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

Multiple CCCH

  • This topic has 16 replies, 1 voice, and was last updated 10 years ago by pix.
Viewing 15 posts - 1 through 15 (of 17 total)
  • Author
    Posts
  • #58296
    awahma
    Guest

    If a cell is having multiple CCCH configuration and there is one other timeslot except timeslot “0” then how many CCCH blocks will be available on this extra timeslot 51 Multiframe as normally four consecutive timeslot make 1 CCCH block so is it 12 CCCH block and three Idle timeslots or is it different?

    Normally 9 CCCH blocks on timeslot “0”(36 timeslots out of total 51) are avaailable while rest are divided between Freq, Synch, BCCH and Idle.

    #58297
    Pix
    Guest

    the second CCCH timeslot is a duplicate of the BCCH timeslot.
    It contains exactly the same number of CCCH blocks, divided between AGCH and PCH the same way as the BCCCH TS.

    Cheers,
    Pix

    #58298
    awahma
    Guest

    you mean that it contains 5 Frequency burst, 5 synchornization burst ,BCCH block (4 consecutive timeslots) and one idle timeslot in addition to 9 CCCH blocks (divided b/w PCH and AGCH)

    seems strange to me, BCCH is being transmitted on timeslot “0” and this extra CCCH timeslot also?

    My follow up to orignal question,
    What advantage I would get from this extra CCCH timeslot in terms of LAC size and Paging Success rate?

    #58299
    Abdel
    Guest

    if ur using combined BCCH(i.e BCCH+SDCCH/4) then u have 3 blocks of CCCH, if u r using non-combined BCCH then u have 9 blocks of CCCH, according to GSM 3GPP specs u can configure extra BCCH+CCCH timeslot on 2,4,and 6, this is the CCCH_CONF parameter.
    u can configure additional BCCH+CCCH slot in timeslot 2 in every Cell of that LAC and will increase ur paging capacity times 2, but remember the PCH load must not exceed 60% if u want to preserve quality. and there might be also some cell parameters that sets the PCH thresold. Ur paging capacity depends on how much CCCH blocks are reserved for the PCH, the number of BCCH+CCCH blocks, and whether ur using IMSI OR TMSI for paging.

    #58300
    awahma
    Guest

    OK PIX and Abdel, Thanks alot for both of you taking time out to answer my question.
    Just to sum it all up,

    “If I have two LACs each with say 125 cells and same number of TRXs currently operating with single CCCH timeslot(CCCH_Conf=0, res_ag_blks=2 and bs_pa_mfrms=2) and if I increase 1 CCCH timeslot for all cells of both LACs then I can combine them into one keeping other variables same.

    Is this assumption cum conclusion of mine is correct?

    Second, as far as I know one Paging request messiage type 3 carrying paging message for 4 MSs will take one CCCH block(4 consecutive timeslots).Is this correct also?

    Thanks again!

    #58301
    pix
    Guest

    awahama,

    the second CCH is a duplicate that double the capacity of RACH, PCH and AGCH, but the FCCH and SCH are not duplicated. The corresponding bursts are set to “idle” instead.

    Regarding the merging of 2 LACs, yes, that is possible. As long as all the cells in both LACs have mCCCH. The drawback is quite obviously that you’re loosing 1 TCH ts in all your cells 🙂

    One Paging Request = One CCCH Block (whatever the number of MS contained in the Pag Req.).

    Cheers,
    pix

    #58302
    awahma
    Guest

    Thanks PIX.

    Your reply has finally cleared all the air of confusion regarding this multiple CCCH subject.

    Just one final question,

    In our network we normally go for LAC splitting after PPS (pages per second) for a particula LAC crosses the figure of “40”.

    From where this figure comes?
    Is it a GSM recommendation?
    Can you please mention ETSI reference for this or exact procedure for calculation of this figure or any thing you people or using in your networks?

    Thanks.

    #58303
    pix
    Guest

    awahma,

    paging per second on which interface ? on the Abis/Air interface or on the A interface ?

