Inexpensive Satellite Does the Job
Sat, 26 Jan 2002 12:11:26 +0100 (MET)
On Fri, 25 Jan 2002, Joseph S. Barrera III wrote:
> > From: Eugene Leitl
> > Latency would suck, though.
> We clearly need more research into increasing
> the speed of light. It's really starting to
> cramp our style.
Heheh. Actually, the latency would come from how rare the flybys are, and
the need for interim store-and-forward. That is, until you come within
the range of the next box which will physically carry your packets nearer
to your destination. Otoh, as people are currently looking at MEMS mirror
galvanometers for realtime beam tracking in terrestrial applications, and
as you already know all the trajectories of pieces of hardware within your
local neighbourhood in the cloud, LoS laser would be a lot better than
wireless, though would require realtime 6DOF position tracking.
A decent geodetically routed protocol with a proper header layout would be
mostly routed cut-through (with none or a minimal depth on header FIFO,
which in photonic switching would be just a short coil of fiber after the
beam splitter -- at 10 GBps bits in vacuum are just 3 cm short), which
will let you see pure relativistic latency. (Unless there's a lot of
traffic, and you have to fall back to store and forward, of course).
Also, superluminal signalling is equivalent to causality violation, so
don't count on it without fancy tricks like spacetime plumbing. If your
biological chronon is so short that you can use light for echolocation,
the only way to keep current is to huddle.