[ARRL-OK] FWD: Fixing the 144.39 APRS Network The New n-N Paradigm

[N7HRT at aol.com]

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

See  the whole page for the rest of the details.
 
 

Implementing this proposal is up to the local sysops, but for the  users, 
they may begin using most of these simplifications immediately. . It is  not the 
intent of this proposal to force anything on local operations, but to  offer a 
consistent APRS set of recommendations that has been well thought-out,  
designed and discussed openly to make the system work better for everyone.  .

SUCCESSFUL APRS: The success of your local APRS is not how many  stations you 
see on your maps, nor how far away, but how reliably your mobile or  handheld 
or portable station can communicate with others in the local area. .  This 
fundamental principal should drive everything we do with APRS in our local  
areas. 

RELIABLE APRS: Said another way, the more stations you see above about  60 to 
100 or so in typical areas, the more packets you don't see due to  collisions 
and the less reliable your network is for local real-time APRS use. 

FIXING THE 144.39 NETWORK: The most important thing we need is USER  
EDUCATION that a local 1200 baud APRS network capacity is greatly reduced by  each 
additional digipeater used. . The area coverd goes up, but the capacity is  
reduced by half. Typical areas where people use 2 hops with four or more digis,  
can only support about 60 to 100 or so users in its RF domain. . Conversly, if  
no digipeaters were used and eveyone could hear each other direct, APRS could  
theoreticlly support about 360 users on a simplex channel direct (with time  
slotting). . But because of many people using too many hops and bad paths that 
 cause too many dupes for surrounding areas, some areas of the country are  
totally saturated with 10 times the amount of attempted packets with only the  
strongest and most abusive getting through. . Net reliability is miserable for 
 those locals whose packets cannot even get into the local digi due to out of 
 area congestion. This is self defeating. There are many things wrong: 

Lack of user understanding of fundamentals and severe limitations of  such a 
small channel 
Too many collisions and lost packets due to too many  Hops and too many 
duplicate packets 
Software that does not place the health  of the Network foremost in the eye 
of the user 
Poor Digipeater settings (HID  not off, UIDWAIT not 0, and bad packet timing, 
excessive paths) 
Improperly  formatted packets See live error list as captured on FINDU.COM. 
Unrealistic  user Expectations (expecting to see hundreds of stations for 
hundreds of miles.  Can't happen on RF) 
Failure to properly set up old PacComm "T" digs in some  areas for WIDE2-2 
support 
Continued use of obsolete RELAY and WIDE paths  that multiply the number of 
dupes. see how bad 
DIGI ROMware that cannot be  updated. See my wish list for future digi 
implementations. 
Fractionalizing  and network disentigration due to local sysop (well intended 
but inconsistent)  hacks.