| From: | Jeffrey Mach <mach@****.CALTECH.EDU> |
|---|---|
| Subject: | Re: The Crash |
| Date: | Wed, 4 Mar 1998 14:20:08 -0800 |
> Sourcebook which then validates a novel that Finley put out. (Of
> course it has been a while since I have read either, so I might be off,
> although I clearly remember them slamming Fiber Optics.)
Anybody actually have a copy of this to post the relevant words (or
synopsis thereof)?
> Basically, Fiber Optics (as strange as this sounds) was only recently
> rediscovered and there was a big shadow war to make sure that no one
> had an advantage over another competitor with releasing the information
> on how to do them again.
Herm.... Since optics and optical diagnostics is one of my fields of
study (the cutting edge Aeronautical research uses non-invasive laser
techniques to measure the kinds of quantities that use to have to be
measured by hardware probes, and in fact, to a large degree _has to_ when
you are talking about hypersonic research) so I am not pulling any of this
out of thin air. Fiber optics, themselves are relatively simple. There
is a little research in materials and materials fabrication that is
dealing with making fiber more flexible, pure (to avoid scattering loss),
etc. but these are relatively minor compared to the work in Digital Signal
Processing for fiber-optic data transmission. Fiber optics are about as
good as they can be and, honestly reverse engineering or even forward
engineering the technology for fiber production is relatively simple. The
truth is that basic Physics puts a cap on the data transfer you can do
through a fiber, even if it was "perfect", since single pulses of light
tend to "blur" due to diffraction as they continue through the fiber, so
for most fiber, improving the fiber is pointless. So, now, the big push
is to come up with newer and sneakier signal processing ways to coax more
data through the same fiber. That is the technology that I can see being
more vulnerable.
BTW, if you like fiber, buy into Lucent Technologies (formerly Bell
Labs) now. (Stock tip for the day)
Case in point: Lucent recently came up with an idea to increase data
transmission through a fiber more than 20 fold, and can be retrofitted
onto existing cables. The idea is to use a single diode laser (similar,
but at a higher power output than the one you would find in your CD
player) as the light source. Interesting property of diode lasers is that
you can modulate their wavelength a little (say +/-1%) by changing their
input current, even at MegaHertz frequencies. Right now, people are
multi-plexing signals by sending up to 10 different laser pulses at
different wavelengths into a single fiber at a time. At the other end,
even though the individual pulses have blurred together, they can be
separated by wavelength and de-plexed back into 10 distinct data pulses.
Breakthrough came when they came up with a much higher resolution
wavelength discriminator for their de-plexer. Then they had the idea to
simply ramp a single diode laser through a band of frequencies chopping
out a bunch of single pulses each at a slightly different wavelength, you
then have a high frequency chopper (think of it as an ultra-high speed
shutter) select out the "off" bits and let through the "on" bits. At
the
other end, all of the pulses come out as one big blob, but can then be
separated out as up to about 200 bits per pulse. (And you though just
sending one bit pulse at a time was cool.) Of course this kind of thing
would initially only be needed for the main transmission lines. _Nobody_
needs this kind of bandwidth for a personal system until say, the
invention of the cyberdeck, though I suppose the CS majors can all drool
over something. I can only imagine the kind of data-throughput they would
be capable of in 205X.
This is the kind of "signal processing" type of thing that might be
vulnerable. And maybe, to a much lesser degree work that is being done on
decreasing the wavelength of light for the lasers being used, which
reduces the blurring and allows you to send your pulses closer together.
Big news recently was that somebody had a working blue-green diode laser.
Unfortunately its power consumption and cooling requirements are pretty
nasty compared to the typical infrared diodes that everybody uses now. In
other words...I can see how a technology like Fiber Optics could be hurt
by something like the Crash, but since most of this stuff is based on
basic Physics, which I don't see being hurt by the Crash, I am doubtful as
to how big the impact would be, other than, as I said earlier, causing a
drought in research funds to move some of these things forward.
Although, I suppose I will have to grudgingly accept FASA's word if they
say "fiber optic technology was lost."
--My two yen
Jeff
