Having a hard time wrapping my head around autocorrelation matrix as it applies to a spectral estimation problem like MUSIC or ESPRIT. If the signal vector contains a summation of sinusoids in noise, how does building the autocorrelation matrix help decipher this information? Seems like a bit of magic that the toeplitz matrix of the autocorrelation sequence can help break out the signal subspace and the noise subspace.



Your Answer

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

Browse other questions tagged or ask your own question.