CQI override, is it possible?

[wolverine]

Hello,

Has anyone come across anyway to override the CQI reported by a UE with a preset value? So lets say make a UE to constantly report CQI 25 regardless of radio conditions etc. I was thinking along the lines of a test mobile, or some proprietary AT commands etc.

It is required for the purposes of validating the CQI compensation algorithm in the nodeB.

[tpkj]

Hm,... what do you expect on the end?
As far as I know NSN is not trusting UEs and their reporting CQIs.
On one side UE is reporting good channel conditions and on another one, if channel conditions are not OK, there will be request for a lot of retransmition from UE.

[tpkj]

additional information:
CQI is compenzated, i think +/-4dB ...it is pending if you obtained ACK or NACK.
It is written in - RU20 docs - Chapter 7.1 - Name: CQI report based link adaptation.;)

[wolverine]

Every UTRAN vendor has a CQI compensation algorithm. But how do you know how well it works?

What I am asking is if we can make the UE report a particular CQI value of our choosing (obviously you cannot do this with an "ordinary" UE).

If we can, then we can use this UE to benchmark CQI compensation algorithms.

[tpkj]

Would you like to test it in lab? Or would you like to test it outside?
As far as I know NSN has internal tool called TTI Tracer. It is for RND. You will instal this to BTS and BTS and will collecting data.

If you will do measurement without this outside, you will see the CQI compensated from Power allocation and TB size for HSDPA, and if there will be change on their value you can only deduct that it is due to compensation of CQI.

Honestly we never did measurement for this in LAB, because there was no eqipment for that.

But anyway, it is interesting topic for comparing of two vendors, in case in the same network.

[tok222222]

Would you like to test it in lab? Or would you like to test it outside?
As far as I know NSN has internal tool called TTI Tracer. It is for RND. You will instal this to BTS and BTS and will collecting data.

If you will do measurement without this outside, you will see the CQI compensated from Power allocation and TB size for HSDPA, and if there will be change on their value you can only deduct that it is due to compensation of CQI.

Honestly we never did measurement for this in LAB, because there was no eqipment for that.

But anyway, it is interesting topic for comparing of two vendors, in case in the same network.

I think the only way is put a trace on the Node B and a test UE connected to a Qualcomm chipset UE connected to Adgilent. Then post process the file with Actix and corrolate the Node B file with the UE file to see the UE CQI vs the Node B CQI as there is measured CQI and Reported CQI :)

[agenov]

I think the only way is put a trace on the Node B and a test UE connected to a Qualcomm chipset UE connected to Adgilent. Then post process the file with Actix and corrolate the Node B file with the UE file to see the UE CQI vs the Node B CQI as there is measured CQI and Reported CQI :)


Good point but i don't believe there is such available trace in the NodeB tracking every 2ms CQI values reported at least i have never seen it. Have you seen such things somewhere and who is the vendor ?

[Tomas Delgado]

Hi, is really intersting what you guys mention. I manipulated the CQI reporting with the measurement power offset. I did this because im told to optimize CQI but with the intention of improving throughput, but let me tell you that it changes nothing not even a little bit of throughput. So does CQI do something or not.

[plannerguy]

hi

small query is CQI entirley depending ON UE ?

So will it very from UE to UE?


regards
Plannerguy

[twenty8]

Hi,

CQI is an UE implementation, and not specified in 3gpp hence it is totally depended on UE vendor. It is possible that 2 different UEs report different CQI at the same location since the algorithm may be different. I have no information on exactly how this CQI value is calculated (need to get this from UE vendors) but it would be based on EcIo , SIR (for orthogonality) and MPO. Since MPO is not always reported to UE from RNC, the RAN makes a compensated CQI based on existing knowledge about MPO. Also LA and PS in RNC can make further compensation based on ACK and NACK being reported by UE.

I'm not sure if its possible to fix the CQI at a certain value, but am guessing ppl like Qualcomm may have tools to carry this out in their R&D labs. As far as I know, there is no tools on the market to report a fixed CQI (ie from TEMS, or NEMO)..

[fon909]

Not sure about NSN system, but in E/// there is a parameter hsMeasurementPowerOffset : The Measurement power offset, sent to the UE and RBS via RRC and NBAP. It is used to offset the CQI in order to utilize the whole CQI range.

[Tomas Delgado]

It exists also for huawei its called HSPDSCHMPOCONSTENUM, you can manipulate it and try but its quite intersting the results you can get im telling cause i already use it.
Now CQI depends on UE category and TB size that the ue can handle. But after test of improving CQI reports with the HSPDSCHMPOCONSTENUM it didnt make any difference in throughput using scheduling method proportional fair and Max CQI. Any ideas why this didnt worked? its supossed that CQI indicates about the radio environment, and some how its mapped to Ec/No.