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

Slow HO in PBGT with ALU

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • 21st June 2013 at 00:52 #69087
    IAN
    Guest

    Dear All,

    Using ALU, Maybe you could help, we had noticed that PBGT are tiggered very late, ie, during drive Test, 3 to 4 seconds after you drive on the next Sector of the Site you do the PBGT HO.

    After checking parameters, HO margin are 5 db, averaging windows are the in the normal range :

    A_PBGT_HO = 8
    W_PBGT_HO=1

    What other parameter could trigger such a late PBGT ?

    Thanks for ur help
    Ian

    21st June 2013 at 03:43 #69088
    pix
    Guest

    Hi Ian,

    A_PBGT_HO = 8 = 4 seconds….

    This explains why it takes 3 to 4s to move to the neighbour cell.

    However keep in mind that a typical A_PBGT_HO should be between 8 and 12. Otherwise, you will suffer from ping pong HO. Drivetests are very particular, as they don’t represent AT ALL how their subscribers are using their phones. They are mostly static or pedestrian.

    Reducing the A_PBGT_HO is valid for highway, or specific micro coverages.

    Regards
    pix

    21st June 2013 at 10:04 #69089
    Lemaure
    Guest

    Hi IAN,

    you should also check your RXLev_Limit_PBGT_HO if it is too low.

    BR

    21st June 2013 at 20:57 #69090
    Ian
    Guest

    Hi Pix,

    Thanks for answer but we are mostly use Driving Test from a car and we had noticed this behaviour. Moreover, when we put the logs into Actix, it confirms that a lot of HO are triggered very late.

    Hi Lemaure ,

    Thanks for ur comments, the RXLev_Limit_PBGT_HO is set to -47dbm.

    BR
    Ian

    22nd June 2013 at 06:29 #69091
    pix
    Guest

    Hi Ian,

    You now say “very late”, and before that, you said “4 seconds”. Which one is correct ?

    A PBGT HO = 8 means it takes 4 seconds to do a PBGT HO. There is nothing mysterious here….

    If your network is “multi-layered” (micro / umbrella, for example), then specific HO strategy is used, changing the triggers of ho. Sometimes you feel like the HO is not occuring during drivetest, but it’s actually normal : the MS is in the umbrella layer and will not go to the microcell, or vice-versa : MS is in microcells and will not go to the umbrella.

    But really, 4 seconds is “standard”. I would not call that “very late”.

    Regards,
    pix

    23rd June 2013 at 08:24 #69092
    Ian
    Guest

    Hi Pix,

    Maybe i used the wrong sentence, i mean very late because after driving over Sector1 (geograhic position), we should handover quickly on PBGT into Sector 2, but we are still on Sector 1 4seconds later, ie, if your frequency planning is tight, you can experience interference when you dragging the opposite sector too long.

    Like i said before, Actix is telling us this by noticing a lot of samples not on the Beam of the antenna but on the opposite side, thats why i said it was like a late handOver.

    I’m talking about Macro site, Intrasite handover.

    Best Regards
    Ian

    23rd June 2013 at 11:34 #69093
    Lemaure
    Guest

    Hi IAN,

    i hope during your drive test you don’t exceed 40Km/h because the more you go faster higher is the probability to drag a cell beyond its normal serving area.
    Also, you should not exceed 32 neighbours because it can extend the processing time of the BSC during HO detection; check on this issue too…

    cdlt;

  • Author
    Posts
Viewing 7 posts - 1 through 7 (of 7 total)
  • The forum ‘Telecom Design’ is closed to new topics and replies.