Matlab - Advantage of filtering a FSK signal before demodulation

Author

Advantage of filtering a FSK signal before demodulation

terence

2005-04-25, 9:00 am

Hi...everyone!!..

I have heard that you can get better BER to SNR performance by first
filtering the FSK signal before demodulating it. I have tried to
implement a model in SIMULINK that has 2 separate filters for mark
and space frequencies. And I wanted to see if the performance can
actually be improved by filtering the signal around the mark and
space signals. However the performance seems to have dropped whenever
I filter the FSK signal. I have read that this works in theory, may i
know how come this scernio cant be replicated in SIMULINK?.

I used FDAtool to implement the filter, and used Average White
Gaussian Noise (AWGN) to simulate interference.

Thanks for your attention!!

terence

AJ \no z\ johnson

2005-04-25, 8:59 pm

"terence" <ho_wei_siang@techie.com> wrote in message
news:ef03dbf.-1@webx.raydaftYaTP...
> Hi...everyone!!..
>
> I have heard that you can get better BER to SNR performance by first
> filtering the FSK signal before demodulating it. I have tried to
> implement a model in SIMULINK that has 2 separate filters for mark
> and space frequencies. And I wanted to see if the performance can
> actually be improved by filtering the signal around the mark and
> space signals. However the performance seems to have dropped whenever
> I filter the FSK signal. I have read that this works in theory, may i
> know how come this scernio cant be replicated in SIMULINK?.
>
> I used FDAtool to implement the filter, and used Average White
> Gaussian Noise (AWGN) to simulate interference.
>
> Thanks for your attention!!
>
> terence

That's not intuitive to me.
Perhaps they are referring to matched transmit and receive filters, to
reduce ISI (Inter-symbol Interference). Try looking at RRC (Raised Root
Cosine) filter theory?
-Aj

Sponsored Links