Satellite communications systems move into the twenty-first century

Wireless Networks 4 (1998) 101 107 101 Satellite communications systems move into the twenty-first century Leonard S. Golding Hughes Network Systems, Inc., Germantown, MD 20876, USA This paper discusses the evolution of communication satellite systems and communications satellite technology from the 1960 s to the 1990 s. The paper identifies the key attributes of satellite communications that has driven this evolution and now drives the future directions such systems will take. The paper then discusses the future direction of communication satellite systems including DBS, MSS, FSS and hybrid satellite/terrestrial systems. The paper points to the continued evolution of the satellite payload to use of spot beams, onboard processing and switching, and intersatellite links, with capability for higher EIRPs. The paper also identifies the earth station trends to more compact, lower cost stations, produced in higher volumes, with the handheld phone for MSS operation being the prime example of this trend. The paper then points to some revolutionary trends in satellite communication networks being proposed for MSS and FSS applications involving fleets of NGSO satellites combined with more extensive ground networks involving new networking concepts, new services (such as multimedia) and new hybrid configurations working with terrestrial networks, involving a host of new network issues and operations. 1. Evolution of communication satellite systems Satellite communications was launched in October 1957 when Sputnik I carried out telemetry communications with the ground for 21 days. The first man-made satellite used for voice communications was Score, launched in December 1958. The early communications satellites such as Score, Echo, Courier, Telstar, and Relay were all nongeosynchronous satellites having altitudes ranging from a few hundred to ten thousand kilometers. Syncom II and Syncom III, launched in July 1963, established the feasibility of geosynchronous communications satellite systems. Intelsat, formed in July 1964 by a United Nations resolution, started the era of the commercial use of communications satellites, which became a reality with the launch of the Early Bird geosynchronous satellite in April 1965. As shown in figure 1, the traffic growth through 1990 was an impressive 15 to 20 percent per year, growing to 120,000 channels by 1990. The Intelsat system was initially viewed as an international trunking system carrying large trunks of voice traffic between big Standard A earth stations, each with 97 foot diameter antennas and costing millions of dollars sort of an extension of international cable systems. In the first 25 years of operation, the capacities of the satellites have increased 200 times over their early predecessors, and the cost per channel for the space segment has shown a corresponding drop in cost. Earth station trends have always been in the direction of smaller, lower cost stations going from the large Standard A Intelsat stations to handheld satellite terminals for new mobile satellite systems to become operational by the year 2000. The trend toward lower cost satellite communications made it possible to operate national and regional commer- Figure 1. Total Intelsat satellite traffic. cial satellite systems, and by the mid-1970s domestic and regional satellite systems began to appear. These systems began to take advantage of some of the inherent characteristics of satellite systems. These inherent characteristics are not only the key to the evolution of communications satellite systems, but are also the key to the future directions of these systems. The two basic properties of satellite systems that lead to its applications are: (1) Ability to cover large areas of the earth s surface at any given instant of time. (2) The communications system is some form of a radio system. These two properties in turn lead to a number of important characteristics: J.C. Baltzer AG, Science Publishers

102 L.S. Golding / Satellite communications systems move into the twenty-first century (a) Ability to provide service and aggregate traffic over wide areas. (b) Ability to allocate resources (bandwidth and power) to different users over the coverage region as needed. (c) Distance-insensitive costs. (d) Ability to provide coverage to mobile users operating over wide areas, including rural areas, water areas, and large volumes of air space. (e) Ability to easily provide point-to-multipoint (broadcast), multipoint-to-point (data collection), and pointto-point communications. (f) Ability to have direct access to users and user premises. These characteristics have led to the use of regional satellites to distribute television signals to cable headends and television broadcast stations throughout the region; maritime and land mobile communications to ships, airplanes, land vehicles, and small transportable terminals; and supervisory control and data access (SCADA) applications for monitoring pipelines, water levels, and other data collection applications. Communications satellites are ideal for broadcast (pointto-multipoint) applications because of the large area coverage capability and distance insensitive characteristic. Another more subtle characteristic that is evident in this application is the quality of the link. Because of the altitude of these satellites over the coverage regions, elevation angles from 10 to 40 degrees can typically be obtained from the user s location to the satellites, which results in a single hop communications link between the distribution earth station and the user earth station, which involves a line-of-sight uplink to the satellite and a corresponding line-of-sight downlink from the satellite. Each link is an additive Gaussian white noise channel, which can be designed to deliver high quality end-to-end service. This single hop access directly to the user can, in many cases, result in higher quality of service than terrestrial links, which may require many hops before the signal is delivered to the user. As higher