Several spectacular changes in the communications scene over the past few years will have a still stronger impact in 1997. Most obviously, the World Wide Web has been expanding at a rate more commonly associated with nuclear chain reactions [Fig. 1]. Showing a nice mixture of optimism and concern, the telecommunications establishment is mulling how best to exploit the Web commercially while deploying such packet-switched networks as the integrated services digital network (ISDN) and such subscriber access technologies as asymmetric digital subscriber line (ADSL) to keep the telecom infrastructure from being overwhelmed.
In the United States, the Telecommunications Act of 1996 has rewritten the rules of the game, so the conventional wisdom that has guided the industry for decades must be largely discarded or at least carefully examined. Although the Act is less than a year old, it has already inspired at least three large mergers and who knows how many other negotiations.
The situation in the rest of the world is similar. Private companies in many nations, especially in Europe, are making deals with one another--both within countries and among them. Government-owned telecom organizations are being privatized. Even countries that rely on a single state-owned enterprise for telecommunications services want to allow competition. This year such countries may figure out how to do so without disrupting relations with their neighbors.
Amid the competitive pressures thus generated, telecom companies have more incentive than ever to squeeze as much performance as possible out of their existing infrastructure. Hence the attraction of local-access technologies like ADSL for boosting the capacity of installed copper subscriber loops, and long-haul technologies like wavelength division multiplexing (WDM) for boosting the capacity of installed optical fiber.
Back at the office, wireless local-area networks (LANs) are benefiting from spread-spectrum technology, just as cellular telephony is. Wired LANs are also deploying new--but not revolutionary--technology this year: Gigabit Ethernet replacing the original 10-Mb/s Ethernet and the more recent 100-Mb/s Fast Ethernet. Asynchronous transfer mode (ATM) technology appears to have lost the LAN wars, although it is playing a large role in long-haul networking.
This year and next, the marketplace must sort out complex issues--technical and competitive--between various local and long-distance telephone service providers, cable service providers, and emerging space-based service providers. As traditional market niches crumble, services available to customers overlap. Interesting barstool bets may be made on how the choice will affect consumers and which services ultimately will dominate.
For years, communications gurus have decried local loops of copper twisted pairs as an obstacle to new services. Designed for voice signals, the twisted pairs were never intended to carry high-speed digital data, which is sensitive to impulse noise and to dispersion--that is, frequency-dependent group delay, which causes closely spaced pulses to spread out and interfere with one another.
With the growing Web and other Internet traffic plus multimedia services such as video-on-demand on the horizon, revamping local-network access has become a top priority. The problem posed by Internet traffic is that its traffic dynamics (burstiness and long holding times) are radically different from voice telephone service (many calls of comparatively short duration). [For more on the differences between voice and data traffic, see the Viewpoint by William Stallings, "Self-similarity upsets data traffic assumptions."] Local exchange carriers (LECs) are challenged to upgrade the infrastructure in their switching centers to handle all the new traffic. The need to upgrade has provoked a fierce public debate as to whether the upgrade costs are just a normal cost of doing business, or an extraordinary expenditure that justifies rate increases.
The special problem posed by the Web is that it encourages lengthy surfing sessions, lasting hours and even days, in some cases. The result is that LECs find their interoffice trunks operating near capacity and ports on their expensive switches are being tied up for lengthy periods while producing no more revenue than a two-minute voice call. (Most local calls, in the United States at least, are not timed.) Thus, LECs are loath to invest in additional infrastructure that is unlikely to generate additional revenue. Moreover, although replacing copper with optical fiber may make sense in the long run, it is no solution for the immediate future: the Geneva-based International Telecommunication Union estimates that the world is now using more than 600 million copper telephone lines.
Several technologies offer quick ways to increase the capacity of copper loops. Among the most promising is the asymmetric digital subscriber line (ADSL), which greatly increases the capacity of existing subscriber loops--up to rates of about 6 Mb/s--without requiring the installation of new cable. Moreover, it accomplishes that feat without affecting the existing plain old telephone service (POTS).
