I have a complex-valued signal $q(t)$ which is sampled at $F_s = 24$ Hz. I am using the matlab periodogram function to plot its power spectral density. Here are the matlab commands I use:

n = length(q);
fs = 24;
[Pxx,F] = periodogram(q, rectwin(n), n, fs)

The result is shown below. The Nyquist frequency should be 12 Hz, however the plot shows frequencies up to 23.875 Hz. The help function says that for complex signals, the function computes a two-sided PSD but doesn't provide much detail on that. Does anyone know how this function is able to compute power at frequencies beyond the Nyquist frequency?

enter image description here

  • $\begingroup$ I've closed this as duplicate, because I believe the question you're really asking is answered in the duplicate I've linked to. TL;DR : When a complex signal is sampled, you're getting two samples per sampling point, so the effective Nyquist rate is double the usual one. $\endgroup$
    – Peter K.
    Jan 6 at 17:19