Normally, when talking about adding noise to simulation data, the choice are AWGN, pink noise and so on. However, in practice, the noise in the experiments changes the locations of frequency peaks as well. How can I simulate this kind of noise?

  • $\begingroup$ Sounds like you might be overdriving amplifier inputs and getting clipping which introduces intermodulation products (peaks at other frequencies). $\endgroup$
    – Andy Walls
    Commented Oct 31, 2019 at 13:49

1 Answer 1


I'm not sure I understand your question.

But usually the noise model is additive.
Namely, due to the linearity of the DFT, at each bin you will have the summation of the signal contribution and the noise contribution at this specific frequency bin.

Now, let's say we have 2 bins.
At the first the signal is 7 at the second is 6.
The noise is 3 and 5 respectively, now the summation will shift the DFT peak though the signal isn't changed.

In some cases the noise might be colored (With time correlation) which means most / some of its energy might be concentrated in a small band of frequencies.

  • $\begingroup$ This is a good point. I know this is a silly question, but in the experiments, the measured frequencies change between several tests. In some tests, some frequencies disappear. Adding white noise to simulation data can not simulate the experimental situation. The SNR is not very low for the experiments, thus for AWGN it is unlikely to have very high peaks. $\endgroup$
    – Kattern
    Commented Jun 21, 2015 at 10:06
  • $\begingroup$ I agree with Drazick; noise is usually additive and has a simple spectrum, so it can't do things like shift a frequency peak present in the data signal. Something else is going on in your experiment, such as fading or Doppler, or even some non-linear effect (for example, non-linear distortion can make your fundamental harmonic to lose power). $\endgroup$
    – MBaz
    Commented Jun 21, 2015 at 15:48
  • $\begingroup$ Frequency could change due to Doppler shift as you've been told, or by frequency mismatch of reference generators if you use two or more boards. If some non-linear things happened it will be additional frequency in the spectrum but the original ones will stay the same. In the case of fading without Doppler spread frequency content won't be affected. You can describe your experiment for more understanding of your interesting phenomenon. $\endgroup$
    – Serj
    Commented Jun 22, 2015 at 7:19
  • $\begingroup$ @Kattern, Could you please mark my answer? $\endgroup$
    – Royi
    Commented Feb 9, 2022 at 6:02

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.