[aprssig] APRS Interstate routing...

[Robert Bruninga]

>>> bruninga at usna.edu 7/30/04 3:47:34 PM >>>
>This email suggests how you can get a total of EIGHT 
>UI Aliases, and FOUR total n-N systems! [by paralleling
> TNC's]...

Wow, and now this really DOES solve the one big problem 
with widescale use of interstate chains of LINKn-N's!
It SOLVES the intersection problem... in the cities where 
Interstates often intersect.  And with that problem behind us, 
I wonder if this INTERSTATEn-N situation could be come an
all new routing philosophy in APRS.?

Think about it:
1) Most of our mobile traffic is on the interstates
2) Most of our commerce and populations are
    distributed along the interstates
3) Most of our infrastructure is dependent on the
    interestates
4) EVERYONE knows where the interstates are
    and wehre they go.
5) The entire country is criss-crossed with them.
6) Its a universal NAMING convention that EVERYONE
    already knows...

Why could we not almost decide that APRS routing 
by interstate numbers is an ideal way to define
APRS traffic paths.  ALl the existing generic calls
would still be supported, but now ANYONE can
with his eyes closed be able to route his packets
to get them just about anywhere he wants to go...

Yes, a few DIGI sites where interstates intersect
may have to add a parallel TNC as I described
before, but also in some cities, there are so many
digis anyway that one can handle the N/S interstate
and a different one can do the EW interstate without
conflict or any other changes...

Sure there will be areas where this may get confusing.
but it is OPTIONAL.  You can still support WIDEn-N
FLOODING in all directions if you want.  But it sure
gives a great tool for the rest of the country!

Remember, by going to LINEAR n-N routing, the 
total number of DUPES is just N and only hits those
you INTEND to hit.  Compared to HUNDREDS of
dupes in all directions to still get only N hops in
one direction..

Wow...
Bob

Original message>>>

The answer is simple.  Just hook two TNC's in 
parallel to the same RF receiver and transmitter.
Let the original TNC support all the original modes
it normally supports, RELAY,WIDE,SS,SAR,
WIDEn-N, whatever.  Then add the new 
n-N's and other aliases into the new TNC (but
making sure to have no dupes!).  Then when a 
packet comes in, it will get digipeated by any of the 
SUM of the possible methods, but the two TNC's
will never collide... (as long as they have all
different settings).  Though you do have a potential
confilct in the NAME Of the digi itself so make one
W3XYZ-11 and the other be a tactical call-11 where
everone will know that the two are at the same
location..

Yes, it costs another TNC, but it DOES solve
the problem!

For more info on the "new n-N paradigm" please see

http://www.ew.usna.edu/~bruninga/aprs/fix14439.html 

And LASTLY, remember, the whole purpose of the
new n-N paradigm is to CUT DOWN on flooded 
traffic, not to look for more ways to bring in more.
So this combination scenario would only apply in a
sparse area where multiple "backbones" or
specific needs cross or converge...  Dont put one
of these downtown unless you eliminate WIDEn-N
entirely from the mix...

Bob, WB4APR