Correlated signal Matlab Code

[Tahid]

Hello there, Thanks for your nice discussion about correlated signal.

I would like to know how to create a correlated signal, by which i want to check detection performance for cognitive radio communication. Can anybody help me to get Matlab code for creating correlated signal by any link or source? I would like to thank in advance.

 

[Mityan]

Correlated signal could not stand alone as it is. Correlation assumes the pair of signals or the same signal with a different time shift. So for you, I guess, this may be any signal (vector or array) for which you definitely know the next value (N values).
Or maybe you should define more accurately your problem.

 

[ashish.mw]

There is not much left to say after seeing this thread:

have a look, this might be helpful to you:

https://www.edaboard.com/threads/260644/

 

[Tahid]

Thanks so much for your reply. In one paper of cognitive radio, it is explained that: if a signal (independent and identically distributed (iid) is passed through the narrow band filter it is converted into a correlated signal. I want to get a matlab code to make iid signal to correlated signal. Help me please....

 

[Mityan]

When you pass the random data through a filter, it is got smoothed, and so you already less or more can predict the next value from the given value. The distribution changes from uniform, and it means that the data signal acquires correlation.
Do the following for example:
Create random data: a = rand([100,1]);
Look at the autocorrelation function: [ACF,~,~] = autocorr(a,[],10,[]); - parameters in it I took intuitively, but they suit the example.
Plot ACF and you will see that all samples but first=1 are close to 0, which means the absense of correlation.
Then I created a low pass filter using fdatool. Parameters: fs= normalized, wpass=0.1, wstop=0.16, Apass = 1dB, Astop=40 dB. (arbitrary values for example)
Then exported the impulse response samples into workspace - vector Num.
then filtered the data: fa=filter(Num,1,a);
And look at the correlation again: [ACFf,~,~] = autocorr(fa,[],10,[]);

You will see that the data has got a correlation. Look also at fa and compare to a.

Adjust this example to your tasks.