Click for next page ( 10


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 9
2 i Opportunities in Telecommunications INTRODUCTION Telecommunications is a big business. The market for network equipment purchased by companies in the United States that offer public network telecommunications services (e.g., local exchange carriers, long-distance interexchange carriers) exceeds $20 billion per year. When one considers the market for all telecommunications equipment in the United States, the number is substantially larger. The total world market is several times greater than the U.S. market. Thus, depending on what one defines as telecommunications equipment, the total worldwide market exceeds $100 billion per year. Up until the late 1970s, telecommunications systems depended on advances in electronics to provide new capabilities, increased performance, and lower costs. Until that time, the media used for transporting information from place to place were copper cables and radio (including terrestrial microwave links and satellites). As f~ber-optic technology emerged from the research lab and field experiments and entered large-scale deployment, the impact of this photonic technology on telecommunications was dramatic. For long-distance (>100 miles) and moderate-distance (5 to 100 miles) point-to-point applications, fiber displaced copper cable and radio as the medium of choice. Today, the deployed U.S. base of long-distance and moderate-distance point-to-point fiber systems has an information carrying capacity (e.g., measured in equivalent voice circuits or in bits per second) far in excess of the total deployed base that existed in the late 1970s. Thus in less than 10 years, this technology has revolutionized many 9

OCR for page 9
10 PHO TONICS transmission aspects of telecommunication networks and has had a very big impact on the competing technologies by displacing them. In this chapter the panel examines a number of different telecommunica- tions applications environments and the potential of emerging photonic technologies to either facilitate those applications or displace existing tech- nologies currently used in those applications. TELECOMMUNICATIONS APPLICATIONS: DESCRIPTIONS AND TECHNOLOGY STATUS Long-Distance and Moderate-Distance Point-to-Point Connections In the late 1970s f~ber-optic technologies emerged from the research and field-experiment phases and were deployed on a large scale in the transmission systems that interconnect telecommunications switching systems and provide the formation-carrying capabilities used by customers attached to networks employing those transmission and switching systems. Fiber offered the ability to transmit large amounts of information over long distances without requiring as many repeaters to remove the effects of loss and distortion. This was in contrast to existing metallic cable and radio facilities, which have limited information-carrying capabilities for a variety of physical reasons. The impact of fiber was twofold. The ability to transmit information for long distances without repeaters or with fewer repeaters (and without the interference and security problems of radio) made it possible to build transmis- sion systems with lower initial equipment costs, lower right-of-way costs, and lower maintenance costs. The large information-carrying capability of fibers is a latent attribute that can be called on in the near future to economically transport video signals that can make use of that capability. Since fiber was first employed in long-distance and moderate-distance applications, the marketplace (purchasers of transmission equipment) has demanded products with ever- increasing information-carrying capability and ever-increasing distances that can be spanned without repeaters. The technology has passed through three generations since the late 1970s, starting with multimode fiber and 800- to 900-nm wavelength light, then moving briefly to multimode fiber and 1300-nm wavelength light, and now using single- mode fiber and 1300-nm wavelength light. The promise of lower light losses points toward the use of 1500-nm wavelength light in the future. In addition, the promise of more sensitive receivers (and therefore longer achievable spans without a repeater) points toward the potential future use of coherent technologies.

OCR for page 9
QPPOR7rUNITIES IN TELECOMMUNICATIONS 11 Long repeaterless spans are important because of the costs of locating, powering, and maintaining repeaters, in addition to the cost of the repeaters themselves. This is particularly true in long-distance terrestrial and undersea cable systems. In undersea cable systems, reliability and power consumption are, in addition, more critical than in terrestrial applications. A figure of merit for point-to-point transmission is the product of repeaterless span and information-carry~ng capability. The product of repeaterless span and data rate in commercially available equipment has increased from 5 km x 45 Mbits/s in 1979 to over 40 km x 560 Mbits/s today. This represents an increase by a factor of over 100 in 8 years (approximately doubling every year) for deployed products. This progress is expected to continue. Commercial products operating at nearly 2 Gbits/s have been announced. In the laboratory, systems have been demonstrated with (simultaneously) data rates of several gigabits per second and repeaterless spans beyond 100 km, for a data rate times distance product more than 10 times that of the commercially available equipment cited above. Local Area Networks In the context of this report, a local area network is defined as a com- munications interconnection system deployed in an office, a factory, or a multi- building campus (distances typically < 1 mile) to allow computing devices and peripherals to exchange information in electronic form via a shared networking facility. Local area networks emerged about a decade ago in response to the proliferation of moderate- and low-cost computing devices (and peripherals) and the need to interconnect them. Early local area networks were based on copper-cable media (e.g., coaxial cable). Because local area networking is a relatively young concept, there is still much to be learned about requirements that local area networks should meet in various applications. For example, even the best choice for the physical layout of the local area network cabling and electronics is still not completely understood, although much has been learned as a result of early installations of various designs. A number of alternative fiber-optic-based local area networks have been proposed, designed, or deployed in recent years. These have various physical topologies (star, ring, bus), various capabilities (peak data rate, delay, number of accessing computing devices that can be accommodated), and various target applications. The components used in these local area networks include many of the components used in point-to-point telecommunications applications plus some additional components unique to the local area networking application. Examples of these additional components are access couplers, which allow light to be partially added or removed from a fiber at an access point; star couplers,

