PDA

View Full Version : 2G Drops due to Radio link failure and T200 expiry



rimoucha
2011-04-18, 02:46 AM
Hello Experts,

One cell in my network has very high drop rate,when investigating the drop causes I found out that it's mainly due to "ERR IND Received on TCHH/TCHF (Traffic Channel) in stable state (T200 expiry)" and to "Conn fail received on TCHH/TCHF (Traffic Channel) in stable state (Radio link failure)".How to investigate further this problem?

Best regards,

m121
2011-04-18, 04:54 PM
Hi,
what vendor are you using?
Problem was allways or appeared recently?
there could be many reasons, first check statistics per TRX (Path ballance, number of drops, Interference on idle etc).
There probably will be differences, post them here for us to investigate further.
If statistics for all TRX's are the same then check Cells's configuration (In Excell or other tool check it against other cell (preferably same site)).

But I think that problem is with one or some TRX's in Cell (bad frequency plan, wrong installation, HW fault, external interference etc., could be even missing ADJC)

rimoucha
2011-04-20, 11:14 PM
Hi,

Thanks for your reply.Below are the answers to your questions

what vendor are you using? =>Huawei
Problem was allways or appeared recently? since it has been in service
there could be many reasons, first check statistics per TRX (Path ballance, number of drops, Interference on idle etc).=> Path balance is ok, no interference on band 3,4 &5
There probably will be differences, post them here for us to investigate further.
If statistics for all TRX's are the same then check Cells's configuration (In Excell or other tool check it against other cell (preferably same site)).

But I think that problem is with one or some TRX's in Cell (bad frequency plan, wrong installation, HW fault, external interference etc., could be even missing ADJC) => neighbors are properly defined,BCCH/BSIC plan seem to be OK.
Any ideas how to investigate this problem?

Thank you :)

stream1973
2011-04-20, 11:21 PM
could you please provide the MAIO offset, MAIO step and MAL of this cell

stream1973
2011-04-20, 11:22 PM
we need information of the MAL MAIO offset and Maio step of all cells of this site thanks

rimoucha
2011-04-20, 11:35 PM
Hi,

Thank you for your reply.the MAIO is 0 (different from the co-sites),the HSN is 7,the MA list has about 24 frequency.

BR,

stream1973
2011-04-20, 11:37 PM
no give me all information:

number of TRXs of each cell in the site.
MAIO ofset of each cell of the site.
MAIO step.
and MAL list.

riyantoyadi
2011-04-20, 11:49 PM
1. restart TRx
2. Check block TCH due to what kind? ig due to Aif please check cell define on MSC/MSS maybe wrong dbase create
3. What kind of alarm happend?

stream1973
2011-04-20, 11:53 PM
please rya let check his problem step by step. I prefer that he give us all information that i requested. thanks

riyantoyadi
2011-04-21, 12:05 AM
Please check first this parameter and LMT for Huawei. Huawei could be monitoring on realtime

Parameters Range Default Value Radio channel management control table Channel distribution algorithm selection Huawei I algorithms and Huawei II algorithms Huawei I algorithms Interference band pre-processing switch YES,NO NO Initial length of interference band filter 0~7 3 Length of averaged interference band 0~31 6 Whether SDCCH is dynamically distributed YES,NO YES Idle SDCCH channel threshold 2 The maximum number of SDCCH channel in the cell specifying the maximum number of SDCCH in the cell. Configured SDCCH in the cell + 8 (too big a value will not be good for system recovery during abnormality) TCH Minimum recovery time (second) 60~3600 600 TCH Idle threshold N1 0~cell TCH timeslot number 2 TCH decision application period T (minute) 0~65535 10 HW II Channel Allocation table SD allocation priority allowed YES,NO YES iInterf. priority allowed YES,NO YES Active Ch Inter.Meas. allowed YES,NO YES AllocationTRX priority allowed YES,NO YES History records priority allowed YES,NO YES Balancetraffic allowed YES,NO YES Interf.of UL Level thrsh. 0~31 31 Interf. of UL qual. Thrsh. 0~70 40 Interf. of DL Level Thrsh. 0~63 31 Interf. of DL Qua. Thrsh. 0~70 40 Filter length for TCH Level 1~32 6 Filter length for TCH Qual. 1~32 6 Filter length for SDCCH Level 1~32 2 Filter length for SDCCH Qual. 1~32 2 Update Period of CH Record 1~1440 30 Update Freq. Of CH record 0~63 2

rimoucha
2011-04-21, 12:48 AM
Hello,


This site has three GSM 900 cells (cell 1,Cell 2 and cell 3) and two GSM 1800 cells (Cell d1 and cell d3).The cell 1,cell 3,cell d1 and cell d3 are among the 5 worst cells in my network with more than 50 drops per day.
Let's focus on DCs cells,we use single MA/HSN per site (RF hopping mode on the second TRx):

Cell d1: two TRxs (TRx1:BCCH;TRx2:Hopping with MAIO=0,HSN=7,MAlist=F1..F24)
Cell d3: two TRxs (TRx1:BCCH;TRx2:Hopping with MAIO=16,HSN=7,MAlist=F1..F24)

I do not find the MAIO step parameter.
Attached are the hopping configs for each cell.