radiated power has been achieved from communication satellites, direct television broadcast to users has become a reality with the ASTRA satellite in Europe and DIRECTV service in the U.S.A. Up to this point in time fixed services has been the main use of commercial communication satellites; however, the trend toward smaller and smaller earth stations has been evident, as has been the trend to provide direct satellite access to the end user. During the last 10 years we have seen a growth in very small aperture terminal (VSAT) networks using user terminals with antenna sizes from 0.75 to 2.5 meters in diameter, providing data networks for corporations, voice networks for carriers, and multimedia services including video broadcast, audio broadcast, teleconferencing, facsimile, and various data services including, more recently, high speed Internet access for corporate and residential users. These VSAT networks have been successful in competing against terrestrial networks because they have offered direct access to end users on the satellite network, multimedia service capability for small incremental costs, high end-to-end service quality, and not only better control of costs but cost-effectiveness when compared to terrestrial costs. The satellite characteristics of large area coverage, flexible allocation of satellite resources, aggregation of traffic over wide areas, and distance insensitivity all contribute to the success of VSAT networks. A recent market survey by Euroconsult indicates that between July 1996 and the end of 2006, 262 to 313 geostationary satellites will be launched, not including the large number of low earth orbit (LEO) and intermediate circular orbit (ICO) satellites that will be launched in the same time frame. Clearly, satellite communications is undergoing continued growth into the twenty-first century. 2. Evolution of communication satellite technology As shown in figure 2, communications satellite technology primarily focused on the geostationary satellite orbit, with a so called bent-pipe communications payload. From geostationary orbit the earth subtends an angle of 17, and the early synchronous (with the earth s rotation) satellites had a global coverage beam, which covered around one-third of the globe, with a bent-pipe payload, as shown in figure 2. The satellite support systems that are needed are identified below the communications payload. These support systems are not part of the communications payload, but affect the payload performance. The satellite configuration shown in figure 2 had limited capacity; however, it was very flexible in terms of the air interfaces and the ground connectivity it could support. One of the most important new technologies is the use of spot beams on the satellite. This technology permits the use of smaller ground terminals by increasing the satellite effective isotropic radiated power (EIRP) and satellite receiver sensitivity. It also permits an increase in capacity by allowing for frequency reuse through spatial isolation between beams, and results in more effective use of limited prime power on the satellite by having the radiated radio frequency (RF) power be directed only where it is needed. The satellite spot beams can be fixed, can be made steerable, or can be made to hop from one location to another in time. The number of spot beams can range from a few to several hundred and typically cover regions on the ground from a few hundred miles to several thousand miles. While satellite spot beams offer many advantages, they make it more difficult to have full interconnectivity among all of the earth stations using the satellite. As shown in figure 3, a spot beam satellite requires some form of interconnection network. This network may be a bank of filters, with each filter having its output connected to a specific downlink beam. In this case, routing between beams is done by selecting the appropriate carrier frequencies. An alternative method may be to have the interconnection done

L.S. Golding / Satellite communications systems move into the twenty-first century 103 Figure 2. Bent-pipe satellite block diagram. Figure 3. Spot beam payload using microwave interconnection. in time. In this case, the interconnection network, at different times, provides different paths between the uplink and downlink beams. The correct connection is made by transmitting the information in bursts on the uplink so as to arrive at the satellite when the interconnection network has the connection to the desired downlink beam. In more advanced satellite payloads, the spot beams are combined with onboard demodulation of the uplink signals, baseband processing and switching of the signals, and remodulation and coding of the information for downlink transmission, as illustrated in figure 4. This allows for full flexibility in both connectivity and air interface characteristics for the uplink and the downlink. The switch shown in figure 4 may be a circuit switch, or it may be a packet (or ATM) switch. The onboard processing and switching may also be controlled from the ground via command and telemetry links so that changes can be introduced to match varying conditions in the network. In addition to the signal processing shown in figure 4, there may be signal processing associated with forming and controlling the spot beams on the satellite using an active phased array antenna. As shown in figure 5, a complete satellite communications system comprises a number of elements in addition to the satellites discussed above. The ground segment generally will include subscriber earth stations, which may be fixed or mobile, and a network of fixed gateway stations, which may be connected to the public switched telephone network (PSTN), public land mobile network (PLMN), or a public data network (PDN). A network coordination and/or an operations center generally is involved to manage the network and handle administrative operations, maintenance, and provisioning functions. Telemetry, tracking, and command (TT&C) stations are provided to monitor and control the well being of the space segment.

