Satellites Key Players in Forming Growing Electronic Global Village.
Ever since 1832, when Samuel Morse introduced the telegraph, one technological innovation has quickly followed another. In fact, the development of telecommunications over the past 150 years might best be pictured as a series of waves, with each successive wave having a greater amplitude than the previous one--the first wave representing the telegraph; the second wave the telephone; the third wave the radio; followed by cables, television, electronic computers, digital vocoders, transistors and, finally, communications satellites and fiber optics. Each of these waves has brougth to our global society a technological capability that has swept the tide of telecommunications to increasingly higher levels of performance.
Another way in which we can chart the remarkable progress of telecommunications over the years is by comparing speeds. One hundred and fifty years ago, for example, a fast sailihg ship could deliver a one-page letter from London to New York in approximately three weeks. Today, a facsimile system operating at 1.5 megabits per second is capable of transmitting that same letter in about one-tenth of a second. Soon, advanced satellites and fiber optics will be sending a page of information in less than one-hundredth of a second. This represents an amazing increase in performance of greater than 100 million times.
These modern systems are not only fast, but super reliable. Modern digital transmission systems already typically operate with an error rate of less than one in a million. With forward error correction and other accuracy-boosting techniques, however, the rate could even fall to one in a billion! That is the equivalent of a single typographical error in the Oxford Unabridged Dictionary.
Costs also reflect the progress of telecommunications. While the cost-reduction figures are not quite as dramatic as the increases in transmission rates and reliability, the progress is impressive. Tariffs for international telecommunications services are now 10 to 50 time less than they were just 25 years ago (allowing, of course, for inflation). Certainly, an economist would agree that that is an impressive bottom line!
Yet, on the eve of the launch of the first commercial communications satellite (Intelsat I) in 1965, not only were tariffs high, but overseas telecommunications capacity was surprisingly limited--just 200 telephone circuits across the Atlantic and much less across the Pacific and Indian Oceans. There was no live TV relay at all across the oceans. Telecommunications to most countries in Africa, South America ans most of Asia was largely confined to highly unreliable high-frequency (HF) radiotelephone service. Even where more reliable submarine voice communications services were available, the charge for a three-minute call between the United States and Europe was roughly $50. Translated into 1983 dollars, that would be over $50 a minute.
Following the success of Intelsat I ("Early Bird"), however, and with the growth of satellite communications and the rapid parallel development of submarine cable telecommunications, marked changes occurred. It is those changes, as well as the current trends in global satellite communications, that I shall focus on here. I shall also explore the past and coming trends in terrestial communications and the relationship between terrestial and space communicaitons.
The major cost components of satellite communications can be divided into three parts: the space segment; the earth segment and terrestial interconnections; and another area loosely referred to as "other costs."
The space segment can essentially be classified as the satellites and their up-keep. Let us first consider the satellite in terms of its subsystems and how these relate to its complexity, cost and performance. The first Intelsat satellite, "Early Bird," was a very small bird indeed compared with the satellite series that followed. Its communications package included only two 25-MHz transponders, while 80 percent of the mass and volume of the spacecraft was comprised of the "housekeeping" portion--the solar arrays, batteries, fuel and the spacecraft itself. From "Early Bird" to the Intelsat VI, satellites have increased dramatically in size, in communications capacity and in general complexity, but the design emphasis has switched from the housekeeping portion (which has shrunk from 80 percent to 56 percent on the Intelsat VI) to the communications portion (such as the antenna-feed structures). These changes make possible greater efficiency in the use of radio frequencies, greater communications capacity and, of course, greater cost efficiency (see Figure 1).
