lsfblp.m

来自「Introduction to Statistical and Adaptive」· M 代码 · 共 44 行

function [a,e,V,FPE]=lsfblp(x,M)
% Linear predictor design using No-Windowing
% forward-backward linear least-squares
% Model parameters [1 a1...aM V]
% V=input signal variance.
% FPE=Akaike's final prediction error criterion
%
% (same as arlsfb.m, which uses "slow" R computation.)
%
% Programmed by: Dimitris Manolakis, 16/3/94
%
%-----------------------------------------------------------
% Copyright 2000, by Dimitris G. Manolakis, Vinay K. Ingle,
% and Stephen M. Kogon.  For use with the book
% "Statistical and Adaptive Signal Processing"
% McGraw-Hill Higher Education.
%-----------------------------------------------------------



Lx=length(x);
R=zeros(M,M);
d=zeros(M,1);
R1=zeros(M+1,M+1);

R1=lsmatrix(x,M+1);

R1=fliplr(flipud(R1))+R1;


R=R1(2:M+1,2:M+1);
d=-R1(2:M+1,1);

c=R\d;

a=[1 c(1:M)']'; 

e=filter(a,1,x);

for n=1:M e(n)=0; end;

V=(e'*e)/(Lx-M);
FPE=V*(1+M/Lx)/(1-M/Lx);