To preserve POTS and to prevent a fault in the ADSL equipment from compromising analog voice traffic, the voice part of the spectrum (the lowest 4 kHz) is separated from the rest by a passive filter, called a POTS splitter [Fig. 2]. The rest of the available bandwidth--from about 10 kHz to 1 MHz--carries data at rates up to 6 bits per second for every hertz of bandwidth.
To exploit the higher frequencies, ADSL makes use of advanced modulation techniques, of which the best known is the discrete multitone (DMT) technology. DMT was pioneered by Stanford University, California., and Amati Communications Corp., San Jose, Calif., and endorsed by the American National Standards Institute (ANSI), New York City, and the European Telecommunications Standards Institute (ETSI), Sophia Antipolis, France.
DMT divides the bandwidth from about 10 kHz to 1 MHz into a set of 256 independent subchannels, each 4 kHz wide. By measuring the quality of the subchannels and then assigning a bit-rate to each based on its quality, DMT customizes the transmit signal for every line. In doing so, it automatically avoids regions of the frequency spectrum that are too noisy or too attenuated to support reliable communications. If the quality of a subchannel degrades enough to affect a system's error performance, the data rate on that subchannel is lowered and the excess traffic moves to a subchannel capable of supporting it. The result is robust communications over single twisted pairs.
As its name implies, ADSL transmits data asymmetrically--at different rates upstream (toward the central office) and downstream (toward the subscriber). Such a technology makes sense for two practical reasons. For one, the typical Web surfer is more interested in downloading large files than in uploading them, and therefore needs more capacity in the downstream (network-to-subscriber) direction.
The second reason is technical: when many wire pairs are squeezed together in a cable, cross talk is inevitable. Signals traveling downstream from the central office are not much affected, because they all are of approximately the same amplitude. On the other hand, upstream traffic originates in subscriber premises, and these buildings may be at different distances from the points at which lines come together in a cable; accordingly, upstream signals can vary greatly in amplitude. If a wire pair carrying a strong signal shares a cable with another wire pair carrying a weak one, cross talk can be all too evident. But since cross talk increases with frequency, the problem can be made tractable by limiting the upstream data rate and keeping it near the low-frequency end of the spectrum.
Meanwhile, cable television providers are not sitting by idly. They want to provide Internet service to PC users over their TV cable systems by means of special cable modems. Such modems are capable of transmitting up to 30 Mb/s over hybrid fiber/coax systems (which use fiber to bring signals to a neighborhood and coax to distribute it to individual subscribers). Further, they are available, and they work.
Cable modems come in many forms. Most create a downstream data stream out of one of the 6-MHz TV channels that occupy spectrum above 50 MHz (and more likely 550 MHz) and carve an upstream channel out of the 550-MHz band, which is currently unused. Using 64-state quadrature amplitude modulation (64-QAM), a downstream channel can realistically transmit about 30 Mb/s (the oft-quoted lower speed of 10 Mb/s refers to PC rates associated with Ethernet connections). Upstream rates differ considerably from vendor to vendor, but good hybrid fiber/coax systems can deliver upstream speeds of a few megabits per second. Thus, like ADSL, cable modems transmit much more information downstream than upstream.
The downstream channel is continuous, but, like Ethernet, divided into packets, with addresses in each packet indicating for which subscriber each is intended. The upstream channel has a media access control that slots user packets or cells into a single channel.
To avoid collisions, in some cable systems, upstream packets are gated onto the network via control signals embedded in the downstream information. Other approaches divide the upstream path into frequency channels and allocate a channel to each user. Still others combine these two multiplexing methods. A few modem companies are proposing techniques like spectrum spreading or code-division multiplexing to reduce susceptibility to interference from antennas and other sources of electromagnetic radiation outside the system. Called ingress noise, it is the biggest difficulty on hybrid fiber/coax networks.
