annotate multi_freq_ddm_int.m @ 4:72c011ed1977 tip

more elaborate example with non-stat. estimate explanation
author smusevic
date Tue, 30 Jul 2013 09:56:27 +0100
parents a4a7e3405062
children
rev   line source
dan@0 1 function mfdi = multi_freq_ddm_int(frqs, mf_ders, sig, N, tol)
dan@0 2 %multi-frequency distribution derivative based estimator for
dan@0 3 % non-stationary sinusoidal analysis. Frequencies may not be at the bins of
dan@0 4 % FFT...
dan@0 5 %
dan@0 6 % Uses the Fourier kernel at specified frequencies. No assumptions about
dan@0 7 %
dan@0 8 % [1] Michael Betser: Sinusoidal Polynomial Estimation Using The Distribution
dan@0 9 % Derivative, in IEEE Transactions on Signal Processing, Vol.57, Nr. 12,
dan@0 10 % December 2009
dan@0 11 %
dan@0 12 %
dan@0 13 %
dan@0 14
dan@0 15
dan@0 16 R = size(frqs,2);
dan@0 17 t = [-(N-1)/2:(N-1)/2]'/fs;
dan@0 18
dan@0 19 t_mat = repmat(t,[1,R]);
dan@0 20 frq_mat = repmat(frqs,[N,1]);
dan@0 21 krnls = exp(1j .* t_mat .* frq_mat);
dan@0 22 krlns_ders = 1j * krnls .* frq_mat;
dan@0 23
dan@0 24 mfdi = gen_ddm_int(krnls, krlns_ders, mf_ders, sig, N, tol);
dan@0 25
dan@0 26 end