COMMUNICATIONS SATELLITES, TECHNOLOGY OF

History

Early Theories. In 1895, a Russian scientist, Konstantin Tsiolkovsky, gave the world its first vision of a stationary satellite. He observed that an object orbiting Earth at 22,300 miles up would match the angular rotation of Earth and thus provide a seemingly stationary ”star” overhead.
Fifty years later, the English science fiction writer and Royal Air Force electronics officer Arthur C. Clarke expanded on this vision. In 1945, he indicated that such an object orbiting Earth at 22,300 miles up must also have its orbit in the equatorial plane to be considered stationary. Clarke called the object a ”satellite” and further noted that providing a satellite in this orbit with a communications repeater could produce a very valuable communications capability. Clarke named such satellites ”geostationary communication satellites.” Clarke postulated that three geostationary satellites spaced 120° apart would provide full global communications coverage for telephone and television service to the world’s population (1,2).
Clarke’s concept was ahead of the technologies needed to make it a reality. In 1945, the science of rocketry was still crude, and radio communications depended on large, unreliable vacuum tubes. However, rocketry and electronics technologies both began to advance rapidly, particularly after the development of the transistor in 1947. The first satellite launches, into a low Earth orbit (LEO), occurred during the Cold War following World War II. The United States and Soviet Union were intent on achieving military and technological superiority over each other. Satellites were seen as potentially valuable Cold War tools for reconnaissance and for maintaining reliable communications among far-flung troops.
The Earliest Satellites. On 4 October 1957, a Soviet satellite named ”Pros-teyshiy Sputnik” (”Simplest Satellite”) was launched into a low Earth orbit and began transmitting telemetric information for what would turn out to be 21 days. Sputnik weighed 184 pounds and was a sphere two feet in diameter. This dramatic Soviet achievement shocked the world, and particularly the U.S. public, which perceived it as a threat to national security. The event was likened to the surprise attack on Pearl Harbor, and it sparked a flurry of U.S. government activity in rocketry and satellites (3).
The rocket that lifted Sputnik into orbit was four stories high. Such heavy-lift launchers used technology similar to that of intercontinental ballistic missiles (ICBMs), so it was perceived that the launch vehicle was at least as important as the actual satellite. In the United States, the public feared that the Soviet Union’s ability to launch satellites translated into the capability to launch ballistic missiles that could carry nuclear weapons from Europe to America. The ”Space Race” had begun (4,5). The Sputnik launch led directly to the creation of the U.S. National Aeronautics and Space Administration (NASA) (6). In October 1958, the National Aeronautics and Space Act launched NASA out of the old National Advisory Committee for Aeronautics (NACA), which had existed since 1915. NASA became responsible for civilian space science and aeronautical research, whereas the U.S. Department of Defense continued to carry out defense work.
The launch of Sputnik II on 3 November 1957, was a second blow to America’s perception of itself as technologically superior to the Soviet Union. A third blow came on 6 December 1957, when the first U.S. satellite, Vanguard, exploded on launch. But on 31 January 1958, the United States launched Explorer I, which successfully transmitted telemetric data for 5 months. Shortly thereafter, in December 1958, America launched Score, the first satellite to transmit a radio broadcast, President Eisenhower’s Christmas message. In 1960, two more satellites followed, Echo I and Courier. On 10 July 1962, NASA launched the first non-government-built satellite, AT&T’s Telstar 1, a LEO spacecraft for relaying transatlantic television and data. In that same year, Telstar was used for the first transoceanic television ever transmitted, a live broadcast that commemorated the first anniversary of the death of U.N. Secretary General Dag Hammarskjold by a link-up of simultaneous ceremonies held at the United Nations in New York and Paris and Hammarskjold’s tomb in Sweden (7). Relay I, launched by NASA in December 1962, was built by RCA, also for transoceanic communications. Meanwhile, in May 1958, Sputnik III was placed into orbit by the Soviet Union. This was a large satellite for the time and demonstrated that the Soviets were ahead of the United States in heavy-lift rocketry.
Geosynchronous Satellites. All of these early satellites were nongeostationary and nongeosynchronous [a geosynchronous (GEO) satellite orbits at 22,300 miles, but is not necessarily geostationary, that is, limited to the equatorial plane—see discussion later]. They were launched into low-altitude Earth orbits because the rockets of the day could not propel the satellites into an orbit 22,300 miles up.
In the early 1960s, when U.S. rockets could at last boost a satellite into geosynchronous orbit, one of the most important questions centered on which was the best orbit to use for a communications satellite, low Earth orbit or geosynchronous. Low-altitude systems had the advantages of lower launch costs,heavier payloads, and relatively short radio-frequency propagative times. The main disadvantages were that many orbiting satellites were required to achieve continuous global communications, and these needed continuous tracking. Geosynchronous satellites, in contrast, had two key advantages. Only three satellites were needed for global coverage, and only minimal tracking was required. Their one primary disadvantage was relatively long radio-frequency propagative times. No one knew how the one-quarter-of-a-second transmission delay would affect the feasibility of using this orbit for telephony (see extended discussion later).
By 1959, a small team of scientists led by Dr. Harold Rosen from the Hughes Aircraft Company (now Hughes Electronics Corporation) was moving ahead, determined to create a geosynchronous communications satellite. By 1960, they had built a satellite prototype (8,9). Meanwhile, at AT&T’s Bell Laboratories, similar work was taking place under the direction of Dr. John R. Pierce and with at least as much zeal. Arthur Clarke would later name John Pierce and Harold Rosen the ”fathers of communications satellites.” Pierce’s team demonstrated the first active communications repeater, but Rosen and his team at Hughes are credited with making it possible, technically and economically, to have a continuous communications capability by satellite earlier than anyone thought feasible (10).
In August 1961, NASA contracted with the Hughes Aircraft Company for the first geosynchronous communications satellite, called Syncom (Fig. 1). Though the first Syncom launch failed in February 1963, a second attempt, the launch of Syncom II in July of the same year, succeeded. Syncom I had just one voice channel and was designed to weigh 86 pounds at the beginning of its life. Though it never attained orbit, it paved the way for Syncom II and Syncom III, satellites that, by August 1964, proved the feasibility and cost-efficiency of domestic and international satellite communications. The Syncom series also validated the concept of geosynchronous satellites, and by 1964, government and other users turned away from LEO satellites for voice, data, and video communications (11).
COMSAT, INTELSAT, and INMARSAT. The 1962 U.S. Communications Satellite Act had a profound impact on international satellite communications. It provided for the establishment of the Communications Satellite Corporation (COMSAT) (12), a privately financed and managed organization that had a minority of U.S. government representatives on its board of directors. AT&T emerged as COMSAT’s largest shareholder. COMSAT was created (1) to govern the operation of communications satellites and ground facilities used to transmit to and from the United States and (2) to develop and manage a new international communications satellite organization, which, in 1964, emerged as the International Telecommunication Satellite Organization, or INTELSAT. COMSAT wasresponsible for the procurement, testing, and launch acquisition of all INTELSAT satellites, and it owned 61% of the organization (13,14). The global, commercial INTELSAT cooperative was formed on the initiative of U.S. President John F. Kennedy, who, at the time COMSAT was created, said, ”I invite all nations to participate in a communications satellite system in the interest of world peace and closer brotherhood among peoples of the world.” INTELSAT was the first organization to provide global satellite coverage and connectivity. In its role as a commercial cooperative and wholesaler of satellite communications capacity, INTELSAT provides service through its signatories in member countries. Currently, COMSAT is the only U.S. signatory, though pending deregulation will encourage others (15).

