parametric frequency representation of a signal.
[spec,freqs]=parafrep(Rx) [spec,freqs]=parafrep(Rx,N) [spec,freqs]=parafrep(Rx,N,method) [spec,freqs]=parafrep(Rx,N,method,q) [spec,freqs]=parafrep(...,'plot')
correlation matrix of size (p+1)x(p+1)
number of frequency bins between 0 and 0.5
can be either 'ar', 'periodogram', 'capon', 'capnorm', 'lagunas', or 'genlag'.
parameter for the generalized Lagunas method.
when called with the additional string 'plot', the output values will be plotted
noise=rand(1000,1); signal=filter([1 0 0],[1 1 1],noise); figure(1);parafrep(correlmx(signal,2,'hermitian'),128,'AR');title('AR (2)'); figure(2);parafrep(correlmx(signal,4,'hermitian'),128,'Capon');title('Capon (4)'); figure(3);parafrep(correlmx(signal,2,'hermitian'),128,'lagunas');title('Lagunas (2)'); figure(4);parafrep(correlmx(signal,40,'hermitian'),128,'periodogram');title('periodogram (40)'); | ![]() | ![]() |