CHANNEL SHARING and COLLISIONS:

The frequency of 144.39 is the only option since it is the existing APRS
channel, is already authorized, and 2 meters provides 9 dB performance
advandage over any other amateur satellite band.  Shared use of the frequency
for both existing terrestrial users using vertical polarization and much
higher power Satellite users is a unique approach to dual use and spectral
effeciency.

The question is how to quantify and calculate  the additive noise
power at the satellite from the MANY weaker terrestrial QRM packets.  As I
showed, a 250W EIRP mobile will probably be able to be received due to the
capture effect with his 11 to 14 dB stronger signal than any other
terrestrial station.  Since the Satellite will have a horizontal dipole
(we hope), this gives us another 8 dB isolation (you cant get perfect 
cross polarization nulling because the difference in Horizontal on both
the east and west coasts).  But, how many such signals must he compete with?
Thus isolation ranges from about 19 to 22 dB (maybe).

By monitoring APRServe (The APRS Internet World Wide Feed at 199.227.86.221
port #23)  one can see an overall throughput of about 2 to 4 packets per 
second.  If we assume that each of these (all dupes have been filtered out) 
represents at least 4 or 5 hops, then one could assume that there are at
least 8 to 20 or more simultaneous packets in the air at any instant 
(as seen from space).  Unfortunately, these will all add as random
noise, raising the threshold.  If we assume the average is about 20
simultaneous packets, then any NATSat station will need an additional
13 dB on average... This assumes noise powers add.

Some have suggested that Noise powers add as the square root of the sum of
the squares.  If this is the case, then the added noise power of these
20 stations would be only a 7 dB or so noise increase?

If so, then our actual SNR of our intended NATSat mobiles may still be
9 to 19 dB above the "average" QRM.  WIth the FM capture effect, this may
be sufficient.  Can anyone model and test this????

This situation is NOT as bad as the SAREX problem where you may have as
many as 100 to 5000 simultaneous transmissions on the channel.  Remember
that the thousands of APRS users are already spread out in TIME over
relative long periods from a few minutes to 30 minutes.
 
What is the right way to calculate the SNR in the presence of multiple
weaker random QRM?  Also, I used in the above calculations an assumed 
receiver noise level of 300 deg Kelvin which I understand is a good
average for what our OMNI antenna will see from Space..

de WB4APR, Bob