Introduction of Consumer Terminals and Applications

Ground segments that serve individual consumers came into their own during the 1990s. Although many U.S. consumers were enjoying direct reception of cable and TV networks using backyard dishes (typically 2m to 4m in diameter), an explosion of demand for home satellite TV did not really start until DIRECTV was introduced. The space segment for this pioneering system consisted of rather standard Ku-band Broadcasting Satellite Service (BSS) satellites operating in the assigned slot at 101 degrees west longitude (the Astra Fixed Satellite Service, FSS, and BSS satellites had already gained eminence at 19.2 degrees west longitude, based on analog transmission).

DTH receivers were installed in American homes by millions of families who either did not have access to cable TV or were unhappy with what the local cable company had to offer. Complementing the user terminals was the world's largest uplinkingcenter for DIRECTV, constructed in Castle Rock, Colorado. This center was the first to employ MPEG-based video compression and had the unique capability to originate 200 simultaneous channels of programming. The proprietary DIRECTV Satellite System (DSS) demonstrated the technical and operational viability of this type of implementation (as with many U.S. innovations, this proprietary standard has been supplanted in later DTH systems by the European Digital Video Broadcast [DVB] series of open standards).

By 2000, approximately 7 million individual subscribers were employing DSS receivers and subscribed to the service (making DIRECTV the United States third largest cable TV operator after AT&T-TCI and AOL /Time Warner). Echostar, a competitor of DIRECTV, introduced its innovative DISH Network and brought the total U.S. DTH subscriber count to over 10 million, the largest for any country in the world. France, Germany, the United Kingdom, Sweden, Thailand, Malaysia, Japan, Mexico, Brazil, and Argentina are among the nations in which consumers have access to a high degree of choice in TV entertainment. Many of these ground segments use the DVB standard, which contains MPEG-2 video compression and multiplexing, digital audio, encryption and conditional access, and online program guides. DVB provides many standard features that permit set-top boxes to be supplied from multiple vendors, but aspects of the conditional access system are very much closed. All together, 30 million subscribers enjoy this digital DTH satellite service at the time of this writing. This is a boon, not just for the operators but for manufacturers of home receiving antennas and set-top boxes, digital video compression and storage equipment, uplink earth stations, and providers of the programming itself (as well as advertisers) who gain a direct connection to viewers.

Returning to the Mobile Satellite Service (MSS) area, Inmarsat did a ood job of bringing satellite communication to the personal level. First with he Inmarsat M and then the M4 (Figure 1.8), it was demonstrated that one ould have reliable communication from a highly portable piece of equipment.Likewise, domestic MSS operators Optus Communications (Australia) nd American Mobile Satellite Corporation (AMSC) adopted similar

Figure 1.8 Examples of compact MSS user terminals for use in the Inmarsat system:
(a) Standard M; (b) Standard M4 (photos courtesy of NERA).

ground segment equipment, including Improved Multi-band Excitation IMBE) voice compression technology from Digital Voice Systems Inc.
(DVSI). While these terminals were not of the handheld variety, they nevertheless emonstrated that individual calls could be placed and subscribers ould be serviced conveniently and relatively inexpensively (e.g., as compared ith the best alternative in a remote place, on a ship, or wherever).The explosive use of the Internet during the 1990s had an impact on satellite communication ground segments. Major corporations that found it an effective way to extend their information technology (IT) systems across a broader business and geographic base adopted TCP/IP and the rest of the Internet suite of protocols. Companies like HNS, Gilat, and STM Wirelessenhanced their VSAT offerings by better supporting TCP/IP. Oddly, much
misinformation was propagated during the 1990s that TCP/IP was incompatiblewith GEO satellite links due to the combination of propagating delayand noise-induced errors. The leading suppliers of VSAT terminals and theassociated hub stations devised effective solutions through protocol spoofing (a technique also used by router vendors Nortel and Cisco to improve
performance) to allow satellite networks to challenge terrestrial solutions.HNS also introduced the first personal satellite Internet solution, DirecPC,to allow subscribers to reach 400 kbps downlink speeds. This service wasavailable in 12 countries at the time of this writing.