    On Air/Abis interfaces:
    max = 33 pag/s without mCCCH
    max = 66 pag/s with mCCCH

    on A interface : it actually depends on your BSC processing capacity. in alcatel, BSC MX = 120 pag/s, BSC LEGACY = 70 pag/s.

    Cheers
    pix

    #58304
    awahma
    Guest

    I was referring to Air-Interface.

    MY question remains there that where these figures are coming from?

    33 Pages/sec according to you and 40 we are using.

    With setting as (CCCH_Conf=0, res_ag_blks=2 and bs_pa_mfrms=2)

    We have total of 28 Paging blocks during an interval of 941.5(235*4)milli seconds which makes that network will repeat paging for same paging group after almost (941.5/28) 33 m seconds.
    (I am making an assumptin that one CCCH block is one page as mentioned by you in your previous post).
    Now if we take inverse of 33 m seconds it will make 29.74 Pages per second.

    Now what do you think of this calculation and why it is different from 33 and 40?Is it logical or I have misunderstood something?

    Thanks!!!!!!

    #58305
    Pix
    Guest

    the difference lies in the number of PCH blocks within the 9 CCCH blocks.
    According to your settings, there are
    2 AGCH blocks
    –> 7 PCH blocks.

    In my computation, there are only 5 PCH blocks.

    Second assumption in “my” calculation : in average there are only 2.5 MS being paged per Paging Request. (ratio of IMSI paging and TMSI paging… type 1, 2 and 3, and all…)

    Third assumption : the 60% load… but here, we use the same value.

    Anyway, 40 and 33 are not so far away. Are you sure you have taken into account your 7 PCH blocks ?
    I will copy here the details of “my” calculation, so you can check it out, on monday.

    #58306
    awahma
    Guest

    I have taken into account all 7 PCH blocks but difference is I have taken 1 MS per 1 CCCH block and 100% load.

    If I take 2 MS per CCCH block at 60 % load,

    PPS = (30*2)*0.6 = 36

    While for 3 MS per CCCH block it would be around 54.

    In our network we are using mixture of Paging request type 1 and 3.

    #58307
    PIx
    Guest

    Hello,

    Here is the computation :
    PCH capacity
    Non-Combined cell
    BS_AG_BLKS_RES = 4 (AGCH blocks) 9 – 4 = 5 PCH blocks (default configuration)

    5 PCH blocks/Multiframe * (3600s / 235 ms) = 76 594 PCH blocks/ hour

    2.5 paging per block due to average usage of the TMSI & IMSI paging methods:
    2.5 paging / Block x 76 594 Blocks =
    191 485 paging/hour at 100% load

    If 60 % engineering limit applied: 114 891 paging/hour
    or, 31.91 paging/s

    #58308
    awahma
    Guest

    Thanks alot!!!!!!

    Thats the end of it.

    Everything is clear now.

    Regards,
    Jawad

    #58309
    Learner
    Guest

    Hello Pix & Awahma,

    Have gone through all your posts in this topic, it’s wonderful that I cleared my confusions really.

    One thing I would like to know is,

    the limit Max – 33 pag/s is the capacity (60% load) that we can plan with our parameters (agblk, use of imsi or tmsi).

    But how should we check what currently is the network doing? Like I want to know the figure in the network so that I can see if it is exceeding this 33 pag/s limit.

    Thanks in advance,

    #58310
    pix
    Guest

    hi learner,

    ccch load is a great optimization topic.
    the “limit” of 33pg/s is more of a threshold, used by design team to define how many cells per LAC should be planned.

    In your network, the BSC is collecting PM counters, and some of them are about Paging load, such as:
    – number of paging requests sent on Air interface
    – number of paging blocks sent on Air interface

    With those 2 indicators you can know:
    – paging load per s. (you will only get the average value, but that’s good enough, thanks to efficient queueing mechanisms in the PCH.
    – number of paging requests per block -> in the cell with a high amount of paging, @BH, you’ll probably see that there are much more than 2 paging requests per block in average.

    Regards,
    pix

Viewing 15 posts - 1 through 15 (of 17 total)
  • The forum ‘Telecom Design’ is closed to new topics and replies.