Re: distributed autonomous networks

[Matthew J Ghio <mg5n+@andrew.cmu.edu>]

> A couple of us here in Australia have been kicking
> around the idea of a cheap network, running via repeaters
> in _very_ small boxes tossed all around a city, which would be
> indestructible through sheer numbers (a la the arpanet).
>   So: anyone got any ideas, schematics, information?
>
> The main idea is a huge number of small, low-power and CHEAP nodes which
> would handle through traffic which anyone with a radio modem could tap into,
> with the nodes disguised and placed in extremely hard to get to places (tops
> of telephone poles, buildings, on bridges and overpasses, etc).

Well, building the transmitter wouldn't be too hard, and you could
probably transmit much larger amounts of data than over a modem if you
used high frequencies.  The most difficult thing as I see it is that
you'd have to build a small computer for each box, and program it to
route data packets, error-check them, and so on.

> The idea being that we would have an untraceable, unbreakable network. Sure,
> you may have an address (perhaps we could set it up as a FQDN domain?), but
> that address would not correspond to any actual physical location. And, it
> would be free, aside from the cost of the equipment. No more phone calls
> to pay for!

A Fully-Qualified-Domain-Name type of network would create
complications.  Since the network will be constantly expanding and
changing, you'd probably need relative domain names. (you'd have to
specify: send packet thru site #10 thru site #34 and deliver to site #19)
To have fully qualified domain names, you'd need to have something on
the net cataloging them.