road, which is the main service area of these two main roads. After testing this section, the base station has not been occupied, and the RSRP has dropped below-1 10dbm, resulting in disorder of the switching chain and disconnection.

problem analysis

The site "Building 23, Yangming Mountain Villa, Changsha (Dihua Meixi Lake was taken away) ZL-B8300462682PT" was temporarily offline due to disputes, resulting in surrounding uninhabited residential areas;

Solution:

It is necessary to restore "Building 23 of Yangming Mountain Villa in Changsha (Dihua Meixi Lake was taken away) ZL-B 8300462682 PT" as soon as possible; Turn on the SRVCC handover of the surrounding cells.

1. 1.4 cannot be switched due to switching.

Problem phenomenon

The problem of over-coverage is mainly manifested as follows:

Under the condition of not occupying the over-covered cell, the SINR of the current cell may deteriorate, and the reported measurement report contains unknown PCI around.

In the case of occupying an over-covered cell, the RSRP changes greatly, accompanied by abnormal uplink signals and missing allocation of neighboring cells, which easily leads to abnormal events such as dropped calls, access failure and handover failure.

As shown in the figure below, the mobile phone occupies PCI= 12 1 cell, enodeid=47 1089, and cannot switch to peripheral PCI= 1 10, which eventually leads to dropped calls;

problem analysis

This kind of problem needs to be combined with the test and analysis of surrounding roads and the topological relationship of base stations to determine which area the problem cell is the main coverage cell in the surrounding area, and then take optimization measures to adjust it;

As shown in the figure below, the last four switch chains in the problem area are 1, 2, 3 and 4 respectively, in which switches 2, 3 and 4 are related to PCI= 12 1. At the same time, it is analyzed that there is no PCI= 12 1 in the main service cell of the surrounding roads. Look at the topology diagram and find that the station is located in a far area around.

Solution:

For the surrounding over-covered cells, methods such as adjusting the pitch angle, antenna hanging height and base station distribution are adopted.

For a base station built in a special scene, compared with this situation, the station actually covers the nearby scenic spots, but due to the terrain, the signal can not be completely controlled, and the neighboring areas of the far base station can be deleted in one direction to avoid the island effect.

1. 1.4. 1 example 1 Cross-regional coverage near the residential area of the second fire brigade in Guanshaling, Changsha City and Elliot Ngok Road Garden.

problem description

The handover of UE from 4625062 to 462093 1 is not timely, which leads to the reconstruction.

problem analysis

1, where enhanced coverage is 4625333 and 4620931;

2. Reduce the inclination of 4625062 by 2-3.

4625333 and 462093 1 The azimuth and inclination of the community have been adjusted to cover the intersection, but the coverage direction is partially blocked, which has reached the maximum optimization adjustment. After adjustment, the test effect is not obvious. After adjustment, the coverage area of the newly opened site -3 community at the intersection of Yuelu Avenue and Elliot Ngok Road in Changsha is enhanced. If we want to completely solve the coverage problem of this area, we need to open the site of this regional planning-the landscape tower of Changsha Oasis Community.

Solutions and retest results

Through the above adjustment, the station switched to 983474 in time, and the signal of cross-regional coverage area weakened.

1. 1.5 The switching parameter is set incorrectly, which makes it impossible to switch.

Problem phenomenon

The vehicle was traveling from east to west in Chazishan, and could not switch to the nearby community in time after returning from the base station 4625 14, resulting in poor RSRP of this section, which eventually led to dropped calls. The following figure

Compared with the previous test, this paragraph is normal, as shown in the following figure:

problem analysis

Judging from the cell occupied by the drive test, even if the station is disconnected, if it can be switched to 462576-2(PCI=259 center frequency 1895Mhz), the signal can be kept at-105dbm, and there will be no dropped call. Look at the reasons for not turning to 462576-2. According to the road test signaling, after the A2 event was reported, the reconfiguration message did not carry the adjacent cells with different frequencies. First, it is considered that the neighboring cell relationship is not configured. But after viewing the neighbor list in the background, I found that neighbors have been added. Further investigation on why the reconfiguration message does not carry the neighbor relation shows that the "EUtranCellMeasurementTDD" pilot in the EutranMeassaParas _ Intercarrifreq table is not configured with 1895, and it can be switched normally after supplementary addition.

solution

1: The main reason is that the handover of the 4625 14 base station is not smooth, so the faulty site is given priority.

