Hi Frank,
The following article below was posted on the LUMM RB board.
I have a question regarding this statement. You will have to excuse my technical ignorance but I?ve been trying to get this question answered and no one has responded to my questions on the RB board.
"The photonics industry still has one major problem to overcome.
Multiplexing is useful now for only extremely high-capacity lines
traveling long distances because it has no effective system for
switching or routing data and telephone calls"
I know for a fact that LUMM can develop devices that can switch (active) photonic devices. I discussed this with them on my visit. A modulator was given as an example and in their web site the following products were included.
"Products include splitters, combiners, star couplers, directional couplers, gratings, interferometers, diffractive micro-optics, interconnects, cross connects, large scale photonic lightwave circuit modules, optical add-drops, dispersion equalizers, optical matrix switches, and modulators."
Notice the optical matrix switches mentioned.
Here's the questions;
1. What is considered switching/routing which the article refers to?
2. Is switching turning bits off and on. 1 for on 0 for off. Or is switching directing the channel to different directions?
3. If LUMM can do switching, is that the same switching that the article is referring to.
I would greatly appreciate you educating me on this matter.
Thanks,
Tito
***************************************************************
ragingbull.com
The electronic computer has 'Intel Inside'
The photonic computer has 'Nothing Inside'
Or so says this article from The NY Times:
www10.nytimes.com
~~~~~~~~~~~~~~~~~~~~~~
Ms. Szelag said the ephemeral nature of photons could mean an odd computer indeed, leading to thoughts of "a computer with nothing inside it."
May 27, 1999
WHAT'S NEXT
Squeezing 80 'Cups' of Data Into 1 Cup
By IAN AUSTEN
After recent crashes knocked out America Online and parts of
AT&T, there were widespread fears that ever-spiraling Internet
use would create network gridlock in the United States. Aside
from some glitches, that hasn't happened.
But the salvation of communications networks has come from the
world of theoretical physics and high-level optics rather than
electronics. Through an alchemy of sorts called multiplexing -- its
even more cumbersome name is dense wave division multiplexing --
telephone companies and Internet service providers can increase
capacity quickly while cutting costs at the same time. The secret
ingredient is light. And the harnessing of the photons of light for such
purposes, a field called photonics, may eventually change the way
computers operate, increasing their capacity as well.
"Without fiber optics and
this method of increasing
capacity, there would be no
World Wide Web," said Dr.
Erich Ippen, a professor of
electrical engineering,
computer science and
physics at the
Massachusetts Institute of
Technology. "It just enables
everything. There's really a
photonics revolution
coming."
The concept behind
multiplexing is both old and
relatively simple -- it is
based on new ways to use
fiber optic cables. The first
fiber optic systems, the backbone of both the Internet and long
distance telephone networks, sent only a single stream of
laser-generated light pulses down each thin strand of glass fiber.
Each cable typically carries 96 to 150 strands. With that much
capacity, the only problem phone companies had at first was filling
their systems. But the popularity of the Web soon eliminated that
worry and left the companies looking for even greater capacity.
To find that capacity, engineers turned to a well-known concept from
radio and television broadcast technology, the fact that a single
strand of fiber can simultaneously carry a number of different
streams, or channels, of information. The trick is that each data
stream must be sent on a different frequency -- in the case of laser
light, that means color.
For fiber optics systems, hundreds of different streams of messages
can be transmitted at once along each of the strands in a cable.
Unlike electrons, the subatomic particles that make all things
electronic possible, photons are packets of light energy.
"You can take a closet and fill it with electrons because they are
particles," said Kathy Szelag, vice president for marketing at the
Lucent Technologies Optical Networking Group. "But if you pump
photons into that same closet, it will never be too full of light. There is
no physical limit." In addition to Lucent, photonics multiplexing
systems are made by Nortel Networks, Pirelli Cables and Systems,
and Ciena.
Like many elegant theories, however, multiplexing was a strategy that
had to overcome a significant problem to be practical: as photons zip
along their cross-country travels, they begin to scatter and slow. So it
became necessary to amplify the signal in fiber optic lines. Early fiber
optic systems used a complex patchwork in which the signal was
converted from light to electricity, augmented electronically and then
changed back into light. The process was awkward and reduced the
system's maximum speed.
Electronic amplification also made running multiple frequencies, or
channels, too expensive and too complex. Each channel in a single
strand of a fiber optic cable would require its own electronic amplifier
every 25 miles or so, at a cost of about $75,000 each. Since there
are more than 100 channels per strand in the most advanced cables,
stringing more fiber optic cable remained a cheaper way to increase
network capacity.
A technological breakthrough, optical amplifiers, changed all that in
the mid-1990's, about the time the Internet began making huge
demands on networks. Optical amplifiers used newly developed lasers
to excite the photons as they passed along. Not only are they neater
in operation, but a single optical amplifier can excite a number of
different channels at once, and the resulting signals it produces travel
farther than ones that are amplified electronically. Suddenly, it was
cheaper to change how signals were sent than to lay more fiber optic
cable.
For AT&T, which installed its first photonic multiplexing system in
1995, the optical amplifiers were an answer to prayers. AT&T expects
the demand on its network to increase by 500 percent in the next five
years -- with nearly all of that increase coming from Internet traffic.
"Thank goodness this technology is growing rapidly," said Dr. George
Gawrys, the planning manager for AT&T's transportation network
division. "The alternative would have been awful."
Not only does multiplexing allow
networks to expand without adding
an inch of new fiber optic cable, but
it also saves money, allowing
further price cuts for long distance
telephone and Internet connections.
Mrs. Szelag, at Lucent, estimated
that the cost of adding capacity
using multiplexing was about
one-fifth the cost of laying more
cables.
The first system AT&T bought ran eight channels of light per strand --
all of it invisible to the human eye. The most advanced systems
currently sold by Lucent offer 80 channels per strand of fiber, which
allows each strand to carry 400 billion bits of information per second.
So each strand is capable of moving five million simultaneous phone
calls, or the contents of 600 CD-ROM's, every second, according to
Lucent.
Impressive as all that may be, Mrs. Szelag said that optical networking
was at roughly the same stage of development as electronics was
when the first Hewlett-Packard calculator came out. "Our job now is to
make photons as important to the 21st century as electrons were to
the 20th," she said.
The photonics industry still has one major problem to overcome.
Multiplexing is useful now for only extremely high-capacity lines
traveling long distances because it has no effective system for
switching or routing data and telephone calls. In a variation of the old
amplifier problem, all that takes place within specialized computers
after the photons are converted to electrons.
Some very small-scale photonic switches have been developed using
"micro-mirrors" that are controlled by electronic-based computers,
and optical switching has been shown to work on a laboratory scale.
But the industry's long-term dream is to create networks that can be
all optical. Recently, an Israeli company, Trellis Photonics, began
work on an optical switch based on crystals. NASA and a
Massachusetts company, Optron Systems, are working on similar
technology.
If such research leads to successful optical switches and routers, that
would be a major step toward optical computers, which could operate,
in theory, at very high speeds.
Many problems have to be overcome to get there. One of the most
significant problems is the fact that there is no photon storage device
to serve the function of a hard drive or floppy disk.
Optical computing, if it proves feasible, would be such a dramatic
change in technology that some people even use fanciful descriptions
to get across the point that its final form might go far beyond the
familiar.
Ms. Szelag said the ephemeral nature of photons could mean an odd
computer indeed, leading to thoughts of "a computer with nothing
inside it."
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Photon storage and photonic switches - a couple of projects for LUMM to contemplate in the future. |
