How do I create a matched filter for signal detection?

Hello, I am doing a project and I need to compare the accuracy of signal detection using cross correlation and a matched filter. I have pulse train with added noise, I have managed to use the xcorr function to correlate my signal with the oulse train and I get a good result! However I am struggling to create a Matched filter algorithm. My cross correlation code is below. Any help will be greatly appreciated. Thank you.

fs = 1.5e6;%sample rate of 1,500,000

twin=0: (1/150)/fs: 0.01e-3;

tpulse = ([1:150]-1)*(1/fs);

fc =130e3;%centre frequency of 130kHz

w = 2*pi*fc;%omega

y = real(exp([-1i*w*tpulse]));%sin wave

a = hann(150);%hanning window to give it click profile

y2 = y.*a';

Nclick=10 %number of clicks;

start=((1:Nclick)*100e-6)/(1/fs); % defines a vector of start positions

y3=zeros(1,length(twin));

for ii=1:length(start);

    y3(start(ii):start(ii)+length(y2)-1)=y2;

end;

SNR=0.2; % signal to noise 20%

noise=rand(1,length(y3));

y4=y3+noise;

c = xcorr2(y4,y2);

Also does the SNR value I've assigned have any effect on my signal?

[Yateesh]

when i am running the above code in my matlab..its saying "the specified module could not be found".. what can be the problem..??

Answer 1

[Honglei Chen]

You can refer to the following thread:

http://www.mathworks.com/matlabcentral/answers/4502-matched-filter

[Ted Shultz]

hyperlinked version of above answer: http://www.mathworks.com/matlabcentral/answers/4502-matched-filter