OCR for page 9
12 PHO TONICS which allow light arriving at an input to the star coupler to be split amongst several outputs of the coupler; and optical switches, which allow remotely controlled reconfiguration of the network for maintenance or rearrangement of connectivity patterns. Low cost and high reliability are required in local area networks with a large number of access ports on the network to which relatively modest cost- computing devices (e.g., terminals) can be attached. Since the computing devices to be attached might cost only a few hundred or a few thousand dollars each, and since there are more places to attach computing devices than there are attached devices themselves, each access port must be very inexpensive compared to the cost of the accessing device. Local area networks with many access ports require high reliability of individual components in order to provide acceptable maintenance costs and acceptable system downtime. Local area networks that have only a few accessing devices or accessing devices that are relatively expensive (e.g., minicomputers and specialized peripherals) can tolerate more expensive components, if that is the only alternative. Metropolitan Area Networks Metropolitan area networks, in the context of this report, are connectivity systems for allowing communication between geographically dispersed local area networks and isolated terminals (typical distances are 1 to 10 miles). They have characteristics similar to those of a local area network. Typically, information carried between points on a metropolitan area network can share transmission paths with other types of telecommunications traffic. Costs of optical components are not as critical in metropolitan area networking because of the larger numbers of computing devices sharing the use of those components (the local area network concentrates traffic onto the metropolitan area network). However, given equally good technical alternatives, the least expensive components will be selected. Because of the potentially large amounts of information traffic carried by metropolitan area networks, higher- speed components and wavelength/frequency multiplexing technologies are increasingly desirable. Metropolitan area networks are early versions of the broadband integrated service digital networks (for business applications) described in the next section. Broadband Integrated Service Digital Networks The concept of a broadband integrated services digital network (BISDN) is to provide a high-bit-rate communications transport capability to a community of unrelated users for voice, data, image, and video communications. It is the

OCR for page 9
OPPORTUNITIES IN TELECOMMUNICATIONS SINGLE MODE FIBER 2.488 Gb/s TRUNK HIGH SPEED _ SIGNAL _ SWITCH (155 Mb/s) 'M _ _ M' . E/O u l r , E/O . l LOW SPEED SIGNAL SWITCH (64 Kb/s) . CENTRAL OFFICE OR UNATTENDED REMOTE UNIT FIGURE 2.1 Setup for a typical office BISDN. SINGLE MODE FIBER 622 Mb/s RADIO PORT 13 u TELEPHONE ~ DATA VIDEO ~ GRAPHICS ~ AUDIO TELEMETRY ~ OTHER PORTABLE VOICE OR DATA CUSTOMER LOCATION E/O = ELECTRICAL-TO~OPTICAL CONVERTER MUX = MULTIPLEXER/DEMULTIPLEXER vision of the future. Fundamental to such a concept is bringing fiber to businesses and residences since copper wire cannot carry the required data rate. Figure 2.1 shows a typical realization of a BISDN. Various high- and low-speed digital signals are combined by electronic multiplexing in a central office or remote (unattended) switching unit and are converted to optical form for transmission to the customer over an optical fiber. At the customer's premises the optical signal is converted back to electronic form, and its various informa- tion components are distributed to appropriate terminals. There are now over 100 million copper-w~re access lines in U.S. telephone networks. If 1 percent per year of these were converted to fiber, this would amount to a demand for millions per year of optical transmitters, receivers, and other optical components as well as millions of kilometers per year of fiber. If the installed cost of a fiber access line was initially about $3000 and declined toward $1000 as production volume increased and technology improved, then converting 1 million copper- wire access lines per year would represent a $1 to 3 billion market. If the changeover rate grew to 5 percent a year, then the market would be $5 to 15 billion in the United States alone. The total U.S. market (100 million lines) for conversion is more than $100 billion. Thus the BISDN marketplace represents a very important application for photonic technologies. Key to this marketplace will be low-cost fiber cables, low-cost transmitter and receiver modules, and possibly novel approaches to the distribution network architecture to reduce the need for active electronic components in the outside plant (in unattended locations). Also key to the timely deployment of BISDNs in the United States is an appropriate regulatory and long-term investment climate that will