Variation in the capacity of cable systems depends less on cable length than on ingress noise and on the number of users seeking simultaneous access to a shared line. (Cable data rates are not particularly sensitive to the length of the coaxial cable; amplifiers in the cable network keep signal power high enough to make length a minor consideration.)
Because cable TV systems use a shared-bus architecture, they may be less expensive to implement than ADSL. But that shared architecture is a double-edged sword. As with any shared medium, as more users go on-line, the capacity available to any one user inevitably falls.
At present, the point is somewhat academic since the top speeds of both ADSL and cable systems will not be usable for years anyway. Internet server speeds, network delays, and personal computer limitations will hold usable rates at or below 2 Mb/s for the foreseeable future. So far, ADSL offers higher security and reliability. Cable modems may offer a less expensive network solution because of the cable plant's shared architecture, but that differential is more than offset by infrastructure costs required to upgrade existing coaxial cable networks to hybrid fiber/coax. The technologies for both ADSL and cable modems are at about the same state of maturity and integration.
ADSL's greatest advantage is that it can make use of existing twisted copper pairs, which are numerous indeed compared with the number of hybrid fiber/coax lines that exist in upgraded cable systems. Today the global ratio is on the order of 600 million to 6 million, or about 100:1. In the United States, it is about 20:1. Even with aggressive cable upgrades, the numbers are not likely to reach parity over the next five or six years.
By a large majority, the U.S. Congress passed the Telecommunications Act of 1996 last Feb. 8. It was a sweeping amendment to the 62-year-old Communications Act of 1934, which created the U.S. Federal Communications Commission (FCC), and it brings full-fledged competition to the U.S. telecommunications industry. Under the new law, services and their providers are allowed or disallowed not by regulation, as in the past, but by contracts negotiated among the participants, none of which is a regulated monopoly. The effects of these revolutionary changes will be felt for years to come.
For the first time, long-distance carriers are permitted to provide local telephone service. Local carriers, being no longer regulated as local monopolies, are free to offer all long-distance services. Cable television companies may offer telephony and information services, and the telephone companies can carry entertainment and information services into homes. Lastly, media companies have greater freedom to acquire and hold assets in any one service area.
The 1996 act supersedes the Modification of Final Judgment (MFJ)--the 1982 consent decree under which AT&T was divested in 1984 of the seven "Baby Bells," the nickname for the Bell regional holding companies. It also eliminates the role of Federal District Judge Harold Greene, who has been monitoring the MFJ since then, in part through triennial Department of Justice reviews of the state of the telecommunications industry. The MFJ had prevented AT&T from entering local telephony markets and had barred the Baby Bells from several activities: certain types of long-distance telephony, manufacturing, and transmitting entertainment and information programs over their own phone lines.
The FCC now has to work out how to implement the 1996 Act's provisions--no easy task. While any competing carrier can ask any incumbent local or long-distance carrier for a connection to the network, the new act recognizes that the incumbent, which has all the facilities, also has all the bargaining power and can thus act as a bottleneck--through unfair pricing or other techniques [see Viewpoint]. If the Baby Bells are able to demonstrate through certain tests that they are not acting anticompetitively or against the public interest, the 1996 Act also lets them into previously prohibited markets. But the FCC may, in its tests, invoke any antitrust test whatsoever, including tests given by the old MFJ, and it may in addition consult with the Department of Justice.
Some major unresolved issues have to be settled in 1997. The FCC must establish a council to handle the assignment of telephone numbers across the nation. At the moment, numbering is administered by the Baby Bells, which have sometimes done it in ways disadvantageous to rival carriers. It has been known, for example, for a new carrier to be given a new area code for all its customers, who thereafter are forced to dial three or four extra digits to call their neighbors. The FCC must also establish a workable plan to make numbers portable--that is, usable by customers throughout their lives, no matter where they reside or which carrier they choose to use.
