Beam Management, Beam Switching and Beam Refinement in 5G NR

5G NR
, , ,
1

Hi Experts,
For beam switching mentioned at L1/L2 procedure, it can be by SSB RSRP or CRI RSRP by parameter report quantity; it’s clear now. (As our discussion here: IE "tci-presentinDCI" in RRC Connection Reconfiguration Message - #3 by RFSpecialist and here: Beam Switching in NR)

One query only without beam refinement option for FR1 ie CSI reference signal beams not there then can we say intra cell beam switching always between ssb beams.

Admin note: this post was updated with image below.

Beam refinement
Beam refinement641×542 97.8 KB

1 Like
2

reportquantity is never missing.
And report quantity decides if beam switching is done based on ssb reporting or based on csi-rs beams reporting.

3

Even I never seen any parameter like report quantity where we can configured cri rsrp or ssb index rsrp in 5g parameter
Only parameter where we can enable or disable beam refinement there ie nrtbeamrefinement true or false
My understanding this parameter only judge refined beam ie CSI-RS beam there or not.

4

@ankgitm, maybe reportquantity is not available for user to change it.
Means hardcoded and cannot be changed.

5

We have CSI-RS based channel measurement ie ssb rsrp for L1 measurement indicates same if its ask to measure.
Then RE occupy by CSI-RS will measure along with secondary synchronization signal.

6

Yes true but interesting how gNB decide value for different CSI report settings.
It’s a huge structure.

7

I also saw like 5 reports for different purposes, some based on ssb reorting, some based on CSI-RS reporting.
They are all 5 reported in uplink for various purposes.

8

Without beam refinement CSI-RS rsrp based beam switching not possible.
I’m getting some clarity but need more brain storming…

9

The approach is different in FR1 and FR2.
mm wave have a different beam management.
For example in mmwave there are not less than 64 CSI-RS beams.
It is a nightmare to read the rrc connection reconfiguration message for it.

And again: everything is vendor dependent and could be much confusion reading several vendors in the same time.

10

As per my understanding beam refinement only valid for FR1.
For FR2 as of now 32 ssb beam only.
In FR1 each ssb consist of 4 CSI-RS beam if beam refinement enabled.
No such concept for FR2.
So from given report quantity different options one thing can be concluded cri rsrp measurement only there for L1 rsrp if beam refinement enabled.
So without beam refinement no switching between CSI-RS beams no cri rsrp.

11

There are different reports with different purposes.
Up to 5 of them.
Some for ssb, some for csi-rs, how could you know which one is for what?

12

Yes that only cri if you look in detail.
Option 6 only mentioned cri rsrp or ssb rsrp.

13

These are the options for report quantity, as per 38.331:

image

14

Point is that cri exists without beam refinement as well but for cri rsrp we need beam refinement to true
As CSI-RS reference signal positions are there in slot above ssb which is linked with cri.

15

Here you have is an example of MAC LL1 CSF Indicator with 3 records of reporting.
2 for csi-rs and one for ssb.

[ 2020  Mar  18    18:11:52.881299 ]  [0xB891] NR5G MAC LL1 CSF Indication  

Chipset timestamp: 2020-03-18 18:12:11.918938, CFN : 94
Dual Mode Index : 1
SIM Index : 255
Version : 65537
Major.Minor Version : 1.1
  log_fields_change
  log_fields_change_bmask : 0x0000
sub_id : 0
num_records : 3
Records
  Record[0]
    report_ul_time
      slot : 65
      numerology : 120kHz
      frame : 615
    carrier_id : 0
    report_id : 2
    report_type : APERIODIC
    num_bits_part1 : 35
    num_bits_part2 : 0
    is_rsrp : RSRP
    num_bits_part2 : 0
    debug_only : Result_from_WP
    report_quanlity : RSRP , SSB_INDEX
    resource1 : 5
    resource2 : 2
    resource3 : 4
    resource4 : 1
    resource_type : SSB
    num_resource_bits : 4
    pad : 0
    rsrp1 : 59
    rsrp2 : 59
    rsrp3 : 58
    rsrp4 : 51
    valid_rsrp : 4
  Record[1]
    report_ul_time
      slot : 65
      numerology : 120kHz
      frame : 615
    carrier_id : 0
    report_id : 0
    report_type : APERIODIC
    num_bits_part1 : 24
    num_bits_part2 : 0
    is_rsrp : RSRP
    num_bits_part2 : 6
    debug_only : Result_from_WP
    report_quanlity : CRI , RSRP
    resource1 : 0
    resource2 : 1
    resource3 : 2
    resource4 : 0
    resource_type : CSI_RS
    num_resource_bits : 3
    pad : 0
    rsrp1 : 67
    rsrp2 : 66
    rsrp3 : 54
    rsrp4 : 0
    valid_rsrp : 3
  Record[2]
    report_ul_time
      slot : 65
      numerology : 120kHz
      frame : 615
    carrier_id : 0
    report_id : 1
    report_type : APERIODIC
    num_bits_part1 : 24
    num_bits_part2 : 0
    is_rsrp : RSRP
    num_bits_part2 : 6
    debug_only : Result_from_WP
    report_quanlity : CRI , RSRP
    resource1 : 0
    resource2 : 1
    resource3 : 3
    resource4 : 0
    resource_type : CSI_RS
    num_resource_bits : 3
    pad : 0
    rsrp1 : 72
    rsrp2 : 59
    rsrp3 : 53
    rsrp4 : 0
    valid_rsrp : 3
16

I need to read carefully difference between cri and cri rsrp.

17

What I noticed so far is that “tci-presentinDCI=true” is missing.
Did anyone found it true?

18

Are you checking at L1/L2 layer info?

19

Nope, this should be present in RRC connection reconfiguration message to setup the scg, so layer 3.

20

Ok, let me check.
Is it related to intra cell beam mobility?