Rapid Developments

The decade of the 1980s saw a big increase in the number of satellites and satellite operators. The GEO began to become crowded in sectors serving North America, Europe, and East Asia. Applications in TV, private data and voice networks, and mobile communications established themselves as the ground segments mushroomed. Private commercial operators such as RCA Americom (now GE Americom), Hughes Communications, and Pan A Sat (now combined as a result of a merger in 1997) grew to profitability onthe foundation of the cable and broadcast TV industry. Television signals relied on these satellites to serve thousands of locations in North America.

Cable TV systems in local communities created the demand for specialized and high-value programming from HBO, Turner Broadcasting (now part of AOL /Time Warner), Disney, and ESPN. Cable networks numbered more than 100 by 1985 and continue to be delivered through 3-to-5-m C-band dishes at the head ends of local cable TV (CATV) systems. Origination of the network feeds requires extensive studio facilities. Ku-band satellites and ground segments also appeared in the 1980s to take advantage of
the smaller dish sizes that this band allows. As a result, VSAT user terminals like that in Figure 1.7 became popular with major retailers like Wal-Mart, Rite Aid, and Kmart, and oil companies like Chevron, Mobile, and Texaco.

The first real DBS system was introduced in Japan in the late 1970s and became very popular by 1985. While offering only two TV channels, the NHK DBS project resulted in several million home DBS installations throughout Japan. The United Kingdom and continental Western Europe saw the explosion of commercial satellite TV during the late 1980s. One could argue that this expansion came about through News Corp's Sky TVand the Astra satellites operated by Société Européene des Satéllites. This powerful combination delivered an attractive programming package directly to subscribers, who bought and installed their own DBS receivers. There is little doubt that this established the viability of direct-to-home (DTH) satellite TV, an application which DIRECTV subsequently took to new heights (but still at GEO altitude).

Inmarsat became the foundation of commercial mobile communications as defense applications of satellite mobile services were already established but with space and ground segments owned and controlled by governments. Several proposals were put forth in the United States and Europe for innovative satellite services using new frequency bands. One was intended for tracking trucks on intercity routes, so that dispatchers knew where their drivers were even if they could not call in. Even though there was a serious shakeout in these early systems, they nevertheless represent an important precursor for the global mobile satellite systems of the next
decade.

We cannot forget to mention the cellular telephone networks, which began to reach critical mass during this same decade. While relying completely on local base stations and switching systems, cellular radio proved the value and reliability of automated radio frequency channel assignment, cell-to-cell handover, and intersystem roaming. Similar capabilities were

Figure 1.7 Examples of VSAT earth stations for two-way communications: (a) 1.2-m data application (photo courtesy of Hughes Electronics); (b) 1-m video-data terminal(photo courtesy of STM Wireless).

demonstrated on the INTELSAT system with the pioneering SPADE SCPC DA system of the early 1970s and the DA system installed in Indonesia in 1977 for Palapa A. These technologies provided a basis for large global mobile personal communications services (GMPCS) projects like Iridium, Globalstar, and ICO.

Need for Broadband and ATM

The International Telecommunications Union (ITU) has defined broadband as any rate higher than primary rate (T1 or 1.544 Mbps). Bandwidth demands of advanced multimedia applications built on the WWW, as well as the Internet, telephony, videoconferencing, and similar networking applications are at last being matched by advances in transmission technology and their application to networks.
Different characteristics of applications drive the need for broadband network support. They include the following:

• High traffic volumes necessitated by increasing use of traditional applications. Increasing use of e-mails that contain documents, graphics, and even moving pictures. Client-server architecture-based applications and large file transfers result in large volumes of traffic, which benefit from broadband networks. Increasing capacity-access networks, especially different forms of digital subscriber line (DSL) circuits, are providing low-cost, high-bandwidth access to single users to their homes, vastly increasing the traffic volumes the supporting
transport networks need to carry.

• Special client-server applications, such as those supported by the WWW, are interactive in nature. Increased bandwidth is needed to provide acceptable latency for increasingly complex graphicintensive contents or to support the increasing use of Web-initiated downloads of large files (such as executables or large audio-video
files).