For the most part, the 1996 Act removed restrictions, but one major exception is the transmission over public media of sexually explicit materials and other services considered unsuitable for children. Internet service providers are prohibited from sending pornography to them, and all new television sets are to have built-in devices that can be configured to block certain channels. The pornography issues are all caught up in complex litigation surrounding the Communications Decency Act.
In October, Deutsche Telekom AG, Bonn, announced that it would go private, putting up for sale US $9.9 billion (thousand million) worth of shares--the world's biggest stock offering ever. Deutsche Telekom is the third-largest telecommunications company in the world, after Japan's Nippon Telegraph and Telephone Corp. (NTT), Tokyo, and AT&T Corp., New York City (at least before their reorganizations, each into three smaller entities) [Fig. 3]. Since Deutsche Telekom is also the biggest telephone operator in Europe, the utility's privatization will reshape Europe's entire telecommunications market as well as Germany's business culture.
A few months earlier, Michel Bon, the president France Télécom, Paris, had announced that the state-owned operator was also to be sold off through a public offering of shares in four months (April 1997), "as long as conditions are favorable." Even after France Télécom becomes a limited-liability company at the end of this year, the French government is to retain a 51 percent stake.
In addition, international phone service from the United Kingdom was opened to competition for the first time last August. Up to then, only the regulated duopoly of British Telecommunications PLC, London, and Mercury Communications Ltd., London, might hold licenses for international facilities. Deregulation should make international calls from the UK much cheaper because it allows competitors to bypass a much-criticized accounting mechanism: up until last August, the fees the two carriers paid each other for delivering international traffic bore no relation to actual costs--a system often described as a cartel. In the years ahead, any carrier with local loop infrastructure in both countries is also free to build its own transatlantic facilities, thereby avoiding the British TelecomMercury accounting rates--a development of considerable interest to Nynex, US West, MFS (Metropolitan Fiber Systems) Communications, and any other U.S. companies having British subsidiaries.
Many countries that rely wholly on a single state-owned enterprise for telecommunications services also want to admit competition. Now they must figure out how to do so in a way that does not create friction or incompatibilities with neighboring countries. Early last year the United Nations' World Trade Organisation (WTO) conducted talks on how best to liberalize (privatize and deregulate) telecommunications. The 48 participating nations tentatively agreed on a draft outline of the basic principles that would be needed to encourage competition in basic telephony services.
Bilateral talks between assorted countries on opening their telecommunications markets went less well. Discussions involving the United States ended acrimoniously when the chief U.S. negotiator, Don Abelson, argued that no consensus had been reached on international and satellite services--and that the FCC should be allowed to refuse operating licenses to companies whose governments keep their markets closed. As a result, the WTO talks missed an April 30 deadline for negotiating a worldwide pact to liberalize post, telegraph, and telephone (PTT) systems.
Australia, Canada, and New Zealand, not to mention the European Union and Japan, all complained that refusal of an operating license on the grounds of nonreciprocity contravenes the WTO's antidiscrimination rules. Early this year, the organization will resume talks for an international accord to privatize the remaining PTTs. Worldwide annual telecommunications revenues are estimated to be worth more than US $500 billion.
Last year, mergers, de-mergers, and joint ventures became the telecommunications giants' preferred method of doing domestic and international business with one another, and the trend is likely to continue through the new year. One blockbuster October announcement was that of the transatlantic merger between British Telecom and MCI Communications Corp., Washington, D.C. The British firm, which already owned 20 percent of MCI's stock, has decided to form a multinational entity called Concert PLC. by purchasing the other 80 percent, thus forming the world's fourth-largest telecommunications enterprise [as shown in Fig. 3]. The new company, based in both London and Washington, D.C., should have the money and the vigor to compete effectively in Europe and Asia. Concert's head, former British Telecom chief Sir Peter Bonfield, has made no secret of his desire to further that goal by forming an alliance with Japan's NTT.
