[aprssig] UIDIGI Preemptive Digipeating (was: Why RELAY, WIDE... is so bad....)

[Tim Schall]

I adjusted the main downtown Seattle, WA digipeater almost six months ago to
act only upon packets it heard "first."  That cut its transmission time down
by probablly half leaving the outlying digis to haul the traffic out of and
around the downtown area.  At that point it was still acting upon
RELAY,WIDEn-N packets in a conventional manner.  That is packets would leave
SEA as SEA*WIDE2-2.  Within the last 24 hours we've set SEA and another
Western Washington digi up in the manner that you describe here.  As we have
tons of users here still using TRACEn-N it also acts upon that but with a
twist.  TRACEn-N (eg: TRACE3-3), with or without the 'RELAY' in front of it
leaves as SEA*,WIDE3-2.  Note that it changed the TRACE to a WIDE to
accomodate the new paradigm.

I also handle problem paths like 'RELAY,WIDE' and paths that are to long
like WIDE5-5 by picking them up and transmitting them out ONCE.  That is,
RELAY,WIDE5-5 would leave my digi as 'WAS5-5*,SEA*'  That gets the user out
to an igate and has the added benefit of letting them know who whacked their
packet and why.  One additional benefit of using a real computer for a digi
is that I can exempt individual users or groups of users from any or all
rules.

We chose to leave WIDE7-7 and TRACE7-7 alone.  Around here people generally
don't use that path in some sort of an accident.  So we chose to leave it
for those times when you absolutly positively gotta do it with RF.

Anyways, glad to hear that there are others experimenting with preemptive
digipeating and the benefits of using things other than Kantronics TNC's for
digipeaters.  If there is any interest I'd be happy to post my digi_ned.ini
file here for inspection.

Tim Schall
KA7PBI

----- Original Message ----- 
From: "Greg Noneman" <greg at clubnet.net>
To: "TAPR APRS Mailing List" <aprssig at lists.tapr.org>
Sent: Friday, April 01, 2005 6:34 PM
Subject: [aprssig] UIDIGI Preemptive Digipeating (was: Why RELAY,WIDE... is
so bad....)


> The renewed discussion relating to RELAY and WIDE got me thinking about
> the local situation in Southern California.
>
> As I stated in another thread, the 4 metro LA digipeaters made a change
> in February to support single hop digipeating.  They do not support the
> normal flooding technique, but rather generate a single digipeat when
> WIDE1-1, WIDE2-2, WIDE2-2 OR WIDE3-3 is received.  Two of the four
> digipeaters (those running UIDIGI firmware) also support RELAY.  This
> change was made to help reduce the congestion in the area, and I
> believe has been relatively successful in doing so.
>
> One of the remaining issues is that many stations unnecessarily use
> RELAY as their first hop in the digipath.  While this can help in some
> situations, it usually just creates additional traffic on the channel
> by adding an extra hop to the path.  This defeats the single-hop
> concept.  Much of the greater LA area is within reach of one of the
> four high-level digipeaters.
>
> Now I've been successfully using UIDIGI firmware since 2000. It has
> been used on one of the SoCal metro digis since 2000 (v 1.8b6) and was
> recently installed on another within the last few weeks (v 1.9b3).  For
> some reason, it just dawned on me yesterday that the preemptive
> digipeating feature of UIDIGI solves our local problem!
>
> Here's our particular situation:
>
>    The N6EX-4 runs UIDIGI firmware.
>
>    UIFLOOD: disabled
>    UITRACE: disabled
>    UIDIGI: RELAY, WIDE1-1, WIDE2-2, WIDE2-2, WIDE3-3
>    Peemptive Digipeating: ON
>    Preempted Calls: WIDE1-1,WIDE2-2,WIDE2-1,WIDE3-3
>
>    Assume the following packet is launched:
> W6XYZ>APRS,RELAY,WIDE2-2.....
>
>   Without preemptive digipeating, if the digipeater heard the packet
> directly, it would respond as follows:  W6XYZ>APRS,N6EX-4*,WIDE2-2.....
>
> This packet would then most likely be heard by the remaining three
> metro digis and be transmitted three more times.  Additionally, any of
> the outlying digipeaters that heard the packet directly from N6EX-4
> would act on the WIDE2-2 and propagate it outward from LA.
>
>    Now with preemptive digipeating enabled, the digi responds as
> follows: W6XYZ>APRS, N6EX-4*.....
>
> One hop, into an IGate and it's done!   This is just what is wanted.
> Should the digi not hear it directly but it is heard by a low level
> (i.e home) digi, it will respond on the second half of the path.
>
> I don't know why it took me so long to see the value in the preemptive
> digipeating feature, but now I'm a believer!   Many of you running
> UIDIGI may already be aware of the benefits of this function, but I
> wanted to pass along my experience with it.
>
> I know this only applies to those running TNC2s with UIDIGI, but for
> situations such as ours, I think is a significant enhancement.
>
>
> 73,
> Greg
> WB6ZSU
>
>
> _______________________________________________
> aprssig mailing list
> aprssig at lists.tapr.org
> https://lists.tapr.org/cgi-bin/mailman/listinfo/aprssig
>
>
>
> -- 
> No virus found in this incoming message.
> Checked by AVG Anti-Virus.
> Version: 7.0.308 / Virus Database: 266.8.4 - Release Date: 3/27/2005
>
>