Figure 1. Hughes engineers Dr. Harold Rosen (right) and Thomas Hudspeth hold a prototype of the geosynchronous Syncom satellite atop the Eiffel Tower during the 1962 Paris Air Show.
COMSAT’s initial capitalization of $200 million was considered sufficient to build a system of dozens of medium Earth orbit (MEO) satellites. These orbit the Earth at about 3000 to 7000 miles up; fewer MEOs are needed for global coverage than LEOs. In 1964, when COMSAT was in the process of contracting for its first satellite, two Telstars, two Relays, and two Syncoms had operated successfully in space. For a variety of reasons, including cost, COMSAT ultimately rejected a joint AT&T/RCA bid for a MEO system that incorporated the best of Telstar and Relay. Instead, COMSAT chose the geosynchronous satellite offered by Hughes, based on Syncom technology. Procured by COMSAT but transferred to the newly formed INTELSAT, the Early Bird satellite (also called “INTELSAT I”) was launched on 6 April, 1965, as the first commercial communications satellite. Built by Hughes, it was launched from Cape Canaveral on a Delta rocket and weighed 85 pounds—still about all that could be lifted to a geosynchronous orbit at that time by American technology.
Early Bird was designed to test the feasibility of synchronous orbits for commercial communications satellites and was a resounding success. The satellite provided 240 transatlantic telephone circuits capable of carrying that many calls simultaneously. This greatly increased telephone capacity across the Atlantic. Early Bird could provide almost 10 times the capacity of a submarine telephone cable for almost one-tenth the price and thereby helped prove the cost-efficiency of communications satellites. Moreover, the public readily accepted the transmission delay. Early Bird was also designed for transmitting television. Though built for just 18 months of life, it could still transmit live pictures of the Apollo Moon landing in 1969. By 1967, three of these first-generation INTELSAT satellites were operating over the Atlantic and Pacific oceans, providing ubiquitous global telephone and television communications for the first time (16).
In February 1976, COMSAT launched a new kind of communications satellite, Marisat, to provide mobile communications services to the U.S. Navy and other maritime and aeronautical customers. Subsequently, in 1979, COMSAT transferred Marisat to the newly formed International Maritime Satellite Organization, INMARSAT. Sponsored by the United Nations, INMARSAT has an intergovernmental structure similar to INTELSAT’s. Each signatory is required to provide an interface with land-based telecommunications networks and is assigned an investment share based on its actual use of the system. Today, COMSAT manages access to the global satellite fleets of both INTELSAT and INMARSAT on behalf of U.S. and foreign telecommunications operators that want to initiate or terminate their satellite transmissions in the United States COMSAT currently enjoys an exclusive relationship with both INTELSAT and INMARSAT in providing this service.
As of June 1999, INTELSAT owned and operated 24 satellites and had a membership of 143 countries and signatories. The highly successful organization, expected to be privatized soon, provides voice, data, video, and Internet services on a nondiscriminatory basis to more than 200 countries and territories. Since its formation, INMARSAT has greatly expanded its services. Today, the organization provides satellite global mobile communications services on land, sea, and in the air. For maritime users, INMARSAT supports phone, telex, fax, electronic mail, and data transmission. Aeronautical applications include flight-deck voice and data, automatic passenger telephone, fax, and data communications. INMARSAT’s land-based customers have access to in-vehicle and transportable phone, fax, and two-way data communications, position reporting, electronic mail, and fleet management for land transport. INMARSAT is also available for disaster and emergency communications, as well as news reporting, where alternative communications links are difficult to access or nonexistent (17). Initially, INMARSAT leased transponders on other organizations’ satellites, but in October 1990, it launched the first of its own satellites, INMARSAT II-F. As of June 1999, INMARSAT had 85 member nations and a global fleet of four geosynchronous satellites.
The arrangement, mentioned earlier, whereby COMSAT enjoys monopoly status as an access point for U.S. domestic and international communications organizations, along with other ”privileges and immunities” conferred on it by the U.S. government, is currently being challenged. In fact, there is a worldwide effort underway to push COMSAT, INTELSAT, and INMARSAT to serve their customers on a fully commercial basis, free of all government ties and protective legislation and regulations. Both INTELSAT and INMARSAT have embarked on this path with the New Skies and ICO satellite systems, respectively. Today, COMSAT is partly owned by Lockheed Martin, a U.S. aerospace corporation and satellite manufacturer. As of June 1999, Lockheed Martin was seeking to buy all of COMSAT, for which it needs regulatory approval.
It is inevitable that INTELSAT will evolve to privatized organizations. These will remain one of history’s best examples of successful international cooperation.
A Worldwide Industry. The use of communications satellites proceeded apace in the United States and other countries. In a number of cases, satellites offered cost savings and improved signal quality, compared with terrestrial and submarine cables. Importantly, satellite transmission was cost-insensitive to distance, as opposed to the way terrestrial transmission facilities were priced. Although the initial commercial launch vehicles and satellites were American, other countries had been involved in commercial and government satellite communications from the beginning. For example, as early as 1961, the German Post Office announced plans to construct a satellite-receiving ground station that could handle up to 600 phone calls simultaneously. The ground station interfaced with Telstar and Relay-type satellites. By the time Early Bird was launched, satellite Earth stations already existed in the United Kingdom, France, Germany, Italy, Brazil, and Japan.
In 1969, the Canadian government created Telesat Canada, a corporation charged with building and operating the world’s first national satellite network. The first satellite, Anik A (Anik means ”little brother” in the Inuit language), was launched in November 1972, and a three-satellite system was completed by May 1975. Anik was notable for introducing new spacecraft technologies, such as the shaped beam. Syncom, Early Bird, and Intelsat II used global beam technology, meaning that the satellites broadcast their signals across the entire third of the world they faced. Anik’s shaped beam permitted a pattern, or footprint, that covered only Canada (18). Following the success of Canada’s Anik system, both Europe and the United States developed domestic satellite systems. In 1974, Europe’s first telecommunications satellite, Symphonie, was launched. In the United States, Western Union contracted with Hughes for three Anik-type satellites to serve the U.S. market. The first of these, Westar I, was launched in April 1974. In the following year, RCA launched its network system, and in 1975, the first live commercial television program was transmitted in the United States. Also in 1975, Home Box Office became the first national cable TV network to be delivered by satellite—and thus began a whole new entertainment industry. By the end of 1976, there were 120 satellite transponders available in the United States; each could provide 1500 telephone circuits or one TV channel (19).
Indonesia, the world’s largest archipelago of more than 13,000 islands, launched its first Palapa satellite in 1976 and followed with another in 1977. Besides supporting telecommunications, Palapa-A1 and Palapa-A2 were intended for distance learning, including teaching the national language (20). By the mid-1980s, many nations had national satellite systems. These included Australia (OPTUS), Mexico (MORELOS), and Brazil (BRASILSAT). In 1999, this tally had grown to include most of the world’s developed and developing countries. In addition, there are numerous private regional satellite systems, such as
Eutelsat (Europe); SES Astra (Europe); Europe*Star; Nahuel (Central and South America); Galaxy Latin America; Thuraya (Middle East and parts of Asia, Africa, and Europe); ACeS (Asia); and AsiaSat, which delivers service to China, Thailand, Malaysia, Pakistan, Hong Kong, Burma, and Mongolia (21).
Today, INTELSAT continues to own and operate the world’s largest global satellite communications system (Fig. 2). However, there is now strong competition from such international entities as PanAmSat, which, it is projected will overtake INTELSAT in the near future. Loral, GE Americom, and Intersputnik also operate global fleets. Despite the early failure of Iridium (Motorola), Globalstar (Loral/Qualcomm) and ICO (a private INMARSAT spinoff) promise to provide worldwide satellite mobile services. Spaceway (Hughes), Teledesic (Boeing, Motorola, Matra Marconi Space), and Astrolink (Lockheed Martin, TRW, Telespazio) are among several companies that anticipate operating global satellite systems for high-speed data exchange, Internet applications, and interactive multimedia.
Satellite manufacturers and launch vehicle providers are found throughout the world, too. Satellite manufacturers in the United States, include Hughes, Loral, Lockheed Martin, TRW, Orbital Sciences, and Motorola. They are joined by non-U.S. competitors such as Daimler-Chrysler Aerospace (Germany), Alenia Spazio (Italy), Aerospatiale (France), Matra Marconi (France) with British Aerospace (UK), Thomson CSF (France), and Alcatel (France). In the United States, launch vehicles are provided by Boeing (Delta and Sea Launch) and ILS-Lock-heed Martin (Atlas). Other prominent launch vehicles include Ariane (France), Long March (China), ILS Proton (Russia), H-IIA (Japan), and Soyuz (Russia). It is likely that there will be consolidation of the many satellite and launch vehicle manufacturers and service providers in the future.