2. From this incident, it is found that if the measurement parameter EUtranCellMeasurementTDD is missing different frequency points, it will also lead to the inability to allocate adjacent cells with different frequency points, even if the adjacent cell relationship is configured, it will be invalid. Therefore, it is necessary to regularly check whether "EutranMeassaparas _ Intercarrifreq" in the EUtranCellMeasurementTDD table contains the frequency points defined by adjacent cells in daily optimization.

1. 1.6 pilot redirection

Problem phenomenon

After the terminal reports the A3 measurement event, the base station directly sends the redirected RRCrelease message, resulting in dropped call.

problem analysis

As can be seen from the figure, after the terminal sends the A3 event, the system directly sends the RRC release message redirected to 37900, resulting in dropped calls.

Solution:

1. Turn on the neighbor cell switching function through the background parameter, and solve the redirection of configuring the neighbor cell but not turning on the switching function.

1.2 interference problem

Problem phenomenon

In the test, RSRP is good and SINR is biased, and areas with serious interference are prone to drop calls, handover failure and other abnormal events.

Problem analysis method

Interference involves many aspects, including internal interference and external interference. Please refer to the survey guidance document for detailed survey methods. Only the cases found at the scene are described here.

1.2. 1 Example1:Reasons for unreasonable PCI planning

In the following figure, it is found that the signal RSRP in the test area is good, accompanied by poor SINR, so the PCI planning problem is given priority, and the same PCI base station is found within a short distance, as shown in the following figure:

1.2.2 Example 2: Interference caused by overlapping coverage

In Grid 17, there are densely distributed co-frequency base stations in the West Second Ring Road, and there is SINR difference in the 200-300m section. The reconstruction probability of this section is high, which is the hidden point of dropped calls. On 23rd, 1 phone was disconnected.

In the following circle, the signals of the four stations are all at -95dbm~- 100dbm. If PCI=66/67, SINR is easy to deteriorate, and the terrain is relatively flat, with a peripheral spacing of 300-400 mm. ..

Solution:

In view of the unreasonable PCI planning, it is suggested to re-plan and modify PCI.

Aiming at the interference problem caused by overlapping coverage, firstly, the coverage is optimized by radio frequency to reduce overlapping coverage, and secondly, the different frequency networking scheme is adopted to solve it. For example, in this west second ring road section, three communities with PCI=66/67/68 may not cover this road section, and the other three communities can cover different road sections well, so it is easy to cause touch 3 interference after introducing PCI=66/67/68.

1.2.3 Example 3: Interference caused by overlapping coverage

The caller received interference from cell 4742 163(PCI254) near Qingshui section 462326 100 meters, which made it impossible to start handover and reconstruction. The base station coverage in this area is too dense.

Solution:

In view of the unreasonable PCI planning, it is suggested to re-plan and modify PCI.

Aiming at the interference problem caused by overlapping coverage, firstly, the coverage is optimized by radio frequency to reduce overlapping coverage, and secondly, the different frequency networking scheme is adopted to solve it. For example, in this west second ring road section, three communities with PCI=66/67/68 may not cover this road section, and the other three communities can cover different road sections well, so it is easy to cause touch 3 interference after introducing PCI=66/67/68.

1.2.4 Example 4: The SINR value on the east side of the intersection of Yuelu Avenue and Elliot Ngok Road is poor and the handover failed.

problem description

On the east side of the intersection of Yuelu Avenue and Elliot Ngok Road, the SINR value of UE occupying 46238 13 is poor, which leads to handover failure.

problem analysis

1, add neighbor relations of 46238 13 and 983474 133, 462526 1.

2. Control the coverage of the intersection of Yuelu Avenue 46238 13;

3. Verify that the newly-opened stations 983474 133( 170) and 4750223(83) are covered to form strong mode 3 interference.

