If csi rs beam there then its on csi rs beam.
Yes this is clear.
Is this a vendor approach or a 3GPP one?
Please state where in 3gpp its is mentioned this.
I’m telling what I read in NR books for 3gpp part.
As we discussed earlier about beam switching on basis of csi rs or ssb.
Yes, it’s in NR Bullet as well 1 SSB = 4 CSI RS beam.
If report quant = csi RSRP then its another indication that csi rs beam enabled.
It may be clear to you but you have not seen in any logs a TCI command to switch pdsch beams, only TCI to switch pdcch beams.
Search again in your logs and you will see I am right.
That’s an example.
Not as per spec.
Ok, thanks.
What is the spec?
How may CSI RS beam can be possible minimum and maximum per SSB?
Is there any TS?
I am not sure about that, I haven’t found any definitive answer from specs.
It’s all up to vendor implementation I think.
One thing it is not clear for me is when the UE needs the PMI and when UE needs best csi-rs beam reporting.
Why sometimes both are needed?
Your query is when UE reports PMI or best csi-rs beam or both right?
What is UE needs?
Can you explain bit more?
I mean why UE is requested to report both?
For gNodeB it would be enough one of them to identify the best beam location.
Why UE needs to report both: PMI reports and csi-rs beams reports?
Not sure but this is my understanding right now - best beam reporting is with rsrp only and PMI reports code-word transmission quality with RI based on SINR.
Indeed PMI would provide the weights needed for each layer for each CSI-RS port whereas csi-rs beam will give the rsrp of the best beam meaning the direction of strongest beam of the UE.
Without PMI or SRS the W precoding matrix cannot be computed.
But I am not clear as both are dealing with beamforming. PMI will influence how each layer is transmitted and received at each RX antenna of the UE whereas csi-rs beam report will give indirectly the weights necesarry to steer the pdsch beams to the right direction of the UE.
Anyway I would like to see a MAC CE DL TB with instructions to the UE to change the TCI for pdsch. I have not seen one yet, only for pdcch.
For PDSCH beam switch UE will get that info in DCI with TCI info.
But that will also depends on the K1/K2 values.
That is my understanding.
Issue is: TCI present in DCI is missing in RRC Connection reconfiguration message.
Then RRC should configure TCI related parameters for pdsch which it doesn’t.
Then NW should send MAC CE Activation Command for PDSCH which it doesn’t.
Only after this new TCI should come in the DCI.
One explanation on sharetechnote is: When tci-PresentInDCI = omit or PDSCH is scheduled by DCI 1_0 then UE uses TCI state for CORESET/PDCCH as the TCI state for PDSCH. it means TCI state for PDSCH = TCI state for CORESET/PDCCH. This is what should happen whentci-present in DCI is missing.
Yes thats correct.
If operator has deployed separate beam for CORESET and separate beam for PDSCH, then this will come into picture, and this is a very complex process and i don’t think this is implemented right now.