PUCCH SNR vs PUSCH SNR

[praghur]

Hello community.

I have set the logging level to debug on the physical layer on gNB configuration file. In the log file, I see different snr in the PUCCH and PUSCH values. See below an example:

2025-02-26T16:16:01.375341 [PHY ] [D] [ 704.9] PUCCH: rnti=0x4601 format=2 prb=[104, 105) prb2=na symb=[10, 12) csi1=0111 t=30.0us
rnti=0x4601
format=2
bwp=[0, 106)
prb=[104, 105)
prb2=na
symb=[10, 12)
n_id=1
n_id0=1
slot=704.9
cp=normal
ports=0
status=valid
csi1=0111
sinr_ch_est[sel]=26.5dB
epre=-14.1dB
rsrp=-14.2dB
sinr=+26.5dB
t_align=0.15us
cfo=-7.4Hz

2025-02-26T16:16:01.380207 [PHY ] [D] [ 704.18] PUSCH: rnti=0x4602 h_id=0 prb=[9, 15) symb=[0, 14) mod=64QAM rv=0 tbs=544 crc=OK iter=2.0 sinr=27.2dB t=246.6us uci_t=0.0us ret_t=0.0us
rnti=0x4602
h_id=0
bwp=[0, 106)
prb=[9, 15)
symb=[0, 14)
oack=0
ocsi1=0
part2=entries=[]
alpha=0.0
betas=[0.0, 0.0, 0.0]
mod=64QAM
tcr=0.92578125
rv=0
bg=1
new_data=true
n_id=1
dmrs_mask=00100001000100
n_scr_id=1
n_scid=false
n_cdm_g_wd=2
dmrs_type=1
tbs_lbrm=159749bytes
slot=704.18
cp=normal
nof_layers=1
ports=0
dc_position=636
crc=OK
iter=2.0
max_iter=2
min_iter=2
nof_cb=1
sinr_ch_est=30.2dB
sinr_eq[sel]=27.2dB
sinr_evm=24.9dB
evm=0.04
epre=-6.2dB
rsrp=-6.2dB
sinr=+27.2dB
t_align=0.04us
cfo=-37.7Hz

The thread gave me details about different SNR calculation in PUSCH. However, can you explain the difference in the SINR in PUSCH vs SINR in PUCCH? I am thinking that PUCCH SINR is calculated from the CSI report while PUSCH SINR is from the PUSCH channel itself. Is that correct?

Thanks in advance for any input.

[dvdgrgrtt]

Hi @praghur,
the SINR reported by the PUCCH is the one you get from the channel estimator - same as sinr_ch_est in the PUSCH. The reason they are different is that they are estimated quantities, with inherent estimation errors. The estimation is carried out independently for each PHY channel, only looking at the DM-RS in the corresponding grant, so it's normal they do not match exactly. Also, take into account that PUCCH grants are very small (1 PRB and two OFDM symbols, in your example), which means that the estimation is based on a very low number of samples and, thus, not as reliable as the one you can get from, e.g., a PUSCH transmission occupying most of the slot.