BR,

rimoucha
2011-04-21, 01:13 AM
Hello,

Other information that may be useful:

- I have noticed that in front of cell1/d1 and cell3/d3 there is a main road interchange (with bridge).
- The HO request distribution is the following:

UL_Quality_Requests_Ratio 4.24 %
DL_Quality_Requests_Ratio 5.78 %
UL_Strength_Requests_Ratio 6.94 %
DL_Strength_Requests_Ratio 0.19 %
TA_Requests_Ratio 0 %
Better_Cell_Requests_Ratio 45.86 %
Load_Requests_Ratio 0.19 %
DL_Rapid_Level_HO_Requests_Ratio 0 %
MSC_Intervention_Requests_Ratio 0 %
OM_Intervention_Requests_Ratio 0 %
Other_causes_Requests_Ratio 36.8 %

Any suggestions please?

BR,

timedomain
2011-04-21, 08:46 PM
this might be transmission problem as well i faced somthing similar before

polo
2011-04-21, 10:43 PM
you should check T200 & RLT setting .

mashi123
2011-04-22, 02:42 AM
Hi rimoucha
1. regarding the two causes Radio link time out and T200 expiry in Huawei most of the drops in the network are due to these two causes + equipment failure + abis failure so you cannot tell what is the agent (agent is the root cause of drop and also the bad guys in Matrix movie ^_^) depending on these counters alone.

2. the most helpful counters in your case is the following counters in Huawei
the UL/DL Rxlevel during the drop
the UL/DL Quality during the drop
the TA measurement during the drop
in M2000 go to MR Measurements --> Radio link failure Measurements per TRX
then u will find the below counters for each TRX
S4300D:Uplink Level during Radio Link Failure (SDCCH)
S4307D:Uplink Level during Radio Link Failure (TCHF)
S4308D:Uplink Level during Radio Link Failure (TCHH)
S4310D:Uplink Quality during Radio Link Failure (SDCCH)
S4317D:Uplink Quality during Radio Link Failure (TCHF)
S4318D:Uplink Quality during Radio Link Failure (TCHH)
S4320D:Downlink Level during Radio Link Failure (SDCCH)
S4327D:Downlink Level during Radio Link Failure (TCHF)
S4328D:Downlink Level during Radio Link Failure (TCHH)
S4330D:Downlink Quality during Radio Link Failure (SDCCH)
S4337D:Downlink Quality during Radio Link Failure (TCHF)
S4338D:Downlink Quality during Radio Link Failure (TCHH)
S4340D:TA during Radio Link Failure (SDCCH)
S4347D:TA during Radio Link Failure (TCHF)
S4348D:TA during Radio Link Failure (TCHH)

if u analyze these Counters u will get one step further in determining the agent for example
if i have very bad DL quality during the drop in certain TRX then i will focus on quality agent or if i have very bad UL level in one TRX i will go and check what is the problem with the level in that TRX.

3. in ur case u have RF hopping activated
if its not carefully activated it will create alot of collisions and alot of interference
please check the RQI measurements ( RQI = 2 * C/I)
if i have low RQI then its either interference problem or very bad UL/DL level


u have to tell us what is the size of problem one TRX, multiple TRXs
this is important , for example if the problem in all TRXs except BCCH then most probably will be related to rf hopping setting

mashi123
2011-04-22, 02:50 AM
also like to tell u
if i have T200 expired this will lead to radio link time out
so the final result is the same for both reasons .. radio link time out

even the transmission outage problems affect these two counters in huawei

mashi123
2011-04-22, 03:09 AM
Hi Remousha
u mentioned in one of ur post that the problem exist in two cells in the same site
i have similar case before it was caused by one feeder swap (not two feeders swap)
to make sure check the following :
1. level and quality during the drop (the measurements that i have mentioned to u before) if u have one or two TRXs with very bad level in each cell then this is one evidence.
2. check the handover attempts per neighbor is it logical
i mean by logical is most of handovers go to the expected neighbor or another neighbor in opposite direction
dont go for drive testing unless u r sure it's swap

rimoucha
2011-04-22, 06:53 PM
Hello,

Thank you for your answers.I've verified: there is no transmission issue.
Mashi123,please find attached the RLF measurement statistics per TRX (for the GSM900 cell ans it's co-cell DCS1800). I actually didn't know how to analyze it.Could you help me please?

