Chappell Brown
Bell Labs is known for revolutions.
wiredbrain.com
In 1947 it was the transistor. Today it is photonics. Called the second silicon revolution, optical fiber systems are in an explosive state of development, reminiscent of the earlier days of the electronics industry.
It's a major revolution riding on a broad-based industry serving the fundamental human need to communicate.
individual.com
Over the past two decades, since fiber-optic communications first began to appear, the carrying capacity of fiber has increased at a faster rate than Moore's law. Now the wavelength-division multiplexing revolution has accelerated that capacity even more, while introducing the flexibility of wavelength-based routing. Forged from an interdisciplinary mix of semiconductor diode lasers, micromachine technology and fundamental advances in optical glass technology, terahertz networking has arrived well ahead of schedule.
"A length of fiber long enough to circle the globe three times is produced every day, and if you extrapolate current trends to 2010, every one of the 6 billion people on earth will have a bandwidth capability equivalent to high-definition television," said Alistair Glass, director of photonics research and development at Lucent Technologies Bell Laboratories. Arriving at Bell Labs in 1967, Glass' career spans the development and implementation of fiber-optic communications systems.
Major breakthrough
"When I arrived, the major breakthrough was the first continuously operating laser, and it didn't run for very long-only a few minutes," Glass recalled.
"This was the time of the early hero experiments and the demands kept increasing and increasing on these devices. There was always that pressure, but the interest in the marketplace represented a dramatic change."
There was always a strong demand to increase the performance of any device.
At first the research arm of AT&T, Bell Labs enjoyed a special status after its founding in the 1920s. Because of the monopoly granted AT&T by the government, in the interests of standardizing the telephone system, the lab could both be part of a commercial operation and play the open role of a national laboratory.
"At that time, there was not much connectivity with business- it was very much intellectually driven. We wanted to be leaders in all the fields relevant to communications," Glass said. But in the early 1980s two developments dramatically accelerated photonics research: commercial long-haul fiber-optic systems began to be installed commercially, and AT&T's monopoly was dissolved by the government, with parts of Bell Labs spun off into other companies as part of a complex divestiture of the telecommunications giant. "We were suddenly handed the mandate to develop commercial products out of our research efforts," he said.
The lab responded with a broad attack on optical communications systems. Innovations in the basic fiber, laser diodes to power them, and integrated optoelectronic components to interface with electronic data systems followed. "Since then, particularly with the founding of Lucent Technologies, optics has been accelerating at an incredible rate," Glass said.
For transporting data over long distances, fiber systems proved to be irresistible. Large bundles of copper wire could be replaced by slender silicon fibers in a process of "demassification" usually associated with the electronics industry. While the debate continues over whether optical interconnect is a viable alternative to electrical wiring inside of computers, the issue has been definitively resolved for long-distance communications. But optical interconnect inside the box may eventually succumb to a long-term trend. Recent developments in metropolitan-area networks suggest that fiber optics is riding a scaling law similar to the shrinking VLSI circuit, and the scaling rate appears to be steeper.
The rapid deployment of fiber optics received an even bigger jolt with a repeat of the '80s scenario in the 1990s. Bell Labs was again transferred in 1996 to another entity-Lucent Technologies-and made the centerpiece of a startup with considerable economic resources. Also brewing in photonics labs was a revolutionary technology called dense wavelength-division multiplexing (DWDM), which has allowed the carrying capacity of optical fiber to ramp up at an astonishing rate. "In the mid-90s it became a fever. We went from eight to 16 to 32 wavelengths on a single fiber and our latest products use 400. Now we have just demonstrated 1,000 wavelengths," Glass noted.
DWDM uses individual segments of the optical spectrum to multiplex signals on a fiber. The idea is recent, considered at first to be a laboratory curiosity since practical systems were already multiplexing channels with a time-division technique. Such synchronous optical networks (Sonet) had been able to extend the capacity of optical fiber and were a welcome development. The wavelength-division multiplexing route has turned out to have far more potential: Bell Labs researchers recently demonstrated a DWDM transmission system capable of sending a terabit of data per second down a fiber. "That represents the entire world's Internet on a single glass fiber," Glass said.
The DWDM revolution has been extremely swift. When Lucent Technologies was established, DWDM was still at the laboratory demonstration stage. While the idea is simple, turning it into practical optical communications systems required a multifaceted development. Multiple-wavelength laser-diode systems and new types of fiber able to carry the multiple wavelength signals without crosstalk had to be developed. And some means of collectively amplifying multiwavelength signals had to be invented. While those problems were effectively solved in a short time, it wasn't easy. Indeed, one outstanding problem has never been solved: how to regenerate multiple wavelength signals.
