The more CSI-RS ports there are higher will be the gain of the AAU and then higher chances will be for MU-MIMO.
This MU-MIMO is very important for FWA users as their antenna locations on rooftop need to be properly placed to allow MU-MIMO usage otherwise is a big loss for MNOs if they can’t exploit MU-MIMO gain properly.
This will be the next challenge for NR RF engineers: placing the CPE of FWA users to allow/boost MU-MIMO.
I think this can be as part of planning tool simulation.
1 Like
Planning using Atoll or other planning tools tough.
User separation key factor.
Is it with 3D users locations and UE antenna facing capabilities?
I don’t think so.
MU MIMO used eigen based beamforming ie SRS reporting used.
May be UE pattern can be imported in tool and tilt azimuth for UE suggested similar to automatic cell planning.
Antenna patterns.
Can you place each user in Atoll with 3D resolution?
Or is it just simulation distribution?
Assumption.
Tried for mMIMO but results not accurate.
In the Ericsson test if there would have been a simulation test they would have use it instead of trying several places for a week.
First time always R&D will help but in live network to achieve no one its hard Truth. 
Good to see, for Ericsson.
Till last year, Ericsson used to counter Huawei support of MU MIMO of 16 layers with the argument that MU MIMO in commercial network may not give good high expectes output because to gain good spatial separation my not be efficient.