CS or PS Busy hour or combined one?

[auto_art]

Hi
i have a concern regarding the PM for UMTS network. should we judge performance of 3G individually for CS and PS busy hours. or should we use combined(both cs+PS) Busy hour for the same...

in my view we should use the combined one as it is a dynamic resource shared system. please share your logical view on the same.!

[boring]

Hi,
to my view, there is not a correct or not answer to your question

everything is a matter of priorities for the operator. for example, in the first 3G days, operators were looking to primarily protect CS traffic, so BH was typically looking for CS BH, and naturally whatever remaining capacity was used by PS users during BH

on the other hand, these days we have lots of PS users so perhaps it makes sense to consider both CS+PS when deciding for the network BH. a couple of points that needs consideration though:

1. 3G RRM (admission and congestion control) are designed in such a way that they always give priority to CS traffic. this means that if we dimension for both CS+PS, then the end result would always be CS traffic experiencing zero blocking

2. Largely, PS users are considered as non-real-time (NRT) traffic class users by current 3G RRM which typically use HSDPA for transport. Because this is not a R99 bearer (like the CS voice) we need to be careful when deciding the BH

hope this helps!

[telwind]

Good thing in mixed CS and PS model may give the frequency and resource efficiency. Different services will interact with each other, affecting the performance of each other. If you want to use this model, you need to use some 'translation factor' to convert erlang traffic each kind of services.

I think it is not much practical to use this model

[auto_art]

can u share how u r judging ur networks performance.?

[cooltyger]

Suppose we have dedicated allocations for PS traffic, then why cant we try to analyze the reports separately???
I feel comparing together in some thing like combined PS+CS is not very convincing!!
It is better to have BH reports analyzed separately as the BH for PS and CS maybe different time of day. In combine it is possible the CS traffic influences our judgement.

[auto_art]

Suppose we have dedicated allocations for PS traffic, then why cant we try to analyze the reports separately???
I feel comparing together in some thing like combined PS+CS is not very convincing!!
It is better to have BH reports analyzed separately as the BH for PS and CS maybe different time of day. In combine it is possible the CS traffic influences our judgement.

didnt u think taking CS busy hour alone will not give optimium results in Busy hours. e.g
if we take CS busy hour; during that hour many HS & R99 connections may be active which can influence the CS performance more intensity rather than CS connections. cuz PS sessions use more power during their operations.