Figure 2. The tiny Syncom satellite (foreground) would fit into one of the INTELSAT VI fuel tanks to which Dr. Harold Rosen is pointing.
Today, communication satellites are a basic element of the world’s telecommunications infrastructure and an indispensable tool of the global marketplace. More than 325 commercial communications satellites—mostly GEOs—now orbit Earth. New applications for these satellites and those in LEO or MEO orbits are constantly evolving, based on new technological developments and increasing demand for space-based communications. Among the newest applications are satellite mobile radio, high-speed broadband data exchange, interactive multimedia, handheld global mobile telephony, and direct-to-home TV featuring high-definition pictures—HDTV (see fuller discussion in the article on commercial applications). These business and consumer services are provided through GEO as well as LEO and MEO satellites.
The early satellites had one television channel, weighed less than 100 pounds, and could require a 30-meter diameter antenna for reception. Today, a communication satellite can transmit more than 200 digital television channels, weigh 10,000 pounds, and deliver signals to one-half-meter Earth stations or palm-sized receiver/transmitter units.

Satellite Technology

Communications Satellites—Yesterday and Today. In a few decades, communications satellites have become an indispensable part of the world’s telecommunications infrastructure. They serve as repeaters in the sky and relay radio transmissions from/to terminals on Earth and in space, much as terrestrial microwave repeaters receive, amplify, and retransmit signals on Earth. During the years since Early Bird, communications satellites have grown in many aspects—size, weight, lifetime, power, and capacity. One way of illustrating this growth is by comparing Early Bird to a satellite designed for the early twenty-first century. The drum-shaped Early Bird was 71 cm (28 in) in diameter and only 58 cm (23 in) high. Its mass in orbit was 34.5 kg (76 lb). By contrast, a late-model, high-powered satellite called the HS 702 towers over Early Bird. In its stowed configuration, this Hughes-built satellite measures 2 m (6 ft 7 in) x 3.15 m (10 ft 6 in), and rises to a height of 3.6 m (11ft 11 in), not including its antennae. Deployed in space, it spans 40.9 m (134.5 ft) in length (Fig. 3). Its mass at launch is as much as 5200 kg (11,464lb), depending on the type offuel. Its dry weight (which excludes fuel) is as much as 3450 kg (7590 lb), two orders of magnitude greater than that of Early Bird. Early Bird was designed for 18 months of life, the HS 702 is designed for missions 15 years long. It can generate as much as 15 kW of direct current (dc) power, three orders of magnitude more than Early Bird. Its transponder capacity is nearly 100 times that of Early Bird.

Figure 3. Artist’s concept of the Hughes-built HS 702 satellite deployed in space shows how the long solar panels collect light that is used to generate 15 kW of dc power.
Orbits. Communications satellites travel in space along a gravity-induced path, or orbit, much as natural satellites orbit celestial bodies throughout the solar system (22-24). For the first 30 years of commercial service, communications satellite designers relied almost exclusively on geosynchronous or geostationary orbits. To observers on the Earth, a geostationary satellite appears to be “fixed” in space, although it is actually traveling at a speed of 11,062 km/h (6875 mph). From a position in geostationary orbit, approximately 35,880 km (22,300 mi) above the equator, a single satellite can illuminate a large oval-shaped area on Earth that is roughly one-third of the globe’s surface. Because the GEO satellite is ”fixed” in space, ground stations within the illuminated area do not have to rotate their antennae to track the satellite, as they historically had to do for satellites in non-GEO orbits. Eliminating this bit of complexity was significant, especially in the early years when the antennae ofground terminals had to be large—of the order of 30 m (98 ft) in diameter—to receive weak signals from low-powered satellites. Large antennae were necessary to compensate for the loss of energy transmitted across the lengthy geosynchronous distances because energy received by a ground antenna is inversely proportional to the square of the distance from the transmitting source. As the power transmitted from GEO satellites has increased over the years, satellite system operators have been able to reduce the size of ground antennae significantly.
At times, satellites can be operated usefully from nonstationary geosynchronous orbits. For example, when onboard fuel is nearing exhaustion, a GEO satellite can be allowed to drift north and south of the equator while still maintaining its longitudinal position. Under these conditions, its orbit develops a small inclination with respect to the equator that grows about 1° a year due to the pull of solar and lunar gravity. For some commercial applications, like TV distribution to cable operators, departure from the satellite’s “fixed” north-south position is acceptable. In this event, the satellite would appear to a ground observer to trace out a figure eight pattern in the sky; the midpoint of the eight is centered on the equator.
Until the advent of personal mobile satellite communications late in the 1990s, all commercial and most military communications satellites operated in geosynchronous orbits (25). The GEO orbit is also favored for government weather, communications, and defense early warning satellites. However, a few of the first experimental military communications satellites and several precursors of today’s commercial communications satellites were placed in much lower altitude orbits. The first version of the U.S. Defense Department’s Strategic Communication Satellite System (DSCS 1) operated from an orbit that combined features of medium-altitude and geosynchronous orbits. A variety of other orbital geometries that range across many combinations of altitudes and inclinations, including widely employed 90° polar orbits, have been popular for weather, navigational, Earth resources, and surveillance satellites. The satellites of the Global Positioning System (GPS), for example, operate in six planes defined by 12-hour circular orbits—each plane separated from its neighbor by 55°—and are located 20,178 km (12,541 mi) above Earth (26).
In recent years, a number of non-GEO orbits have gained currency for such applications as personal mobile satellite communications (27,28). These include LEO and MEO orbits. The altitude of LEO orbits usually ranges from about 500 km (311 mi) to 1500 km (932 mi) above Earth, whereas MEO orbits are roughly 5000 km (3,107 mi) to 12,000 km (7457 mi). Orbital configurations at these lower altitudes are carefully chosen to avoid exposing a satellite’s electronic components to life-limiting radiation in the Van Allen belts, unless the operator is willing to bear the cost and weight of radiation shielding. The belts are between about 2000 km (1243 mi) and 8000 km (4971 mi), and also above MEO, but well below GEO altitude. Still another type of orbit, the high Earth orbit (HEO), a form of the highly elliptical, 12-hour Molniya orbit extensively employed by reconnaissance satellites during the Cold War, has been adapted for broadcasting CD radio from space.
The historical attraction of GEO orbits is that they can provide service with few satellites and operate through fixed antennae on the ground. Moreover, they provide ubiquitous service for users throughout their broad coverage area. A satellite operator can offer service across nearly a third of the globe using one satellite—and can serve the entire Earth, excluding sparsely populated polar regions, with three. As the altitude of an in-orbit satellite system decreases, more satellites are needed for continuous global coverage because of the shrinking area they can illuminate on Earth at any given time. However, the satellites themselves can be smaller and lighter, and therefore less costly than modern GEO satellites (29). The typical LEO satellite is shorter lived because of the effects of atmospheric drag and radiation at low altitude. Hence, it needs to be replaced more frequently than its GEO counterpart, adding to system cost. The cost of launching LEO satellites, however, is lower because they require less energy to boost them into low-altitude orbits. But the lower the altitude, the greater the need for extensive ground networks or complicated signal handoffs from one satellite to another, as the satellites pass within view.
