Hello Andy, Jim, Rik and all,
yes it is just that what I wanted to point at:
In a beacon mode we need not communicate information
other than the confidence on the identity of the
station "heard".
Three years ago I implemented into my CWP-system
exactly what Andy describes. Finding very weak
constant non-drifting carriers is easy. But
modulated, and heavily drifting signals as in my
case of EME, that really is difficult.
In the EME case the problem is home-made. Since
the relative motion of Earth and Moon is known,
and every operator knows his location on Earth,
he could precisely compensate the drift such that
the Moon hears him (and sends his signal back) on
a constant frequency. And the receiver could
compensate for it's own motion relative to the
moon. But sometimes it seems to me that amateurs
do not like to go the easiest way.
Indeed, the WSPR system, i.e. the infra-structure
as seen by the operator is fine. Unfortunately,
the subsystem for communication via the radio
waves could be better by more than 6 dB.
Also finding a BPSK-signal is easier than WSPR.
73!
Klaus, DJ5HG
> Automatic monitoring of the signal strength / frequency of CW beacon
> signals, including those that drift by a small amount, can be achieved
> using signal processing techniqes borrowed from the radar and Comint
> worlds. Referred to as CFAR, or Constant False Alarm Rate, the
> principle is to take an FFT of the wanted band, then obtain an
> estimate of the noise level and all CARRIER type signals present by
> using statistical techniques, like averaging over adjacent time
> samples, searching adjacent bins and voting.
>
> I described the process in detail in the RadCom Data Column in three
> parts over April/June/August 2008 and it also appears in "Command" [
> http://www.rsgbshop.org/acatalog/Online_Catalogue_Computing___Radio_39.html
> (still some copies left at bargain prices:) ]
>
> The S/N of any carrier that exceeds the noise level by a certain
> minimum amount can be determined accurately by the monitoring process,
> its frequency determined and drifting signals can be tracked.
> Multiple carrier s can be tracked if needed The process does not work
> so reliably where signals are modulated, so clearly are not of much
> use for situations where the beacon callsign etc is not known in
> advance. However, it can be made to work sucessfully on the
> microwave bands, where beacons freqeuncies and locations are known in
> advance.
>
> Microwave Beacon Monitoring software using CFAR techniqes and
> including all source code can be found at :
> http://www.g4jnt.com/dspsw.htm
> with an overview at http://www.g4jnt.com/UWBCNMON.pdf
>
> This particular software is too wideband for LF use, but the code
> could be modified for lower sampling rates and bandwidths, and used
> for auto monitoring of QRSS signals, once their origin has been
> determined by other means
>
> Andy G4JNT
> www.g4jnt.com
>
>
>
> 2009/1/27 Rik Strobbe <[email protected]>:
> > Hello Klaus,
> >
> >> On the other hand, WSPR has the advantage of automatic
> >> recording. But that is not an advantage of the mode, it
> >> simply is the lack of a corresponding simple program that
> >> does the same with a QRSS signal.
> >
> > the automatic recording/reporting ability is clearly an advantage of WSPR in
> > regard with propagation studies. And I agree that QRSS or DFCW would serve
> > as well or even better if a similar automatic recording/reporting would be
> > provided.
> > But I am afraid that a "waterfall display to ASCII" conversion is not simple
> > at all ;-)
> >
> >> Nevertheless, WSPR is not very near to the Shannon limit.
> >> I spent nearly all my leisure time of the last year with
> >> the design of a new digital ham radio QSO-mode HD43 that
> >> comes as close as possible to the theoretical limit.
> >> I will send a preprint on request.
> >
> > I am interested
> >
> > 73, Rik ON7YD - OR7T
> >
> >
> > Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
> >
> >
> >
>
|
