From: Mike Coslo on Aug 7, 9:24 am

an old friend wrote:
Dave Heil wrote:
Mike Coslo wrote:



BTW, CQ has an article on HF digital transmission. Seems that they have
got it all wrong too. They have a method that works, but it is pretty
slow for images (or files) of any appreciable size.

HF will never be the place for high speed digital transmission. There
is too much noise and signals are subject to the vagaries of wave
propagation phenomena.

Why do you keep beating this Dead Horse on "rapid transmission
of high speed digital transmission?"

Shannon's Law is absolute, proven by experiment, embraced by all
the OTHER radio services. Shannon's Law is 58 years old and
mature, both de facto and de jure.

"Vagaries of wave propagation phenomena" [at HF due to ionospheric
changes] has been well-known to academia, the commercial and
government users of HF since the 1930s. The most well-known
(to non-amateur communications on HF) is "selective fading,"
a relatively short-duration change of BOTH amplitude and phase
over a relatively narrow band-span. In commercial SSB (using
12 KHz bandwidth format of four 3 KHz separate voice-grade
circuits) this was much reduced in its effect by simply
sending TWO tone-pair sets for each AFSK TTY circuit, the pair
separated by about 1 to 2 KHz; voting circuitry at the receive
end picked out the "best" received signal. That reduced at
least 95 percent of the error effects of selective fading.

The "charge" that high-speed data transmission is "impossible"
due to some (unquantified) "phase changes" has been thrashed
and discarded by at least two HF digital voice transmission
methods, one of which (DRM or Digital Radio Mondial) has been
in test for five years and is now heard over two dozen HF
broadcasts. DRM can carry binaural ("stereo") audio. The
bandwidth occupancy is NO GREATER than a conventional monoaural
AM transmission. The key to such untilization of a relatively
narrow bandwidth for higher throughput lies in examination of
the environment and using coding theory to fit that environment.

Will it send "high-quality" digital pictures in data mode?
YES, but at relatively slow transmission rates depending on
the bandwidth used/allowed/allocated. Shannon's Law is
irrefutable. It will NOT carry "live, feature-length
motion pictures." [that seems to be your goal but that goal
is not possible to attain] It can send high-density image
data without problem, with a minimum of error, but a cost
of waiting a relatively long time for each image.

There are OTHER forms of on-the-fly determination of "the
vagaries of propagation" [on HF]. The U.S. government has
been using a standardized method for about five years now
called Automatic Link Establishment or ALE. While perhaps
not adapatable to amateur radio HF applications, it is a
system for continuously monitoring signal quality, a scan
of other predetermined frequencies to check their quality,
and automatic changeover to whichever predetermined
frequency signal quality is best. ALE on HF has been
devised and tested expressly for HF beginning about two
decades ago.

Both of the above mainly-HF systems have been little
publicized in the amateur radio press. That is a fault of
the amateur radio publishers, not the system. It is NOT
the simplistic basic-level radio theory (and coding theory)
which can be easily digested in a single reading sit-down.
They require THINKING, "non-traditional" thinking away from
what had been theory of a half century ago and propagated
as "state-of-the-art" long after its first appearance.

On pushing throughput to much greater rates in relatively
narrow bandwidths, consider the advances in amateur radio
HF techniques over the last half century. In the 1950s the
"standard" RTTY frequency-shift was 850 Hz for 60 WPM
TTY 5-level coding. Today it is 170 Hz for 100 WPM 8-level
TTY coding. AM voice used to take at least 6 KHz bandwidth
but single-channel suppressed-carrier sideband cut that in
half plus reducing the old AM heterodynes from CW carriers.
PKS31 innovated and devised in the UK, tested in Europe,
was a 30 WPM RTTY system using a bandwidth no larger than
a conventional on-off keyed "CW" (morse code) signal. Those
are now "accepted" methods because the amateur radio press
has publicized it.

However, in OTHER areas, look at the common, ordinary PC
modem operating with Plain Old Telephone System (POTS)
lines of 3 KHz bandwidth. The data rate over the "56K"
modem is about NINETEEN TIMES FASTER than "straight"
analog AM. MILLIONS are used each day in the USA alone.
You don't look into the WHY of such a large increase in
throughput and I don't understand why you don't. The
answer lies in combinatorial modulation PLUS some rather
simple coding theory to increase the data bit rate to
a top of 56 THOUSAND bits per second. The modulation of
the modem's audio tone carrier is a combination of AM
and PM. It does NOT violate Shannon's Law. Such an
EQUIVALENT system could be applied to HF (it is in the
TORs or Teletype Over Radio outboard boxes) but that can
be incorporated into an HF transceiver as an integral
part. All of the radio amateurs, duly licensed as part
of the "nation's service" and complete with federally-
authorized call signs, seem to be satisfied with the LAZY
way out...let someone else do the innovation and design.
They won't "accept" it until the product ad appears in
QST ready for shipment, has reviews from the "ARRL Lab"
and all can argue over the ad specifications. Packaged
innovation ready to go. Done by OTHERS. Everything for
"the bands" (meaning only HF). "Standardized." :-(

I've brought up "scaling" of data rates before but that
seems to be a non-understanding topic. It isn't in a
convenient QST or CQ or QEX article so it isn't "accepted."
Yet SCALING is done (has been for decades) in antenna
testing as well as data rate. Look at high-definition
television broadcasting that is now phasing in to
consumers. The image throughput is more than doubled PLUS
extra data is sent for quadraphonic sound (not just "stereo")
and closed-captioning in a channel space NO LARGER than
(in the USA) 6 MHz. It's a three times REDUCTION in
bandwidth...PLUS more than double the amount of video
data. The secret is in the MPEG (Motion Picture Experts
Group) digitized video data coding and compression. Not
only that, the image/sound quality is very nearly FREE
of all the "propagation vagaries" due to phase and analog
changes from moving reflections, greater immunity to
random noise (such as from tools or appliances). SOMETHING
EQUIVALENT might be done for audio on HF...perhaps scaling
down the present-day SSB bandwidth of about 2.1 KHz to just
700 Hz or maybe 1 KHz...more than double the band
occupancy for voice signals and with much greater immunity
to "flutter" and selective fading effects.

Look at the WLANs (Wireless Local Area Networks) now in
use by the hundreds of thousands daily, perhaps a dozen
or more in the local vicinity of other WLANs...AND in the
same band as cordless telephones (dozens more) and high-
data-rate systems such as CCTV monitors or music distribution
(dozens more). Each is NOT INTERFERED WITH by all the other
local systems, all can operate as if the others did not
exist. The secret to their success is Distributed Spread
Spectrum techniques plus coding theory. Every system
EXISTS in the same bandwidth yet each is separate and
undisturbed by others. No "heterodynes," no need for
fancy, expensive filters-in-the-IF, or ultimate
refinements of decades-old conventional techniques.

But, you don't seem to care about such possibilities or
even getting a hint of what might be possible. You, like
way too many others will only "accept" something if
someone else has worked it out and it is a PRODUCT on
the market. Then you can sit around and natter about the
advertising phrases and argue someone's "lab reviews"
and sound like "expert radiomen" of "extra" class when
you don't know dink about its insides. Intellectual SLOTH.
LAZINESS. All you wanna do is play wid yer raddios and
pretend to radio greatness. shrug

dit bit