Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors

For Pix-ALU SDCCH Drops(Radio)

  • This topic has 17 replies, 1 voice, and was last updated 11 years ago by MehmedO.
Viewing 15 posts - 1 through 15 (of 18 total)
  • Author
    Posts
  • #68667
    Aakriti
    Guest

    Hi pix,
    How are you.Soory i am raising a new post for my concerns,I have a Problem of High SDCCH Drops in ALU.
    In My Network there are 2 Vendors- ALU & ZTE, ZTE is not impacted but ALU there are High SDCCH Drops since begining.
    Now on Analysis from NPO, All the drops are in the Bucket of SDCCH Radio Drops, There are no drops with SD HO(as it is disabled), Hardly 10-15 Drops sometimes in BSS Drop bucket.
    Can you suggest how i can further analyze SDCCH Drop in ALU system.SD Assignment is not that bad, Neither TCH Congestion is the issue. TXR,Signalling & HW are Ok.
    Also, Can you suggest some Timers-parameters to Reduce SDCCH Drop.

    #68668
    Aakriti
    Guest

    Hi Pix,

    Please guide.

    #68669
    pix
    Guest

    hello,

    I suspect that the SDCCH drop is happening only in some cells ? not the whole alu network.

    At what hour is the peak of sd drops ?

    what is the SDCCH Mean Holding Time ?

    can you get sdcch drop for location area updates only ?

    The idea is to prove that those drops are due to a storm of location area updates from faulty mobiles.

    On the opposite, if the drops happen on all cells, then you may be facing a wrong parameter configuration. Could you try to put all your SDCCH timeslots on the BCCH TRX ?
    (ie. no sdcch ts on TRX other than TRX1)

    Are you using frequency hopping “sfh” ?

    cheers
    pix

    #68670
    Aakriti
    Guest

    Hi PIX,

    Taking 1 BSC – SDCCH Drops are 1800 approx in Busy Hour for all the days, Traffic for this BSC in Busy Hr is 1400 Erl approx.

    1.Yes the Problem is not related to all the cells, But 20-40 % of cells of all the BSC have this Problem, Though if compare with ZTE the problem is there with only 2-3% cells.

    2.Drops are Increasing when Traffic is Increasing, Highest in the Busy Hour i.e 19th Hour.
    Drops Start getting Double from 300 Drops to 600 & Increase so on from 7:00 AM till 22:00 Hrs in night.

    3.Mean Holding Time i.e(SDCCH_occy_total / SDCCH_allocated) from NPO is 4.1 to 4.2 every day in Busy Hour.

    4.Also, SDCCH Is mainly Configured on BCCH TRX, But will check for few cases & if configured on TCH, will shift & check the case.

    5.Checked with many cells, SD Assignment & TCH Assignment are good for these Poor SD Drop cases, Even TCH Cong is not there.

    6.We are Using Adhoc Freq Plan, Based on Optimized Frequency Plan generated from Measurements.. No degradation Observed because of this.

    7.I am not able to get the SDCCH Drops due to Location Update failures, Can you help me with counters/indicators.(Also I learned somewhere that LOC Updates does not happen on SDCCH in Dedicated Mode) so how it will Impact in SDCCH Drops.

    8.These are the drops, Happening due to radio Link Timeout or Lapdm timer expiry, So as Immediate remedy , Are there any Timer or parameter which can be increased to resolve this slightly.

    Thanks,
    Aakriti

    #68671
    pix
    Guest

    hi akriti,

    I’ll copy & paste your post here, and reply within each bullet point.

    1.Yes the Problem is not related to all the cells, But 20-40 % of cells of all the BSC have this Problem, Though if compare with ZTE the problem is there with only 2-3% cells.

    — OK

    2.Drops are Increasing when Traffic is Increasing, Highest in the Busy Hour i.e 19th Hour.
    Drops Start getting Double from 300 Drops to 600 & Increase so on from 7:00 AM till 22:00 Hrs in night.

    — that seems to indicate a high frequency load due to high traffic load. In other words, interference. Check out TCH drops on the same TRX than the one carrying SDCCH. If AMR is enabled, then TCH will be much more resistant to interference than SDCCH, and will face less drops, rending the comparison useless. So you can check the average DL & UL RXQUAL through RMS on each TRX. For the 40% cells having high SD Drop, do you see a higher average RXQUAL ? (.8 is ok, 1.2 is really bad.. decimals are important!) Check out the UL FER too.

    3.Mean Holding Time i.e(SDCCH_occy_total / SDCCH_allocated) from NPO is 4.1 to 4.2 every day in Busy Hour.

    — That is very acceptable. How high is the SD Drop % on those cells ?

    4.Also, SDCCH Is mainly Configured on BCCH TRX, But will check for few cases & if configured on TCH, will shift & check the case.

    — ALU system will preferably use the SD timeslot located on the “highest” TRX. Only when this SD TS is full, then it will use a “lowest” SD TS. At busy hour, that is probably the case, and more and more SDCCH TS are being used. I let you find out the best methodology to change TRX location. There must always be at least 1 SD TS on TRX0.