• Applications that are inherently broadband, such as moving digital images, digital photography, music, and VOD, necessitate a broadband infrastructure.

• Convergence of telephony and data networks, at the core transport level transmitting transparently circuit-switched data, or in the use of voice-over-packet technologies in packet-switching networks, will result in the need to transfer large volumes of voice/data necessitating increasing deployment of broadband networks.

Audio bit-rates range from CD-quality sound, which yields a 1.411 Mbps at 44.1 KHz sampling with 16-bit quantization, to 64 Kbps for digital telephony. With compression algorithms, MPEG layer 3 produces CDquality voice stream of 128 Kbps with 12:1 compression rates, and typical vocoders produce 8-Kbps compressed speech rates in telephone networks.

Video streams produced by MPEG-1 to support 720 pixels sample density per line, 576 lines per frame, and 30 frames per second TV sizes have variable rate bit streams ranging from 1.86 Mbps to a maximum of 15 Mbps.

The mix of different services, in addition to having different bit rates, have different traffic characteristics and real-time needs that require the network to guarantee a defined QoS generally defined in terms of sustained bit rate, excess bit rate, and burst size.
To handle such traffic volumes and QoS guarantees, emerging networks have used latest transmission, switching, and protocol technologies.
While narrowband transmissions generally operate over copper wires or coaxial cables, broadband transmission technologies use the large bandwidth provided by optical fibers. Currently, the wireless transmission technologies have also been developed to carry DS3 rates (45 Mbps) and above to form important segments of access networks in a broadband architecture.

Broadband rates are described in optical carrier (OC) rates:
OC-1 51.84 Mbps
OC-3 155.52 Mbps
OC-12 622.08 Mbps
OC-24 1244.16 Mbps
OC-48 2488.32 Mbps
OC-192 9953.28 Mbps

Advances in optical transmission technologies have recently produced dense wavelength division multiplexing (DWDM) technology. DWDM allows splitting of light into multiple wavelengths (colors), each of which can carry data at 10 Gbps and enables a single fiber to carry up to 100 different wavelengths giving an aggregate bandwidth of more than 1 Tbps per fiber.

Similar to T1 and T3 framing in traditional data/voice networks, the Synchronous Optical Network (SONET) standard, as part of a larger telephony standard called Synchronous Digital Hierarchy (SDH) by the Comite Consulatif Internationale de Telegraphie et Telephonie (CCITT), defines the framing for transmission of digital data over fiber. The basic building block in SONET is a synchronous transport signal Level 1 (STS-1) frame, which is organized as a 9-row by 90-column byte array, transmitted row first.

The basic frame timing used in T1 circuits (i.e., 8,000 frames per second) gives an STS-1 rate of 51.84 Mbps (90 × 9 × 8 × 8,000). SONET framing allows multiplexing of low-speed digital signals such as DS1 or DS3 and is able to integrate services such as ATM, which is made up of fixed length cells of 53 bytes in length.

ATM is a technology linked to the development of broadband Integrated Services Digital Network (ISDN) in the 1980s. As a packet-switched high-performance technology, ATM can support multiservice applications, including multimedia. Established standards for network and user interface, signaling, and traffic management have facilitated ATM’s rapid growth and adaptation to different networking needs. ATM architecture provides switched and permanent categories of virtual circuit connections [i.e., permanent virtual circuits (PVCs) and switched virtual circuits (SVCs)] between end systems and promotes optimized utilization of bandwidth by defining different classes of services.

The explosive growth in the broadband needs of emerging applications dictate the need for fiber-based infrastructure development. The ATM networking technology and fiber synergy has encouraged network service providers to increasingly deploy ATM-based transport networks.
The following sections provide a broad overview of the technologies and protocols that provide the framework for the detailed coverage on ATM interworking.

ATM Interworking in Broadband Wireless Applications

Wireless networks for data and voice communications systems are establishing themselves as important components of traditionally fully hardwired networks. Initially popularized by the mobility they offered to data and voice users, wireless networks are increasingly being deployed as competition against wire-line and optical leased line provisions to businesses, infrastructure (cellular backhaul) applications, and as local loop for new entrants to the telecommunications market. These networks alleviate costs involved in hardwiring communication lines to user premises and allow fast deployment.

