From: Keith Heinemann (khmann_at_performance-data.com)
Date: 2002-01-23 10:45:50 UTC
> mallum wrote:
> Forgive my lack of knowledge, but what is an 'AP-AP mesh' ?
near as I can tell, it's a technique to let APs log into each other. It would appear similar in function to an Ad-Hoc environment where all the nodes run bridging software (and can repeat packets received from wireless back onto wireless network). A dynamic topology is created whereby any node can use any other node as a gateway. instead of a point-multipoint topology, we wind up with multipoint-multipoint. The advantage of the 'AP-AP mesh' over ad-hoc, is that it adds AP mode connectivity for wireless user STAs, which allows an ordinary windows client to roam.
In my case (rural community Internet co-op), this is huge, because I will be able to create a client-access micro-cell (repeater) by simply popping a dual-antenna card into a USR2450 or laptop. one omni antenna for client access, one directional (helical) for backbone uplink to the nearest mountain top. Prism built-in diversity should make this work just fine.
> -- mallum
> on Wed, Jan 23, 2002 at 06:44:36PM +0200, Jouni Malinen wrote:
> > Just a quick update on this.. I've started to merge changes from
> > OpenAP project. I've merged most of the changes and fixed couple of
> > bugs during this, but there is still some work (and testing) to be
> > done.. It looks like OpenAP changes do not care much about
> > Ad-hoc/Managed mode operations (i.e., they would be broken) etc. I'll
> > need to fix these and also make AP-AP mesh code configurable with a
> > module parameter etc. so that it is not used unless needed. My current
> > work version seems to work in Host AP mode. I'm thinking of releasing
> > some sort of test version after some minimal testing.
> > --
> > Jouni Malinen
> > SSH Communications Security Corp
> > jouni.malinen_at_ssh.com