[aprssig] time for APRS second generation network?

Jason Winningham jdw at eng.uah.edu
Wed Jan 5 17:08:36 EST 2005


On Jan 5, 2005, at 2:56 PM, Robert Bruninga wrote:

> Jason, you are missing two things about APRS:
> 1) it is a one-to-all traffic system.
> 2) Real-time data flows outward to one's ALOHA limit

How is designing a network layer to support these requirements missing 
either of these points?

> 3) Thus there are no LAN boundaries

Not even the horizon? I submit that there are boundaries (at least 
practical ones) that can be defined and analyzed, even if no one has 
done so before (which I doubt).

>  and there are no "routes"
>     other than flooding outward.

... and I propose a potential design whereby data can flow outward, not 
loop back and forth, or die at the source due to congestion.

> 4) You want to route from LAN to another

This is exactly what APRS does now - it just does it very badly on a 
busy channel with dumb routers (aka smart digis).

>  but this is just not
>    APRS.
>    a) it assumes you know exactly where you want your pakets to go
>        (in that case use a directed path)
>    b) Or it assumes you can define for the network so that it knows w
>        where you want your packets to go...

You have already defined both of these in 1) and 2).

> None of that suits APRS.

"This just isn't APRS."  "None of that suits APRS."  Not to be too 
flippant (stick some smileys in here), but is APRS a moody woman or a 
network with defined requirements and characteristics that we can 
analyze and perhaps improve upon?  In reality, probably both...

> Remember, with an ALOHA limit of 60 and with 500 stations
> all trying to use the same channel with digis at 10,000 feet
> trying to define LANS is an effort in futitility.  In many cases
> if not all in SOCAL, one digi covers a aFULL ALOHA lan.
>
> Thus "routing" is meaningless.

OK, you've cited an example that is so broken _nothing_ will fix it.  
Does that invalidate the router concept?  New N-n won't fix this, do we 
throw New N-n out, too?


In case I've been wasting thought on invalid assumptions, I'll state 
some of them:

Assumption:  traffic has grown on the APRS network to the point where 
congestion has broken it in some areas, otherwise you wouldn't be 
proposing New N-n to fix it.

Assumption: traffic volume will continue to grow on APRS.

Assumption: if we don't keep improving the network, it will break 
completely and APRS will die.

Assumption: we ought to be exploring ways to improve the network so 
that it does not break completely.

-Jason
kg4wsv





More information about the aprssig mailing list