I am working on radar received signal which is $15000\times 2008$ matrix. Here each row corresponds to one chirp or one signal, this way there are 15000 signals. The difference among them is just the time difference and doppler shift.

If I plot the PSD of one signal(sweep) C(1,:) with MATLAB's psd I get a distinct peak single sweep psd,

If I take the PSD of ten signals, i.e ten signals combined in one row [C(1,:) C(2,:)...C(10,:)] then take the psd, the plot is similar the only difference is that the gain increasing, I assume it is the processing gain, also as I increase the number of signals and take the psd the gain increases even more many signals combined

I want to know what is the reason of the gain in the plots as the lenth of the signal is increasing, the fft length is not altered. I Would like to know any mathematics proving this.

x=[C(1,:) C(2,:) C(3,:) C(4,:) C(5,:) C(6,:) C(7,:) C(8,:) C(9,:) C(10,:)];
Pxx = periodogram(x);
hpsd = dspdata.psd(Pxx,'Fs',Fs); % Create a psd data object.

In the code for the periodogram this occurs:

if (n > nfft)
    Pxx = 0;
    rr = rem (length (x), nfft);
    if (rr)
        x = [x(:); zeros(nfft-rr, 1)];
    x = sum (reshape (x, nfft, []), 2);

Where nfft is the length of the fft and n is the length of your vector x.

Note how if n> nfft (as in your case) the vector is reshaped and summed together.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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