[Lowfer] Time coherent scheduling, was: Group of Four

[Stewart Nelson]

Hi Mike and all,

There is a system of LF beacons that all operate on the same
frequency.  There are no IDs; each station transmits at known,
different times.  It can be used to simultaneously monitor
propagation conditions from many locations to your receiver.
Distant signals can be received, even if you are less than a
mile from a transmitter.  And the receivers enhance the S/N
by not processing data when a signal should not be present.

This system has about 70 transmitters worldwide, and has
functioned well for many years.  It's called LORAN :)

There are many advantages to a time division multiplex scheme,
but I see several potential problems:

First, if your transmitter is "on" only 10% of the time, then
with a given PEP, remote detection (at a given confidence level)
will take ten times as long, compared with PSK, FSK, etc., or
five times as long as with QRSS.  Aside from the inconvenience,
good propagation may not last long enough for a successful capture.

A possible workaround for the above:  47 CFR 15.217(a) says
"The total input power to the final radio frequency stage
(exclusive of filament or heater power) shall not exceed one watt."
The final "stage" includes not only the active device(s), but also
passives such as decoupling capacitors.  If a very large capacitor
is used, it should be possible to feed one watt DC into the final,
and transmit almost 10 watts PEP, as long as the duty cycle is no
more than 10%, and the repetition rate is fast enough relative to
the capacitor size.  To be practical, the time slots would need
to be much shorter than one minute, but there are other advantages
to faster rotation anyway.

The second major problem is that efficient decoding of the ensemble
will require specialized software.  As a minimum, you would need to
integrate multiple pulses from a given source.  Taking full advantage
of a pulsed system would involve considerable additional complexity.

But the result could be quite fruitful, yielding propagation delay
and phase stability information, robust rejection of birdies and
other carrier-like interference, and reliable detection of lower
signal levels than possible with QRSS.

73,

Stewart KK7KA

----- Original Message ----- 
From: "Mike Staines" <Mike_Staines@Email.com>
To: <lowfer@mailman.qth.net>
Sent: Wednesday, December 11, 2002 12:24 AM
Subject: [Lowfer] Time coherent scheduling, was: Group of Four


> Dave wrote:
> >   Alternatively or as well-There's a good case here for you all to develop
> > GPS synchronized element timings-and then sort out who goes when!
> 
> Actually, I have some code floating around here (never finished) that would
> generate QRSS keying based on time. My original plan was to do something
> like the propagation beacons on HF where each beacon could have a segment of
> time on the frequency.
> 
> I did not use GPS for timing however. I just used the PC clock. But with the
> various time protocols on the internet the PC clocks can be accurate enough
> for this kind of work.
> 
> Thinking about it, I might just explore the thought a little more. Since the
> time slot for each station would be known there is no need for "IDs". If a
> carrier shows up at 12:01 pm on a given frequency then it *has* to be a
> particular station. If we provide one minute slots for each beacon, and the
> beacons transmit carrier for that one minute, then in 10 minutes we could
> potentially see the propagation conditions for 10 different areas of
> listening.
> 
> This would also address the issue of nearby beacons "clobbering" listeners.
> The local beacon (even your own beacon) would only be on, say, 1/10 of the
> time leaving 90% of the time available for looking.
> 
> A final thought... This might also help with post-processing of signals
> (i.e. recording to wave files, etc) since you could index to a particular
> time that a signal should be present and not waste time processing data when
> you know the signal should not be present.
> 
> My contribution for the year.
> 
> Thoughts?
> 
> Mike
> wa1ptc