[aprssig] New n-N Paradigm FACTS

[Robert Bruninga]

I posted my Packet Analysis program that I used to get
these numbers below.  Just capture an hour of packets
with your TNC and run APRSTEST.EXE and it will generate
these results.   Here is a

Screen shot:
http://www.ew.usna.edu/~bruninga/aprs/aprstestpix.gif

and you can Download APRSTEST.EXE:
http://www.ew.usna.edu/~bruninga/aprs/aprstest.exe

>>> bruninga at usna.edu 1/22/05 4:16:21 PM >>>
>This is the real truth about the New n-N Paradigm.
>WITH NUMBERS!

%users   %pkts   path used
  7%          3%       RELAY or WIDE only
35%        27%      RELAY,WIDE or WIDE,WIDE
26%        50%      RELAY,WIDE,WIDE or WIDE,WIDE,WIDE
14%          5%      WIDE2-2
24%         10%     WIDE3-3
  3%           7%     WIDE4-4

Notice that one fourth of the users are responsible for HALF
of all packets.  They are using the very inefficient W,W,W
or R,W,W path.  We must get away from this kind of path
in APRS.  

What is immportant above is the ratio of packets to the
number of users of a given path.  The paths with the
smallest ratio of packets per user is, of course, more efficient.  
Re-arranging the table from most efficient to least is revealing:

%Users  %pckts  Factor  Path
14%          5%      2.8    WIDE2-2
24%         10%     2.4   WIDE3-3
  7%           3%    2.3    WIDE
35%         27%    1.2    WIDE,WIDE
26%         50%     0.5   WIDE,WIDE,WIDE
  3%           7%     0.4    WIDE4-4

Notice the factor column.  W2-2 is almost 3 times more efficient
than W,W.  W3-3 is 5 times more efficient than W,W,W!

Now compare the New n-N Paradigm receommendation of
W2-2 to the most common east coast W,W,W and the factor 
is a SIX TIMES REDUCTION in volume of packets.  Imagine how 
much more reliable our network can be if it efficiency improved 
by a factor of SIX!

Lets DO IT.  Use WIDE2-2 in WIDEn-N areas and SS2-2 in
states moving to the SSn-N scheme.

Lets move out on the New n-N Paradigm NOW.

de WB4APR, Bob


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