Wireless systems have also facilitated provision of new data/voice services to multitenant units (MTUs) and to businesses in built-up city centers in shorter deployment times, thus increasing the competitiveness of businesses using wireless facilities. Figure 1.1 illustrates a point-to-multipoint (PMP) wireless network deployed in an urban environment.

While edges of voice and data network experience the fast deployment of the wireless component, asynchronous transfer mode (ATM) technology is increasingly becoming the central technology in the workgroup and enterprise network environments and as the transport technology of choice in nationwide wide-area networks (WANs). ATM provides scalable bandwidths and quality-of-service (QoS) guarantees at attractive price performance points, facilitating a wide class of applications that can be supported in a single network. The bandwidth hungry applications spawned by the World Wide Web (WWW), as well as the general use of the Internet, telephony,

video-on-demand (VOD), and videoconferencing have given rise to the need for simultaneously supporting the services on the same network. ATM is serving as the main catalytic technology in promoting the convergence of multiprotocol multiple networks into a single network providing multiple services.

ATM is designed to meet the requirements of both network service providers and end users. With ATM, service providers and end users can establish priorities based on the real-time nature of the traffic. Delaysensitive voice and real-time video traffic are often given the highest priority,and nondelay sensitive traffic such as e-mail and local-area network (LAN)traffic are given a lower priority. These priority levels allow the service providers to charge according to the QoS.
Because of ATM’s popularity and its potential for providing orders of magnitude of additional bandwidth for user traffic, traditionally popular networks are being forced to use ATM in internetworking arrangements. Such arrangements and standards that allow Internet Protocol (IP) networks and Frame Relay networks to coexist with ATM provide the current deployers of these networks the flexibility to evolve networks toward the most costeffective solution. Many such interconnected network arrangements already exist, and product offerings facilitating such arrangements are available today. Systems Network Architecture (SNA) and X.25 networks have internetworking arrangements to expand their life span in the predominantly IP,
Frame Relay, and ATM networking infrastructure.

Another important trend is the move toward voice systems to use packet networks. As the rate of growth of packet data networks began to exceed that of the circuit-switched telephony networks, the need to deploy packet networks to carry voice has overtaken the need to build circuitswitched networks to carry data. Voice-to-data network interworking and the need to carry Voice over Frame Relay (VoFR), IP, and ATM networks is increasingly important to the convergence of data and voice networks into an all packet network.
Convergence to an all ATM network requires that proper consideration be given to maintaining the same or better QoS that users expect from ATM networks while the data is transported via non-ATM networks. The QoS aspect is further complicated when a wireless segment is introduced at the edge of the ATM network. Air-protocols may often alter the traffic pattern of the ATM cells that are carried over it. Impact on QoS arising from such situations must be addressed in the design of wireless systems in broadband interworking environments.

This book focuses on the following technical aspects of the use of ATM technology in wireless broadband networks:
As current data networks evolve into ATM networks, and as voice and packet data networks converge toward a single multiservice network, the development of interworking solutions becomes critical. Different access methods, and the required interworking need for using IP, Frame Relay, and, ultimately, ATM as the transport network, are identified and detailed.
Interworking functionality that is currently employed, and the necessary interworking arrangements between the access protocols and the transport protocols will be described with detailed reference to authoritative compliance standards developed by organizations that are key players in the networking field. This book will also deal with how the mapping of traffic parameters between the components in an interworking arrangement is implemented to maintain the QoS that the network user expects the service provider to deliver.

Transition of the wireless component from an expensive segment into an integral part of the voice/data network will also be explored, and the performance impact on end-to-end voice/data sessions will be assessed. The effect of traffic parameters on QoS aspects in different protocol segments of the network will be detailed, and practical solutions on setting traffic parameters to accommodate the wireless component will be discussed. ATM will be used as the main network protocol in these discussions.

