Japan`s emerging technologies: femtosecond lasers and plastic fiber
Japan`s government and major electronics companies are currently pursuing two emerging technologies--femtosecond lasers and plastic optical fiber, as well as related light sources. The first technology is a long-term national project supported by the Agency of Industrial Science and Technology, its national research laboratories and the electronics industry.
The second technology is supported by members of the ATM Forum--a worldwide industrial group that supports the standardization of asynchronous transfer mode (ATM) links--who predict a major market for low-cost plastic optical fiber in private ATM local area networks within the next few years. Japanese members of the ATM Forum include NEC Corp., Asahi Chemical, Mitsubishi Rayon, Sony, Toray and Toshiba. Among the Forum`s U.S. members are Boeing, Cisco Systems, Packard Hughes and AMP Inc.
Basic research into femtosecond technology--the creation of ultra-fast optical pulses in the femtosecond region of 10-15 to 10-12 seconds--began in fiscal 1993, as part of the industrial science agency`s Frontier Research Program. To date, it has been a lackluster effort to raise the level of basic research in Japan. The Japanese government, however, quickly detected the potential application and market for the technology and began a full-scale 10-year project.
The ultimate aim of this project is to realize terabit-per-second fiber-based information systems. According to the Agency of Industrial Science and Technology, currently, there is no promising technology for super large-volume communications that are expected in the "multimedia era."
Meanwhile, optical multiplexing technology has been developed for large-volume and high-speed fiber-based communications, but widespread implementation of this optical technology is considered difficult because response in the time domain depends entirely upon conventional electronics technology. Because various properties of light--including ultra-fast response--have not yet been fully utilized, femtosecond technology has emerged as a contender.
National and industrial research centers will develop materials and tools for this technology and will develop femtosecond laser technology to realize ultra-short pulse generation, pulse control and expansion of wavelength range. They will also develop femtosecond opto-electronic technology by researching the principles of new materials. In addition, they will develop application technologies, including ultra-fast information transmission and ultra-fast measurement by applying light-pulse and signal-processing technology in the time domain.
During the first six-year term of the project, the Agency of Industrial Science and Technology plans to allocate 12 billion yen (U.S. $120 million); in fiscal year 1995, the agency will spend 20 million yen (U.S. $200,000). The project will be handled mainly by major electronics manufacturers such as NEC Corp., Hitachi Ltd., Hamamatsu Photonics, and by national universities and research centers, including the Electrotechnical Laboratory in Tsukuba.
Historically, government-inspired research rarely provides concrete results that can be readily shifted from the laboratory to the factory floor; it does, however, improve the technical level of all participants.
Plastic optical fiber
The second technology being pursued is plastic optical fiber. The ATM Forum is considering the use of a plastic-optical-fiber-based physical media-dependent sublayer to cut costs for users who install a private ATM local area network with a new transmission line. This proposed sublayer uses a plastic optical fiber for the transmission medium and a red light-emitting diode (LED) or laser diode for the transmitter light source.
One advantage of this type of fiber, according to the Forum, is that its dielectric nature ensures electromagnetic compatibility (that is, no radiated emissions and susceptibility) regardless of data rate. This characteristic provides built-in upgradability. With copper medium, such as unshielded-twisted pair (UTP) cable, however, data rates increase with increased radiated emissions.
The second advantage of plastic optical fiber is its low total system cost. The desktop local area network environment is characterized by transmission distances of less than 100 meters and a large number of connections. So, total link cost is dominated by connectivity and transceiver costs. To date, glass fiber has not significantly penetrated the desktop network market because of its high transceiver and termination costs; Category 3 and Category 5 UTP cable has dominated the market because of lower transceiver and termination costs.
Standards based on UTP links will not provide the best cost and performance for future performance demands. In particular, the Forum explains, complicated coding and filtering circuits for suppressing electromagnetic interference and for compensating frequency characteristics of twisted pairs tend to increase the cost of the UTP interface card, if the line rate is set to the same range as that of ordinary high-speed networks.
Fortunately for Japanese suppliers of plastic optical fiber and supporting equipment, Japan and other Asian countries are well-placed to install plastic optical fiber because they have far fewer installed copper-wire-based local area networks than the United States. These countries can move directly to plastic-optical fiber-based network installation once specifications have been set.
Link designs for these fibers incorporate a large-core (0.5- to 1-millimeter) fiber that permits connector tolerances to be relaxed without sacrificing optical coupling efficiency. This fiber permits the use of low cost, plastic injection-molded connection components that substantially reduce connectivity cost over glass fiber systems.
Reducing connectivity cost
Connectivity cost is further reduced by simple cable installation procedures. Also, the proposed plastic optical fiber solution uses low-cost plastic packaged 650-nanometer LEDs (to match the low-loss window of the fiber) and silicon photodiodes that are mass-produced. At present, polymethylmethylacrylate-based plastic optical fiber has two low-loss windows at 570 and 650 nm; at 650 nm, both LEDs and laser diodes are available; however, at 570 nm, only an LED is available.
In Japan, a multivendor environment exists for laser-emitting diodes, laser diodes and silicon photodetectors, and another is emerging for plastic optical fiber. Mitsubishi Rayon Co. Ltd. (Tokyo), for example, has recently developed two types of high-bandwidth plastic optical fiber--graded-index and step-index. The graded-index technology was licensed from Dr. Koike of Keio University, in Tokyo, and has reached the stage of sample production.
Mitsubishi plans to provide samples to local area network vendors and manufacturers to enable them to review the characteristics required for data link design and to determine the final specifications. Meanwhile, the company expects that its step-index plastic fiber will meet present demand when combined with visible LEDs; it started commercial production of this fiber last June. The fiber demonstrates 156-megabit-per-second transmission over 100 meters. When a 650-nm light source is used, the fiber`s attenuation is 130 decibel/kilometers over a maximum transmission distance of 100 meters at 300 Mbits/sec.
NEC Corp. has developed high-speed light sources and high-speed detectors in the 650-nm bandwidth. Unfortunately, the frequency response of semiconductor optical devices working in the visible wavelength region is poorer than that of long-wavelength optical devices used for telecommunications systems. NEC, therefore, developed 650-nm aluminum gallium indium phosphide multi-quantum-well laser diodes for application to gigabit-per-second plastic optical fiber transmission. It also developed high-speed and low threshold-current 650-nm AlGaInP laser diodes and high-speed gallium arsenide/AlGaInP PIN photodiodes for plastic optical fiber data links. The company has also demonstrated 4-Gbit/sec pulsed modulation and detection in fiber.
The ATM Forum is now considering a two-pronged approach for standardizing the plastic optical fiber physical media-dependent layer--step-index (the more mature technology) and graded-index. Progress on graded-index plastic optical fiber, however, will depend on industry moves toward mass-production. q
Paul Mortensen writes from Tokyo.