[aprssig] Precedence Bit and IP Encapsulation

[Iain R. Learmonth]

Hi,

I am planning to begin work on an AX.25 over IP interface for the HamBSD
system. This will be based on RFC1226 [0] however I am interested in
updating this document to additionally cover quality of service and
security considerations. I have a draft published and would be very
happy to receive reviews and/or comments [1].

A question that I do not have an answer to is what to do about the
precedence bit when carrying APRS over IP networks. I'm talking here
about putting APRS frames directly into IP datagrams, not using APRS-IS.

It is possible to specify the quality of service that a packet will
receive by setting the differentiated services codepoint field in the IP
header. By default, IP traffic is "best effort". In my draft, I
currently set APRS traffic to be prioritized above other traffic, except
for that which has set the precedence bit. APRS traffic setting the
precedence bit will receive best effort treatment on the IP network, and
will be dropped in preference to dropping other APRS traffic.

An alternative is to treat APRS traffic with the default best effort
treatment, and APRS traffic with the precedence bit sit as "lower
effort" [2].

The decision here will rest on what this bit is actually used for in
practice. If it's for traffic that is of little value then perhaps it is
better to drop this traffic before dropping other traffic sharing an IP
link. What is the value of APRS traffic with the precedence bit set
compared to other "routine" IP traffic?

Thanks,
Iain MM0ROR.

P.S. If you know of someone that would be interested in commenting on
this draft, please feel free to forward this message on to them. I'm
happy to take reviews/comments directly about general AX.25 usage if
they are not relevant for this mailing list.

[0]: https://www.rfc-editor.org/info/rfc1226
[1]: https://tools.ietf.org/html/draft-learmonth-rfc1226-bis-03
[2]: https://tools.ietf.org/html/rfc8622

-- 
https://hambsd.org/