104 L.S. Golding / Satellite communications systems move into the twenty-first century Figure 4. Onboard processing payload. equipment (CPE). Much of the added sophistication is realized in software, where more and more of the effort is centered in the design of new systems. Greater effort has been exerted not only in developing satellite system standards, but making sure satellite systems can interoperate with terrestrial network standards such as ISDN and ATM. As we move into the twenty-first century we are seeing the rate of change in satellite communications systems increase dramatically. Figure 5. Elements of a satellite communications system. In addition to the space segment and the ground segment, there is a third component of the satellite system, namely the air interface. This interface is primarily involved in physical layer issues, such as identification of modulation and multiple access, forward and return control channels, power control, timing and frequency synchronization, and fading and shadowing control. In addition, the air interface is involved with Layer 2 functions such as medium access control and link access control, and some Layer 3 functions such as call control, radio resource management, and mobility management (in the case of mobile systems). Both the ground segment and the air interface have been undergoing significant changes over the past decade, with an emphasis on increasing the performance capability of the satellite system to compete more effectively with terrestrial systems, while at the same time driving the size and the costs down. The subscriber terminals in all the application areas have been continually reducing in size, going toward handheld terminals for either voice or data applications as proposed for future mobile satellite systems (MSS), or fixed antennas approaching 1 foot in diameter. The electronics in the subscriber earth stations are becoming more integrated, with the use of more MMICs and applicationspecific integrated circuit (ASIC) chips, as the volume of the earth stations is increasing and the prices are decreasing. More sophisticated signal processing and intelligence are being incorporated in the satellite system, in the air interface, in the management of the system, and in the interfaces to the terrestrial networks and customer premises 3. Future direction of communication satellite systems The basic inherent characteristics of communication satellite systems referred to in section 1, determine the future direction of these systems. These future directions are as follows: (1) Direct Broadcast Systems (DBS). (2) Mobile Satellite Systems (MSS). (3) Fixed Services Systems (FSS). (4) Hybrid Satellite/Terrestrial Systems. Satellites have always been well suited to provide broadcast services. The capability to provide high radiated power from the satellite combined with digital source coding and transmission have made it feasible to provide television and program audio signals directly to a subscriber having an earth station with an 18-inch diameter antenna or less and costing only a few hundred dollars. With such a small earth station, the subscriber is able to receive a large number of channels (over 100 TV channels in the DIRECTV service), with high quality signals. DBS television systems already exist in Europe, the U.S. and Japan, and will be extended to Latin America and other parts of Asia. Direct Broadcast Audio will see a rapid expansion with at least two systems nearing completion and getting ready for launch in the next couple of years. Category (4) above refers to the Hybrid Satellite System. This is a system that uses the terrestrial network to carry out some functions and the satellite network to carry out other functions being provided, taking advantage of the

L.S. Golding / Satellite communications systems move into the twenty-first century 105 Table 1 Representative MSS. Name Organization Features Start date AMSC Hughes 1 GSO Sat. 1996 Iridium Motorola 66 LEO Sat. 1998 GlobalStar Loral/Qualcomm 48 LEO Sat. 1998 ICO ICO 10 MEO Sat. 2000 Odyssey TRW 12 MEO Sat. 2000 ACes 1 GSO Sat. properties of each network to provide the overall services being provided by this hybrid network. The DIRECTV DBS system serves as an example of such a hybrid system with terrestrial telephone lines being used to provide return data links from each satellite subscriber receiving unit to the Satellite Control Center. These return links are used to forward subscriber requests, usage data, and other information. In this case, terrestrial telephone lines are a lower cost solution than providing a satellite return link from each subscriber location. One of the largest new growth areas in satellite communications is the Mobile Satellite Systems (MSS) area. Table 1 shows some representative MSS that