[hfsig] Is there anyone out there?

[Charles Brain]

'Simple'

You can estimate the frequency error from a repeated preamble.
Then apply the frequency error compensation using a complex mixer.

The frequency error estimator I use can estimate the error to
a fraction of a Hz even on a multipath channel. If is a Gaussian channel it
can get it down to about 0.001Hz.

For wide frequency errors you run multiple preamble detectors in parallel.
4285 use 3 channels to detect the preamble with up to 75Hz error.

My general purpose frequency error estimator multiplies the received preamble
by its complex conjugate (so all the symbols are then 'identical')  then accumulates
the phase error across a number of symbols. It does that for the entire preamble
takes the average and then from that calculates the frequency error.

You can then apply the frequency error correction to the received signal using a complex
mixer to the quadrature signal.

Something like 4285 transmits repeated 80 symbol preambles through out the
entire transmission. Any Doppler shift between  the preambles can be taken care
of by the adaptive equaliser.

That is just one way of doing it for a particular waveform.

I recently bought a book called 'Synchronization Techniques for Digital Receivers'
ISBN 0-306-45725-3
It has sections on just about all known carrier recovery techniques.

- Charles


----- Original Message ----- 
From: "Trevor ." <m5aka at yahoo.co.uk>
To: "TAPR HF Modes SIG Mailing List" <hfsig at tapr.org>
Sent: Tuesday, October 13, 2009 9:05 AM
Subject: Re: [hfsig] Is there anyone out there?


> Somethink I've been thinking of is a Digital Mode that could be used for QSOs through the Amateur Radio satellites.
>
> Currently good old CW is the only effective digital mode for random two-way contacts.
>
> The reason for this is Doppler shift.
>
> For a satellite in a Low Earth Orbit the Doppler can between 10-30 Hz per second depending on the bands in use and orbital height.
>
> Now I know you can use a PC to automically tune the Tx and Rx to compensate for Doppler but in practice most people seem to 
> compensate for Doppler by manual tuning. Now this works fine for CW where it doesn't matter if a signal is 200 Hz off tune, you 
> still tell what's being sent but currently none of the available Digital Modes can handle that.
>
> 73 Trevor M5AKA
>
>
>
>
>
>
> _______________________________________________
> hfsig mailing list
> hfsig at tapr.org
> https://www.tapr.org/cgi-bin/mailman/listinfo/hfsig