21Mbps per RBS or per Cell (E//)

[Muhammad Imran Rafique]

Hi Experts,
I have one question which really need answer quickly from E// experts here on this forum.

If we have all features activated, how much maximum theoretical throughput we can achieve in one CELL & in one RBS. To make it simple I have only one carrier (F1) &
codes:15
Cells:3
MIMO:NO

logically I can get 21mbps easily, this throughput is possible on any one cell at one time OR we can get 21MPbs from each cell simultaneously?
What is the reason?

Don't confuse, I am repeating we don't have any limitation of features from E//, every feature is available.

Hope you will come up with solid justification, its urgent as I have plan few Iub related capacity stuff.

BR//
Imran

[electron]

Hi ,


At first let me make something clear that is "U will never reach to 21M practically" as this value is the most ideal and it can be reached theoretically or in lab" the point is u did not consider the biggest problem which is radio interface and with its all problems(interference, re transmission, assigned modulation scheme,...).

Also this Thr can be reached in peak which means in a cell there should not be any other user except one to assign all resources (i mean Power) so we can say it is applied for cell.


Cheers
Hi Experts,
I have one question which really need answer quickly from E// experts here on this forum.

If we have all features activated, how much maximum theoretical throughput we can achieve in one CELL & in one RBS. To make it simple I have only one carrier (F1) &
codes:15
Cells:3
MIMO:NO

logically I can get 21mbps easily, this throughput is possible on any one cell at one time OR we can get 21MPbs from each cell simultaneously?
What is the reason?

Don't confuse, I am repeating we don't have any limitation of features from E//, every feature is available.

Hope you will come up with solid justification, its urgent as I have plan few Iub related capacity stuff.

BR//
Imran

[Muhammad Imran Rafique]

Yes no need to repeat I..e ideal radio conditions, single user etc

Question remains there.

21Mbps is possible for one cell only or 3 cells from same RBS can give 21mpbs(theretical) simultaneously.

BR//
Imran

[electron]

CELL LEVEL (Please read my previous post again )
Yes no need to repeat I..e ideal radio conditions, single user etc

Question remains there.

21Mbps is possible for one cell only or 3 cells from same RBS can give 21mpbs(theretical) simultaneously.

BR//
Imran

[Philippe]

To get the theoretical peak of 21 Mpbs in HSDPA+, you need to use the 15 codes of the OVSF tree (Spreading Factor 16 used for HSDPA) and 64QAM.
So if you have 15 codes per site, you will only be able to reach the theoretical peak on 1 cell.
If you want to reach the theoretical peak of 21Mbps on each cell, you would need to have 15 codes per cell, so the 45 codes per site.

Of course all these peaks are theoretical and obtained in good radio conditions with low probability to get peak at the same time on the 3 cells (I know you know but I repeat for people who would not read other posts)

[wolverine]

From a theoretical point of view you also need to consider if your baseband (DUW or or older HS-TX boards) is dimensioned to provide 45 codes and 63Mpbs of throughput.

[electron]

Yes it is good point which u mentioned; The point is in no any condition user will not reach to 21M in reality even through indoor cell in non peak hour. I personally tested in the best condition and i reached up to 18M .

If there is another observation please share .

Cheers.
From a theoretical point of view you also need to consider if your baseband (DUW or or older HS-TX boards) is dimensioned to provide 45 codes and 63Mpbs of throughput.

[Jerry]

Yes it is good point which u mentioned; The point is in no any condition user will not reach to 21M in reality even through indoor cell in non peak hour. I personally tested in the best condition and i reached up to 18M .

If there is another observation please share .

Cheers.
It is physical layer data rate. Application layer data rate is approximately 85% of that, due to the inclusion of IP headers (overhead information) etc.
you also need very good rf condition, cqi is about better 28