[Olsr-users] Malformed HELLO message?

[Henning Rogge]

Am Donnerstag 25 Februar 2010 16:15:21 schrieb Pavel Boyko:
> On Thursday 25 February 2010 05:49:22 pm Henning Rogge wrote:
> > >   Now I'm trying to predict network behavior using simulation and my
> > > 
> > > simulator, unfortunately, doesn't support link quality extension.
> > 
> > If your simulator says "it works", don't believe it. ;)
> 
>   Could you provide some references?

I can provide you with the practical experience/feedback of the 
Freifunk/Funkfeuer OLSR networks (which are some of the largest mesh networks 
running real traffic) and with a similar advice from meeting most of the 
creators of the OLSR RFC.

*** Hopcount metric DOES NOT WORK well for wireless mesh networks. ***

They work well in simulator with Tworay-Ground models, but that's only because 
this model is not good.

> > OLSRd cannot handle MPR selections because of a problem in the dijkstra
> > algorithm. You will not get well behaving routes.
> 
>   Do you talk about a bug (where is a bug tracker after all?) or about some
> fundamental limitation of the dijkstra algorithm?
It's a bug in the dijkstra implementation of OLSRd 0.5.6. It is solved in the 
development branch, but it's too complex to port it back easily.
 
> > Thats why all Freifunk/Funkfeuer networks use "MprCoverage 7" at the
> > 
> >  moment.
> 
>   It is sad. Am I understand right that even with LQ enabled I need to set
> MprCoverage to 7?
Yes.

Henning Rogge
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
URL: <http://lists.olsr.org/pipermail/olsr-users/attachments/20100225/dc5baa68/attachment.sig>