- This topic has 36 replies, 1 voice, and was last updated 11 years, 3 months ago by Coolvepowly.
8th February 2009 at 12:23 #46018PixGuest
I’m not a big fan of TA filtering neither, unless for troubleshooting purpose (for example, try to filter the faraway MS for few hours, see if it enhances the QoS or not).
you’ll always face phantom RACH. There are two facts that I gathered about those ghosts:
– they are usually produced by the BTS itself (electromagnetic noise) –> you can’t fight it.
– they don’t matter, as long as you don’t face SDCCH congestion. Think about it… a ghost rach will just load your system a little bit more. If it’s below a certain amount, then consider it as acceptable and move on to other topics.
Looking at your problem, can you give some numerical values per hour during a standard day?
sdcch assign unsuccess (# and %)
requests and success ?
split of sdcch requested for calls, sms, location update, or imsi detach, etc.9th February 2009 at 21:01 #46019Azzazi_RFoptimization_engGuest
there are 3 cases can occure,
1- increase n location update, increase in sdcch blocking, and sdcch failures and low cssr (not phantom RACH)
2- same as above but with no sdcch blocking. (phantom RACH)
3- increase in sdcch failures with no increase in location updte or sdcch blocking at all. and ths is wieird phenomena !!!
all cases are solved by site reset or change BSIC or BCCH.
all cases may occure any time during the 24 hours.
we r fighting the ghosts of the networks :-S9th February 2009 at 21:51 #46020PixGuest
Those are not ghosts, but just hard/software malfunctions. Sudden increase of sdcch assign fail due to radio can usually be fixed by a reset.
Changing the BCCH or BSIC is another way of resetting a cell, that’s why it’s also fixing it.9th February 2009 at 22:30 #46021GSM_OptimizerGuest
there will be no changes in timing advance because in the phantom RACH phenomena the problem is in MS (not the BTS) which not recognize the served BCCH/BASIC cell but will recognize anothere BCCH/BASIC cell with different LAC and also the down tilting is not solution because it will affect the coverage area and also will not solve the problem and i traied to minimize the RX level access but also it didn’t solve the problem, the only solution is to make a new BASIC/BCCH plan,,,,,,, i am facing now a new phenomena which is phone RACH, phone RACH is caused due to faulty mobile hand set which goes into loop, mobile request RACH and after that the BTS give the SDCCH and wait for ack. but the mobile didn’t respond and begin to request RACH once again, this phenomen affects THE sdcchs POOL and cause SDCCH blocking, it may be related to location update and may not so call setup decreased, the only solution i have till now is to reset the site!!! but it is not a solution.10th February 2009 at 12:32 #46022PixGuest
First scenario is NOT a ghost rach. It’s an effect of duplicate frequencies, and the RACH is real.
Second scenario is strange… Are you certain the MS goes into loop ? Couldn’t it be the BTS that loops the received RACH ?
On which interface do you see the RACH repetition ?10th February 2009 at 19:33 #46023AAKGuest
Thanks PIX and TNS for the response
Now in my scenario (GHOST / PHANTOM RACH) following points has been observed.
1. The Location Update attempt do not go to the MSC but is discarded before BSC sends it to MSC. Discarded in BSC with Timers expire. At NSS end LU success rate is fine and around 96 %.
2. No SD blocking is observed.
3. Particular area gets affected and all of its cells covering that area have this issue.
4. The issue is always LU failure thus effecting CSSR. LU failure goes to around 80-85 % of all those cells.
5. Mostly the issue raises at night at around 3-4 am
6. Cells are not any LA border cells.
7. There are around 400-500 attempts and 85 % failures and the attempt and failures scenario goes for around 3-4 hours. sometimes also goes for 6 hours.
8. MOC / MTC / SMS SDCCH attempts are OK and success rate is around 95-97 %
Again what confuses me that “No SD blocking is observed”
and all LU failures are due to timer expiry10th February 2009 at 19:58 #46024GSM_OptimizerGuest
in first scenario, the duplication in frequancies between neighbour cells or what and also the RACH is not real RACH because this RACH is caused due to the MS which recognize anothere BCCH/BASIC different from its serving cell, to far from its serving cell(in different LAC) and try to access it but ofcourse there is RACH failure which increase the SDCCH failure rate and decrease the call setup.
in the second scenario, MS goes into loop and not the BTS because if the BTS goes into loop this will reflect on TCH_Traffic_Carried which will be equal 0 because the loop in this phenomena between RACH and SDCCH and if BTS make this loop there will be no traffic, our problem is to now the kind of hand set but we cann’t trace because the call didn’t completed and enter MSC to now the Kind from AUC.10th February 2009 at 20:40 #46025AAKGuest
What i think is that that the issue we are facing is same and also GHOST RACH / Phone RACH is also the same scenario but we cannot confirm that RACH command is due to interference or by a particular mobile phone
(if i would come to know the mobile manufacturer, i will surely sue it for making my office life so difficult) 🙂
But i still think that if it would have been mobile issue than it would be very common in every cell. The issue is that the problem comes in some cells and then other day some other cells
and in some cities, this issue do not exist at all.
so what do you people say
Mobile issue or Interference issue ??10th February 2009 at 20:53 #46026PixGuest
I faced something similar a loooong time ago:
a BSC from competitor is down during the night, so all the BTS from BSC are down.
