Long-range global communications trends point to escalating fiber-optic usage
During the next decade, worldwide voice, video and data applications and products are anticipated to experience double-digit growth via fiber-optic components and networks
jeff d. montgomery
The worldwide lightwave market forecast for 1994 to 2004 calls for fiber-optic network deployment in video, voice and data communications applications to expand rapidly and extensively. In fact, global consumption trends for advanced communications photonic products indicate a strong 45% average annual growth rate from $312 million in 1994 to nearly $2 billion in 1999. Even though the growth rate is expected to slow to 29% during the next five years, photonic products are predicted to become a $7.2 billion market by 2004.
North America is estimated to increase its market share of fiber-optic component consumption from 47%, or $3.1 billion, in 1994, to 55%, or $17.5 billion, in 2004. By contrast, European regions are forecast to lose market share, dropping from 24%, or $1.54 billion in 1994, to 15%, or $4.7 billion, in 2004. Far East regions are also expected to lose market share, dropping from 23%, or $1.52 billion in 1994 to 13%, or $4.2 billion in 2004.
In addition, the deployment of asynchronous transfer mode, or ATM, switches is expected to increase methodically to both telephone company central offices and premises local area networks. Consequently, an important telephone company market is predicted to develop for virtual local area networks. Based on this evolution, instead of investing in ATM switches and related products, corporate enterprises are projected to seek connectivity to ATM-based switching services via fiber-optic cable.
Beyond the year 2000, voice, video and data communications are expected to be melded into a universal digital data network. The global telecommunications industry is figured to purchase the major share of this network by leveraging from their entrenched base. Virtually all types of communications and images are anticipated to be digitized, packetized and transported worldwide, all indistinguishable from conventional data.
As part of the information transmission increase, a frenzy of partnership formation is foreseen among telephone companies, interexchange carriers, cable-TV multiple system operators, competitive access providers, and others, all aiming to achieve a long-term position on the global data network. Over the next twenty years, these partnerships are projected to increase, re-form and aggressively compete for rapidly growing revenues. Compelling economic issues should force numerous competitors toward consolidation. The final form of this competition and the identity of the winning groups are not yet apparent. Most likely, however, AT&T should become the leader in establishing powerful partnerships.
With strong global competition and common communications technologies, however, the global data network is figured to evolve into a commodity business. Greater long-term profits are calculated to be won by organizations who prefer not to have network ownership, but instead opt to derive transport packages and shop among network operators for the best toll rates. Telephone companies, in general, are historically weak in this type of operating mode, which is deemed mandatory for market success in dealing with the future global data network.
Business drives bandwidth
Enterprises are becoming global competitors with their rapid proliferation of foreign-based offices. To remain competitive, enterprise managers need daily transfers of data in greater volume (bandwidth) than was used a decade ago. Consequently, businesses have been the early drivers of demand for more bandwidth and connectivity. Government agencies and educational institutions are joining this trend.
Whereas business bandwidth demand is driven directly by economics (return on investment), residential/consumer demand depends on a choice among various uses of discretionary income. To date, consumers have shown little interest in offerings beyond conventional cable-TV services. However, as costs decrease for services such as multimedia and videoconferencing terminals, consumers are expected to join the market.
The local loop, which connects the telephone company central office to the subscriber, consists of feeder, distribution and drop links. The outward expansion of fiber toward the subscriber has progressed rapidly to business enterprises driven by bandwidth demand and supply. However, fiber has moved slowly toward residential areas. The expansion of fiber further into residential loops is estimated to be paced by both the rate of increase in bandwidth demand and the reduction of the cost premium of fiber compared to other media.
Telephone companies have not yet detected strong consumer demand for new broadband services. On the other hand, the cable-TV industry is expanding its residential bandwidth capacity and entering voice-service competition. Interexchange carriers and others are also entering this market. Caught in this bind, telephone companies have tried to create the impression of aggressively moving forward with broadband deployment while actually investing much less than implied.
From 1990 to 1993, several major telephone companies promoted the concept that acceptable video could be delivered to consumers via twisted-pair copper cable, perhaps even over the installed voice-grade cable. That premise was shelved because of transmission problems, and planners turned to coaxial-cable drops that copied cable-TV architectures. From 1993 to 1995, the installation emphasis has increasingly focused on hybrid fiber/coaxial-cable networks.
However, the twisted-pair copper cable solution has just been revived by a three-chipset-based high-speed technology introduced by AT&T Microelectronics and Broadband Technologies (see Lightwave, November 1995, page 1.) This copper-based approach appears to offer a low-cost, low-risk solution for offering near-term broadband services.
Telephone companies are increasingly aware that neither twisted-pair nor coaxial cable will accommodate the long-term bandwidth-distance product requirements to support residential services. They also surmise that hybrid fiber/coaxial cable compared to fiber installed all the way to the home provides fewer long-term cost savings than earlier predicted. Accordingly, telephone company enthusiasm for installing hybrid fiber/coaxial cable to tens of millions of homes is, in general, fading.
AT&T, not a factor in early fiber-in-the-loop, or FITL, competition, aggressively attacked this market in 1993. Initially, the company responded to customer demand by promoting hybrid fiber/coaxial cable. Its portfolio has subsequently included the unshielded twisted-pair cable/switched digital video architecture, which takes fiber to the curb. AT&T also has developed, and has been engineering for high volume production of fiber-to-the-home, or FTTH, products that will be cost-competitive with hybrid fiber/coaxial-cable types. It is expected to ultimately promote all popular broadband-to-subscriber solutions, but will advance fiber to the home wherever feasible.
