[aprssig] Future Concept for APRS

Robert Bruninga bruninga at usna.edu
Mon Sep 21 14:42:14 EDT 2009


>> Please, understand that APRS digis RARELY 
>> hear every packet, because FM collisions 
>> occur MOST OF THE TIME.  Hence, Viscous 
>> digipeating is a BAD idea, because it 
>> multiples the number of SLOTS taken by 
>> each packet, instead of having ALL digi-
> peaters respond at the same time which is the 
>> way APRS was designed. Viscous digipeating 
>> will really slow down a network and multiply QRM!
> 
> It is an EXCELLENT idea for fill-in digipeaters, 
> which only need to digipeat if the digipeaters 
> at higher elevations don't catch the packets. 

I think you are missing the problem.  On the majority of fill-in
installations, such fill-in digi's often hear two digis at the
same time.  Therefore this fill-in will NOT HEAR half of all
packets on the air because any that are simultaneously
digipeated by the two BIG digis will collide and the fill-in
will NOT see the packet, so then it waits a while (using
viscous) digipeating and it then sends it.  Taking up another
full slot (blocking the inputs of the big digis)

> It does not increase the number of slots 
> taken by a packet in this case, 

I respectively completely disagree.

The ONLY time that viscous digipeating can work is when the digi
where it is implemented can hear one and only one other digi, or
at least one digi 15 dB stronger than all others.  THen and only
then is there a net gain.  All other conditions the viscous
algorithm ADDS QRM to the network.

And my experience is that very-very few APRS people that put up
digis make any effort at RF analysis and congestion perofmance
at the digi, before setting up parameters.  The result is that
most such installations that would enable viscous digipeating
would be detrimental to the network instead of benefit.

Look at how many systems screw up even the simplest TNC
parameters..

My conclusion is that Viscous digipeating can only be a benefit
under a very narrow set of conditions and that in all other
cases, it is a net-negative and should be avoided.  Since we
have no way of enforcing proper digi set up, it is safer to
avoid these kinds of loose canons that can be both bad in most
cases though good in some.

Bob, WB4APR






More information about the aprssig mailing list