Hi Experts,
In case of MU-MIMO 5G, is it possible that the UE1 get 1 layer (using same PRB, say PRB0 to PRB20) and at the same time UE2 get 2 layers (using same PRB, say PRB0 to PRB20)
Yes, it is possible because users are in 2 different beams.
Thanks You.
In LTE as well?
Don’t know for LTE.
That means gNB has served 2 UEs using the 3 layers at the same time using same PRBs.
So any combination of number of layers and UE is possible untill I don’t cross max limit of 16 layers and RI = 2?
What do you think?
As I remember in the case of mMIMO, 16 users can share resources in DL per TTI, RBs will be assigned as per UE radio conditions.
In Huawei, if I see 3 MU MIMO layers, that means 3 UEs with RI = 1?
I understand that in R16 this 2 layers per UE in MU-MIMO does not exist anymore and 4 layers can be scheduled.
I have yet to find this official 3gpp spec for this…
No.
You have special counters for MU-MIMO.
Plenty of them.
How many PDSCH layers, and how many UE paired.
And so on.
MU-MIMO with one layer means at least 2 layers I need, right?
Yes, it is possible 1 layer for MU-MIMO for one user.
But it is not common for scheduler to work this way.
Ok, that means it is not actual MU MIMO.
Yes it is MU-MIMO. User 1 gets 1 layer, user 2 gets 2 layers, user 3 gets 3 layers, and so on.
Each of them using a different beam.
Each of them using same 100 PRBs.
It is MU-MIMO.
Just that they are spatially separated.
Each one in a different beam.
Different SSB beams?
And/or different csi-rs beams.
I am specifically asking about MU-MIMO with one layer.
Take N.ChMeas.MIMO.DL.MuPairing.1Layer.RB
Those counters sum up the PRBs when only 1 layer was used for a single user although multiple user are paired for MU-MIMO.
Pairing is an activity that is changed periodically: some UE gets in the pool some UE gets out of the pairing pool.
There could be times when only 1 UE is scheduled with data with only 1 layer although 4-5 UEs are paired for Mu-MIMO.
If there is no pairing, that is not MU-MIMO, right?
Yes, no pairing means no MU-MIMO.
anybody knows whether pairing is done when the selected PMI matrix for UEs are different ? What is the triggering point of gnb decides “uncorrelated users” in spatial domain to judge UEs can be paired or not…
Thanks