4. Adjust the inclination angle of 46238 13 by 2 to 4, the azimuth angle by 300 to 290, and the inclination angle of 983474 133 by 2 to 0.

Repetitive result

By controlling the coverage and optimizing the neighbor relationship, the interference phenomenon of this section disappears.

1.3 base station version problem

1.3. 1 TM3\8 dropped the line after switching.

Problem phenomenon

According to the foreground signaling, the mode of the terminal is TM3, and then A2 measurement is sent, and then the terminal receives the reconfiguration message controlled by different frequency measurement, and the downlink is out of synchronization after sending, and the call is dropped after initiating reconstruction. In this process, the wireless environment revolves around RSRP-100 and SINR 3. From the background signaling analysis, the base station side receives A2 sent by the terminal, then sends the measurement reconfiguration message, and then sends the reconfiguration message of TM8 mode switching. When the reconfiguration of TM8 can't be sent, the user plane reports RLC error of SRB 1, resulting in release and dropped call.

problem analysis

The reconfiguration of 12:09: 14 seconds is in TM3 mode.

12:09: 14 seconds, send the A2 measurement reconfiguration, and then broadcast the message. 12:09:26 seconds after receiving the reconstruction request, it was rejected.

12:09:26 seconds bye-bye fell on the terminal.

A2 measurement is received at the base station side 12:09: 18.

12: 09:18second TM8 mode conversion reconfiguration.

12:09:25 seconds, an error mark appears, and the reconfiguration is not sent, reaching the maximum number of retransmissions.

12:09:25 seconds text release.

Solution:

This problem is a known version problem, which has been upgraded to R5p version in the field, and the problem has been verified.

1.4 core network related issues

1.4. 1 QCI=5 is not valid.

Problem phenomenon

The caller initiated a session request, but there was no response, resulting in no connection;

problem description

Caller 12:27:57 initiated the invitation, but it was not connected after 12:28: 13 did not respond. After checking the DRB bearer, it is found that there are two bearers with priority 9, as shown in the following figure:

solution

HSS deletes redundant APN contracts and then returns to normal, as shown in the following figure;

Pagination in the process of 1.4.2 TAU

Problem phenomenon

10:28:22, QCI= 1 was established after the caller initiated the invite request, but the called party did not receive the page; This phone call;

problem analysis

When the called party 10:28:23 moves, cell reselection occurs, TAC change and TAU update are initiated, but the disconnection event caused by this call paging is not received;

The failure to receive the page this time is a process conflict. When paging, we will treat the TAU message as a paging response. If the TAURequest message does not carry the activeflag, the user plane tunnel cannot be established and the message cannot be delivered.

solution

The core network will be modified in the next patch, and the user plane tunnel will be established regardless of whether the UE carries the active flag in the TAU process.

1.4.3 Return from 2G/3G to 4G core network without paging.

problem description

Calling 4G starts calling, called 3G returns to 4G, and the core network does not send paging.

problem analysis

The caller 16:23:39 initiated the inviterequest, and the called party returned from 3G that the 4G 16:23:49 ims registration was successful, but this page was not received. When the user is in 3G, HSS has already made domain selection, and at this time, the user has re-selected to 4G, which is impossible to reverse. There is no specification support for this scenario.

solution

The core network replied that there is no protocol support for the time being.

QCI= 1 was deleted after crossing TAC.

problem description

After crossing TAC, after receiving two instructions (QCI= 1/QCI=5) released by ERAB, the call was dropped on the core network S 15: 46: 24 of 462696-2 cell.

problem analysis

SGW mistakenly deleted the session, which caused the eNB to receive an error indication, and then initiated the release. There is a known problem with the existing SGW. In the process of "4 cuts, 3 cuts and 4 cuts in a short time under the condition that the MME has changed and the SGW has not changed", the SGW will delete the context and cause the user to drop the line.

solution

You need to upgrade the SGW version to solve it.