Meanwhile, in the wake of the 1996 Telecommunications Act, the United States saw several major mergers. Two of the Baby Bell regional holding companies, SBC Communications Inc. (formerly known as Southwestern Bell), San Antonio, Texas, and Pacific Telesis Group, San Francisco, announced their merger worth US $17 billion. Some $2 billion of their joint annual revenue is expected to flow from calls between Mexico and the United States. Half of all U.S. calls to Mexico originate in the territories, especially California and Texas, covered by those two companies, while 80 percent of all Mexico's international calls are made to the United States--the greatest interchange of phone traffic between any two countries.
Bell Atlantic Corp., Philadelphia, and Nynex Corp., New York City, are also busy completing their merger this year. Between them, these two Baby Bells cover the eastern United States from Maine to North Carolina--a lucrative area crowded with corporate customers, where some 45 percent of U.S. phone calls both originate and terminate, and where 30 percent of the United States' $14 billion in international phone calls originate. Even before the merger, the two regional companies had begun offering long-distance service through a cellular joint venture.
Yet another merger involved two of the world's fastest-moving private telephone operators: MFS Communications Co., Omaha, Neb., and WorldCom Inc., Jackson, Miss. The fourth-largest U.S. long-distance carrier, WorldCom is to swap 2.1 of its shares for every MFS share, a transaction that values MFS at around US $14 billion. Since 1987, the Nebraska firm has built networks in 49 U.S. cities as well as in London, Paris, Frankfurt, and Stockholm, and it plans to build additional local fiber networks in another 45 financial centers. WorldCom chiefly leases bandwidth from other operators for long-distance and local connections.
Truly competitive telecom operators must have their own alternative local loops, so that the combination of WorldCom's long-distance network and MFS's local access ones should make MFS WorldCom a serious global contender.
Furthermore, before the merger, MFS had bought Uunet Technologies Inc., Fairfax, Va., one of the leading Internet service providers--and some analysts think that the Fairfax company is one of the key components that could transform MFS WorldCom into a major player.
Nevertheless, bigness is not the only game in town. Last year, AT&T completed its voluntary breakup into three independent companies, first announced late in 1995 [for details, see "Communications," IEEE Spectrum, January 1996, p. 41].
The years' biggest de-merger news, however, was the announcement made last month by Japan's Ministry of Posts and Telecommunications: NTT is to be broken into three pieces. NTT was till then the world's largest telecommunications company. It will become two regional companies for local telecommunications (one for the eastern part of the country and one for the western) plus a long-distance carrier, which for the first time will be allowed to enter the international market. All three will be organized under a single holding company. As of Spectrum's press date, the exact timing of the breakup was still under discussion.
Although NTT was privatized in 1985, shortly after AT&T's original divestiture, two-thirds of its shares are still owned by Japan's Ministry of Finance, and the company has a monopoly of some 70 percent of the country's domestic markets--both local and long-distance.
Moreover, Japan's telecommunications industry is still highly regulated. To date, NTT has been prohibited from offering international services, and Japan's international carrier, Kokusai Denshin Denwa Co. (KDD), Tokyo, has been barred from playing a domestic role. To Japanese policy-makers, a greater concern is domestic long-distance rates, which are as much as four times those prevailing in France, Germany, the UK, or the United States; the country's telecommunications sector has suffered from a lack of innovation; and the per capita penetration of such services as e-mail and mobile telephony in Japan has trailed that of the United States by a factor of 313 [see table].
Even telecom companies not overhauling their organizational structures are rushing to do business through multinational joint ventures. British Telecom and two of Germany's largest industrial concerns, RWE AG, Essen, and Viag AG, Munich, have committed US $1.5 billion to $2.0 billion over the next five years to build a domestic phone network connecting both business and residential customers. The alliance wants to capture 1015 percent of the country's fixed telephony market over the next decade and to apply for Germany's fourth cellular license. Its main rivals are the incumbent Deutsche Telekom, and a partnership consisting of German companies Veba AG and Mannesmann AG, both in Düsseldorf, with AT&T and the UK's Cable & Wireless Plc., London.
