[aprssig] New map for New n-N Paradigm

[Robert Bruninga]

>>> mwrobertson at comcast.net 2/8/05 9:30:20 AM >>>
>Seems to me we are now wanting to isolate every community 
>in their own little LAN, 

Yes, it is a fact, that a 1200 baud channel can only handle
so much traffic.  It is time to realize that trying to use APRS
beyond that 100% sturation point is pointless, ruins APRS
integrity and reliability.  Yes, it puts lots of people on the map
but the channel cannot be used for real-time reliable communications.

>and if WIDE2-2 mobiles are in between, too bad...

Remember a mobile generates 5 to 15 times as much of a
load on the network as a single user station.  In most areas,
the network cannot afford the load of mobiles using WIDE3-3
or 4-4 or even 5-5 paths.  Remember, that going from 2-2 to
3-3, triples or quadruples the number of packets and the
area covered.  Going to 4-4 from 2-2 is a 9 to 16 times
increase on the network for just that one mobile.

Even a WIDE2-2 can still get into an IGate even it the
IGates are only spaced every FOUR digipeaters.  I'd love
to see a plot of the USA showing all IGates.  If there
is any area where there is not an IGate within 4 hops
of another IGate, then lets FOCUS ON SOLVING THAT
problem, not crushing the network everywhere becuse
of that one fault in one area...

> As I understand, you want to do away with RELAY altogether, 

Yes, because it causes a multiplication of dupes that there
is no existing digipeater code (other than UIDIGI ROMS) that
can filter those dupes out.  Eliminating those dupes can
DOUBLE or TRIPLE network capacity.  Thats a big improvement!

>and make the  fill-in stations honor WIDE2-2. Some of these 
>fill-ins would eat one of the WIDE2-2, and the next 
>WIDE2-1 probably is not in range of an I-GATE..

Yes, there are two things here:
1) A lot of the reason we need fill-in digis is because of
   the QRM that causes our existing digis to never hear a 
   clear channel and therefore to never hear a mobile out 
   in the fringes due to colliions.   THe RF path is there, but 
   the mobile has to be 10 times the strength of the QRM 
   to get in.   This problem will be drastically reduced if we 
  clean up the whole mess with the New n-N Paradigm

2) If an in-fill digi is still needed, then lets solve it properly
    with a WIDEn-N digi (which has perfect dupe cancelation)
    instead of adding to the problem by putting in RELAYS
    with all their dupes.

>Are we attempting to do away with the tracking of travelers 
>cross  country that use to use RELAY,WIDE2-2 or 
>RELAY,WIDE3-3 out in the open  country? 

Not at all.  We are tyring to make it work better for them by
making the network several factors better in efficiency.  THis
means muliplying RANGE by reducing QRM.  RIght now the
network is QRM limited, and no digi in most areas can hear
to its full RF range potential because its receiver is listening
to 100% QRM channel.

But continuing to cripple the efficiency of the network by
continuing to rely on the original RELAY,WIDE process
invented 13 years ago and known to have sever DUPE
problems is holding us back.  In 1994 we first proposed
the WIDEn-N system.  Kantronics gave it to us in 1998.
But due to legacy concerns we have never taken the
action necessary to get rid of the old dupe-multiplying
RELAY,WIDE process...

>I surely am missing something in this new scheme of things..

The new scheme is finaly realizing that the only way to get
to network efficiency is to do it "right" and this means 
having to give up some old crutches..  an to let the network
help enforce some of the network discipline that is needed
for an efficient network.

And this is not something that "Bob" is trying to force down
everyone's throat.  It is the culmination of years of the best
input from many of the best network folks and digipeater
sysops, and in fact, I was part of the problem in not wanting
to give up the old RELAY,WIDE because of all the digis
in the East that are still WIDE only.  

A year ago when I started this New n-N paradigm stuff, 
HALF the digis in thhe Maryland, VA, and PA area were
not WIDEn-N capable.  And back then we did not know
how to mke a TRACE digi be compatbile.  Now we do.
Also, looking at my map last night, I notice that a lot
of folks have gotten the message and now only 4 
digis (compared to 16 last year) are still TRACE.

THus, in my mind, its time to move onward.

Hope that helps.

Bob, WB4APR

Robert Bruninga wrote:
> Just to give something to throw rocks at, I changed the
> map on the New n-N Paradigm page to show those
> areas of the country that might have so much volume that
> they may want to restrict WIDEn-N flooding to the number 
> of hops shown.  If I guessed wrongly, let me know.
> 
> http://www.ew.usna.edu/~bruninga/aprs/fix14439.html 
> 
> about 45% down the page.
> 
> I know of no areas that would be inviting WIDE5-5 or
> larger hops.  If there are such areas, let know and I 
> will add them to the map.  Remember though that
> those areas would need to be 5 hops or more away from
> any population densities that *don't* want to see them.
> 
> Please help me fill in the map so we can all see what
> drum we are marching too.  Thanks.
> 
> de Wb4APR, Bob
> 
> 



_______________________________________________
aprssig mailing list
aprssig at lists.tapr.org 
https://lists.tapr.org/cgi-bin/mailman/listinfo/aprssig