- This topic is empty.
5th June 2012 at 09:30 #68297slim omraneGuest
Please to clarify that point :
In case on SDCCH congestion , Is there a repetition of channel request?Only in case of location update and call reestablishment allowed?
The KPI SDCCH_congestion rate is thus wrong because it contains redunduncy?
Please the link for 3GPP clarification7th June 2012 at 18:34 #68298pixGuest
if the sdcch is congested, the BSS will behave accordingly to your setting “Enable Immediate Assignment Reject”.
If “ImmAssRej” = DISABLE then the MS will repeat the Channel Request almost immediately
If “ImmAssRej” = ENABLE then the MS will repeat the Channel Request only “WI_xx” ms later, which provides more chance for a SDCCH channel to be available.
The MS can send up to MAX_RETRANS+1 channel requests.
Yes, the SDCCH Congestion is artificially increased by retransmission. However, if you activate the “ImmAssRej”, you can apply the formula:
SDCCH_Cong – Imm_Ass_Rej / Channel_Request
As a result, you are counting only the last congestion failure. If the SDCCH is setup at the 3rd attempt :
SDCCH Cong = 3
Imm Ass Rej = 3
–> sdcch cong rate = 0%
I hope it makes sense… please test the formula carefully because I don’t remember having tried it.
(i didn’t care much about artificial SDCCH congestion.. it should always be zero IMO… really and artificially !)
pix8th June 2012 at 21:11 #68299SDCCH congestionGuest
Further clarifications please:
The SDCCH requests are repeated till max retrans+1 whatever is the type of the service needed (LU,SMS,call establishment)?
In case of LU,it will stop retransmissions by max retrans+1 or it will do another cycle?
Please the link of 3GPP doc for repetition of SDCCH requets in case of SDCCH cong9th June 2012 at 11:04 #68300pixGuest
yes, retransmission works for any kind of request.
Location update will indeed start a new cycle, because the MS has to do the LU anyway.
44.018 -> the channel request procedure
23.122 -> the location update process (i’m not sure whether they say how long to launch a new cycle in this doc).