Toward the end of September, British Telecom announced a strategic alliance with Compagnie Générale des Eaux (CGE), Paris, the French utilities, property, and communications concern. The UK operator is to invest about $1.5 billion for a 25 percent stake in Cegetel, a telecom company newly established by CGE in the hope that it will be France Télécom's greatest rival.
In the Netherlands, British Telecom and NV Nederlands Spoorwegen (NS), Utrecht, the national railway company, have formed a joint venture called Telfort, which is to offer communications services to businesses. Telfort, based in Amsterdam, hopes to be granted an operating license this year. Initially, it will offer data, corporate voice, and virtual private networks (VPNs) and will sell Concert's international voice and data services. The Dutch telecom market is the fifth-largest in Europe, valued at annual revenues of about $7 billion.
Meanwhile, Deutsche Telekom, Sprint, and France Télécom have formally announced a joint venture--Global One--to provide full communications services to multinational corporations. Regulatory permission was granted by both the European Union and the United States' FCC. Global One has two rivals: Concert and the WorldPartners Association, Murray Hill, N.J., consisting of AT&T and Unisource (itself a consortium of the Swedish, Dutch, Spanish, and Swiss PTTs).
Just as ADSL promises to extend the usefulness of existing wire plant, wavelength-division multiplexing (WDM) promises to extend the usefulness of existing fiber. WDM is basically old-fashioned frequency-division multiplexing at optical frequencies. Researchers at each of three laboratories have used it to attain a sort of Holy Grail of digital communications: data transmission at 1 terabit a second--yes, a full 1000 Gb/s.
Fujitsu Laboratories Ltd., Kawasaki, Japan, transmitted 1.1 Tb/s over 150 km of conventional fiber; 55 optical carriers were used, each modulated at 20 Gb/s. Along with AT&T Labs, Basking Ridge, N.J., and Lucent Technologies, Murray Hill, N.J., AT&T transmitted 1 Tb/s over 55 km; in this experiment, outputs of 25 diode lasers were split into two orthogonal polarizations, and each of the 50 signals then modulated at 20 Gb/s. NTT multiplexed 10 lasers, each carrying 100 Gb/s, and sent the combined 1-Tb/s signal over 40 km of dispersion-shifted fiber. (The 100-Gb/s rates of the individual optical carriers were achieved through time-division multiplexing with delay lines--itself no mean feat.)
All three experiments were conducted at 1550 nm, the wavelength of optical fiber's minimum attenuation and the only one at which erbium-doped fiber amplifiers function. Broadband optical amplifiers are key to WDM systems since they amplify the entire spectrum of interest in a single device. Repeaters that regenerate pulse trains, rather than amplify them, admittedly permit longer transmission distances; but they are more expensive since they must demultiplex the combined signal, regenerate each pulse train separately, and then recombine them.
Meanwhile, WDM is proceeding apace in the real world, as well. At last June's SuperComm '96 trade show in Dallas, Ciena Corp., Savage, Md., announced that Sprint had deployed Ciena's MultiWave 1600 system along a 320-km route in the U.S. Midwest. The multiplexing system transmits all 16 channels at slightly different wavelengths of infrared over existing optical fiber. Each of the carriers, placed 2 nm apart in the region around 1550 nm, supports a data rate of 2.5 Gb/s--the same rate transmitted without multiplexing. As a consequence, the multiplexed system has a capacity of 40 Gb/s. Ciena next plans to introduce a 40-channel dense WDM terminal with a capacity of 100 Gb/s.
In a related development, the world's biggest undersea fiber-network project--the $1.5 billion Fiber-optic Link Around the Globe (FLAG)--is due to be completed this year. Starting with a 27 300-km link between Britain and Japan, the repeatered optical fibers will run through the Mediterranean and Red seas as well as the Atlantic, Indian, and Pacific oceans. The 5.3-Gb/s network will multiply the bandwidth between Europe and Asia by a factor of five. Managed by Nynex, the project is being built by KDD Submarine Cable Systems and by AT&T Submarine Systems for operation this September [see "The glass necklace," Frank J. Denniston and Peter K. Runge, Spectrum, October 1995, pp. 2427]. Another ambitious project is Africa ONE, the 35 000-km WDM Sonet ring around Africa.
