I'm currently looking into the Rayleigh fading implementation, but, unfortunately, don't get some implementation details.
For example, the generation of the sinusoids [1]
float s_i = scale_sin*cos(2*M_PI*d_fDTs*d_m*cos(alpha_n)+d_psi[n]);
AFAIS, alpha_n is uniformly distributed between [-pi; pi], so FDTs * cos(alpha_n) is a random walk through the doppler spectrum.
However, multiplying that with d_m (the absolute sample number) will amplify the effect over time and change autocorrelation properties.
I did some simulations that seem to confirm that.
Since one comment in the code mentions a table, I wondered whether there is a reference to a description of the algorithm. (I hope I didn't miss it in the docs)
Best,
Bastian
[1] https://github.com/gnuradio/gnuradio/blob/next/gr-channels/lib/flat_fader_impl.cc#L76
_______________________________________________
Discuss-gnuradio mailing list
Discuss-gnuradio@gnu.org
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
No comments:
Post a Comment