Projecting ahead, one may foresee future spacecraft that are, in effect, very large antenna structures and very large power systems connected by very small and compact electronic systems. The "black box" that constitutes the active electronic communications part of a satellite is thus following the same trend toward smallness that we have seen in the development of digital computers and very-large-scale integrated (VLSI) devices. It is easy to see the areas where significant cost economies might be achieved, such as the development of: high-efficiency power systems that produce higher rates of energy while requiring less mass; advanced antenna structures that can concentrate beams more directly to specific earth stations in order to achieve greatly expanded transmission capabilities and make possible much higher degrees of frequency reuse with minimum sidelobes; and monolithic integrated signal processing units of miniature size that can process and regenerate vast amounts of digital communications. Projections of some of the major technological advances that will help satellites become more efficient in future years are provided in Figure 2.
One thing is clear, however, and that is that there is no consensus as to "the" single best strategy for the telecommunications satellite of the future. Some see a bright future for massive multi-purpose space platforms, others for radio-frequency or optically connected clusters of satellites of more modest dimensions. But the trend is toward higher-capacity systems that operate at even lower unit cost (see Figure 3). A recent technical paper, for example, has stated that $350,000 per year per 36-MHz transponder is a figure of merit for a contemporary Fixed Satellite Service (FSS) spacecraft, and about $2 million per year per TV channel for a direct-broadcast satellite (Broadcast Satellite Service, or BSS). In both cases the figures relate to development and production costs, and not to amortized costs reflecting the cost of money or any operating, monitoring or in-orbit sparing costs.
The challenge over the next 15 years will be to reduce those figures to somewhere between $50,000 and $100,000 for FSS and to below $500,000 per year for a BSS TV channel.
Any discussion of space-segment costs must, of course, include launch services. The cost of "Early Bird" in orbit was about $7 million (in 1965 dollars), about half of which was for launch services. In 1983, that would be the equivalent of about $18 million. Despite the fact that Intelsat VI has 170 times the communicatioins capacity and 55 times the mass of Intelsat I, the cost for launching Intelsat VI is in the range of just seven to 10 times more. The development of reusable launchers such as the Space Shuttle and Ariane 5, of private launch vehicles and of improved perigee stages to move satellites from low-earth to geostationary orbit, are all indications that improved and highly flexible new launch capabilities can be expected in coming years.
Earth-segment costs also represent a significant portion of the overall satellite circuit cost. New and simpler earth station designs, based on offset feed and phased array devices, as well as improved manufacturing techniques, new materials and continued breakthroughs in solid-state electronics have made lower costs possible. Solid-state electronics--the same technology that has been applied to computers and to satellites--has probably been responsible for the greatest cost reduction, but a closelly related factor has been the development of advanced modulation and coding techniques that allow more and more communications services to be transmitted through the same channel.
How much have earth-station costs been reduced? One example is the 30-meter Standard A Intelsat earth station antenna, which began operation in the mid to late 1960s. At the time, an investment of $4 million to $10 million US (an equivalent of perhaps $10 million to $25 million US today) was required. A Standard A station in 1983 dollars now costs about $3 million to $5 million or, in "adjusted" dollars, the equivalent of two to three times less than it did 15 years ago. But this is only the tip of the iceberg. In addition to the Standard A, Intelsat has developed a whole new range of earth stations, including the 10-meter Standard B, and 13-meter Standard C, the 5 to 10-meter Standard Z for domestic services, and 3.3 to 8-meter Standard Es. A Standard E earth station is expected to cost between $100,000 and $500,000 (US).
Moreover, Intelsat is working to develop a Standard D earth station for what we call "basic user" services, where one or two voice circuits would be provided through a 4.5-meter terminal that possibly could be installed at a cost between $25,000 and $40,000 (US) per terminal. This trend will continue, stimulated in part by the new direct-broadcast antennas and "intelligent" date broadcast terminals, such as the new 0.6-meter terminal that the Intergovernmental Bureau for Informatics (IBI) recently tested in Italy using Intelsat satellites.