One of the most daring ventures before the worldwide telecom industry is being launched before the end of the century. Indeed, several international consortia are planning to introduce satellite services with the ultimate telecommunications solution: global connectivity to and from any spot on the earth's surface.
Transmitting mobile voice and data communications services via satellite is hardly a strange concept; Inmarsat, the largest provider of mobile satellite communications services worldwide, established the first mobile satellite system for the international maritime community in 1979. Indeed, Inmarsat, London, jointly with Comsat Mobile Communications, Bethesda, Md., is hoisting four new Inmarsat 3 satellites to offer global phone service called Planet 1.
What is unfamiliar is the planned provision of satellite-based, global mobile voice and datacom services to pocket-sized, hand-held devices. Because mobile satellite services (MSS) would provide instant infrastructure, along with a portfolio of enhanced services, it should cash in on demand for telecommunications services in developing countries where wireline and cellular systems provide inadequate service or do not even exist
Among the first literally to get off the ground is Iridium, the system of 66 satellites encircling the earth in six low-earth orbital planes at 780-km altitude, being launched throughout this year for a global wireless personal communications system. Seventeen launches from three launch sites (Vandenberg Air Force Base near Lompoc, Calif.; Baykonur Cosmodrome in Kazakhstan in the former Soviet Union; and Taiyuan Satellite Launch Center, 650 km southwest of Beijing, China) will deploy the entire constellation, which will begin operation in 1998. A hand-held terminal will allow telephone calls between any two points on earth at 1616.01626.5 MHz. The Iridium system is being developed by a partnership of 17 companies headed by Motorola Satellite Communications Division, headquartered in Chandler, Ariz.
Another planned MSS service hopeful is Globalstar LP, San Jose, Calif., which will deliver satellite telephone service through a constellation of 48 low-earth-orbiting (LEO) satellites. Final assembly of the satellites began last October, and the first ones will be launched this year; the initial services are scheduled to be delivered next year, in 1998.
Ultimately, this worldwide digital telecommunications system will offer wireless telephone and other digital services, such as data transmission, paging, facsimile, and position location. Globalstar L.P., based in San Jose, Calif., was founded by Loral Space & Communications Ltd., New York City, and Qualcomm Inc., San Diego, Calif., and has strategic partners AirTouch Communications, Daimler-Benz Aerospace, Elsag Bailey, France Télécom/Alcatel, Hyundai/Dacom, Loral Space & Communications, and the Vodafone Corp.
On Dec. 15, 1995, the FCC issued the first in a series of decisions to allocate thousands more megahertz for licensed (high-power) and unlicensed (low-power) high-speed wireless communication and vehicle anticollision radar. The FCC's action is expected to create novel markets for manufacturers and novel business opportunities for service providers.
The newly allocated spectrum is in the extremely high-frequency (EHF) band, which spans 30300 GHz, or 110 mm. Since antenna gain increases with frequency, an EHF antenna can be quite small and still provide high gain and excellent directionality. The high carrier frequency easily supports data rates in excess of 5 Gb/s.
According to Bennett Z. Kobb, editor of SpectrumGuide, Falls Church, Va., that data rate is what makes millimeter waves so attractive for such multimedia applications as videoconferencing, the World Wide Web, and wireless cable TV. Other potential applications include high-speed wireless interconnections between wide-area networks and long-haul networks as well as between individual users and the backbone of the anticipated gigabit-per-second National Information Infrastructure.
