Usually one can find the Jn(b) by bascially using the summation function and expand out with given n and b.

But what if you're given the sideband spectrum and have to work backward to find beta (b).

I've search the internet and here is what I've been using.

Jn+1(b) = (2*n/b)Jn - Jn-1

So one would iteratively working backward for all the Bessel sideband and solve for b until one get the least rms error from the actual Jn(b) expansion.

Is this a good way of doing it?

Does anyone have a better way I would love to know?

I am working to find the best algorithm to find the periodic FM jitter from looking at the spectrum analyzer.

Thanks in advance.