This is so because lower altitude systems remain in view of the mobile satellite system user for very brief intervals, perhaps only minutes. By contrast, higher altitude satellites pass much more slowly through the viewer’s field of vision. Consequently, at lower altitudes, a typical call will have to be handed over from one satellite to another more frequently than at higher altitudes. The greater the frequency of handovers, the greater the chance of a phone disconnect. Often, low-altitude satellites will appear close to the observer’s horizon, increasing the likelihood that an obstruction such as a building will interrupt the call. This in turn means that the higher the satellite’s elevation angle (i.e., the angle from the observer to the satellite) at any given time, the less likely that the call will be lost due to surrounding obstructions. Lower altitude satellite systems add to the cost and complexity of the ground infrastructure. Communications Links, Frequencies, and Bandwidths. The communications link between satellite and ground station can be represented by a power balance equation expressed logarithmically in decibels (dB). This relationship takes into account the predictable factors in the communications loop. It indicates that the power received at the receiver is equal to a summation of all of the gains and losses in the link. Thus, power received equals the transmitted power, minus the waveguide losses, plus the gain of the transmitting antenna, minus the propagative path losses, plus the gain of the receiving antenna, minus the receiver waveguide losses. The free-space path loss is substantial and is fixed for a given frequency and distance. At GEO altitude, it amounts to 200 dB at the 12-GHz transmission range of a direct broadcast satellite and 196 dB at the 4-GHz band of a video distribution satellite (30). The measure of the difference between the power actually received and the threshold power required for reception is called the link margin. System designers try to provide sufficient link margin to ensure successful communications.
Frequencies allocated for use or potential use by satellites extend through a number of bands, or range of frequencies (31). Within those bands, separate portions are allotted for communications uplinks to and downlinks from the satellites to keep the two separated from one another. The satellite’s transponder translates signals from the uplink frequencies into downlink frequencies. The bands are identified by letter designations, a practice derived from the World War II lettering scheme for military electronic equipment. Thus, satellite frequencies extend from longer wavelengths at L band up to very short wavelengths at Q band (32). Satellite communications service for mobile users, such as those aboard ships, barges, and oil rigs, are in the L band (1-2 GHz), where they have replaced unreliable short-wave radio. Other mobile services, including personal mobile communications, are in the S band (2-4 GHz) as well as the L band. Fixed satellite service (telephony, data, and facsimile communications through Earth stations at fixed locations) is in the C (4-8 GHz), Ku (12.5-18 GHz), K (18-26.5 GHz), and Ka (26.5-40 GHz) bands. Broadcast satellite service (direct-to-user TV) is in the Ku and K bands. Military communications satellites use the X band (8-12.5 GHz) for fixed satellite service and Ka and ultra-high-frequency (300 MHz to 3 GHz) bands. Other bands in the high millimeter range, V (40-50 GHz) and Q (above 50 GHz) have promising applications for transmitting large quantities of information to small antennas.
Frequencies in the Ku, L, and C bands are regarded as prime real estate for communications satellites. C-band frequencies are most commonly employed by commercial communications satellites. Among available frequencies, the C band is least affected by man-made noise and atmospheric attenuation. But where broadband and high capacity are needed, the higher frequencies located in the Ku and Ka bands become more desirable. Portions of the Ku band, in particular, are widely used for business communications through small Earth terminals and for high-power TV broadcasting. The shorter frequencies (longer wavelengths) below the L band are subject to ionospheric disturbances that cause fading and other random signal disruption. This explains why L-band mobile satellite communication service quickly supplanted short-wave communications for ship-to-ship and ship-to-shore applications in the 1970s.
Bandwidth, the capacity of a satellite communications link, is the spread of usable (i.e., as allocated by regulation) frequencies available for transmitting intelligible information. Available bandwidth is divided among the transponders contained in the satellite’s communications payload. Transponders (or repeaters) are electronic devices that receive radio signals, amplify them to increase their strength, and transmit them on command at different specific frequencies (33). At the heart of the transponder are either traveling-wave tube amplifiers (TWTAs) for high-power, higher frequency applications or solid-state power amplifiers (SSPAs) for lower frequency, usually lower power applications. Communications satellites that orbited in the late 1990s typically carried 48 transponders. The newest satellites entering service at the turn of the century offer as many as 94 transponders, plus backups. By the late nineties, the satellite fleets of major international satellite service providers such as INTELSAT and PanAmSat each provided service via about 800 orbital transponders.
How bandwidth is divided among transponders reflects a balance of factors, including the satellite’s power resources, the number of TWTAs that can be carried, and the desired power per channel, or power per transponder. The following is an illustration. A satellite that can generate 8000 W of dc power might supply 7000 W to its communications payload, mainly for the TWTAs. If the TWTAs’ portion of that figure—say, 6800 W—is converted to radio-frequency (RF) power at 60% conversion efficiency, there will be 4080 W of RF power available from the TWTAs. If satellite designers want 100 W of power per channel, there is sufficient power for 40 channels. For 200 W per channel, 20 channels can be created. If the choice is 20 channels—10 channels for each of two polarizations—the total available bandwidth will be divided by 10. If the bandwidth (capacity) available within the allotted ”usable frequency” range is 250 MHz, there will be 25 MHz available per channel. Some of that bandwidth, perhaps 10%, has to be reserved for guardbands that separate the channels.
Consequently, the usable bandwidth will be 0.9 times 25, or 22.5 MHz per channel.
Onboard Antennas and Spot Beams. The area on Earth illuminated by communications satellites, or coverage area, is commonly referred to as the satellite’s footprint. To improve efficiency, satellite designers usually configure antenna patterns to illuminate specific high-revenue-producing areas on the ground, or a particular country or group of countries for a domestic or regional satellite system. Traditionally, shaping beams to match the contours of specific areas on Earth requires directing energy from the transponders at a sculpted parabolic reflector by an array of feedhorns (Fig. 4) controlled by a beam-forming network. The reflector then collimates the energy on Earth. This is a complicated, costly process, made even more so as customer requirements have grown. As an alternative, designers can use mathematical techniques to contour the otherwise smooth parabolic reflector, so that it can produce the desired pattern when fed by only a single horn. The reflector surface is mechanically shaped and creates a rippling or dimpled appearance. The dimples are precisely positioned by extensive computation to produce the desired pattern, thereby supplanting complex multiple feed arrays for directing radiated energy in the desired pattern. Elimination of the feed arrays reduces weight and expense.