The FCC allocated the 5964-GHz band to unlicensed devices operating at a maximum power density of 9 µW/cm2 at a distance of 3 meters from the antenna. This band is at an oxygen absorption region of the spectrum, where path loss is very high, making it unsuitable for long distances. Even so, manufacturers are excited by the possibilities since fixed-service licenses are in great demand just below 40 GHz, and until last year the FCC had not adopted rules to permit general use of the spectrum above that frequency. Fortunately, the high path loss at 60 GHz reduces interference over substantial distances.
Although allocated late in 1995, the 5964-GHz band is being made available only in 1997; the waiting period was intended to let the electronics industry develop a "spectrum etiquette" governing the devices' access to the airwaves. Now required in the 19201930-MHz band for wireless private branch exchanges and the 19101920- and 23902400-MHz bands for wireless data devices, spectrum etiquette prescribes orderly spectrum-access methods that do not rely on data exchange between products. Apple, AT&T, Hewlett-Packard, Metricom, Motorola, Rockwell, Sun Microsystems, and other companies developing spectrum etiquette for the new frequencies have formed a Millimeter Wave Communications Working Group.
The FCC deferred considering additional bands for unlicensed use and for the Licensed Millimeter Wave Service (LMWS). LMWS licenses will almost certainly be auctioned, with the FCC and potential bidders expected to focus on broadband wireless cable TV applications similar to the Local Multipoint Distribution Service (LMDS), both at 28 GHz. LMWS and LMDS licenses will probably be auctioned this year.
Meanwhile, cellular telephony--analog and, increasingly, digital--continues to grow at a rate of about 50 percent a year in the United States [Fig. 4].
Intel Corp., Santa Clara, Calif., has launched a free software program that lets users make long-distance phone calls over the Internet. The company's clout, and the inclusion of telephony in software from Microsoft and Netscape, could expand this market, which has so far been the preserve of techies. Intel's software has already been embraced by some 120 companies, including Microsoft, which has a compatible product called NetMeeting.
Last April, the Massachusetts Institute of Technology, Cambridge, announced the formation of an Internet Telephony Interoperability Project. The project is based at MIT's Research Program on Communications Policy (RPCP), a multidisciplinary research group focusing on infrastructure interoperability and Internet economics as well as technical, economic, regulatory, and business issues. Why MIT? The school has repeatedly demonstrated leadership in the development of the Internet, most recently by forming the World Wide Web Consortium at MIT's Laboratory for Computer Science, which has played a key role in the development of protocols for the Internet and will be advising the Internet telephony project. More information about the project can be found on the Web at http://itel.mit.edu.
Meanwhile, British Telecom and MCI announced they were close to completing the "biggest Internet network," creating additional capacity between Europe and Asia. The two have invested "multimillions" on a network that went live last July--not competing with the existing infrastructure but interconnecting with it to improve its capacity and network management. Until then, Internet traffic between locations in Europe or Asia had often been routed through the United States because of the lack of transmission capacity across those continents. According to British Telecom, the partnership will increase Internet capacity by up to 30 percent by deploying at least 12 new fiber-optic linked switches in major cities in the United States, Europe, and the Asia Pacific region to act as hubs.
The network, which is to be operated by Concert, offers corporate customers faster and more efficient Internet service at a premium tariff, guaranteeing levels of quality and reliability comparable to those the developed world expects of its voice circuits. The partners reckon that the new Internet network will break even either in the current fiscal year or the next one.
In case anyone was worried that all this expected Internet growth will cause an address crisis, that possibility was removed for the foreseeable future by the development of version 6 of the Internet Protocol to replace version 4--the designation v5 having been assigned in the interim to the Stream Protocol.
IPv6 changes the length of the address field in IP packet headers from 32 to 128 bits, increasing the number of possible addresses from a mere 4 billion to something around 3*1038--a truly mind boggling number representing about 6*1028 addresses for every human being on the planet. Bill Stallings said it well in his paper "IPv6: The New Internet Protocol" [IEEE Communications Magazine, July 1996, pp. 96108]: "Even if addresses are very inefficiently allocated, this address space seems secure."