FPGA Central

Anyone engaged in SNR calculation in the presence of colored noise.

H(f) --> + --> H2(f) --> MF -->
^
|
AWGN

H2(f) makes the AWGN colored at its output!
How to calculate SNR at MF(Matched Filter)output OR H2(f) output?

On Mar 10, 4:41 am, "saqib Kaleem" <ul98...@yahoo.com> wrote:
> Anyone engaged in SNR calculation in the presence of colored noise.
>
> H(f) --> + --> H2(f) --> MF -->
> ^
> |
> AWGN
>
> H2(f) makes the AWGN colored at its output!
> How to calculate SNR at MF(Matched Filter)output OR H2(f) output?

There are many methods for estimating SNR. Often they depend on the
characteristics of H(f).

Google on "SNR estimation" and any known characteristics of H(f).

Expect there to be many more results than useful results. Be patient
and thorough in your examination of them. The more care you use in
specifying your search, the better your results will be.

You will also find that there are many results that come from
responses to the same question having been previously asked on
comp.dsp.

Good luck

Dale B. Dalrymple

On Mar 10, 7:41*am, "saqib Kaleem" <ul98...@yahoo.com> wrote:
> Anyone engaged in SNR calculation in the presence of colored noise.
>
> H(f) --> + --> H2(f) --> MF -->
> * * * * *^
> * * * * *|
> * * * * AWGN
>
> H2(f) makes the AWGN colored at its output!
> How to calculate SNR at MF(Matched Filter)output OR H2(f) output?

What are the given observations? Only the output signals?
Do you know H and H2? Are you looking for an analytical answer?
When is this homework due?

On Mar 10, 7:41*am, "saqib Kaleem" <ul98...@yahoo.com> wrote:
> Anyone engaged in SNR calculation in the presence of colored noise.
>
> H(f) --> + --> H2(f) --> MF -->
> * * * * *^
> * * * * *|
> * * * * AWGN
>
> H2(f) makes the AWGN colored at its output!
> How to calculate SNR at MF(Matched Filter)output OR H2(f) output?

You'll need to know H2(f), but if you can isolate the noise power at
the MF output, you can then use the noise bandwidth of the H2(f) -> MF
cascade to determine the amount of noise power at H2(f)'s input. It
just so happens that isolating white noise's PSD parameter No is more
straightforward due to its circular symmetry. You could use the model
that you have for the H2 -> MF cascade to determine what those
convolutions do to the noise term's PSD, then use that knowledge to
try to estimate the variance of the noise at the matched filter
output. Given that variance, you can back up through the chain to
estimate the AWGN's variance also.

Jason