Figure 4. Feedhorn array that helps generate spot beams for INTELSAT VI satellite is checked out in simulation laboratory.
There is increasing use of reconfigurable spot beams, narrowly shaped beams that illuminate specific areas, thereby increasing the power concentrated in those areas. By using onboard digital processing, modern satellites can generate many, even hundreds, of reconfigurable beams simultaneously. This has created the opportunity for satellite operators to obtain greater bandwidth by reusing the limited amounts of assigned spectrum. Using the same frequencies a number of times does not create interference if these frequencies are confined within spot beams directed at geographically separated regions (34). Special Challenges: Attenuation and Echo Cancellation. Attenuation of electromagnetic energy becomes increasingly severe at higher frequencies above 12 GHz, as signals are absorbed by the atmosphere. Some spectral points experience more severe effects than others (35). At 22 GHz, energy is almost totally absorbed by water vapor, and at 60 GHz by oxygen, rendering these and some other higher frequencies useless for communications between satellites and Earth. Attenuation of microwave signals by rain droplets in the atmosphere can become a severe problem at higher frequencies because of the relatively large size of the droplets compared to the diminishing size of wavelengths at these frequencies (36). The droplets absorb and scatter the microwave energy. Rain attenuation in the Ku band can account for about 2 dB in signal loss and three to four times that amount in the Ka band. The presence of rain attenuation is not accounted for in the power balance equation because of the random nature and unpredictable magnitude of rain cells. Systems have to be designed with sufficient link margin to accommodate rain attenuation. This could mean increasing transmitter power and/or the gain of the transmitting antenna. The amount of attenuation will vary with geographic area and seasonal rain activity. A system designed for 99.5% availability—where attenuation is less than 2dB in 99.5% of the time—could experience about 4 hours of outage in the months of heaviest rainfall. Under these conditions, a viewer’s TV picture could freeze, and the sound would stop. A large telecommunications installation typically would need a 6- to 8-dB link margin to be certain that it could offset the effects of random rainstorms. Or the attenuation could be averted by using spatial diversity, transmitting from the satellite to two widely separated Earth stations (37).
Despite rain attenuation in space-to-Earth or Earth-to-space links, Ka-band frequencies are successfully used in space-to-space communications as intersat-ellite links. As cross-links for communicating between satellites, they are free from rain attenuation. Such links are used by some military satellites and new commercial personal mobile communications satellite systems. Intersatellite cross-links are transponder links that transmit and receive information between satellites, thereby establishing connectivity among satellites in a system. In military communications applications, cross-links permit authorities to pass information securely from one GEO satellite to another in achieving global or near-global connectivity without fear that potential enemies will eavesdrop on or jam these links. In the commercial satellite-based personal mobile communications systems coming on-line, cross-links are a key to providing global coverage. For example, each satellite in the system’s six orbital rings is equipped with four Ka-band (23.18-23.38 GHz) antennae, two fixed and two gimbaled. A satellite can communicate directly with the satellite before and after it in the same orbit through the fixed antennae. The gimbaled antennae on a satellite allow communicating with the two neighboring satellites on either side in adjacent orbital planes. A user’s call can be relayed via the cross-link from satellite to satellite until it reaches a satellite within reach of its destination. At that point, calls are downlinked directly to the desired parties, if they are system subscribers, or to local gateways and through the public switched telephone network to the intended individuals. Ka-band frequencies are expected to have widespread applications in broadband, high-power commercial satellite systems in the future.
Digital echo cancellers have been developed and installed in digital phone exchanges and satellite networks to eliminate the once-troubling problem of voice echoes on satellite-relayed telephone calls (38). The echoes were produced by the reflection of a party’s speech created by a hybrid in the telephone network. The hybrid’s function is to route energy between the two-wire phone pair connecting a caller in an office or at home to the four-wire trunk line on longdistance calls. Additional hybrids that may be present between the two parties in the network add to the problem. The difficulty is further compounded in a GEO satellite connection by the nearly 36,000-km separation between the speakers and the satellite. The round-trip propagative time for electromagnetic energy across the lengthy distance creates a 260-ms signal delay. In a satellite hookup, telephone parties could be disturbed by echoes of their own voices delayed by 260 ms in the midst of phone conversations. Digital echo cancellers in the form of software or firmware solve the echo problem by sampling a digitized version of speech and mathematically removing the echo. The echo is suppressed by a factor of 10,000 (39). Though the echo is essentially gone, the delay caused by the signal transit time to and from a GEO satellite is not. Some people are annoyed by the delay, but various studies indicate that 90% of telephone subscribers have no serious objection to it.
Ground Antennas (40). As satellites have grown in size and power, the size of Earth stations for accessing a satellite’s communications capacity has dramatically declined, giving users greater flexibility, easier access, and lower costs. The antennae of large communications facilities today are about 3-9 meters (9.8-29 ft) in diameter for operation in the C and/or Ku band. At Earth stations, terrestrial signals typically are multiplexed, encoded, and modulated in the baseband section of the ground terminal. The signals are upconverted and amplified before transmission through the antenna to a satellite. Weak signals downlinked from the satellite are amplified through a low-noise amplifier, down-converted, and passed to baseband equipment for demodulation, processing, and interfacing with the terrestrial networks.
Customer premises antennae for offices or homes measure about a meter (3 ft) or less in diameter. Individual subscribers can receive TV programming broadcast to their homes by satellites operating in the Ku band through antennae as small as one-half meter (18 in) in diameter for some systems (e.g., Hughes’ DIRECTV—see discussion in the article on Commercial Applications of Communications Satellite Technology). Subscribers to satellite-based personal mobile communications systems (e.g., Globalstar, Thuraya, and ICO) can speak to other parties via L-band satellites through handheld, cellular-like phones that have small omnidirectional antennae. The appearance of small, compact ground terminals, called very small aperture terminals (VSATs), has led to the widespread use of satellite networks by businesses, schools, governments, and telecommunications organizations throughout the world. Typical VSATs vary in size, depending on application, but generally range from 0.85 m (2.8 ft) to 1.2 m (3.9 ft) in width or diameter for Ku-band systems.
Arranged in what is called “star,” or “hub-and-spoke,” network architecture, VSATs provide multipoint communications between a hub station at an entity’s headquarters or data center and a multiple number of remote sites. A computer at the hub handles switching for voice or video. The hub station broadcasts data via satellite to the VSATs at various remote sites. The terminals of individual remote sites are smaller and less powerful than the higher power, more expensive one at the hub. The remote sites can complete a return link to the hub through the satellite, thereby bypassing any terrestrial link, but their signals are too weak to be received by, or to interfere with, other sites in the network. (More expensive, “mesh” networks enable all sites to communicate directly with each other.) The “star” concept enables users to balance the higher cost of the hub against that of multiple lower cost VSATs. Manufacturers have reduced the size and weight of VSATs and increased their capability by adding more processing power. Size reductions help in two respects: the smaller the size, the lower the cost of installation on a customer’s premises, and the smaller the area occupied by the VSAT.
For all satellite applications, satellite systems are controlled and monitored from the ground by skilled technical personnel through one or more tracking, telemetry, and command (TT&C) stations. INTELSAT, for example, operates six such sites internationally to ensure reliable, line-of-sight contact with its entire orbital fleet (41). The exact positions of the satellite are determined from data secured by on-site tracking antennae. Personnel, consoles, and data processing equipment are housed in a satellite control center, which may or may not be collocated with a TT&C station. The satellite control center is manned and operated 24 hours a day throughout the year.
Spacecraft Design. Communications satellites have two fundamentally different designs. One is the spin-stabilized satellite, or spinner; the other is the three-axis, or body-stabilized satellite (42). The spin-stabilized satellite has a cylindrically shaped body that is spun about its axis, typically at a rate of 60 rpm to achieve stability inertially, much like a toy top becomes gyroscopically stable by spinning. The attitude and orientation of the satellite can be altered by a number of onboard thrusters. Almost the entire satellite spins, with the notable exception of its antennae, which are despun and pointed in a fixed direction toward Earth. Subsequent innovations in this concept led to a dual-spin design in which the entire payload, not just the antennae, is despun. To achieve stability, the body-stabilized design relies either on an internal gyro, called a ”momentum wheel,” or a set of reaction wheels. The satellite’s control system compensates for changes in attitude by applying small forces to the spacecraft’s body.
The spinner design is the simpler of the two, and that simplicity translates into relatively low cost and long life without much ground intervention. The spinner employs a novel control system to maintain satellite attitude, or orientation, in space. A single axial thruster pulsed in synchrony with signals from an onboard Sun sensor controls the attitude of the satellite’s spin axis. When that thruster is operated in a continuous mode, it provides velocity control around the spin axis. Pulsed radial jets control velocity in a particular direction. The control concept is covered by what has become a seminal patent, referred to by the name of its inventor, Donald Williams, and validated in extensive legal proceedings. Prime power for the satellite comes from solar cells that cover the circumference of the spinning cylinder. More and more power has been squeezed from this design by enlarging the cylinder and by such steps as adding a second concentric solar-cell-covered cylinder that telescopes out from the first in space. But covering the spinning cylindrical surface with solar cells means that, at any given time, the Sun is illuminating no more than a third of the surface solar cells. The remaining two-thirds are not converting solar energy into electrical energy. This relatively inefficient use of energy resources becomes less acceptable when higher satellite power levels are necessary. As the power required exceeds 1 or 2kW, the body-stabilized configuration becomes the preferred choice.
Because the three-axis stabilized spacecraft does not rely on a spinning body for stabilization, the satellite can assume any convenient shape, usually a box, for the convenience of its communications function. One surface of the box, on which antennae are mounted, is oriented toward Earth. Solar cells are mounted on flat panels that extend in wings from either end of the body and are oriented toward the Sun. The solar panels can be kept pointing at the Sun by motors aboard the satellite, as it revolves about Earth.
In addition to more efficient use of its solar cell resources and its potential for higher power levels, the body-stabilized satellite provides better pointing accuracy. This permits more accurate pointing of the satellite’s antenna beams. Otherwise, the size of the antennae would have to be enlarged to ensure sufficient gain across the coverage area. When its solar wings and antennae are folded in a stowed position, the body-stabilized configuration lends itself to more efficient use of the volume within the shroud of a launch vehicle. And the box-like shape has more surface space for mounting antennae, permitting designers to employ more complex antennae. The drawbacks of the three-axis design stem from its relative complexity. Its momentum wheels have to be controlled in speed and pivoting. The satellite has to be commanded frequently, requiring an onboard computer. A typical spin-stabilized communication satellite can now be constructed in about 60% of the time required to produce a body-stabilized version. The difference in time is a measure of both the greater capability of the latter and the greater simplicity of the former.
Spacecraft Subsystems. An active satellite consists essentially of two parts—a payload and a bus. (A passive satellite is one that does not generate energy; an example is the Mylar-coated reflective balloons considered promising candidates for communications relay in the early 1960s. The payload contains the satellite’s communications equipment and antennae that create an infrastructure for communicating with users throughout a continent or in regions or countries where service is supplied. The bus has the task of protecting the pay-load during the demanding launch period, placing the payload into its assigned orbit or orbital slot, and maintaining it there. The bus supports and maintains the payload throughout its lifetime (43).

