How can we determine the suitable coverage scenario for 3D beamforming 5G?

Hello Dears.

Please how can we determine the suitable coverage scenario for 3D beamforming 5G?

Admin note: this post was updated with image below.

There is a Huawei table for this.

Yes, it is pretty clear based on that.

For example scenario 6 is for most coverage but no adjustment possible in azimuth.

Those adjustments are important later, as you want to tune your coverage towards where are most 5G users.

Thanks!

Please, for scenario 6, the Horizontal 3 dB beamwidth is 110° that means, the beam will be fixed in this direction.

That is the beam will be static not dynamic tracking the user (for traffic beams).

No, it means that beam can cover any location within that 110 degrees. Beams will be of course dynamic.

Huawei have dynamic beams, not static like Nokia

N1O1 is number of possible beams locations in H plane, while N2O2 is number of possible beams locations in V plane. If any N=1 then the corresponding O is 1 otherwise O has the value 4.

Thank you so much!

Please do you have any documents explains that in details? :pray:

It is well explained here:

5G_CSI_RS_Codebook

But you can also read in 5G in Bullets about it.

This is where you can find what N1 and N2 has your antenna if you do not know:

In which message can we find this?

Iin the RRC setup fo 5G SA, or in the RRC reconfig to setup the 5G leg in NSA.

Here is another example:

With csirs = 2 * N1 * N2 = 32 csi-rs ports maximum possible.

As per as I know, there are 3 types of Beams: SSB, CSI and traffic .

I think you are talking about CSI beams used for channel estimation?

These are the traffic (PDSCH) beams. Total no of beams depends on your CSI RS port congratulations and is equal to O1N1O2*N2

Tips to remember:

  1. No of CSI ports = 2 * N1 * N2
  2. No of Horizontal beams = O1 * N1, where O1 = 4
  3. No of Vertical beams = O2 * N2, where O2 = 1 if N2 = 1, otherwise O2 = 4
  4. i11 = 0 to O1 * N1- 1
  5. i12= 0 to O2 * N2 - 1
  6. i13 is the offset to the beams ( when RI >= 2)
  7. i2 is the phase shift and/or amplitude based on the codebook mode 1 or 2

Just to be clear for all:
N1 * O1 * N2 * O2 represents total possible positions of a beam orientation.

That means, for example N1 = 8 and N2 = 2: 8 * 4 * 2 * 4 = 256 distinct position of a beam on the ground.

N1 N2 possible configuration for single panel antenna are limited to this: :point_down:

In picture above each antenna element in logical configuration is in fact a group of real antenna elements having same phase shift.

Does this also mean 256 antennas or antenna elements?

Or total antennas are = 2 * N1 * N2 = 32, and we generate 256 different beams using these 32?

No, it does not mean 256 antenna elements.

Yes, we generate 256 different positions of a beam using those 32 csi-rs ports.

This is an easy diagram to understand how PDSCH layers data are sent to antenna elements: :point_down:

Do we also have information on physical antennas while generating csirs beams?

Like in this case, 32 CSI-RS ports are used.

By using N1 and N2, we get to no the number of physical and vertical alignment of antennas?

Each AAU can have different real antenna elements.

But csi-rs can be maximum 32 no matter if 256 or 384 or 1024 real antenna elements.

UE does not need to know how many real antenna elemenst has the AAU.

UE only needs to know CSI-RS ports and their configuration to report the PMI.

Everything is on sharetechnote… just it is complicated to put bits and pieces together.

Ok, Thank You!

Thank you all so much, very informative!

Just need your attention.

Can you point the beam position for a instances where UE is reporting RI = 3,
and first beam position as i11 = 7, i12 = 6, and second beam position as i11 = 7, i12 = 10.

Can you please point the beam position in possible 64 beam position?
Given-( N1, N2 = 2,2 )
no of csi rs port= 8,
i13=1 and i2=0.