So i have a time series which decreases in frequency as time increases. I dont think this meets the criteria for a FFT that requires:: the system must be linear; and the data must be strict-ly periodic or stationary; otherwise, the resulting spectrum will make little physical sense.

So i read that a joint pdf should stay constant through time and that it must be stochastic for the system.

The average frequency can be barely picked with the fft, so my professor tells me to do that. But id like for a way to justify using the fft simply in my paper. Does anyone have any recommendations?

Best,

Abid