A communications satellite contains seven subsystems (44):

• The communications subsystem (“payload”) contains the satellite’s radio-frequency equipment. A wideband receiver at the front end of the subsystem accepts incoming communications channels that occupy a specified band of frequencies. Then the channels are separated according to frequency by a multiplexer, or bank of filters, and apportioned among the payload’s various transponders. After amplification in the transponders, the channels are re-combined by another multiplexer for retransmission to the ground.
• The power subsystem generates, regulates, and controls power obtained from the solar arrays and onboard batteries primarily for use by the communications payload. This subsystem also maintains operation of the satellite during periodic solar eclipses.
• The attitude control subsystem senses any deviations from proper pointing directions and keeps the spacecraft and the antennae pointing in the correct directions—the solar arrays pointing toward the Sun and the radiators away from the Sun.
* The propulsion subsystem generates thrust to place a GEO satellite into a desired orbital slot and to adjust its position periodically to offset movements in the (1) north-south direction due to solar and lunar gravitational attraction and (2) east-west direction due to the oblateness of Earth’s poles. The last-named function is called station keeping. GEO satellites contain either a solid rocket apogee kick motor (AKM) or a liquid bipropellant (separate fuel and oxidizer) system. The function of the AKM (45) and in part that of the bipropellant system is to insert the satellite into geosynchronous orbit when it reaches the apogee of a geosynchronous transfer orbit. The satellite is placed in an elliptical transfer orbit by a perigee kick motor during the final phase of the launch sequence. Besides performing the apogee kick function, the bipropellant system also helps raise the perigee of the transfer orbit to coincide with its apogee in geosynchronous orbit, a process called orbit raising. It also handles the station keeping duties. On satellites that have an AKM, a monopropellant system performs orbital positioning and station-keeping duties. The tankage, valves, lines, thrusters, and fuel of this subsystem account for a significant portion of the mass of a satellite at launch and even after initial insertion into GEO orbit.
* The thermal control subsystem radiates the heat generated onboard the satellite into space. The thermal environment inside the satellite is kept at room temperature and all excess heat has to be radiated from the satellite. For this purpose, the subsystem uses such devices as thermal blankets and reflective mirrors. If the satellite has an AKM, as spinners do, an insulating wall and thermal barriers protect components from heat generated by the motor firing. New satellite designs are adding more TWTAs that have higher power outputs to their payloads. Despite increases in TWTA efficiency to as much as 70%, the remaining 30% is generated as useless heat that must be removed.
* The TT&C subsystem enables ground personnel to monitor the health and status of the satellite and issue commands to the satellite. It telemeters to the terrestrial TT&C station information regarding satellite temperatures, remaining fuel, TWTA performance, and pointing directions. The command portion accepts signals from the ground for controlling housekeeping functions, recharging batteries during solar eclipses, and dumping the energy buildup from the momentum wheels ofa body-stabilized model. When the satellite comes to the end of its mission life as onboard fuel nears depletion, it can be commanded through the TT&C subsystem to deorbit and turn off its communications subsystem.
• The structures subsystem is the chassis that provides physical support and protection for sensitive equipment during launch when the satellite must survive the effects of severe acoustic and vibrational forces. It employs a truncated cone or trusswork with panels for mounting bus and payload electronics.
Reliability, Lifetime, and Cost. Satellite designers constantly seek to improve reliability so that service delivery by communications satellites is ensured throughout their contracted mission lives. Satellites have to operate in space without physical intervention for as long as 15 years in the case of newer GEO models. And the space environment is unforgiving; such phenomena as ionizing radiation pose hazards for sensitive electronics. Satellites are usually designed with redundancy for all critical components to prevent catastrophic failures. The increased reliability achieved through the years is attributable to numerous factors. Among them are better designs, fewer parts, improved manufacturing processes, and fewer electronic interconnections due to the application of semiconductor chips that have high circuit density. Potential life-limiting spacecraft elements are batteries, lubricants, thrusters, and TWTAs. Mechanical deployments of antennae and solar panels are another source of possible difficulties. Once a matter of great concern, TWTA technology and manufacturing know-how have advanced to the point where the chances of losing a transponder because of a TWTA failure are slim. Like other portions of a satellite, transponders are designed with redundancy to permit continued operation in the event of a failure. Now, transponders need only a single TWTA to back up three or four active TWTAs, not the one for two ratios of two decades ago. The TWTA suffers from only a single wear-out mechanism—its cathode. But the design of this element is well proven by millions of hours of actual operating experience. The infrequent failures that occur result from processing errors, not design inadequacies. This compels the tube manufacturer to maintain constant vigilance over production processes.
Commercial GEO satellites are actually designed for a lifetime as long as 18-20 years, but this life span, or design life, is limited by practical matters, such as the inability to replace degraded components and battery cells in an orbiting satellite. The satellite’s design life is also limited by the onboard fuel supply. Therefore, the actual mission, or operational, life, of necessity, is shorter than the design life—no more than 12 to 15 years. Satellites in LEO orbits have shorter operational lifetimes—perhaps 5 to 8 years—limited as they are at very low altitudes by the effects of atmospheric drag and radiation. The contracts of GEO satellite manufacturers with their customers generally contain performance incentives that reward the builder for achieving specific operational lifetimes. Just as Early Bird surprised its owners by unanticipated longevity, many other commercial satellites have as well. For example, Marisat 2, one of INMARSAT’s satellites for mobile communications, had a 5-year mission life when placed in orbit in 1976. Twenty-two years later it was still providing service for the international maritime consortium, albeit from an inclined geosynchronous orbit.
Buying a satellite system involves a large investment by a communications entity. For instance, one major mobile communications satellite system, Thuraya, that was built in the late 1990s has an estimated value of $1 billion. The figure includes the cost of two GEO satellites, the launch of one, ground facilities, training, and user equipment. (See further discussion of Thuraya in the article on Commercial Applications of Communications Satellite Technology.) The customer’s cost of communications satellites, as measured by several key parameters, has been declining as a consequence of enhancements in satellite power, bandwidth, and lifetime. Thus, the satellite’s price per kilowatt of power, price per transponder, and price per transponder year are all declining.
The costs of launching a satellite, the associated launch services, and insuring the launch and in-orbit satellite performance are a significant portion of the total price of acquiring and orbiting a satellite. The cost of insurance varies with the satellite, launch vehicle, recent claim experience, and available underwriting capacity, so that it varies to reflect current circumstances. During a period of 20 years, launch insurance premiums have varied from 6-20% of the insured value; the average is in the 10-13% range. In general, acquisition of the satellite represents slightly less than half the cost of a satellite in orbit; launch services and launch insurance account for the remainder.