are in operation or will be in operation over the next 5 years. While current MSS have geosynchronous satellites and ground systems that use mobile terminals which vary from the size of an attaché case to Standard A Inmarsat stations with 1-meter diameter steerable dish antennas, and compressed digital voice or low speed data services, the future MSS will use handheld telephone ground terminals, battery operated or equivalent wireless personal digital assistance (PDA) devices. Many of the MSS use a fleet of nongeosynchronous satellites either in (LEO) low earth circular orbits (400 to 1,000 miles in altitude) or in (MEO) medium circular earth orbits (5,000 to 7,000 miles in altitude) with either polar or inclined orbits. In addition to the handheld terminal these MSS will also have other ground terminals including airplane, shipboard, land vehicles, and fixed paging and messaging-only terminals. We have already indicated that hybrid systems would be employed in DBS systems. More elaborate use of hybrid systems will occur in MSS. Mobile terminals with dualmode capability will be provided, which can be used with terrestrial cellular systems (when in range of such systems) or operate through the satellite. Registration, mobility management, and advanced intelligent network features all may involve terrestrial network mobile switches working in conjunction with satellite network switches. Billing, administrative functions, call control, routing and related roaming functions will also involve terrestrial switches working with the satellite network switches and stored databases such as HLRs and VLRs, where these databases may be part of terrestrial cellular systems. Another characteristic of nongeosynchronous MSSs is that they provide worldwide coverage. The fixed ground network of gateway stations needed to support worldwide coverage of mobile subscribers itself may involve an extensive terrestrial network interconnecting gateway stations, involving terrestrial transfer of signaling and bearer traffic as well as satellite transfer of these signals. Fixed Satellite Services (FSS) have been the primary application for communication satellites since commercial operation started with the Intelsat System and more recently leading into large numbers of VSAT (Very Small Aperture Terminals) networks throughout the world. However, the future FSS show significant new changes from the existing C- and Ku-band FSS in operation today. From proposals, filings, and papers made public, a large number of organizations appear to be pursuing broadband multimedia fixed services using new Ka-band satellite systems. A large number of these new systems are following an evolutionary process from the Ku-band satellites. Examples are Spaceways by Hughes and VoiceSpan by AT&T, which are Ka-band geosynchronous systems. These systems are considerably more advanced than their Ku-band counterparts, having many more spot beams, wider bandwidth channels, on-board processing and switching, higher capacity yet small, low-cost subscriber ground stations, and a capability to provide not only voice, but all types of data, video, facsimile and advanced Internet services. Interoperability with terrestrial networks and terrestrial network standards such as ISDN, ATM and SONET is a primary objective as is interoperability with a variety of customer premises equipment, including LANs, routers, videoconferencing terminals (MPEG standards), workstations, digital PBXs, and other broadband equipment. More sophisticated network management may be deployed providing various amounts of bandwidth on demand as needed, different operating modes (packet, circuit or asynchronous random access), and different Quality of Service. In addition to the evolutionary proposals, there also are some revolutionary proposals for new Ka-band FSS. These include: Teledesic, M-Star (Motorola), Sativod (Alcatel). These proposals are for large fleets of non-geosynchronous (typically LEO) satellites providing worldwide broadband multimedia service. The M-Star system is a 20 50 Mbps service, primarily targeted to providing a feeder link network for terrestrial cellular systems throughout the world. Teledesic is talking about a satellite fleet of over 800 satellites, all having both on-board processing and intersatellite links. Clearly, the above FSS are completely different from previous FSS and will generate many new challenges that have not been faced before. The use of intersatellite links has also been proposed for some nongeosynchronous MSSs such as Iridium. While intersatellite links have been used in some government satellite systems, they have not as yet been used in commercial systems. Most of these more revolutionary systems are in their early stages, with the system architectures still in the formative period.

