Why to use format3 and format4 of PUCCH in my gNB of 5G NR SA?

Hi Experts.
Why do I need to use format3 and format4 of pucch in my gNB of 5g nr sa?
Why not just using format2?

We have short and long pucch format for NR.
As of now short pucch format used.

So when is useful to use with long pucch format?
You said what is used now format2.
So what give me format4 (long pucch) wheras not using format2?

If format2 is giving me all requirement, so why do I need to use format3 or format4?
Here is my confusion…

Is there any use case that forcing me to use format3/format4?

I think it depends on your cell settings for pucch parameter.
What is current value?

What do you mean?
I mean could you give an example for using format4 and not using format2?
Even if depends on pucch parameter.
Just to understand how do I choose the suitable format…

I think:
PUCCH format3/4 can have max 14 symbol and 16 RB for format 3. (1 RB for format4) .
PUCCH format 2 Max 2 symbol/max 16 RBs.

And in short

PUCCH format 2 - Max REs = 2x16x12 = 384 REs only.
PUCCH format 3 - Max REs = 14x16x12 = 2688 REs

So we have more REs in PUCCH format 3.

More REs large coverage area.

It’s reasonable, thanks alot for your explanation! :wink:

Might be using long pucch :grinning:

1 Like

Any reference to how many bits of PUCCH actually usuable able and how much are overhead (like DMRS)?

What do you mean?
It depends on your cell settings.
So I think there is no pre-determined bits that are considered as overhead or useable bits.

I think the answer is:
How many resources are allocate to DMRS is relatively straight forward.
At least for one of the formats (was it 2?) this is 50%, the rest could be easily found.
However, this only answer the number of avilable REs for UCI, to get to the number of UCI bits you could carry is more complicated as it is a function of the allocation size, and the coding rate which in turn should be a subject of the channel conditions.
In an optimized implementation you’d need to constantly estimate the channel conditions of each UE and in the extreme, per PRB.

Make sense ?

Yes, thanks.