[aprssig] On flow control/smart beaconing...

[Jim Lux]

Lots of interesting stuff on packet transmission timing strategies.

But, aside from the desire to not send packets every 30 seconds while 
parked at work, there IS a problem that's not addressed by the usual smart 
beaconing strategy (beacon rate proportional to rate of position change), 
and that's dealing with getting into the system at all with a low powered 
tracker.

I can beacon every minute on a 20 minute drive across the San Fernando 
Valley, where my speed is essentially constant, as is my heading 
(therefore, making predicting my position from an old packet quite easy), 
and get maybe 1 or 2 packets (out of 20 or so radiated) in to the N6EX digi.

Now, my beacon is (arbitrarily, for the purposes of argument) fixed output 
power.  To a first order, the probability that I will get a packet in is 
proportional to the energy I radiate (that is, power times number of 
packets in a given time interval).  Therefore, the optimum strategy, for 
me, is to beacon almost continuously, and hope that I get in.  (mind you, 
such a strategy would be viewed as abusive by some)

One might argue that this is an anomaly.. I'm running a 5W transmitter, and 
the digis see hundreds of miles, and dozens of 50W transmitters, so the 
"APRS network architecture" in the area is flawed: there have been 
suggestions that Wide1-1 digis need to be sprinkled throughout the area, 
for instance.   However, it's not all that peculiar.  It's not like the 
greater Los Angeles area is somehow special, and even if it were, then the 
APRS system should accomodate it (beaconing guidelines, etc.) as it 
sits.  Heck, if all those 50W mobiles running around were able to digi my 
feeble little 5W packet to a high digi, it would work better.

And, of course, as someone else pointed out, my interest is NOT in getting 
the packet onto the internet (although it IS useful for diagnostic 
purposes), it's really to get my packet distributed widely over the RF 
channel, in particular, over the hills beyond my line of sight.  However, 
because of the "congestion management" policies, this doesn't happen.  I 
can have line of sight to one high digi (on Oat mtn), and my friends in 
Arcadaia can have line of sight to another high digi (both members of the 
N6EX system), but he can't see my packets, nor I his.  This is not 
particularly useful in a tactical sense, if we are both converging on a 
common location from opposite directions.

Jim, W6RMK