QCI= 1 is released due to handover during call initiation and loss of acceptance message;

problem description

With the handover of the terminal during the call initiation process, the ACT message sent by the terminal to the core network was released because it was not received by the core network over time.

problem analysis

2 1:0 1:23, the calling UE initiates paging, and the called UE initiates ERAB bearer after receiving it, and the establishment is completed; 2 1:0 1:26, the called UE received the RRC reconfiguration message requesting to activate the QCI 1 bearer, and then the called UE reported INVITE580 (the precondition failed), which led to this disconnection.

solution

DT message didn't wait (lost), whether the core network has a protection mechanism needs to be solved by the core network;

1.5 terminal exception

1.5. 1 Terminal hangs up abnormally, resulting in disconnection.

Problem phenomenon

The called party sends a ringing message of 180 to the calling party, and the calling party also successfully receives the ringing message of 180. However, after the called party sends the message of 180, it sends the busy tone of 406 to the caller three seconds later (see the figure below), and the core network plays the sound to the caller after receiving it, and then releases it. The same phenomenon, both phones are not connected.

From the signaling point of view, the reason why the called party sent the 486 user busy message was that the terminal actively refused, which had nothing to do with the network. As for why the called party sends the message that the user is busy and refuses after ringing for 3 seconds, the terminal problem needs to be solved by the terminal.

1.5.2 The terminal does not report TAU request.

Problem phenomenon

After a normal call, the caller switches from PCI= 17, TAU=29580 to PCI=64, TAU=29482, then releases RRC and reconnects to other cells. After repeating this process for three times, the terminal actively sends BYE, and the called terminal TAU is normal. Because the mobile phone terminal needs to report a TAU request after switching different TAU, the terminal has never initiated TAU here, which is the reason of the terminal.

1.6 test software statistics

1.6. 1 Abnormal statistics decrease.

Problem phenomenon

2G manually releases the called party, reports disconnection, and the hang-up process ends. At this time, the caller receives the BYE sent by IMS under 4G, deactivates the QCI= 1 bearer, and marks it as normal call clearing, but it will still be counted as dropped call. At this time, the caller continues the normal release process, which is a statistical problem of the software.

Analysis of 1.7 eSRVCC switching problem

1.7. 1 GSM neighbor parameter error leads to dropped call.

Problem phenomenon

Mobile phones cover a weak area of LTE. After receiving the reconfiguration message of B2 measurement, the mobile phone initiates the Measurementreport(B2 event) and receives the RRC connection release issued by the network, then redirects to 2G and drops the call.

problem analysis

When the UE reports the A2 measurement report, the eNB sends a B2 reconfiguration message to the UE. According to the B2 reconfiguration message, the UE measures the B2- 1 and B2-2 conditions and reports the B2 event. The reported B2 event includes the target 2G cell BCCH/NCC/BCC to be switched, as shown in the following figure:

1. Under normal circumstances, after receiving the B2 event measurement report, the eNB sends a mobilityFromEUTRACommand message to the UE and switches to the GSM neighboring cell;

2. In an abnormal situation, the network sends an RRC connection release message to redirect the UE to a GSM cell with a BCCH of 5 12, as shown in the following figure:

Then the caller is redirected to the GSM network, and the mobile phone is idle in the 2G network, so the call is dropped, as shown in the following figure:

Through the analysis of the above phenomenon, it can be seen that the handover of the UEVoLTE service eSRVCCC to the G network neighbor of BCCH 525(BSIC 12) failed. After checking the parameter configuration of the neighboring cell of G network in the network management, it is found that the BSIC configuration of the neighboring cell is 7, which is inconsistent with the actual BSIC 12 measured by UE. In the network management, after changing the BSIC of the adjacent area of G network to 12, it can be switched to this cell normally.

Solution:

Synchronize LTE--> GSM network neighbor definition and actual GSM network planning data, as shown in the above example, LTE->; The BSCI configuration in the GSM neighbor cell definition is 7, while the actual BSIC measured by UE is 12. After changing the BSIC of GSM neighbor defined by LTE to 12, eSRVCC is normal.

1.7.2 handover preparation failed.

Problem phenomenon

UE is empty talk, and eSRVCC can't be carried out after initiating B2 measurements for many times, which will eventually lead to re-establishment and dropped calls.

