- Jul 13, 2005
- 481
- 0
- 71
I'm attempting to implement my own periodogram function in matlab (for a project), to estimate the PSD of a discrete signal of length 2048 given to me. I noticed something interesting with Matlab's Periodogram function. When I call [Pxx w] = periodogram(v1), there is something funny with Pxx. According to the formula for calculating a periodogram, Pxx(1) (the sample at freq w =0) should be given by (abs(sum(v1)).^2)/2048 (I wrote this in matlab notation, the formula can be found in any spectral analysis book). But that isn't the case. The function I wrote matches the value.
Does this mean matlab's periodogram function is wrong (which I don't think is the case). So what could be reasons for this mismatch? Spectral Analysis gurus please help....
Does this mean matlab's periodogram function is wrong (which I don't think is the case). So what could be reasons for this mismatch? Spectral Analysis gurus please help....