I am reading a research paper on NOMA BER performance. For the simulation the author chooses oddly high awgn power -40dbm ( considering noise spectral density -174+10logB this would result in unreasonably high bandwidth) for simulation. He mentions that this complies with 3GPP standardized scenarios. Also he chooses transmit power range from -30 to 24dbm as x axis in BER plot.. how would this translate in SNR? Is it just transmit power in dbm + noise power in dbm ? Because that would result in high SNR in range of 60dB. Please clear up my confusion,

Thanks in advance Here’s the papers link in PDF: https://arxiv.org/abs/2301.03872 enter image description here


1 Answer 1


That's -40 dBm transmit power, not noise power. The AWGN power is stated explicitly to be -30 dBm for this figure. That's relatively much, but remember that what matters here are SNR or SINR, and so with transmit powers between -35.2 dBm and +29.8 dBm (attention, the transmit power is emitted thrice!), this corresponds to an SNR range of -5.2 to 59.8 dB, which is reasonable; note how the curve shows a bit error rate of 0.5 for all methods up to roughly -15 dBm transmit power, corresponding to an SNR of 15 dB; that is quite close to the lower bound for which any kind of coding could still help to recover $U_1$:

The three users are at 50 m, 75 m and 10 m distance, and the loss exponent is 2, they transmit the same power. So, the received power of $U_2$ at the central node is $\frac 49$ as much as that of $U_1$, and of $U_3$ it's only $\frac 14$, so for $U_1$ the signal-to-interference ratio is $\text{SIR}=\frac 1{\frac 49 +\frac 14} = \frac 1{\frac{25}{36}} = \frac{36}{25}=1.44 \approx 1.6\,\text{dB}$, which explains the high BER.

  • $\begingroup$ thank you for the comment. however i still have some mixup: 1. note that this is for uplink, i think you was explaining thinking this is downlink 2. I am not sure what you mean by ((attention, the transmit power is emitted thrice!), I get that the three users are sending each with P, but in your calculation for SNR you only considered once, why does that matter ? 3.shouldn't the SNR be calculated with received signal power rather than transmit? in this case shouldn't we subtract the pathloss from it? 4. as you said it's higher than normal noise, even though he says it's 3gpp standard sc $\endgroup$ May 14 at 14:04
  • $\begingroup$ re: 1. no. 2. no, the SNR calculation especially takes 3×P into account. 4. 3GPP standards are several thousand pages of different scenarios. If someone cites "3GPP" without saying what specifically, you need to ask them, not me, what they meant. $\endgroup$ May 14 at 15:26
  • $\begingroup$ So if I want the equivalent SNR for let’s say P=10dbm is SNR=3*10dbm-(-30dbm)= 60dB? Also you seem to have forgotten to answer for my 3rd question 3.shouldn't the SNR be calculated with received signal power rather than transmit? in this case shouldn't we subtract the pathloss from it? Thanks again for your time $\endgroup$ May 14 at 18:02
  • $\begingroup$ Apologies for the oversight, so 3*0.01W to dbm ~14.5dbm and SNR=14.5dbm+30dbm =44.5db? Also please answer my second question $\endgroup$ May 14 at 18:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.