Does a Fourier Transform of a white noise exist? If so, what is its general form?

It is possible to compute the fourier transform of a Power Spectral Density?

My problem in detail. I have to compute the response of a dynamic system given the PSD of an input ($ \Phi_{vv} $).

I thought to solve the problem in three step:

  1. Build a shape filter (system which, given a white noise input ($ n $), gives a colored noise as output)

  2. Express both the shape filter and the system dynamic with a fourier transform

  3. If i can manage to express also the input (the white noise) in the fourier domain I should have solved the problem

1 Answer 1


The Fourier Transform for the autocorrelation of stationary white noise exists: in fact "white" here refers to the characteristics of that Fourier Transform, in that the power spectral density will be constant over the entire spectrum, and also implies that each sample in the time domain is independent of all other samples.

The distribution of the magnitude of the samples in time is also usually defined, such as being "Gaussian" or "Uniform" depending on what distribution is actually used. For white noise samples the phase is also distributed uniformly $\in [0, 2\pi)$. Due to the central limit theorem, the Fourier Transform of any distribution in time that is white will tend toward Gaussian in frequency. The resulting Fourier Transform will be Gaussian distributed in magnitude, uniformly distributed in phase, and each sample in the frequency domain is independent of all other samples.

The Power Spectral Density can be computed from the Fourier Transform for time limited signals as the complex conjugate product, and not the other way around. The reason is once computed all phase information is lost so there is a many to one relationship in the computation.


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.