HTTP Throughput

[kadir]

Hi fellas,

Long time past i did not write to forum untill i really need help from you guys :)

I need some idea to improve HTTP throughput, especially for ramping up for service start.

thanks to all of you
br

[perimedes]

Hi,
Could you please add some information about the issue? I mean, is an optimization, is something you need because a new trouble has arise? Vendor and version of the network will help.
br/

[kadir]

Hi,
Could you please add some information about the issue? I mean, is an optimization, is something you need because a new trouble has arise? Vendor and version of the network will help.
br/

Hi,

after the swap E// to H// http throughput is lower this is the case, and ran version is ran16 for huawei. But i need general idea to improve http throughput, i mean what can be the parametric changes may improve the http, or how i can investigate this kind of issue.

if you can suggest me as general steps, would be good.

br

[lamhd_bk]

HTTP throughput is application throughput, so that you must re-calculate the MAC throughput to make sure that the problem from your networks, then find out the solution.
Anyway, some problem from the Post processing tool that you use to calculate HTTP throughput, may you must consider the 0 kbps for every sesson

Rgs

[kadir]

HTTP throughput is application throughput, so that you must re-calculate the MAC throughput to make sure that the problem from your networks, then find out the solution.
Anyway, some problem from the Post processing tool that you use to calculate HTTP throughput, may you must consider the 0 kbps for every sesson

Rgs

Thanks,

i think this can be good start to go forward about this case. I will try to identify if this is the case than i will update you. But incase if this is not the reason i mean tool or calculation, do you have any idea to ensure some improvement on this. Like any parameter that we can play with to get more throughput as we usually do for some other KPIs.

regards
kadir

[giegala]

Hi,

Does it stay in Cell_Fach longer in HW than in E///? did you check if there is channel switching at the start of the session? what is the number of HS-DSCH codes assigned for each vendor?

[kadir]

Hi,

Does it stay in Cell_Fach longer in HW than in E///? did you check if there is channel switching at the start of the session? what is the number of HS-DSCH codes assigned for each vendor?


First question, IF you mean this parameter BeF2HTvmThd, In E// it is 512 and H// 1024
Code number is the same because we mapped E// settings to H//.
second quesiton i will check and reply you back.

br

[giegala]

It is not about parameter mapping. it is about measurement. You have to find with TEMS or similar where is the delay. From your settings it looks that E// will get to HSDPA quicker than HW, and that might give you substantial delay. But DT test is needed.

[kadir]

It is not about parameter mapping. it is about measurement. You have to find with TEMS or similar where is the delay. From your settings it looks that E// will get to HSDPA quicker than HW, and that might give you substantial delay. But DT test is needed.


Thanks friend,

i will get this advice it is very logical. When solve the issue i will update you guys as well.

br