The eNB side displays that it has received the B2 measurement reported by the mobile phone and initiated the handover request, but it has received the message that the handover preparation of the core network failed;

problem analysis

Under normal circumstances, after receiving the B2 event measurement report, the eNB sends the MobilityFromeTracCommand message to the UE, and the UE will receive the MobilityFromeTracCommand to realize handover;

In an abnormal situation, the UE initiates multiple B2 events without receiving the mobilityFromEUTRACommand. At this time, it may be related to the deterioration of the air interface wireless environment, resulting in the B2 event measurement report not being reported to the eNB, which needs to be analyzed in combination with the signaling at the eNB side. As shown in the figure below:

When the eNB receives the B2 measurement report, it sends a handoverrequire message to the MME (preparing resources for eSRVCC handover), but then it receives a reply that handover preparation failed. See the figure below:

The reason for this failure is usually that the core network does not configure the relevant functional parameters of eSRVCC for the target cell, so the core network needs to check whether the relevant parameters of the target network cell are valid or configured correctly.

Solution:

The reason for this failure is usually that the core network is not configured with SRVCC function, the target MSC is not configured, and the TAU is not configured. Need to check with the core network and the target network whether the relevant configuration is effective.

1.7.3 GSM adjacent frequency point configuration is incomplete.

Problem phenomenon

After the UE reported the A2 event, the network sent the B2 reconfiguration message and successfully reported it to the network, but the mobile phone RSRP met the B2- 1 decision threshold but never reported the B2 event measurement report, which eventually easily led to redirection and dropped calls.

problem analysis

In the actual network, due to the change of wireless environment, the synchronization after G network parameter optimization or RF optimization is not timely, and the frequency allocation of GSM adjacent cells in LTE cells is not accurate enough. After the terminal measures, the GSM frequency point sent by A2 reconfiguration does not meet B2-2 event, which leads to the failure to trigger eSRVCC. As shown in the figure below:

Other possible reasons: it is necessary to check whether the neighboring cells of different systems have been set to "support handover", as shown in the following figure.

Solution:

Improve and update the GSM neighbor relationship and parameter definition in LTE neighbor definition in time;

For the problem points, it is suggested to scan the frequency of GSM network or analyze the GSM test data to check whether these frequency points have been included in the background configuration.

1.7.4 Due to the mobile phone, it is impossible to switch SRVCC.

Problem phenomenon

The calling and called mobile phones are in the same cell, and the reconfiguration message after the calling mobile phone reported A2 contains B2 threshold and frequency information of different systems, while the reconfiguration message after the called mobile phone reported A2 does not contain configuration information of different systems, resulting in that the called mobile phone does not perform SRVCC and the calling mobile phone is normal. The following figure

problem analysis

Due to two different behaviors in the cell, firstly, it is necessary to check the reporting ability of the mobile phone. From the UE attachment request message or TAU(TrackingArea Updates) message, it is found that the UE capability reported by the called mobile phone does not include the SRVCC capability message, and it is described that GSM frequency band is not supported. Including the caller, so it's important to check what's wrong with the phone. Because there was no such problem in the previous test, it was suspected that the mobile phone was mistakenly set to lock LTE during the test, so the mobile phone was set to lock LTE and supported 2, 3 and 4G modes for comparison and verification.

The following figure is the signaling description of the TrackingArea update without locking LTE, including the indication that the mobile phone supports SRVCC.

The following figure shows the TrackingArea update signaling after the called party locks the LTE network, indicating that the mobile phone does not support E-GSM mor R-GSM. SRVCC flag is not supported.

Solution:

This kind of problem needs to check the mobile phone settings and actual support ability to ensure that the support ability is reported.

1.8 Other

Parameter configuration problem of 1.8. 1 led to abnormal return to GSM.

Problem phenomenon

After the calling mobile phone occupies the 462502 base station to initiate INVITErequest and servicerequest, the mobile phone enters the GSM network and initiates the subsequent access process.

problem analysis

As shown in the figure below, the calling party and the called party occupy the same cell, and the wireless environment is good. The calling party can't stay in LTE to make a phone call, and the called party is normal, so the problem of occupying other base stations does not exist. After the mobile phone is changed, there is still an abnormal call process, which has little to do with the mobile phone.