Sei sulla pagina 1di 17

SDCCH Access KPI Problem

Ramdom Location Update Increase


25-April-2010
Problem Description and Conlusions
SDCCH Access KPI drop for one or more hours for a BTS randomly (KPI drop more than 20%
Some times drop more than 60 % )
It happen in all BSC6000 and BSC32.
it happen at any hour with low or high traffic
Problem has increase in last two months but it is happening from more than one year ago
There are not congestion in SDCCH Channels at that hours
It is not consequence of recent BTS and BSC Software Upgrade because it continue happening
after and before upgrades.
Fail is random in time, so there are not high KPI BSC impact per hour because problem occur at
different hours at different days in different BTS but KPI in BTS is bad during these hours
There are not BSC/BTS/Sector/TRX alarms at these hours
Channel Request amount for MOC and MTC, (including average Location Update Success per
hour) are normal at that hour, so service is normal for most users at these cells during the problem
Channel Request amount for Location Updates Increase dramatically (In fact it increase up to
1000% but MOC/MTC keeps normal. All SDDCH Inm asignment for those additional LU fails in MS
Some examples for this problem are not border LAC sectors .Problem happens in rural and urban
areas or isolated BTS, so it is not a coverage or border LAC ping pong problem.
It is not happening in 3G Network for periodic registration, so it is just 2G problem.
After further investigation from Huawei it also happens in other countries like Venezuela,
Rumania. After Brightcomms Investigation It also happens in at least two 2G Ericsson Network.
So conclusion: problem is in an especific MS but is not a Huawei Network Problem.
Next Slides show the analysis and conclusions on this issue
% SDCCH Access vs SDCCH Request Times x Service Type

EXAMPLE: Aranjuez
Location Update
Problem Location
Update (green line)
increase more than
1000%. And KPI SDCCH
Access decrease (Purple
line)
% SDCCH Access vs SDCCH Request Times x Service Type

OTHER EXAMPLE:
BTS Lazaro Location
Update Problem (All Febrero-Abril
BTS Sectors)
% SDCCH Access vs SDCCH Request Times x Service Type

Other Example:
BTS GAB Loc Update Febrero-Abril 10
Problem

Octubre-Diciembre 09
RBS VMEA
BSC6000 LA PAZ
Other Example: BSC6000 VMEA
History from Jan10 (3 Months)
Problem Has Increase 1 Month ago

RBS TAR1A
BSC32 TARIJA
Other Example: BSC32 TAR1
History from Jan10 (3 Months)
Problem Has Increase 2 Month ago
RBS TARA
BSC32 TARIJA

Other Example: BSC32 TAR


History from Jan10 (3 Months)
Problem Has Increase 3 Month ago
SDCCH ACCESS BSC6000SCZ2 x Sector . EXAMPLE for a Cluster

Other Example: All BTS from a cluster


History from two weeks. All Cluster has
same behaviour
Problem Happen randomly for almost all
BTS at different days and different hours
Other Example: SRO
Last Two Weeks

RBS SRO Santa Rosa


BSC6000 SCZ2

BTS SRO
Six Months (Last Year) June-September 09
BTS SRO Santa Rosa BSC6000 SCZ2
History for 3 Months just between 1am y 7am morning.
Period of 6 Months: Jan-Apr10.
It happen during not peak hours too, so it is not Peak hour problem
* RBS SRO Santa Rosa BSC6000 SCZ2 Same behaviour in A,B,C
* No alarmas, no sdcch Congestion no drop calls. Traficc normal
* Rural Isolate Site .No interacion with other cellls
* Mobile Originating and terminating calls (MOC, MTC) are OK,
(red and blue lines) but Location Update increase about 1000%
for some hours and SDCCH Accesability decrease Green vs
Purple Line
When
problem
appear
This part is
OK

Here is the
problem
1) RED LINE: LU REQ increase up 700%, (100 to 700 per hour)

2) BLUE LINE: Network Respond correctlly sending all aditional


Normal Behaviour: InmAss Commands at same quantity as LU increase. So SDCCH is
Almost 100% reserved and ready to use (there are not congestion on SDCCH).
SDCCH Success =
Purple Line / Blue 3) ORANGE LINE: For all of this Aditional LU Req, Time Out
Line Ocurred (T3101 = 3Seconds). It means MS can not get SDCCH
channel correctly, so something is missing in air interface for an
specific MS
4) GREEN LINE: In average normal 100 location update per hour
are OK at the same Time.. It means most MS are doing location
update correctly. It means it is not a problem in Location update
process in CORE NETWORK side. It is just in a Specific MS

5) YELLOW LINE: KPI SDCCH ACCESS, drop sharply Because is


calculated as : Total Establish Indication (PURPLE) / Total
Inmediate Assigment (BLUE) and all additional LU REQ are failling
Other Example in CBBA: TDOB
It last 8 hours. LU increase 1000%
Other BTS: Tilata C
Same Problem Two times at
diferent days. Diferent hours
One of them at low traffic
hour. So it is not a
traffic/load problem

SEE next slide Um Trace


Trace Um Layer 3 Message Detail in BTS Tilata at 3:14 to 3:22 PM

Channel Request Increase too much, but


network Respond without problem with similar
quantity of Inmediate assignment. But one or
Some MS doesn,t respond any more with SABM
/ Call indication Mesasge

Location Update Request Vs Location


Update Accept are Normal during problem
hour. In Fact, values are similar than on hour
before and One Hour after

13 14 15 16 17 TIME/HOUR
Trace In Tilata
During the problem.
Too many Channel request
within some seconds. It is
an abnormal behaviour

Potrebbero piacerti anche