[aprssig] Impossible to be digi+something else? (DWAIT=0)

Wed May 11 11:13:39 EDT 2016

Lynn has it right.  Here is my summary:

The DWAIT=0 intentional priority for all APRS digiepaters for maximum
network throughput is based on the fundamental assumption that all such
digipeaters are at HIGH SITES.  That is, their TNC’s are in the best
position in the area to hear anything that is going on and to ONLY transmit
when the channel in the entire area is clear.

This does not matter if the transmission contains the DIGI’s own beacon, or
weather, or frequency objects or user packets.  It does not matter.  The
DIGI which has the best ears in the area is in the best place to make the
decision to use the channel to maximize throughput, avoid collisions and
only transmit when the channel is clear.

There are two related corollaries:

1)      ALL digis that have the same pending packet will ALL transmit at
the same time (intentional fratricide) so that particular APRS data will
take up only one time slot, no matter how many digis heard it.  Yet
surrounding next-hop digis WILL hear it.

2)      ALL user systems and any system that is NOT at the digi site or
other such highest-in-the land site will set DWAIT to 1 or other value so
that the DIGI gets first chance at the clear channel and the user will not
get stepped on.

Notice that #2 really does nothing to help the network, it only helps that
user avoid getting clobbered every time by the digi!

We have not really stressed that so much, since the network does not care.
But it is useful to users who want to maximize their throughput to make
sure to NOT have DWAIT=0 and choose some other value.  I have never studied
what that optimum value is.  It only needs to be milliseconds since once
the DIGI does transmit, then the user TNC is going to hold off until the
DIGI packet is finished anyway.

Hope that helps.

Bob, WB4APR

*From:* aprssig [mailto:aprssig-bounces at tapr.org] *On Behalf Of *Lynn W.
Deffenbaugh (Mr) via aprssig
*Sent:* Wednesday, May 11, 2016 10:23 AM
*To:* TAPR APRS Mailing List
*Subject:* Re: [aprssig] Impossible to be digi+something else?

I have asked this question several times and the answer is always the
same.  Digipeaters are not supposed to delay their digipeated packet, but
ARE supposed to wait for a clear channel.  So they're "semi-polite" as it
were.  From http://www.choisser.com/packet/part03.html (a definition of
DWAIT):

*DWAIT*: Used to avoid collisions, *DWAIT* is the number of time units the
TNC will wait after last hearing data on the channel before it transmits.

So a DWAIT of zero will not stomp on other packets, but will jump in as
soon as they "last heard data on the channel".  Not really monopolizing in
my book, just being aggressive with getting their data on the air.  As long
as the digi's other packets are not too long, too frequent, or too
numerous, I wouldn't see a need to modify the DWAIT by packet type.

Lynn (D) - KJ4ERJ - Author of APRSISCE for Windows Mobile and Win32

On 5/11/2016 9:44 AM, spam8mybrain via aprssig wrote:

I was re-reading some of the old postings about digipeating, where they
state that digipeaters (only) should be "impolite" and use a DWAIT of zero
to avoid channel clutter. This would seem to imply that no station can be a
digipeater and something else (weather station, etc.) unless the full
functionality is built into the TNC, because there is no way for an
external application to tell a KISS TNC on a packet-by-packet basis which
packets are DWAIT 0 and which are to have normal competition for airtime.
Either the digi would be "polite" and cause additional airwaves clutter, or
the non-digi functionality would be "impolite" and dominate the channel.

So, do any of the "soft" TNCs (AGWPE, DireWolf, etc.) have the ability to
specify priority (DWAIT=0) transmission on a per-packet basis? How does
this jive with the proposed sat-gate mode?

I'm curious because I'm doing some research on an idea for improved
digipeating, and it requires using an external program with a KISS TNC. As
such, I'd like to be able to selectively decide whether my station is
"polite" or "impolite" on a per-packet basis (assuming impoliteness is
still preferred digipeater behavior).

Any feedback would be welcome.

Andrew Pavlin, KA2DDO

_______________________________________________

aprssig mailing list

aprssig at tapr.org

http://www.tapr.org/mailman/listinfo/aprssig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tapr.org/pipermail/aprssig_lists.tapr.org/attachments/20160511/602b7844/attachment.html>