Thank you in advance

mohyedeen_alkousy
2011-04-22, 07:20 PM
@Rimoucha
can u send the below counters also they exist in RLF measurements sub-function , they have same function of previous counters but averaged values

AS4307D:Mean Uplink Level during Radio Link Failure (TCHF)
AS4327D:Mean Downlink Level during Radio Link Failure (TCHF)
AS4308D:Mean Uplink Level during Radio Link Failure (TCHH)
AS4328D:Mean Downlink Level during Radio Link Failure (TCHH)

AS4317D:Mean Uplink Quality during Radio Link Failure (TCHF)
AS3337D:Mean Downlink Quality during Radio Link Failure (TCHF)
AS4318D:Mean Uplink Quality during Radio Link Failure (TCHH)
AS4338D:Mean Downlink Quality during Radio Link Failure (TCHH)

AS4347D:Mean TA during Radio Link Failure (TCHF)
AS4348D:Mean TA during Radio Link Failure (TCHH)

can u send it in excel sheet

rimoucha
2011-04-22, 10:09 PM
Hello,

Thank you mohyedeen please find attached the RLF measurements in excel format.

BR,

mohyedeen_alkousy
2011-04-23, 02:39 AM
@Rimoucha
from the measurement that u have sent i can conclude the following :

1. the drop is occurring at distance less than 1 km from the site and that exclude many possibilities like overshooting ,...

2.the drop is occurring on strong level on uplink and downlink in all TRXs and that also exclude other possibilities like power problems

3. the uplink quality is acceptable in all TRXs but the downlink quality is bad in 1542 , 1544 , 303 and 305

bad downlink quality on strong levels in more than one TRX
i think u have to check the frequency plan again
try to change MAIOs this could lower the collissions and check the BCCh frequencies.

one important point why the handover cause distribution is not affected by the bad quality at downlink.
maybe the answer is that u have lot of repetitive handover that increased the percentage of PBGT handover.

rimoucha
2011-04-23, 03:05 AM
Hello,

Thank you very much for your answer.
- Could you please tell me how to analyze those RLF statistics?how to judge whether the UL/DL level is good or not?UL/DL Quality is is good or not?
- So as far as I have understood it's probably an interference issue => I will review my frequency plan and try to change the BCCH and hopping parameters.

BR,

mohyedeen_alkousy
2011-04-23, 06:17 AM
@Rimoucha
u r welcome any time
for the RLF measurements to analyze them u have to note the following :
1.the mean value of UL/DL Level/Quality=
the accomulated UL/DL Level/Quality values divided by number of radio failures
why?
the accumulated value is the summation of rxlevel or quality taken from all the measurement reports that arrived to bts or mobile just before the drop happened
for example if i have the following three MRs taken before 3 drops happend in three different mobiles :
1st MR contain the following:
serving cell Rxlevel Sub = 10 (10 means 10-110=-100 dBm)
2nd MR contain the following:
serving cell Rxlevel Sub = 4 (5 means 4-110=-106 dBm)
2rd MR contain the following:
serving cell Rxlevel Sub = 7 (7 means 7-110=-103 dBm)

then the accumulated value of Rxlevel will be 10+4+7=21
for us the value 21 doesnt mean anything unless we divide it by the number of measurement report which is equal also to number of drops so
mean rxlevel=(10+5+7)/3=7 or -103 dB

and the same scenario goes for rxquality and TA measurements

2. to get the most accurate result we should take large period of time and sum all the accumulated values and divide it by number of drops and by that way we can get accurate rxlevel or quality or TA uplink or downlink

3.now i think u know the rest
if i have bad quality and bad levels it means something
and if i have bad quality it means another thing
if i have very large TA and bad level it mean another thing
if i have very bad level at only one TRX it means something wrong with that TRX


one last important thing
if the baseband hopping (not rf hopping ) is activated then all the measurements per TRX will be wrong
that goes for all vendors not only huawei

rimoucha
2011-04-26, 04:09 AM
Hello,

The drive test revealed that both feeders for those sectors are crossed in 900 & 1800 which may cause interference.We are about to correct this and I will let you know about the result :)

Br,

rajmeena06
2011-07-22, 05:40 PM
Dear all,

To analyze RLF Fail report Below mentioned formula is useful .
If some one have good macro or idea Pl share.

IF(J2>4,"TA",IF(I2<10,"UL LEV",IF(G2<10,"DL Lev",IF(H2>5,"UL Qual",IF(F2>5,"DL Qual","Fast Fading")))))

The formula shows due what reason drop happen.