OCR for page 9
14 PHO TONICS encourage the large capital investments required and will allow BISDNs to be deployed in an economically efficient manner. There is a worldwide race to develop the necessary technology. Photonic Technologies Within Equipment As telecommunications moves toward the higher data rates associated with BISDNs, the ability to perform functions such as multiplexing, switching, and internal component interconnects at these high data rates becomes a limiting factor in cost and practicality. Higher-speed electronic circuits consume more power, which creates the thermal management problem of keeping components from overheating. Higher-speed electronic circuits also experience bottlenecks in the flow of electronic signals between circuit elements because of the limited bandwidth of metallic interconnects, even very short-distance interconnects between circuit elements on a monolithically integrated circuit. A question often raised is whether photonic (optical) devices can help to resolve these problems in a way analogous to f~ber-optic technology's opening up almost unlimited bandwidth in transmission systems. Various scientifically interesting photonic components can perform a switching or routing function in the laboratory, and there has been much speculation as to how these components might revolutionize the capabilities of switching and computing systems. However, the large technological gap between these laboratory devices and practical systems needs to be filled by scientific breakthroughs as well as by engineering. When these gaps will be filled is an open question. There are, however, some near-term possibilities. Very often it is the interconnects between components that represent the most troublesome engineering problem in equipment operating at high data rates. Here optical interconnects in the form of optical waveguides (fibers) on plug-in boards and backplanes, accessed directly by components with optical input-output capabilities, show much promise. Optical interconnects between plug-in boards or arrays of components based on free-space optics (arrangements of lenses and mirrors) also appear attractive. Fiber interconnects between backplanes are becoming increasingly popular. Small numbers of incoming and outgoing optical fibers may be rearranged by interconnecting them with optical switching devices that act as a remotely controlled patch panel. Such remotely controlled patch panels have possible applications for protection switching, which allows a network of interconnected fibers to recover from a fiber or equipment failure, or for rearranging the connectivity pattern of fiber networks in response to the physical movement of users.

OCR for page 9
OPPORTUNITIES IN TELECOMMUNICATIONS Photonic Networks 15 Today's telecommunications networks have architectures that evolved to conform to the capabilities of the technologies--copper cable and radio transmission systems, and discrete electronic components--available when these networks were being designed and developed. With the advent of fiber optics, advanced optical components, and highly integrated high-speed electronic cir- cuits, one can ask whether a telecommunications network should retain the existing architecture and use modern technologies to substitute for older technologies or should adopt a new architecture tailored to derive the maximum benefit from the new technologies. New architectures that have been suggested tend to capitalize on the ability of fibers to carry very large amounts of information with very little marginal cost once a fiber is in place. They tend to push switching toward the edges of the network: large amounts of information are distributed throughout the network, and the desired information is selected, as needed, by the terminals connected to the network or by gateways at the edges of the network that stand between the network users and the network. These approaches often build on optical wavelength-division multiplexing (WDM) or optical frequency-division multiplexing (coherent techniques) to deliver these large bundles of information throughout the network. It remains to be determined whether these architectural concepts will be preferable to architectures with more switching internal to the network. Military Telecommunications Military telecommunications applications for fiber-optic systems closely parallel the commercial applications discussed previously. For example, there are fixed land-based applications for high-data-rate, long-distance, point-to- point links. There are land-based, shipboard, and airborne applications for local area networks for computer interconnects and for BISDNs to support combinations of voice, data, video, and other signals. Certain attributes of optical fiber systems have greater importance in military applications than in commercial applications. Among such attributes are immunity to electromag- netic interference, relative security from eavesdropping, spanning of long distances without electronic repeaters, and low cable weight. Military applications also impose additional requirements on fiber-optic systems, e.g., wider operating and storage temperature ranges; ability to withstand severe vibration, shock, and other mechanical stress; and robustness of system performance in the presence of multiple simultaneous subsystem failures. Many of these additional requirements are satisfied with appropriate packaging and other physical design measures as well as with appropriate system design,

