[Cu-wireless] new network design

[stephane at shimaore.net]

>         Stephane will begin an implementation for Linux and for his
>         Soekris board right away.

Yup. And the following is a first attempt at the write-up I had the silly
idea to propose to do for Sascha. Have fun!

S.


Copyright notice:

The following original text is hereby placed in the public domain to allow
for its free use by the CU-Wireless group. (The author accidentally
assumes no responsabilities for its (mis)use, nor does he claim any
representation of facts. :] )


------------------------->8 Start Here

The idea

The Champaign-Urbana Wireless project (also known as "CU-Wireless" or
"CUW" project) is until now a volunteer-based, grassroots effort aimed at
providing the foundation for a freely-accessible, affordable, and
standard-based networking solution for the entire community.


The differentiators

Although similar efforts are already underway either at the local
(PrairieNet) or national (Seattle Wireless, etc.) levels, our approach and
goals constitute a breakthrough on multiple aspects:

- Contrarily to the different grassroots wireless projects already started
in the country and abroad, our aim is not to provide access at given
so-called "hot-spots", which act as disconnected entry points to the
Internet, but to provide a dense wireless network. This approach is more
in line with so-called Ad-Hoc networks, in the line of the terminology
adopted by the IETF MANET working group.

- Beside the technology shift this new paradigm represents, the community
impact is also completely different. Where only lucky neighbors can
benefit of the service proposed by one in traditional wireless projects,
we propose to challenge the entire community to provide and support its
own development. The network will become stronger as it grows, each new
node on the wireless map becoming a supporting pillar of the overall
infrastructure.

- Finally, this project exploits the best technology available today in
terms of both performances and price. For the same cost, individuals can
be linked to the backbone at access speeds exceeding by three orders of
magnitude those of a modem. Although the infrastructure will require some
investment, its price per end-station is extremely low compared to the one
required to support a modem, DSL or cable solution. More importantly, the
use of FCC unlicensed spectrum means that no recurring charges are
applicable. The infrastructure maintenance costs are limited, while the
density of the deployment implies that we can offer no "single-point-of-
failure" up to the end-user, something which is unthinkable in wireline
environments.

We strongly believe that the technology, cost, and community choices of
the project are clear differentiators that make it a precedent at the
international level.