[Premium-Rx] Filter phase response

[John Reed]

----- Original Message -----
From: "CHARLES HUTTON" <charlesh3 at msn.com>
To: <premium-rx at ml.skirrow.org>
Sent: Wednesday, October 08, 2003 2:57 PM
Subject: Re: [Premium-Rx] Filter phase response

Ok, let's clarify this a little bit.  I should have said "commercially
available DSP filters" in the comparison.  I haven't found one yet that was
much good.  If I was programming to receive weak CW, I would use a FIR
matched filter.  It can be proved that this filter is optimum in the
reception of weak carriers buried in Gaussian noise.  I have programmed
these up for offline processing of weak signals, and they do work better
than anything else. Check the references for the ZRO weak signal tests on
the web.

John Reed

> And of course that is because DRM uses OFDM, so phase accuracy is
important
> to prevent symbol errors.
>
> Going back to the theme of AM and CW reception, I think it is very
incorrect
> to label "DSP filters" as third best. "DSP filters" says nothing about the
> filter characteristics and rather only tells us the implementation.
> Furthermore, a "DSP filter", in theory and practise, can exactly duplicate
> an analog filter given the same type (IIR) and the number of taps that
> matches the order of the desired analog filter.
>
> But of course the major advantage of DSP is the easy implentation of an
FIR
> filter.
> >