    5.Checked with many cells, SD Assignment & TCH Assignment are good for these Poor SD Drop cases, Even TCH Cong is not there.

    — You could try to work on power control here. Try to inhibit power control for a longer time (SDCCH PC INIHBIT, or something similar). Instead of 3s, put 6s….

    6.We are Using Adhoc Freq Plan, Based on Optimized Frequency Plan generated from Measurements.. No degradation Observed because of this.

    — Not until now, isn’t it ? Were the SD Drops fine when you first deployed adhoc fp ?

    7.I am not able to get the SDCCH Drops due to Location Update failures, Can you help me with counters/indicators.(Also I learned somewhere that LOC Updates does not happen on SDCCH in Dedicated Mode) so how it will Impact in SDCCH Drops.

    — Look for counters which are named *LU*. Then drag and drop counters into the indicators tab, to get all related indicators.

    8.These are the drops, Happening due to radio Link Timeout or Lapdm timer expiry, So as Immediate remedy , Are there any Timer or parameter which can be increased to resolve this slightly.

    — Yes, there is a “timer”, it is the same than for TCH (radio link timeout). Not really a timer, but i guess you are familiar with that.
    But it reminds me about ‘lapdm”. Could you check the LAPD RSL congestion ? (can be seen in the NPO object “lapd bts-bsc”, very few counters there). If RSL is congested, that might lead to delays, and possibly SDCCH drops. It is a very long shot 🙂

    Rgds
    pix

    #68672
    Aakriti
    Guest

    Hi Pix,

    AMR Is enabled, Also below are my pint wise reply :

    1.TCH Drop is High on BCCH TRX, (i.e 10-15 Drops on BCCH TRX, 2nd TRX Only 2-4 Drops).

    2. SDCCH Drop I am not able to find at TRX Level- Though SDCCH Is defined On BCCH TRX Only for the cell verified.

    3. UL_RxQual Samples are almost same on both the TRX(Calc : sum(0-5)/sum(0-7)*100), i.e Rx-Qual UL- TRX -1 = 98.28. TRX-2 = 98.30 (for 1 Day).

    4.DL_RxQual Samples are Poor in BCCH TRX (Calc : sum(0-5)/sum(0-7)*100), i.e Rx-Qual DL- TRX -1 = 95.95. TRX-2 = 96.21 (for 1 Day).

    5.Also I have checked “Number of calls with bad quality measurements and with bad FER measurements on the uplink path” : TRX 1 = 225 on an avg. & TRX 2 = 50 on an avg for 10-15 Days.

    6. UL & DL Interference is also High in BCCH TRX.

    7.SDCCH Drops are High Since begining,No Impact of High SD Drops seen after Frq Plan Implementation.

    Thanks,
    Aakriti

    #68673
    pix
    Guest

    ok, so just try to put the 2nd -> nth sdcch ts on the TRX 2 -> n
    only keep 1 sdcch ts on trx1.
    That should help.
    As the stats seem to indicate, the bcch freqs seem quite interfered.

    #68674
    Aakriti
    Guest

    Ok,I will try & let you know..

    Also Please let me know if Changing TCH Priority from BCCH to TCH will be helpful in any case ??

    #68675
    pix
    Guest

    TCH priority will not change SDCCH allocation algo 🙁

    #68676
    MemedO
    Guest

    Hi,

    There is really long SDCCH duration before dropped like 44 seconds. All these drop durations are same around 44-45 sec. I wanna be sure that it should not be radio fails.

    T200: 360 ms, RLT: 48 SACCH period.

    Most SD drop occurs in less than 25 second.

    But long duration fails are aound 44 second.

    Do you have any idea ?

    #68677
    pix
    Guest

    hi MemedO,

    Interesting problem… Were you able to trace those SD drops ? It looks like it is the only way to find out more about them.

    cheers
    pix

    #68678
    MemedO
    Guest

    Hi Pix,

    There is no fail for Abis and A interface. Because BSC release SDCCH channel after 42 sec and it is pegged in SD_ABIS_FAIL_CALL.
    I found that there is some USSD service which waits answer from UE on SDCCH.
    I just tried to increase T9105, but there is no response from operator tı implement it

    Fyi, MehmedO

    #68679
    pix
    Guest

    hi memehdo,

    Obviously, the MS is just stopping the communication for no apparent reason, and the BSC is waiting until timer expiry (45s).
    It sounds a lot like the “crazy” MS situation. Are you able to check whether it is a SD drop during a LU procedure ? (some vendor provides such counter)

    Since the fault comes from the MS, i’m not sure which action would fix the issue… A trace would allow to know which IMSI(s)/IMEI(s) are behaving poorly.

    Regards,
    pix

    #68680
    MemedO
    Guest

    Hi Pix,

    The SDCCH drop has been improved after timer change as I mentioned before.

    Br, Ahmet

    #68681
    pix
    Guest

    T9105 monitors a BSC-MSC procedure, which should be ultra-fast. What were the old & new values you set for this timer ?

    it’s very intersting, i’m thankful you are sharing this info.

    thanks,

    pix

Viewing 15 posts - 1 through 15 (of 18 total)
  • The forum ‘Telecom Design’ is closed to new topics and replies.