New Technologies

Digitalization. The latest communications satellites incorporate greater amounts of digitalization to enhance their capabilities and flexibility (46). Digital signal processors of increasing capacity began to appear in satellite payloads during the 1990s. They can perform such roles as routing signals to any one of scores of individual spot beams and controlling and forming the beams. The beams are generated by a planar phased array antenna that contains multiple elements. In the process, called digital beam-forming, each element of the array captures a portion ofthe signal. The processor computes and controls the relative phase and amplitude of each element and can electrically bend the antenna beams into any desired shape. In this digital beam-forming antenna, the outputs from the array are sampled by an analog-to-digital converter and are stored in a processor that computes the phase changes needed to bend the antenna beam into the desired shape. A nearly unlimited number of these virtual antennae can be created. Each of them, shaped differently and pointed in a different direction, emanates from the one multielement phased array. They can be repointed quickly because the process of generating them is mathematical and there is no movement of mass involved in repointing. The ability of the processor to generate as many as 200 or 300 spot beams, to reuse limited frequencies extensively, and to support thousands of voice channels simultaneously makes possible such new satellite services as personal mobile telephony, broadband data exchange, and multimedia communications. Digital processors give operators the freedom to satisfy changing customer demands by reconfiguring the power level, frequency, and beam shape after a satellite is in orbit (47).
Four separate generations of digital processors evolved in rapid sequence during the initial 10 years after they were introduced into communication satellites. They demonstrate successive improvements in circuit density, decreasing power consumption, and compactness that led to a progression in capability. The first generation, for example, for a military payload, used 13 different types of application-specific integrated circuits (ASICs); each circuit had about 15,000 gates. By the third generation, for application in a personal mobile communication satellite (Thuraya), the number of ASIC designs dropped to nine, but their density rose to 1.5 to 3.5 million gates per ASIC—a 100-factor jump over the first-generation level. By the fourth generation, for application in a broadband, multimedia satellite (Hughes’s Spaceway, see discussion under ”Commercial Applications of Communications Satellite Technology”), the ASICs count climbed to the four-to-seven million-gate range. Processor power consumption declined by approximately 50% from generation to generation. Consequently, the power required by the fourth generation was merely 13% of that of the first. The fourth-generation design provides 960 times the processing capability of the first, measured in raw processing power, or number of tera operations per second.
In a satellite application, a digital processor functions as a massively parallel digital computer with analog inputs and outputs. It converts analog radio signals received from individual terminals into numbers, performs the necessary computations entirely in the mathematical domain, and reconverts the resulting data into analog signals for retransmission to the terminals. This is called a “demod/remod” design. A demod/remod digital processor also performs filtering, switching, demodulation, modulation, beam-forming, and other functions without introducing distortion or drift. It can correct errors, reconstruct corrupted signals, and eliminate noise. The demod/remod design relies on known characteristics of the user’s signals. An alternate design, the transponded system, knows nothing about the format of incoming signals and is insensitive to changes in them. Processors of this design perform little or no filtering. Also, they cannot flexibly allocate downlink power to accommodate different sizes of user terminals.
Power. Several technical innovations have contributed to the ongoing trend toward increased satellite power. The introduction of gallium arsenide solar cells marked a turning point in efforts to improve the efficiency with which photovoltaic cells convert solar light into electrical energy. Gallium arsenide cells achieved efficiencies of 21.6%, nearly doubling the 12.3% figure obtained from traditional silicon cells. The increase enabled designers almost to double the dc power output from solar arrays of comparable size and mass. The greater power can be translated into an increase in the capacity ofthe communications payload. Alternatively, when the extra power is not needed, the size and mass of the satellite solar cell arrays can be reduced, thereby decreasing satellite mass and possibly launch vehicle cost.
The initial gallium arsenide cells were dual-junction devices that have two semiconductor junctions—a gallium arsenide layer on a single crystal germanium substrate and a gallium indium phosphide layer atop the gallium arsenide. The improved efficiency stems from the ability of each layer to convert a different part of the light spectrum into electrical power. When the Sun’s rays strike the top layer of the cell, shorter wavelengths are converted to power. The top layer is transparent to the longer wavelengths, which penetrate the gallium indium phosphide layer and strike the gallium arsenide layer, where they too are converted to electrical power. The cells are grown in an epitaxial chamber by a gas reduction process. Continuing work has led to triple-junction gallium arsenide cells. These enhanced the efficiency of earlier gallium arsenide cells by 20%, and brought conversion efficiency up to 26.8%. Four-junction cells are expected to reward satellite manufacturers with still higher solar cell efficiencies in the early years of the new millennium. The solar arrays of the new body-stabilized Hughes HS 702 satellite employ gallium arsenide cells, as did several other satellites in the late 1990s. Gallium arsenide cells are an important contributor to the satellite’s ability to generate as much as 15 kW of end-of-life (EOL) power. Satellite builders generally specify power levels expected at the end of the satellite’s lifetime, as opposed to higher power levels at the beginning of life (BOL). The difference reflects anticipated degradation in the solar cell output due to the deleterious effects of X rays, solar protons and electrons, and other particles encountered in different orbits.
Innovative solar concentrators mounted along the satellite’s solar wings add significantly to the power generated by the HS 702 satellite’s solar cell panels. The angled solar reflector panels add 50% to the power output from the solar cell panels on the wings by concentrating more solar energy on the solar cells. The reflector panels are angled outward along both sides of the wings to form a shallow trough, and the solar panels are at the bottom. Sunlight that otherwise would not strike the flat panels is reflected back from the reflectors onto the solar panels. The HS 702 is configured to enable designers to tailor the power output from the arrays to satisfy specific customer requirements. This is done by choosing any one of six different solar array configurations that have up to five panels per wing.
Power subsystems of today’s communications satellites rely primarily on nickel-hydrogen batteries for powering payloads during eclipses. But batteries are heavy and costly; they weigh of the order of 500 kg. The next battery technology expected to provide capacity for higher power satellites early in the new century was in the development stage in the closing days of the 1990s. Lithium ion batteries hold the promise of halving the mass of nickel-hydrogen satellite cells while offering a less expensive alternative. One possible version built of plastics contains a nonliquid electrolyte within the plastic. High-speed flywheels that extract the energy of a spinning mass are another promising technology but are further removed from application (48).
Propulsion. Beginning in the early 1960s, cesium and mercury electric ion propulsion systems flew aboard U.S. Air Force and NASA satellites. Since then, the performance of new commercial communications satellites has been getting a dramatic lift from a newer spacecraft propulsion system first flown on a noncommercial European Space Agency satellite in 1992. Also a form of electric propulsion (49), the xenon ion propulsion system (XIPS) makes possible reductions of as much as 90% in the mass of a satellite’s fuel (Fig. 5).

