Talk:Optimized Link State Routing Protocol
||This article may be too technical for most readers to understand. (September 2010)|
what is olsr? how olsr works ? what about ad-hoc ?
Small computers - problems or not?
The article suggests that OLSR doesn't work on smaller computers because of the overhead of continuously checking routing etc. Yet it is part of the Freifunk firmware that has been designed to run on Linksys WRT54G's , and seems to do so successfully. Can any experts offer some more information? --mgaved 21:44, 27 June 2006 (UTC)
While OLSR can have issues with limited computer resources in large networks with fast update rates, OLSR should work fine on less powerful machines within smaller networks (~10-15) without issues. The TC and HELLO interval rate can be slowed down in slower moving networks to reduce the computational power and overhead needed to maintain network connectivity. The biggest differences with OLSR when compared to AODV is overhead distribution. OLSR's, and other proactive protocols, overhead is relativity fixed based upon number of MANET nodes, and TC/HELLO timer intervals. AODV, and other reactive protocols, overhead depends almost exclusively on application traffic patterns. Bebemaster 14:00, 15 May 2007 (UTC)
I found this article hard to follow. It needs more basic information before the detailed discussion of critiques and responses. Any objection to my reorganizing it? Thefamousmred (talk) 20:03, 26 August 2008 (UTC)
I agree. To me a reorganization with more details/history and less its good/no it sucks content would be helpful. I would be willing to contribute to some sections but don't have time to rework the layout. —Preceding unsigned comment added by 184.108.40.206 (talk) 17:09, 26 September 2008 (UTC)
The OLSR-NG section must go
The article is about the OLSR routing algorithm. The OLSR-NG section is about a particular implementation of OLSR (the unik-olsrd daemon). This section must go.Jec (talk) 03:40, 16 December 2008 (UTC)
Jec: I would like to have you rephrase the OLSR-NG section instead of removing it altogether. The OLSR-NG project has shown that it does run with hundreds of nodes on linksys WRT54GL. So instead of denying it (by deleting the section) I would have rather seen a rephrasing of the article that a particular implentation can indeed withstand the regularly repeated criticism and therefore it is not a fundamental flaw of the protocol but rather of prior existing implementations that led people to the critique. Can you rephrase the OLSR-NG section so that you feel happy with it? thx! --220.127.116.11 (talk) 17:05, 11 May 2009 (UTC)
- I stand by my opinion that the OLSR-NG section doesn't belong in this article. OLSR-NG is just an implementation of OLSR that doesn't use completely brain-damaged data structures, and hence scales moderately well (a few hundred nodes in a single routing domain is really not that big a network, and a WRT54-GL has a fairly powerful processor). --Jec (talk) 22:03, 29 August 2009 (UTC)