Therefore, all MS from competitor will attempt to get a SDCCH and perform an IMSI ATTACH. It generates a huge amount of SDCCH congestion.
Ok… it’s nothing like your current issue 🙂 but perhaps it can give you some ideas?
Anyway, can you tell me which timer expires? Could you write down the full message flow of what you’re seeing (on the abis for example)?
Bts – BSC : CHANNEL REQUIRED
BSC – Bts : CHANNEL ACTIVATION
Bts – BSC : CHANNEL ACTIVATION ACK
BSC – Bts : IMMEDIATE ASSIGNMENT
Where does it stop?10th February 2009 at 21:16 #46027AAKGuest
Following is the Signaling flow analysis
1.MS sends a CHANNEL REQUEST (OK)
2.BTS sends a CHANNEL REQUEST message to BSC (OK)
3.Upon receipt of the CHANNEL REQUEST message, BSC allocates signaling channels, and sends a CHANNEL ACTIVATION message to BTS (OK)
4.Upon receipt of the CHANNEL ACTIVATION message, BTS selects specified channel, and sends a CHANNEL ACTIVATION ACKNOWLEDGE message to BSC (OK)
5.BSC sends an IMMEDIATE ASSIGNMENT COMMAND message to MS via BTS(OK)
6.MS sends an SABM frame for establishing link with BTS. (NOT OK)
After step 6, when MS do not send the SABM, the timer T3101 expires and we get an immediate assignment failure
then at step 7 UA frame is sent by BTS if step 6 is OK and then further steps are OK.
well one more question i want to ask that i have seen in some operators the LU success rate is subtracted from CSSR formula. Is it due to this reason.
Is LU success rate subtracted in your company CSSR formula11th February 2009 at 09:08 #46028PixGuest
This happens just in one area…? My guess is that there is a powerful military radar in the area. But it is strange that it’s always the same “LU coded” Random Access Ref that is heard.
Crazy thing is that it happens only at night.
Each ghost “Channel required” has a different Random Access Reference ?
Could you check it ?
So, as a sample:
What are the RA Ref of the ghost Channel requireds ? It should be 8 bits (i’m actually asking for the binary values…)
What about the timing advance (access delay) ? (no bits here please)
Ok, well, don’t give me one thousands example, but just a few ones.11th February 2009 at 10:13 #46029AAKGuest
Actually, it is not any particular area everyday.
It is just like it hops to areas to areas in the city. Today Area 1 then next day Area 2 and then goes on.
but i have not found any sequence yet.
But the issue i have found in all these cases is same. LU failures
Now what u r asking, i am not getting
from where can i find RA Ref of the Channel requireds
is it in any Interface TRACE. Um or A
Thanks11th February 2009 at 20:04 #46030PixGuest
Different areas every day… Mobile RF jammer ?? It happens in some countries… where the king goes, the jammer goes.
The RA ref can be found in the channel request (Air) or the channel required (Abis).
How can you trace the Air (Um) interface for all users ? Do you know such a tool ?
In my mind, the trace i ask you is on the abis. The Air is IMO impossible to trace (but am i wrong ?)12th February 2009 at 16:01 #46031AAKGuest
Well one thing i know is that
We can trace the UM interface at BTS end , but the issue is to understand and pick the required thing.
for RA reference i will let you know soon.
But for the king part, we have already checked for any VVIP, VIP or IP movement in that city 🙂
and we have found none13th February 2009 at 05:27 #46032GSM_OptimizerGuest
we are facing the same problem and this what i called phone RACH, AAK you are coorect the MS didn’t respond to the SABM and goes into loop and begin RACH request again so i want to ask you AAK if you face high SDCCH blocking rate because this phenomena affect the SDCCH pool, i was thinking that this problem is related to location update but i face one case without increasing in LU, after investigation the problem is in mobile hand set but i’m not sure from that because we cann’t prove it,
please AAK i want to ask you if your network is 2G/3G net. or 2G only and if 2G only you are making any rooming on 3G net.