Adaptive PMI weight

Hi All,
Question: Huawei TDD 5G has an adaptive PMI weights based either on SRS(UL)-based weights for beamforming or on CSI_RS (DL)-based weights for beamforming.
But if you check values the threshold to use this value is -20 dB for SRS SINR (recommended by Huawei).
Name of parameter is SrsPreSinrJudgeThld.
That means that as long as UL SRS SINR is higher than -20 dB the gNodeB will use SRS(UL)-based weights for beamforming.
Just curious how Huawei is dealing with SRS of -20 dB SINR.
I think there’s a lof of interference at -20 dB and signal of SRS is not much useful.
Has anyone tunned this value to higher values like -10 or -5 dB?

Can you please check again, -20 or -2 db?

It is -20.

Ok, parameter = - 20.
What is SINR value?

This is a very weird value of SINR.
What usage of SrsPreSinrJudgeThld = - 20 dB can be used for?
Nothing canm be decoded at -20 dB.

Report Value 6 = -20 dB (SNIR).
0 means = - 23 dBm…1 = - 22.5 dBm…

But we’re are talking about uplink sinr of SRS.
Then again, my question is, what good is a value of srs sinr = -20 dB?
Why Huawei recommends -20 dB for it?
UL srs sinr is the first to degrade in poor link budget. How come beamforming weights are calculated Based on UL SINR fo SRS when this is -20 dB?
This is valid only for TDD.
For FDD beam weighting is not based on UL SRS, it is based only on CSI-RS PMI.

What I am yet to become clear is the srs based weighting of beams.
How on earth Huawei works with srs sinr of -20 dB to calculate beamforming weights with such a low sinr of srs?

So the threshold for srs-based weights for beamforming is -20 dB.
Wehenever srs sinr is higher than -20 dB at gNodeB then the beamforming weights in TDD are based on srs UL measurements.
Only when srs sirn is lower than -20 dB then beamforming weights are based on csi-rs measurements in DL
And so based on PMI reporting when srs-sinr < -20 dB.
Also about srs ports for 4x4 mimo in dl we need 4 tx antenna at the UE ( so the UE transmitts 4 SRS and gnodeB measures those signals at each of the csi-rs ports and so can calculate the PMI based on UL srs).

It’s new features changing dynamically.

It’s static parameter and can be configured between -20 dB and +20 dB.
Default value is -20 dB that means srs is used up to srs sinr=-20 dB. After that pmi is used
Name of param is SrsPreSinrJudgeThld
So it goes like this: gnodeB measures peridoicall srs sinr in uplink. As long as it is greater than -20 db then it will use srs measurements for beamforming weights.
As soon as the srs is lower than this param (SrsPreSinrJudgeThld) plus 3dB threshold it will use PMI reported by UE.
So for example when srs sinr is -24 dB it will use PMI reported by UE.

What is that feature Adaptive PMI weight?
could you please explain about it and why it’s used and its purpose?

Thanks alot

Hi @RFSpecialist , please could you explain why at low SINR we are using pmi weight and at HIGH sinr we use use SRS weight? any good article I can read research measurement for that decision of that switch?

why it’s recommended at -20db which means if UL sinr is smaller than -20db then SRS weight used, else pmi weight used …

Any good explanation why it’s used like this or any motivation for?

thanks alot