OCR for page 9
16 PHO TONICS without any direct implication for the photonic devices. However, certain re- quirements place the fiber and photonic components in the critical path of viable applicability and lead to photonic technology challenges. An example is the design of optical fibers and cables that are resistant to radiation, can operate over a wide temperature range without excessive attenuation increase due to a phenomenon called microbending, are capable of sustaining high strains without breaking (e.g., for fiber guided missiles), and can withstand severe chemical environments (a requirement not necessarily more stringent than some for com- mercial applications such as oil well logging cables). Another example is the design of optical emitters and detectors that are radiation resistant and can operate in high-temperature environments. On balance, however, the requirements of the military applications for fiber-optic systems are being addressed successfully. The only key enabling technology to tee emphasized, one that is needed for commercial applications as well, is a high-reliability optical transmitter and receiver module that can operate over standard military specification ranges of temperatures with failure rates of less than 1 per million device hours. For commercial applications low cost is also critical but should follow from high-volume automated production. ENABLING TECHNOLOGIES All of the important technologies that are required for successful exploita- tion of the markets identified in the section "Telecommunications Applications" can be divided into two categories: (1) technologies requiring continued or increased emphasis on development in the next 5 years so that U.S. corpora- tions can maintain or strengthen their competitiveness in existing but evolving markets, or can ensure that they will compete successfully in emerging markets, and (2) technologies that are enabling but are underdeveloped so that increased or continued research is appropriate at this point. In this report these two types of technologies are referred to as enabling technologies ripe for development and enabling technologies requiring continued research. Long-Distance and Moderate-Distance Point-to-Point Connections Enabling Technologies Ripe for Development Fiber cables with low loss and low dispersion (variation of group velocity with wavelength) at 1.3-micron wavelength, and with low microbending loss and high strength. Note: to date the United States has demonstrated competitive- ness in this technology, and there are several U.S. companies currently

OCR for page 9
OPPORTUNITIES IN TELECOMMUNICATIONS 17 marketing high-quality, low-cost fiber cables. Continued development of manufacturing technology is needed to reduce costs in this increasingly competitive market. ~ Fiber cables with low loss and low dispersion at 1.5-micron wavelength, and with low microbending loss and high strength. Note: U.S. fiber companies appear to be competitive in this emerging technology. Transmitter modules with high performance in terms of output power into an integral single-mode fiber, single-frequency (longitudinal mode) operation, narrow line width, low chirp (frequency shift under modulation), long lifetime in an unconditioned ambient temperature, high-modulation-speed capability, tolerance to elevated ambient temperatures for short durations without disruption of performance, and low power consumption. Note: although U.S. companies have thus far been able to produce transmitters containing lasers for systems deployed to date, the increasing performance demands of the marketplace combined with the generally acknowledged lead of the Japanese in laser technology suggest that escalated development effort will be necessary to retain U.S. competitiveness. It is believed that this is a development challenge rather than a research challenge. (In this context "development" focuses on manufacturing technology as well as on design for low-cost manufacturing.) Receiver modules with high performance in terms of sensitivity, dynamic range, bandwidth, and linearity. Note: although U.S. companies have thus far been able to produce receivers for their systems that result in reasonably competitive system performance, non-U.S. companies (e.g., the Japanese) have been aggressive recently in developing high-sensitivity, high-dynamic-range receivers that appear to outperform U.S. counterparts. Since much of the research in low-noise, long-wavelength avalanche photodiodes (APDs) and in low-noise amplifiers originated in the United States, it is suggested here that increased development effort is needed to retain U.S. competitiveness in this key technology. ~ Passive components for wavelength multiplexing and demultiplexing and for related technologies such as transmitters with stable and predictable wavelengths. These are needed to deploy wavelength-multiplexed systems with moderate numbers of concurrent wavelengths. Enabling Technologies Requiring Continuing Research Transmitter and receiver subsystems incorporating very narrow line width, single-frequency lasers for coherent communications systems (coherent detection). New types of ultralow-noise APDs.

OCR for page 9

OCR for page 9
OPPORTUNITIES IN TELECOMMUNICATIONS Photonic Technologies Within Equipment Enabling Technologies Ripe for Development 19 ~ Low-cost, high-reliability transmitter and receiver modules for short- distance point-to-point interconnections between boards and shelves to replace metallic cable connections. Note: although a number of U.S. companies successfully market transmitter and receiver modules for point-to-point interconnections within equipment and between pieces of equipment, continued diligence to increase reliability and reduce cost is required as this market grows. Enabling Technologies Requiring Continued Research Optical traces ("conductors") on circuit boards and backplanes to replace conventional metallic traces. Optoelectronic assemblies to interconnect to optical traces on circuit boards and backplanes (transmitter and receiver components integrated with electronic devices). Free-space optical interconnects. Optical transmitter and receiver arrays for parallel interconnects. Photonic Networks Enabling Technologies Requiring Continued Research Tunable optical transmitters and receivers. Optical amplifiers with gain over a wide wavelength range. Photonic switching devices and subsystems. Coherent communications technologies. Military Applications Enabling Technologies Ripe for Development Ruggedized versions of commercial technologies with increased toler- ance to physical abuse (e.g., temperature range for storage and operation, vibration, shock, corrosive environments, and radiation).

