01.m

基于LMS算法的自适应滤波器的MATLAB仿真

% filter parameters
M=20;      % number of taps
mu=0.05;   % step-size parameter
e_max=500; % maximum #of epochs
% constants
pi=3.14;Fs=0.02;   % signal frequency
Fn=0.05;           % noise frequency
% initialize
w=(randn(1,M)-randn(1,M))/100;
d=zeros(1,M);u=zeros(1,M);
u_out=zeros(1,e_max-M);
f_out=zeros(1,e_max-M);
% Generate desires signal and input(signal+noise)
for t=1:M-1    
    d(t)=sin(2*pi*Fs*t);    
    u(t)=d(t)+0.5*sin(2*pi*Fn*t)+randn;
    endt=M;
    epoch=0;
while epoch<e_max    % generate new input    
    input=sin(2*pi*Fs*t);    % shift new input into array    
    for i=2:M        d
        (M-i+2)=d(M-i+1);        
        u(M-i+2)=u(M-i+1);    
    end    
    d(1)=input;    % add undesired freq & random noise    
    u(1)=input+0.5*sin(2*pi*Fn*t)+0.09*randn;    
    u_out(t-M+1)=u(1);       % compute filteroutput    
    output=dot(w,u);   
    f_out(t-M+1)=output;       % LMS algorithm    % compute error    
    e=d(1)-output;       % update weights    
    for n=1:M        
        w(n)=w(n)+mu*u(n)*e;    
    end       %    
    in(t-M+1)=u(1);    
    out(t-M+1)=output;    
    err(t-M+1)=e;       
    t=t+1;    
    epoch=epoch+1;       %  plot noise and filtered signal   
    figure(1)    
    subplot(211);    
    plot(t,u(1));
    axis([0 e_max -2.5 2.5]);    
    title('滤波器输入')';    
    hold on;    
    subplot(212);    
    plot(t,d(1));
    axis([0 e_max -2.5 2.5]);    
    title('滤波器输出')';    
    hold on;    
    figure(2)    
    subplot(211);    
    plot(t,output);
    axis([0 e_max -2.5 2.5]);    
    title('滤波输出信号')';    
    hold on;     
    subplot(212);   
    plot(t,e);
    axis([0 e_max -2.5 2.5]);   
    title('误差信号')';    
    hold on;
end