The third cost element that I mentioned--"other costs"--although frequently overlooked, is perhaps the most significant of all, for unlike the per-unit cost of satellite and earth stations, TTC&M (Tracking, Telemetry, Command and Monitoring) and "other costs" are going up--at least on a relative scale. As satellites become more complex and operate in multiple frequency bands and in a large number of different beams, TTC&M networks must become more sophisticated. For this reason, Intelsat has initiated research and development studies to determine to what extent spaced-based, rather than ground-based, systems might be able to perform these functions more reliably and at lower cost.
"Other costs" are not restricted to TTC&M, however. The so-called terrestrial tails" (telephone, data circuit and television channels from the earth station to end user and from end user to earth station) are a very significant part of the service.
The charges for the "terrestrial tail" often exceed the combined total for the earth station and space-segment charges. Then there are the billing, overhead and administrative costs; the cost of related plant and office investments, depreciation, local international switching and signaling facilities, repair and maintenance; as well as the "subsidy factor" for postal or domestic communications services. These vary greatly (depending upon the country involved).
In coming years, the largest cost component in international communications will be that of human labor, namely for marketing, billing, operations and maintenance. Because of the labor costs, as well as the cost of local switching, exchange equipment and other increasing "terrestrial tail" costs, Intelsat has introduced a new business service that provides a more complete end-to-end and fully automated service capability. This should result in lower costs for overseas telecommunications service to the end user, particularly as use of an equivalent voice circuit increase from five to six hours per day to 12 or more. I should encourage innovative new service, such as digitally compressed videoconferencing, digital voice, off-peak-hour electronic mail and store-and-forward voice mail, to develop much more rapidly.
It would, of course, be difficult to talk about the economics of communications of overseas communications without saying a few words about submarine cables and fiber optics. We at Intelsat, in fact, have a keen interest and healthy respect for this technology. Submarine cables, like satellites, have shown remarkable progress over the last 20 years. The first three transatlantic cables, with a combined capacity of 210 telephone circuits, were installed between 1956 and 1962. Then came Early Bird, with a capacity of 240 voice circuits, and the technology race was on. Since then, each successive submarine cable, up to the current SG-type of cable with a capacity of some 4,000 telephone circuits, has resulted in marked capacity increase, and increased economic efficiency.
The TAT-8 fiber-optic cable, to be installed in 1988/89, is an especially important development in overseas communications. We should, therefore, consider it an important benchmark, both technically and financially. It will cost in excess of $400 million (US) to instal and will have a digital throughput capacity of 560 megabits per second, or about 8,000 channels at 64 kilobits per channel plus a spare optic cable capable of sending 280 .megabits per second. It will have up to three multiple termination points in Europe and probably two in North America. With DSI and digital compression techniques, significant service capability increases of up to 36,000 voice circuits will be possible. With 28 participants in this extensive undertaking, the TAT-8 cable is the most ambitious submarine facility ever conceived and implemented for international communications.
In spite of these impressive statistics, a comparison of performances is enlightening. For instance, the Intelsat VI satellite, to be launched in 1986, if operated in an all-digital mode, would have a throughput capacity of approximately 3.5 billion bits of information per second, or on the order of five to seven times the capacity of TAT-8 (depending upon how you credit the space capacity). An in-orbit Intelsat VI satellite will cost well under half as much as a TAT-8 (namely, about $180 million versus $425 million--but, with an estimated in-orbit life of 10 years, will have only about half the lifetime of TAT-8. The Intelsat VI will provide hundreds of different pathways, both east-west in the North and South Atlantic regions and north-south in the Americas and Europe and Africa, rather than a single stream across the North Atlantic.
Fiber-optic cables are indeed an important new technology whose characteristics and capabilities make their installation and use quite logical. As in the past, cables and satillites will continue to be both competitive and complementary. Figure 4 provides some rules of thumb to help identify where the use of one of these technologies might be more appropriate than the other. What the chart does not address, however, is the increasingly important issue of communications retoration strategies.
