[Olsr-dev] Some reflections on link state routing and metrics

[Markus Kittenberger]

i don`t think logarithmic scaling really solves the problems, it just makes
log(problem) *g

of course multiple things must be considered in a metric

but i see 2 main goals

- independent of what else the metric is based on packetloss on olsr packets
MUST not occure on relatively low cost links,..

taking this into account, multiplicating packetloss other values (og
potentially higher ranges) is no very good idea

otherwise we WILL for SURE have BIG routing problems, but of course the
severity of problems can be reduced if cost range of metric is smaller,..
(like ETX for example, which only quite seldomly HAS routing problems due to
this)
or we make sure there is no OLSR-packetloss (on low cost links)
(IMHO the better approach)

- metric cost should reflect real word usability of the link (wtf is reals
world usability *g)
otherwise people will be unhappy again
of course there are many ideas to throw here in (which als take into account
airtime, and so on)

one attempt may be hypotethical tcp bandwidth (which takes into account
availiable netto bandwidth, latency, and unicast packet loss)




On Mon, Feb 9, 2009 at 1:45 PM, Henning Rogge <(spam-protected)> wrote:

> Am Monday 09 February 2009 13:12:57 schrieb L. Aaron Kaplan:
> > So, I believe it is an interesting discussion which was started with
> > the range of possible metric values.
> > What about using metric values from different sources such as
> > effective bandwidth etc but on a limited range?
> we could use something like a logarithm on the available bandwith before
> apply
> it the the metric value. This way larger bandwidth will mean better
> linkcost,
> but highspeed links would not dominate the metric easily.
>
> (a logarithm base 2 would be easy to implement in integer arithmetic)
>
> Henning
>
> *************************************************
> Diplom Informatiker Henning Rogge
> Forschungsgesellschaft für
> Angewandte Naturwissenschaften e. V. (FGAN)
> Neuenahrer Str. 20, 53343 Wachtberg, Germany
> Tel.: 0049 (0)228 9435-961
> Fax: 0049 (0)228 9435-685
> E-Mail: (spam-protected)
> Web: www.fgan.de
> ************************************************
> Sitz der Gesellschaft: Bonn
> Registergericht: Amtsgericht Bonn VR 2530
> Vorstand: Prof. Dr. rer. nat. Maurus Tacke (komm. Vors.), Prof. Dr.-Ing.
> Joachim Ender (Stellv.)
>
>
> --
> Olsr-dev mailing list
> (spam-protected)
> http://lists.olsr.org/mailman/listinfo/olsr-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.olsr.org/pipermail/olsr-dev/attachments/20090209/04d8e358/attachment.html>