Large areas
One consequence of that missing solution is the fact that DWDM can only be implemented on campus-wide or metropolitan areas. By doping fiber with the rare-earth element erbium, it is possible to build a simple light amplifier that is essentially a laser. When a multiple wavelength signal is passed through an erbium fiber loop and optically pumped, it emerges unchanged except that it is at a higher energy level. One nice aspect of this operation is that the actual content of the wavelength channels is irrelevant to the amplification process. Unfortunately, to recondition optical signals, it becomes necessary to decode their content and relaunch them. Thus signal regeneration, which is essential in long-haul networks, is still unavailable to DWDM.
Balancing this deficiency in very long transmissions is a new wave of all-optical switching elements that are able to add or remove a wavelength channel from a fiber. These add-drop multiplexers offer a high-speed switching function that could not be duplicated with electronics, and have made metropolitan-area networks into a unique flexible, high-throughput communications medium.
This essentially new form of photonics technology is spawning an industry in optical switching components. "Now people can invent a novel device that relates to communications and it will find its way into products extremely rapidly-less than a year," said Glass. "We are now in a situation of 'invent on demand' where as soon as a problem is perceived, someone immediately comes up with a solution."
This explosive growth poses a formidable challenge to electronics technology. "If you compare the speed of silicon chips versus the capacity of optical fiber communications, fiber optics is going significantly faster than electronics, and where the fiber ends-that becomes a significant bottleneck." Glass is convinced that fiber to the home office and then fiber to the home are just around the corner. "We have a demonstration project going with Bell South where we have wired up a suburban neighborhood with little fiber-optic network units on the side of each house," he said.
Dealing with the high volumes of data that are coming off optical fibers will present a big challenge to electronics. Fortunately, wavelength-division multiplexing eases that task since each wavelength can be processed simultaneously by different circuits. Ultimately, electronics and optics technologies offer complementary abilities: "Optics is ideal for transporting data from point A to point B, but it is weak in the area of logic and switching," Glass pointed out. "That is where we will need electronics."
Copyright c 2000 CMP Media Inc.
By Chappell Brown
The world economic summit is less interesting because the big and powerful
are less interesting.
The rate of technological has multiplied on itself because computers can work faster and communications are better therefore computers and communications becomes faster and faster. My guess is that optic fiber to the door will make on-air or cable broadcasting uneconomic - video on demand will replace it - the program producers will distribute directly to the consumer - like in MP3 - the video store goes on line - The move producer - such as Blair Witch could be sold directly - same with any show or news or whatever - so there goes networks - maybe even magazine writers with direct sales -
Wireless systems can get up to 400 kps to a million somehow - wiredbrain.com for a lot of applications that is fine - and OS chip technology will make greater use of less and less with less energy and heat - more light and lighter -
code division multiple access (CDMA) technology.
HP is investing $2 million in New Media Venture Partners (NMVP) and will
provide up to $15 million in debt financing to help the company fund and
incubate e-commerce start-ups. In return, subsidiaries of NMVP will use HP
products and services.
If I were a high technology company - in information systems, computers,
communications or any part of the 25 % of the economy - and almost all the growth sector - now including networks - broadcasting - publishing - entertainment - music - video - electronics - service - I would have a venture capital connection so I could send people out and find out what is going on. The battle for the airwaves is not just about broadband but the content - software and services. If you put a few hundred thousand in interesting technologies you gain access to information. There is almost a certainty that something will come from left field and change all the rules again.
Cable is too slow and greedy. The telephone companies too slow and bureaucratic. Both have shown a preference for short term gains rather than long term survival. Microsoft is showing the same brain arthritis - inflexible - such as IBM was - GM and other big and rich - missed every important technology - but could buy it after it had been proven. That may or may not be possible. .
wiredbrain.com
The most common wireless transmission standard, GSM, which stands for
Global Systems for Mobile communications, is particularly prevalent in Europe and Asia. According to market research firm Dataquest, nearly 157 million GSM-based mobile phones will be shipped worldwide this year,compared with shipments of about 43 million CDMA cell phones.
But many industry observers say CDMA, strongest in North America, is moreefficient and can handle Internet-based transmissions better.
There is also time division and dense systems - I do believe the key is China - the PLA and post telegraph - along with the EU will set the standards.
--
|
