[CUWiN] advice,

[Sascha Meinrath]

Hi Bill,

On Wed, 7 Jul 2004, Bill Comisky wrote:

> hi all,
>
> We (CNT in Chicago) are going to try rolling out a CUWiN mesh for our
> community project in North Lawndale.  We've been able to get access to
> the old Sears Tower in the neighborhood, which is a 14 story tower with
> the roof at about 260 feet, see photos:
>
> http://wcn.cnt.org/photos/2004/07/old-sears-tower/

I'd read the press release you sent out -- I'm very excited to hear that
the CUWiN software will be utilized.  We're happy to help with it.

> Our plan at this point is to use a point-to-point 802.11a link from the
> roof of the Homan Square community center which is providing the
> internet connection (about 600 feet away) to the tower.  Then we're
> going to use CUWiN with sector antennas from the tower.  We have good
> line of sight to a large area from the tower.  We hope this setup will
> ultimately give us the best user capacity for the tower as opposed to
> using it as just a big access point.  Of course we will add more
> internet connected nodes as needed.
>
> I'm looking for any problems you can think of with this kind of setup.
> Is there a big hidden node problem here if several nodes can see the
> same tower antenna but not each other (forgive my ignorance of the HSLS
> algorithm!).  Also I wanted to ask about your network and if there are
> any rules of thumb you have learned for things like optimal mesh density
> (how far apart are your nodes? how many other nodes can each "see"?) or
> any other best practices we can learn from before we start.

It shouldn't be a problem if the nodes can see the tower but not each
other.  Our network has all sorts of different configurations -- our
office testbed is composed of 6 nodes that can all see each other; our in
vivo networks are composed of 2 to 6 nodes in various configurations.

The CUWiN software is stable and has been used by us for about two years.
However, HSLS and ETX (route prioritization) are not fully implemented yet
-- we expect to complete both by December 31.  Nodes will still mesh fine;
but we don't expect that the system will scale past several dozen nodes
with the current software.  I'll check with the development team tomorrow
and see if they expect any problems.

It would be helpful to know the specs of the nodes and radios you plan to
use.  We'll want to be sure that everything is compatible.  I can be
reached much of tomorrow at 217-278-3933.  Let's keep in touch so we can
ensure a successful launch of your network.

--Sascha

-- 
Sascha Meinrath
Project Manager & Pres.  *  Project Manager        *  Project Coordinator
Acorn Worker Collective *** Eggplant Active Media *** CUWiN
www.acorncollective.com  *  www.eggplantmedia.com  *  www.cuwireless.net