In the past, Intelsat satellites have been an effective, readily available, cost-effective source of restoring service during submarine cable outages. At no time since 1969, when Intelsat service became truly global, has it been unable to restore failed cable service. During that same period, Intelsat has had to rely on its own satellite facilities to fully restore its own service capabilities simply because there was not sufficient spare capacity on submarine cables to effect restoration. No one knows for certain how sparing will be accomplished in the future, but the increasing size of both space and terrestrial systems seems to suggest some change will occur.
These guidelines for optimal utilization choice between satellite and terrestrial facilities, however, are not to be regarded as static, since future decisions regarding satellites versus submarine cables will have to rendered in an ever-changing environment. Several factors will favor the increased use of satellites: higher-powered satellites that can operate to increasingly smaller terminals located closer to the user; the combination of technological improvements described earlier that will serve to bring down the initial capital cost of a wideband transponder to perhaps $50,000 per year of lifetime; and the introduction of on-board signal processing regeneration that will allow more complex "networks" to be established. These new signaling and switching-systems-in-the-sky can make possible more cost-efficient satellites, particularly for "network" telecommunications.
As these on-board signaling and regeneration capabilities increase and as computing power capability on board the spacecraft starts to resemble the world's largest four-wire electronic exchanges, some of the old guidelines will be out the window. When on-board processing capabilities begin to reach a billion calculations per second or greater, the network architecture can be radically changed. Perhaps cryogenic Josephson junctions with super-conductivity will be used in electrical processing of on-board information. Perhaps, alternatively, light switching and processing, in effect a photonic computer, might be designed for on-board information processing.
With such breakthroughs, truly remarkable capabilities become possible. Already a dynamically switched TDMA matrix has been developed to handle 80000 telephone circuits through on-board processing. In the future, however, it is conceivable that one could even see networks as large as 10,000 earth stations interconnected within an interactive network that would make it posssible to service 50 million earth-station-to-earth-station pathways.
The future of telecommunications could, of course, also move the other way--particularly digital communications. Many national telecommunications systems are moving toward the concept of the "wired city" and "teleports" connected to these optical-fiber electronic highways. In the United States, AT&T has its Advance Communications System, while West Germany has its Broadband Integrated Glass Fiber Optical Network (BIGFON) project, Japan, France and other countries have comparable plans. These new terrestrial systems will be able to handle broadband information services, including videoconferencing, as well as other sophisticated business communications requirements.
The next 20 years, therefore, should be a very dynamic time in the field of telecommunications, with breakthrough after breakthrough in both terrestrial and space communications systems. It perhaps will not be possible to simply say that a terrestrial system can do a certain task to the exclusion of a space system, or vice versa. The question very likely will be: "What margins are required for terrestrial systems, for space systems, for hybrid systems?" "What are the relative costs of those various systems?" Or, "Will it most likely be a cost-effective hybrid space and terrestrial communications system that ultimately provides the greates redundancy, reliability and flexibility for the end user?"
Putting this discussion into proper perspective, one overriding conclusion is clear; namely, that regardless of whether one talks of space or terrestrial communictions pservices, costs will continue to drop--and to drop significantly. We hope the Intelsat's record of tariff reductions over the last 18 years can be duplicated again by the end of the 20th century.
In summary, the future is bright both for space communictions communications and terestrial communications. We shall see diversity, overl
aps and multiple solutions as to which facilities will be used. These solutions will vary from country to country and from region to region. If the Integrated Services Digital Network (ISDN) standard-making process, now underway within the International Telecommunication Union, is even reasonably successful, we will ultimately achieve a workable and cost-effective digital global telecommunications network that includes both satellites and fiber optics and, ultimately, this diversity will culminate in the long-heralded electronic global village, which will at last become a reality rather than simply a catchy slogan.
|Printer friendly Cite/link Email Feedback|
|Date:||Mar 1, 1984|
|Previous Article:||Satellites Launch New Telecom Opportunities.|
|Next Article:||Broadband Cable-Satellite Connection Key Link for Business Communications.|