variance equation of a white noise

Hi everyone,

I hope someone can answer this difficult question for me:

I am trying to implement a STFT transform algorithm on a noisy speech. I read a white noise wave and after preforming the STFT, I called D. Now, I need to solve this equation in matlab:

λd(k) = E{|Dk|^2}

I searched the net and found that this E is the variance function in matlab. But how do I calculate this function? Do I just take average of my D matrix and then square it and then subtract the kth element from it? or do I take the average and subtract the kth element squared?

Please help!!!

[Star Strider]

What are ‘d(k)’ and ‘Dk’? How do they relate to each other?

What are you solving for?

Answer 1

[Image Analyst]

If λd(k) depends on k, then what are you averaging over when you take the expected value? If you average over k to get the expected value, then λd(k) can no longer depend on k. Maybe you simply want

d = var(D);

[Mona]

No, I don't think my Lambda is dependent on K... I think I just need var(D)....Let me try..

Thanks to all

Answer 2

[Youssef Khmou]

you can program E or use var function,

 %var(x,0)
 %var(x,1)