106 L.S. Golding / Satellite communications systems move into the twenty-first century 4. Challenges created by the next generation of satellite systems There are many challenges that are created as a result of the new directions of the next DBS, MSS and FSS systems. Some general trends can be observed in all of these systems. First, greater complexity in the spacecraft, with the use of more spot beams, more on-board signal processing and improved r f system performance. This leads to challenges in building more complex microwave and digital integrated circuits that are also efficient in terms of the use of prime power and can cope with the spacecraft environment. The added sophistication built into the spacecraft also has implications in support systems such as attitude control, prime power generation, battery support and thermal control, requiring improvements in performance in these systems as well. In the subscriber ground terminals, the challenge is to build these terminals smaller in size and at much lower cost and higher production volume, with the ultimate terminal being handheld dual mode telephones for the MSS. This means that higher levels of integration are necessary, leading to more complex microwave and digital integrated circuits. Another important technical challenge is low-cost, compact tracking antennas, which would be very important for the non-geosynchronous MSS and FSS. One of the most important challenges going into the 21st century is the design of the optimum network for the satellite systems that will use a fleet of satellites to provide service on a worldwide basis, proposed both for MSS and FSS. Having satellites moving relative to a location on earth and then relying on multiple satellites being visible to maintain reliable continuous communications, raises a number of new problems that need to be solved in order to have these systems provide high performance and low cost. Some of the new problems that need to be dealt with include: Satellite resource management. Network resource management. Handover strategies. Power, frequency, timing and synchronization control and management. Optimum fixed ground network configuration to support this fleet of satellites, minimizing terrestrial backhaul costs and optimizing the use of satellite resources. Optimum use of intersatellite links in conjunction with the ground network. Security and authentication in this worldwide system. Optimum routing and configuration management of traffic. Optimum use of control channels. Mobility management, registration and paging for MSS. Optimum use of integrated functions such as position location and communications in MSS. Interference management. Failure mode analysis and management of a fleet of satellites. Tradeoff between signal processing employed on satellites versus signal processing incorporated into the ground network. Terrestrial network interfaces. The above list of problems is not a complete list, but it identifies some of the key problems for these proposed nongeosynchronous systems. Another challenging area in satellite communications going into the 21st century is hybrid systems. How to best use combined satellite and terrestrial networks to provide optimum performance for a variety of existing and new services. A perfect example is the proposed new MSS, where the satellite service is integrated with terrestrial cellular, PCS and paging networks. To provide worldwide mobile service to the user, there are a number of network architecture questions which need to be solved including: where do HLR and VLR databases reside for particular subscribers; which mobile switches provide billing information, mobile registration, AIN features such as call screening and call forwarding. How is interoperability maintained between various terrestrial systems such as GSM or AMPS-D and the satellite system? In the network management area, one may have separate network management systems for given terrestrial cellular systems and the satellite system. In a hybrid system architecture some integration of both network management systems needs to be considered. There are also questions of how interoperability is maintained between different functions in a hybrid system. For example, operation in-building from the satellite system may be difficult in the case of an MSS. Under what conditions should operation be changed from the satellite to a terrestrial system needs to be resolved in this example. Many similar questions dealing with which portions of the network handle which functions need to be resolved in such hybrid network configurations. In the new broadband worldwide FSS systems being proposed, extensive terrestrial networks connecting gateway stations are likely to be involved. Such networks will be used for network management, signaling, billing and administrative functions, and possibly backhaul of traffic. The integration of such terrestrial networks to ensure both functional and standards interoperability will be one of the key problems which needs to be addressed, especially for worldwide operation. One of the challenges for these FSS systems as mentioned in the previous section will be to determine what functions will be carried out on satellites and what functions will be carried out in the ground network. This may be especially important in the broadband FSS, where millimeter wave frequencies that have high rain losses are used. For those applications requiring high availability in regions where rain is more predominant, new solutions need to be found for the FSS to provide the required quality of service. This may require alternate routing via

L.S. Golding / Satellite communications systems move into the twenty-first century 107 intersatellite links, use of diversity ground stations, or some other solution. Low delay transport may also be needed for certain applications. This may require routing some traffic over the terrestrial network and some traffic over the satellite system to meet the requirements for that application. Broadband multimedia requirements will present many new performance challenges that did not have to be faced for more conventional voice and low-speed data applications. As we enter the 21st century, satellite communications systems are moving to provide new services in the DBS, MSS, and FSS, with new network architectures, new technology, and new levels of integration with terrestrial networks. In addition, integration of functions is taking place such as combining navigation and communications and surveillance and communication. It is an exciting time with many new systems being proposed resulting in many new and exciting challenges that need to be met. Leonard S. Golding is currently a Vice-President at Hughes Network Systems, a subsidiary of GM- Hughes Electronics Corporation. He is in charge of Systems Engineering at HNS, and has been involved in directing and working on projects in DBS, MSS and new multimedia FSS programs. He also has been involved in terrestrial wireless communication projects including PCS systems such as PACs and broadband WLL systems in the millimeter wave bands. He also oversees HNS standards activities and is involved in new business development activities. He is a fellow of the IEEE, has served on numerous professional and industry committees as well as having over 40 publications and 10 patents. He received his doctorate at Yale University and undergraduate degrees at Columbia University.