Resistance to mixed media
Telephone companies historically have deployed fiber only when forced by increased bandwidth demand and favorable economics. They have successfully resisted the use of multimode fiber in applications where this fiber type would be adequate and economical. These companies currently face the added complexity of coaxial-cable distribution and drops. Committed to fiber-feeder networks, they would welcome a cost-competitive fiber alternative to coaxial cable. Meanwhile, they seem comfortable with the proposed increases in bandwidth using existing voice-grade unshielded twisted-pair copper.
The hybrid fiber/coaxial-cable approach, which runs fiber to a neighborhood node, comes at a time when telephone companies, in general, have substantially reduced their engineering/planning and crafts staff. These companies would like to avoid the added problems of dealing with coaxial cable, which include investing in inventory, training and more suppliers.
Residential consumers have shown little willingness to spend tens of dollars per month for broadband services beyond the entertainment programs available via cable TV. Hundreds of surveys, focus group investigations and field trials have failed to uncover consumer enthusiasm or a "killer" application. For several economic reasons, this situation is forecast to change over the next decade.
In contrast to analog broadcast limitations of frequency spectrum allocated per channel, the information content that can be digitally transmitted over fiber is virtually unlimited. The first technology of expanded performance is represented by high-definition (digital) television.
Unlike analog receivers, however, future digital television sets are predicted to provide expanded performance through increased bandwidth at little or no cost increase. This low-cost factor follows the trends established for computers and other digital devices. Larger screen size, higher resolution, computer-generated simulation and animation, three-dimensional effects and other visual enhancements are expected to attract millions of consumers into this market.
Global communications strategies
The installed global communications networks are presently inadequate to handle future broadband service demands, especially in Eastern Europe and emerging nations. To meet expected demands, national telephone companies, regional Bell operating companies, interexchange carriers, major cable-TV and wireless system operators, and others are entering multi-regional negotiations aimed at forming partnerships that can produce and support long-term global broadband networks. Numerous operating networks are expected to be operational by the year 2000.
Among the companies jostling to form global networks, AT&T offers a commanding combination of established networks, undersea fiber-optic links and foreign telephone company partnerships; technical expertise; strong capital position; and global reputation and recognition. The recent split by AT&T into three separate companies (see Lightwave, November 1995, page 1) permits the company to pursue global leadership positions in communications and competition at all levels, without crippling its growing and profitable hardware business.
In local area networks and other premises data networks, fiber-optic cable installations are projected to increase rapidly. By 2003, fiber is estimated to account for approximately 80% of the entire premises data interconnect bandwidth and should continue to increase thereafter.
Multimode fiber cables dominate in data communications network applications, and this situation should continue. Singlemode fiber cable use, however, is gauged to grow more rapidly, and by 2004, should represent almost 25% of all data communications fiber interconnects.
ATM switching (including ATM embedded in hubs and concentrators) is foreseen as the base of future local area network architectures. Switch fabrics should advance from OC-3 rates at 155 Mbits/sec to OC-12 rates at 622 Mbits/sec and much higher by 2004. The ATM product field, however, is highly volatile in terms of standards, interconnectivity, suppliers and other market variables. Because many network planners have been slow to commit to ATM, telephone companies are offering local area network ATM switching services. Instead of investing in evolving ATM technologies, local area network managers can, instead, interconnect to a telephone company`s ATM switching service via fiber-optic cable. Known as virtual local area network service, this telephone company business element is expected to impact future premises local area network planning.
Several fiber optics and related photonics technologies are calculated to markedly affect telephone company and local area network interconnect architectures during the next decade. They include photonic switching, parallel optical interconnects, optical amplifiers and dense wavelength-division multiplexing, or WDM, technology.
Semiconductor optical amplifiers, nonlinear optical polymeric films, crossbar optical switches and nonlinear optical crystals are electronics-driven solid-state technologies expected to be developed in the future.
Electronic switch matrices--at least in 8x8 configurations--are estimated to be available by late 1996 and should develop into a major market (more than $100 million) within a decade. Demands abound for high input/output (such as 192x192) fiber matrices in central offices, disaster recovery, built-in test scanning and redundant opto-electronics switchover.
During the past few years, particularly for long links (tens of kilometers), network planners have gradually evolved to WDM over 1310- and 1550-nanometer channels on one fiber. During 1992 to 1994, this usage expanded to three channels at 1310 nm and at the low and high ends of the 1550-nm range.
The advent of optical fiber amplifiers, which have relatively flat gain over a 30-nm range in the 1530- to 1560-nm range, has increased interest in high-density WDM for more channels in one band. AT&T Bell Laboratories has developed WDM units and the required high-stability laser diodes, aimed at one channel per nanometer in the 1550-nm band. The economic payoff is attractive for undersea fiber-optic cables, where capacity can be increased by 30 times by adding only a WDM assembly and high-stability laser diode transmitters and receivers. AT&T is also using this technology to upgrade its terrestrial interexchange networks.
High-density WDM adoption is expected to be implemented along traditional economical directions. Undersea links should be the first users, followed by long-haul terrestrial interexchange links, and interoffice trunks and feeders. High-density WDM is not likely to be economically feasible in premises data network use within the next decade. u
Jeff D. Montgomery is chairman of Electronicast Corp. in San Mateo, CA.