CCWN 75:23
                                               
LIMITS OF COMMUNICATION DUE TO PROPAGATION CONDITIONS Chas
Woodson

        It may have occurred to some of you, if we can gain 23.5
dB by  reducing the bandwidth from 2000 cycles to 9 cycles by a
CCW filter, why not  reduce it by another factor of 100 and get
another 20 dB. There are at least  three problems with this.      
   1. The required frequency stability is beyond our present 
capabilities. At present we need about 2 X 10-7 and can obtain
about 10-7  with care. Another factor of 100 would require 2 x
10- which is beyond our  present techniques. In my judgment this
could be achieved.                                                
                       e                                          
                                                                  
                                             2. Narrower
bandwidth will reduce the information rate. Our current CCW 
signals, using a .1 second pulse length, give 10 bits of
information per  second which results in about 12 wpm. A decrease
of bandwidth by 100 would give  .1 bit per second or about .12
words per minute. Several years ago I built a  coherent binary
code system which worked at .1 bit per second and operators 
found it frustratingly slow.          3. Modulation of the signal
by the propagation path. Experiments I did  with the above system
lead me to estimate that on the average each skip adds  about .5
cycle FM modulation at 14 MHz, although there are times when
little  modulation occurs. Thus a typical signal going from the
U.S. West Coast to  Japan might make 4 skips and therefore have 2
cycles of modulation added. This  causes large losses with a .1
bit per second filter. It looks like the  propagation modulation
of skip signals makes the limit of useful filters about  the same
as the convenient-for-humans CW rate of 10 bits per second. Space 
communications are another matter as the modulation of space
signals is far  less          Therefore, it is my estimate that
we are near the point where  decreasing information rate and
increasing propagation loss due to  modulation meet for the most
efficient communication between points on  earth by means of
propagation skip. On the other hand, under favorable  propagation
conditions, I am sure we can get another 20 Db if we accept 
lower information rates. If you stop and think about it, even
hams who a~e  sending at 10 bits per second (12 words per minute)
are actually  communicating at only about 1 bit per second when
you leave out the  information lost from QRM, sending errors,
simplex operation, repeated  information, receiving errors, and
preambles.          I predict we will see a new type of ham
station in the next decade  which will be communicating at 1 bit
per second in duplex operation. A  small computer on each end
will receive information from the operator  at a rate and by an
input method convenient for him, e.g., keyboard or  code key.
Transmission between the station.s will be QRP. One watt will  be
plenty as with 40 Db gain it will be 10 Db more effective than a
KW  with current techniques. Perhaps 10 watts will be customary
for hand  held antennas. The computer will automatically switch
transmission to  the highest information rate which can be
handled by the propagation  path. It will receive information at
the rate sent by the other  station, but display it for the
operator in a way convenient for him,  e.g., by audio CW, CRT
character display, typewriter, or spoken  characters (I will
leave spoken words for the next decade although it  is now
technologically possible).          I further predict that one of
you reading this will be  the first to build Such a station.