Figure 5. Thruster for xenon ion propulsion system (XIPS) protrudes from PanAmSat PAS-5 satellite. The electric propulsion system complements the satellite’s less efficient bipropellant propulsion system to reduce substantially the amount of fuel required by that system .
Normally, at the beginning of orbital life, chemical propellants required by a satellite’s bipropellant propulsion system for attitude control and station keeping throughout the mission account for at least 25% of satellite mass. This could amount to 500 to 700 kg (1102 to 1544 lb) for a conventional communications satellite. The comparable amount of fuel for the more efficient XIPS propulsion system would be only 75 to 105 kg (165 to 232 lb) of inert xenon gas. At a cost of about $30,000 per kilogram to place that mass in orbit, millions of dollars could be cut from launch costs. The mass reduction can also be translated into an increase in satellite payload, or communications capacity. Or, with additional xenon, a longer satellite mission life can be obtained because the quantity of onboard fuel is the principal limiting factor on the satellite’s operational life. Alternatively, the XIPS technology could produce benefits that combine these savings. A XIPS-equipped satellite uses the impulse generated by a thruster assembly and ejects electrically charged particles at high velocities. The entire system consists of a source of pressurized xenon propellant, a power processor, and the cylindrically shaped thruster assembly. Thrust is created by accelerating positive ions generated within the assembly’s ionization chamber through a series of gridded electrodes at one end of the assembly. The electrodes create more than 3000 tiny beams of thrust. The beams are prevented from being electrically attracted back to the thruster by an external electron-emitting device called a neutralizer. XIPS generates a very high specific impulse, expressed in seconds as the ratio of thrust to the rate at which the propellant is consumed. The higher the specific impulse, the less propellant required. The specific impulse of a XIPS system is 2600 to 3800 seconds, depending on the satellite configuration—rough-ly 10 times that of a bipropellant system. (Note that the 1964 NASA SERT I satellite, which used a combination of mercury and cesium ion, had a specific impulse of about 5000 seconds, but this was not proven sustainable or available all the time.)
The thrust level of a modern XIPS system is very small. For an HS 702 satellite, it is 0.165 newton, or 0.037 pound of force. In a satellite such as the HS 702, the XIPS system can also augment the satellite’s bipropellant propulsion system in orbit-raising. Using XIPS to help circularize the elliptical transfer orbit further reduces the amount of chemical propellant needed on the satellite, thereby again shrinking satellite mass. The extra mass reduction gives customers even more latitude in choosing among the benefits of a less expensive launch vehicle, additional payload, or longer mission life. The HS 702 XIPS is designed for a maximum orbit-raising duration of 90 days, compared to 3 days for the higher thrust but less efficient bipropellant system. The trade-off for satellite customers is one of mass reduction benefits versus delayed arrival on station. Systems. The power per kilogram of satellite mass has grown steadily. Early in the 1980s, when the current generation of spinners was introduced, the figure stood at about 1W per kilogram. It climbed to 2 W per kilogram for the HS 702′s predecessor, the three-axis HS 601, in the early 1990s. Later in the decade, the figure reached 4W per kilogram for the HS 702. It is likely to rise for later generation satellites. The availability of higher power from a satellite like the HS 702 enables operators to increase transponder effective isotropic radiated power (EIRP), a measure of performance that takes into account transmitter power, antenna gain, and the losses in the transmitter waveguide. This is the satellite portion of the previously cited power balance equation. A high EIRP permits customers to use smaller Earth antennae such as the 45.7-cm (18-in) and 60-cm (24-in) home antennae for direct-to-home TV in the United States and Europe, respectively. Or operators can apportion available power among a larger number of transponders, effectively doubling the transponder complement and reducing cost per transponder in orbit. Still another option is to choose a combination of both performance benefits.
The newer, higher power satellites are the backbone of high-speed, broadband satellite services (e.g., Teledesic, Spaceway, Astrolink—see later) coming on line early in the new millennium. Because switching functions are handled by digital processors in the satellite rather than on the ground, whole new vistas will be opened for corporate and other VSAT users. The small terminals will be able to broadcast at data rates of multiple megabits per second, compared to the recent 128-kbps rate. And hundreds of megabits per second will be down linked to them.
Digital compression techniques account for a panoply of new satellite-delivered entertainment and business offerings. Digital video compression effectively increases the use of communications satellite capacity by factors of 4-8, and could double that by the early 2000s. Thus, a 16-transponder satellite could transmit 256 TV channels. The doubling could reduce an operator’s operating and capital costs by an order of magnitude. Before the recent application of compression techniques, only one analog broadcast-quality video signal could be transmitted through each transponder on a satellite. Consequently, the number of video programming channels was limited by the transponder count. Now, using digital compression, broadcasters can convert analog video channels into digital data. The digital signals are compressed and transmitted by the satellite to end users. At the end user’s set-top receiver, they are decompressed and converted back into the customary analog form for viewing. Each video channel occupies only a portion of the transponder’s bandwidth, so additional channels can be squeezed into that transponder. The compression process involves representing video signals by a series of numbers, eliminating redundancy in the scenes, retaining only frame-to-frame differences, and reconstructing the video at the receiving end from the sequence of numbers.
A comparable transformation is occurring in the transmission of CD-quality audio by satellite. The entire contents of a compact disk amounting to 620 megabits can be compressed into 32 megabits and replaced by a ”flash” random access memory (RAM) chip. The contents of the flash chip can be downloaded in 1.5 s from a satellite to a consumer by a direct-to-the-home transponder. There, it can be stored in the hard drive of a TV set-top box. A 10-gigabit hard drive has sufficient capacity to store the equivalent of approximately 300 CDs. Now the set-top box with speakers becomes the storage medium and solid-state ”music player” for the home, replacing the stereo and compact disk player. Similarly, digitally compressed audio can now be transmitted by a satellite to a moving automobile through a ”whip” antenna and stored in memory. Drivers can then select the music or audio programming they wish to hear. In another application of the same technology, pictures taken by a digital camera now can be sent around the world in seconds. A typical digital camera can store about 100 pictures on a flash chip. The chip can be removed and inserted into a computer to uplink the contents to a satellite for retransmission.
In recent years, there has been a resurgence of interest in both LEO and MEO orbits for communications satellites (50). The non-GEO orbits have won adherents among operators of global mobile personal communications systems. Those such as Globalstar, for example, permit subscribers who have handheld cellular phones to speak to others beyond the range of normal cellular transmission towers almost anywhere in the world. Note that the now-defunct Iridium handheld global mobile telephony system did not fail because of its LEO orbit, nor for any technical reason. Missteps in marketing and financing are generally cited. LEO orbits also are being used for two-way data communications and messaging systems (e.g., Orbcomm). All of these systems require that the user maintain line-of-sight contact with the satellite. Unlike terrestrial cellular phones, they have insufficient link margins to permit subscribers to place or receive calls from inside a building or other structure.
The interest in non-GEO orbits was stimulated by explosive growth in worldwide cellular phone usage and an accompanying upsurge in demand for greater telephony infrastructure. Designers of LEO systems rejected GEO orbits on two counts (51). They believed that a GEO satellite system could not provide enough link margin for a satisfactory link between satellites and on-the-go mobile subscribers using small handheld phones that have omni-directional antennae. They also maintained that the high latency, or 260-ms signal delay, from GEO orbit (plus any processing and speech compression delays) would be unacceptable to cellular phone subscribers. Other designers, however, concluded that an attractive compromise was offered by a MEO global mobile communications satellite system: global coverage that uses fewer satellites at high elevation angles; fewer handovers than a LEO system; and, at 100 ms, a shorter round-trip propagative delay than the 260-ms delay from geosynchronous orbit.
GEO mobile communications satellite builders resolved the margin issue by resorting to higher power satellites using very large deployable parabolic antennas of about 12.5 m (40 ft) in diameter (52). The combination of high spacecraft power and high gain of the large antenna ensures acceptable margins for satellite-based personal mobile communications. GEO systems of this type (e.g., Thuraya (53), and ACeS (54)—operated by PT Asia Cellular Satellite) offer mobile service to properly equipped subscribers anywhere in a region within the footprint of a single GEO satellite. But they do not offer global coverage. To do so would require additional satellites and, possibly, more than a single hop.
As for the latency issue, GEO satellite designers are convinced that users can accept as much as a 400-ms delay in voice communications. Consequently, GEO mobile systems are configured to keep the burdensome and unalterable round-trip propagative delay, plus compression and processing delays, within the 400-ms limit. At low orbital altitudes, LEO systems pay only a small penalty for shorter propagative delays, but they still must deal with appreciable processing, handover, and compression times, especially for geographically separated parties on long-distance calls.

Conclusions

Amidst the post-World War II Cold War environment, President John F. Kennedy urged all nations ”to participate in a communications satellite system in the interest of world peace and closer brotherhood among people of the world.” By 1967, three INTELSAT satellites were operating over the Atlantic and Pacific oceans, providing ubiquitous global telephone and television communications in hundreds of languages. By the early to mid-1970s, new domestic satellite systems were created in many countries and regions, such as Canada, Europe, the United States, and Indonesia. By April 2000, the date this chapter was written, sophisticated communications satellites became a fundamental element of the world’s telecommunications infrastructure and an indispensable tool of the global marketplace. These satellites have enabled many new space-based applications to be introduced to businesses and consumers around the world. These include direct-to-home television broadcasting, handheld global mobile telephony, and high-speed broadband data exchange.
These sophisticated communications satellites have evolved since the mid- 1960s. In 1965, the Early Bird (INTELSAT I) satellite weighed 76 pounds. The mass of today’s satellites can exceed 12,000 pounds. Strides in satellite technology have also extended the lifetime and transponder capacity by several orders of magnitude. The comparative sophistication of today’s communications sat ellites derives from the application of several advanced technologies. These include digitalization, which is used for numerous functions to enhance their capabilities and flexibility. For example, by using onboard digital signal processing, modern satellites can generate hundreds of reconfigurable spot beams simultaneously. Among other things, this enables satellite operators to generate maximum revenue and to obtain greater bandwidth by reusing the limited amounts of assigned spectrum. Digital echo cancellers have also been developed and installed in digital phone exchanges and satellite networks to eliminate disturbing voice echoes on satellite-relayed telephone calls. Digital compression techniques have already increased satellite use capacity by a factor of 8, and soon a 16-transponder satellite might transmit 256 TV channels. This will greatly reduce an operator’s operating and capital costs. Before the advent of digital compression, only one broadcast-quality analog video signal could be transmitted through each satellite transponder. Digitalization is also revolutionizing the transmission of CD-quality audio by satellite, thus enabling the start-up (to date) of three new satellite digital audio services: Sirius, XM Radio, and World-Space (see discussion under ”Commercial Applications of Communication Satellite Technology”).
Other advanced technologies are making their imprint on a satellite’s power and propulsion. As gallium arsenide solar cells replaced traditional silicon cells, satellite power doubled and enabled satellites to generate as much as 15 kW of power at the end of life. As a replacement for mercury and cesium electric propulsion systems, xenon ion propulsion makes possible reductions of as much as 90% in the mass of a satellite’s fuel requirements. At a cost of about $30,000 per kilogram to place a satellite in orbit, millions of dollars can be saved in launch costs. This reduction can also be applied to an increase in satellite payload, longer satellite lifetime, or any combination.
These and all ongoing technological innovations are intended to provide maximum available bandwidth through frequency reuse; maximum flexibility for operators and users; lower launch and in-orbit operating costs; more space-based communications options for consumers; extended satellite lifetimes and greater reliability; larger satellite footprints, including those created by reconfigurable spot beams; higher power levels resulting from increased use of solar power; and reduced cost per transponder.