[Laser] Troposcatter and Information Theory-2

[John Matz]

Hello,

I haven't jumped in until now, but here's my two cents.

Most of the digital modes are within a few db of each other in
signal-to-noise performance and are usually within a few dB of the Shannon
Limit.  With error correcting codes and a Viterbi demod one can get closer
to the Shannon Limit.  To make a noticeable effect in sensitivity, one must
slow down the data, reduce the thruput.  For example, FSK RTTY is running at
60 wpm and 45 bps.  PSK31 BPSK runs at about 40 wpm and 31 baud with 1
bit/symbol.  CCW OOK runs at 12 wpm and 10 baud and 1 bit/symbol.  WSJT JT44
runs about 7 wpm and 6 baud with about 5.5 bits/symbol.  As you might have
guessed, JT44 has the best sensitivity, but also the lowest thruput.  Also
see my website at my.core.com/~jematz.

What does it all mean?  You can get a slight improvement in performance by
choosing the best mode, but the dominant effect in sensitivity is from
dropping the thruput rate.  Slower can be more sensitive.  That's why people
like to try really slow CW or data.  How slow?  Slower means more narrow
band which means less noise, but higher stability is required.  In laser
systems that use a modulated subcarrier, like CW on 900 Hz or PSK31 on a
soundcard output tone, stability of the audio tone should be pretty good, so
we can run slow and sensitivity should be good.

As for the improvement by the averaging of many repetitions ... if the
signal is detected first and many copies (N) are averaged after, the
improvement in detectability is about 5 log N .  If the N copies are
averaged before detection (coherently), the improvement is about 10 log N.
I think.  This means you could repeat the data at a higher speed N times,
predetect average coherently, and get the same sensitivity as at a lower
speed and sending once.  But the JT44 system detects first, then averages,
which gives half as much improvement from averaging.  I think.  I haven't
thought about this enough lately.

73
John Matz KB9II

----- Original Message -----
From: <stjohn at ocsnet.net>
To: <laser at mailman.qth.net>; "Tom Upton" <StJohn at ocsnet.net>
Sent: Thursday, July 01, 2004 3:59 PM
Subject: Re: [Laser] Troposcatter and Information Theory-2


> try http://aintel.bi.ehu.es/psk31.html
>
> "PSK31 is a new digital mode:
>
> designed by Peter G3PLX
> that betters SLOWBPSK, an idea and implementation of Pawel SP9VRC
> based on the RTTY mode of operation,
> useful for live keyboard to keyboard QSO
> that works at 31.25 bauds,
> that uses varicode character coding what gives 50wpm,
> easy to use and monitor,
> that give very good copy under low Eb/No numbers and is thus suitable
> for
>           QRP,
> that instead of using FSK or on/off keying uses BPSK or QPSK with a
>           Viterbi decoder,
> that is available for free for many platforms, including Windows (c)
> with
>           SoundBlaster type Soundcard,
> and uses advanced DSP and narrow band (31 Hz!!) techniques."
>
>
> PSK31, also known as "warbler", is very narrow band, and is used by the
> QRP (Reduced Power) Hams to talk amazing distances (1500 miles) on much
> less
> than a tenth of a watt of radio frequency power.
>
> It is quite able to be used on telephone freqs.  It
> is easily detected below the noise threshold of receivers.
>
> Should we not be looking at it in addition to long duration dits and
> dahs?
>
> Tom Upton  AD6N
>
>
> TWOSIG at aol.com wrote:
>
> > I am still trying to absorb the ideas on posted on this topic (and
recovering
> > from Field Day).
> >
> > I see the benefit of slow transmission rate at 136 KHz, it is obviously
a
> > bandwidth limited channel.  Similarly a phone modem cannot transmit fast
data.
> >
> > What is not clear to me, is the reason why very long closely spaced
tones
> > have an advantage sending information through air on a laser beam, which
is not a
> > bandwidth limited channel.  It seems to me sending the same message with
> > shorter tones, but repeating the message, so long as the total
throughput is
> > constant, should provide the same probability that the message is
correctly
> > received.
> >
> > I think that part of the reason I am not understanding is that things
are
> > described in radio communications terms or techniques that have been
used for
> > weak signal radio reception and Very Low Frequency work.  I know that
there are
> > similarities, but there are also hugh differences when we are talking
about
> > light beams.
> >
> > Here is an example of a difference.  The laser diodes that I have been
> > working with are notoriously non-linear.  I use a 555 audio frequency
square wave
> > generator to modulate it, and with CW, I start and stop the generator.
If I did
> > that on any RF frequency, I would get terrible key clicks.  Are they on
my
> > laser signal?  Probably.  Do they interfere with my communication or
with anyone
> > else's communication?  No.  So I probably do not need to find a way to
> > suppress the key clicks.
> >
> > Another example.  In theory, I could key a crystal oscillator with the
> > control line of my Ramsey laser transmitter.  It would be a simple
on-off keying of
> > the oscillator at about 18 kHz, pulse width modulated.  I could then
tune a
> > receiver to the crystal frequency and using AM mode, listen to the
sounds picked
> > up by the microphone.  For that matter, at close range, I could use an
> > un-tuned wire for an antenna, a chunk of galena, a cat-whisker, and high
impedance
> > headphone.  Would it work?  Yes.  Should you use it on HF?  No.
> >
> > My point is that the differences between laser work and VLF, MF, HF,
VHF,
> > UHF, SHF, weak signal, and all the rest, are not being communicated, and
maybe
> > are not being considered.  If a technique works for VLF weak signal,
another
> > technique might suit laser weak signal better.
> >
> > James
> > N5GUI
> > _______________________________________________
> > Laser mailing list
> > Laser at mailman.qth.net
> > http://mailman.qth.net/mailman/listinfo/laser
> _______________________________________________
> Laser mailing list
> Laser at mailman.qth.net
> http://mailman.qth.net/mailman/listinfo/laser
>
>