OCR for page 9
20 Enabling Technologies Requiring Continued Research PHO TONICS ~ Ruggedized versions of commercial technologies indicated previously as candidates for continued research. Higher-performance versions of commercial devices (e.g., ultralow-loss optical fibers, higher-power optical transmitters, ultrasensitive optical receivers) that may not be practical for high-volume or widespread commercial deploy- ment but that may be practical for specialized military applications with high tolerance to cost. THE IMPORTANCE AND IMPACT OF STANDARDS AND MODULARIZATION in commercial point-to-point telecommunications applications of f~ber- optic technologies, network providers have procured various components of these systems separately. By taking advantage of standard or open interfaces (e.g., interfaces with publicly documented and stable specifications) between major components of a system, network providers can purchase optical fiber cables from one manufacturer, optical line and span-terminating repeaters from another manufacturer, and multiplexing equipment from yet another manufac- turer. This approach allows more competitors to enter the marketplace, since potential competitors with specialized capabilities need not develop capabilities in all aspects of the system in order to offer products. This approach will likely be followed as BISDN (fiber to the home and business) procurements are made. It is likely that large-volume purchasers of local networks will want to have the option of procuring fiber cables separately from terminal equipment. From a purchaser's viewpoint, this approach typically results in lower cost due to increased competition. From a seller's viewpoint, it is a two-edged sword. A seller offering a complete range of system technologies might prefer that procurements be made on a system basis so that fewer competitors are capable of making an offering. On the other hand, a seller who lacks one or more key technologies (or who does not desire to invest in one or more key technologies) would prefer the disaggregated procurement approach. From a U.S. competitiveness viewpoint, disaggregation can prevent U.S. companies from being locked out of a large market because of weakness in one or more key technologies that individually represent only a small part of a much bigger system.

OCR for page 9
OPPORTUNITIES IN TELECOMMUNICATIONS SUMMARY 21 Although fiber optics has become the dominant technology for point-to- point long- and moderate-distance telecommunications applications and has emerged as a multibillion-dollar-a-year business in the 1980s, the largest applications of fiber and related photonic technologies in local area networking, metropolitan area networking, and BISDNs are yet to come. As existing and emerging markets evolve, low cost, high reliability (quality), and high perfor- mance will become increasingly important. Whereas research and engineering that precede transfer of technology to the factory will continue to be important, increased emphasis on perfection of manufacturing processes (manufacturing engineering) will likely be critical to U.S. competitiveness in these markets in the future. Manufacturing success will require the close cooperation of people concerned with materials, devices, and systems. Certain key technologies ripe for development have been identified, all of which emphasize development of lower-cost, higher-quality versions of existing technologies. Numerous areas requiring continued research activity necessary to enable certain new markets or to reduce costs in existing or emerging markets have also been identified. The impact of standards and open interfaces on increasing competition and preventing particular technologies from becoming bottlenecks to competition in larger system procurements was also discussed. (Refer to Appendix D for additional technical data.) REFERENCES 1. Personick, S. D. 1981. OpticalFiber Transmission Systems. New York: Plenum Press. Personick, S. D. 1985. Fiber Optics Technology and Applications. New York: Plenum Press. Midwinter,J. E. 1979. OpticalFibers for Transmission. New York: Wiley and Sons. 4. Kao, K. C. 1982. Optical Fiber Systems--Technology, Design, and Applications. New York: McGraw Hill. IEEE. 1983. Special Issue on Fiber Optic Systems. Journal on Selected Areas in Communications SAC-13 (April). IEEE. 1985. Special Issue on Fiber Optic Local Area Networks. Journal of Lightwave Technology LT-3~3) (June). 7. IEEE. 1984. Special Issue on Undersea Cable Fiber Optic Systems. Journal of Lightwave Technology LT-2~6) (December). 8. IEEE. 1986. Special Issue on Broadband Communication Systems. Journal on Selected Areas in Communications SAC-4 (July).

OCR for page 9
22 PHO TONICS 9. IEEE. 1988. Special Issue on Fiber Optic Local and Metropolitan Area Networks. Journal on Selected Areas in Communications SAC-6 (July). 10. IEEE. 1988. Special Issue on Photonic Switching. Journal on Selected Areas in Communications SAC-6 (August).