High SDCCH Fail on VSAT cells
pix - 19th June 2008 (06:02 GMT)
|ayan, the subscriber does one dial, and the MS will send 3 channel requests in a row (with some time between them in order to wait for an answer).|
Ayan - 19th June 2008 (00:08 GMT)
But it should be 2 otherwise sdcch assignment failure rate will not be 50% right? And this 50% failure does not mean MS need to make two tries? It is not related.
pix - 18th June 2008 (08:06 GMT)
i would say the number of channel requests retransmissions you authorize for the MS is equal to 0.
it should be equal to 2.
this is a BSS parameter.
Ayan - 18th June 2008 (03:26 GMT)
In that case, what will be the possible reasons if a MS sometimes needs to make a few dials before he get through to a receiving MS if A.bis is over satellite? As first call will always get destination MS is unreacheable or out of coverage area? Which timer cause this problem?
pix - 4th June 2008 (14:01 GMT)
|was gonna ask you the same question :) IMO it is depending on the SW implementation in the MS. some vendors might put it to 5s, fixed.|
let's forget about it, it's really not important at all !! it doesn't impact the BSS in any way.
Peyo - 4th June 2008 (13:52 GMT)
|Pix: I've made a mistake :)|
yes it's minimum for T3126= 2.15 secondes
I don't deeply know T3126 but from 4.08:
The minimum value of this timer is equal to the time taken by T+2S slots of the mobile station's RACH. S and T are defined in section 188.8.131.52. The maximum value of this timer is 5 seconds.
to be honest i don't understand this maximum value of 5s , it cannot be reached by tuning S and T. If you have any explanation to provide it would be great.
pix - 4th June 2008 (10:05 GMT)
ok, long time i didn't check out these parameters... T3126 cannot be tuned, it is automatically set to a value computed by the MS, according to the value of S and T, correct?
value of T3126=2S+T = 2.1s (you did an error in your computation ? or i misunderstood ?)
well, anyway, T3126 doesn't help here, because it doesn't control the repetition of Ch. Req. but just how long it should wait after the last Ch. Req.
thanks a lot for your in-depth analysis, it's a good "memory refresh" for me :)
Peyo - 4th June 2008 (08:41 GMT)
|Pix: yes, S=217 is the maximum value you can have (for tx_integer =32 and non_combined BCCH)|
the only thing that may make a difference here is that this setting also increase T3126 minimu value (which is the timer started by MS after last chan_req is sent . min T3126 = 2S+T =249 *4.62 = 1.15s which is only 6 ms longer than maximum time between retransmission and it can be up to 5s so there's is probably no gain.
I agree with you that there's nothing to do with this.
pix - 3rd June 2008 (19:55 GMT)
thanks for the calculations :) i put these settings on a VSAT cell, but it didn't change anything. It looks like 1.5s is still too short for the reply. But wait a second... the value for S cannot be as big as 217, can it ??
And anyway, you're right. The problem lies on the RACH repetition on the air interface. The channel request will then be transmitted on the abis interface. So your proposition is 100% valid !
But conclusion is still the same : for VSAT cells in Alcatel systems, a SAFR radio = 50% is N.O.R.M.A.L.
It's not my conclusion, tests have been done by TAC. it's worldwidely accepted.
peyo - 3rd June 2008 (15:53 GMT)
I'm talking about radio interface and note Abis. Forget that
Post a reply to this article