Commercialization of the Ground Segment

The creation of the Communications Satellite Corporation (COMSAT) and the subsequent launch of Early Bird in 1965 ushered in the era of true commercialization of this medium. It has largely gone unnoticed that one of COMSATs greatest legacies is its introduction of major earth stations for telephone and TV services to the general public. The very first transmissions over Early Bird used the same earth stations that were constructed for Telstar; however, these facilities had very limited service capability, both in terms of capacity and physical location. The intention of Early Bird and all of the INTELSAT satellites that followed was to improve the international telecommunications network. This demanded that earth stations be located in essentially every country of the world.

COMSAT worked in cooperation with AT&T, RCA, and others in the United States, and the major post, telegraph, and telephone (PTT) agencies of countries around the world, to standardize the design and operation of these large earth stations. They helped found the International Telecommunications Satellite Organization (INTELSAT), a treaty-based cooperative of national entities around the world. (INTELSAT was created as a quasigovernmental body but spun off its transponder-leasing business to a Netherlands-based satellite operator called NewSkies Satellites.)

A typical INTELSAT earth station of the 1960s, such as the COMSAT facility in Etam, West Virginia, represented a substantial investment yet offered connectivity with other earth stations through the GEO satellites operated by this consortium. While begun and nurtured by COMSAT, INTELSAT set out on its own in the mid-1970s and initiated a broad range of other services and applications of their space segment. The initial Standard A type of station (initially requiring 30-m antennas) was joined by more cost-effective Standard B earth stations (at 15m to 20m) that allowed countries to provide domestic satellite communication networks. Similar stations were installed as part of domestic satellite (DOMSAT) systems, such as the Palapa A network.

The INTELSAT system established its preeminence through the 1960s and early 1970s as the number of earth stations grew from hundreds to thousands.
Using technology and standards originally developed by COMSAT and improved along the way by INTELSAT and its members, the earth stations interoperate effectively. At all times, they maintain quality and order (no small task for a system used by a wide range of operating organizations on every continent).
Headquartered in Washington, D.C., INTELSAT manages the system using its control center, which is connected to TT&C and monitoring stations strategically located around the world. Individual earth stations, which are owned and operated either by members (e.g., the domestic telephone companies or PTTs) or other users who obtain their authority from these entities, are under the direction of INTELSAT’s control staff.
While INTELSAT saw its global system grow rapidly in terms of the number of satellites and earth stations, an important new phase of satellite communication appeared in 1971 when Canada launched its first GEO 
DOMSAT, Anik A (Anik means brother in the native Inuit language). Telesat Canada established the first domestic satellite system, which would become a model for more than twenty other countries. It pioneered applications like rural telephone service to remote regions and national TV broadcasting directly to major cities and small communities far from terrestrial transmitters. While INTELSAT required antennas of at least 15m in diameter, the performance of the Anik satellites allowed Telesat to employ 10m and even 4.5m antennas for these services (see Figure 1.6). This innovative system also spurred Canadian industry, allowing several companies to gain
a viable foothold as international suppliers of satellite and earth station equipment.

The United States, while a pioneer of commercial satellites and earth stations, took a back seat to Canada and only produced its first domestic network in 1974 with the launch of Western Union’s Westar 1 satellite. Being a purely commercial company, Western Union implemented its ground segment to augment its existing terrestrial microwave network. Westar earth stations were located in major cities around the country to support telephone, telex, and data communications. Later, the Public Broadcasting Service (PBS) became the first U.S. television network to use satellites for program  distribution and backhaul (e.g., the point-to-point transmission of video from temporary sites and remote studios at affiliated TV stations).

Integrated ground segments of the day, primarily used analog technology for telephone, telex, and television service, and depended on human operators to control access and manage services. The first domestic satellite system outside of North America was introduced in Indonesia and its neighboring Southeast Asian neighbors in 1976. The Palapa A satellites (named for a mythical fruit that Gajah Mada, an ancient king of Java, refused to eat until all of Indonesia was united) used the same design as Anik A. The ground segment was provided on an integrated basis by an international team consisting of the Indonesian PTT, Hughes Space & Communications Company, Ford Aerospace, ITT, and the TRT division of Philips. Internal to every earth station was a singlechannel per-carrier (SCPC) system that provided telephone service on a demand-assigned (DA) basis. This was directly integrated into the Indonesian direct-distance-dialed (DDD) automated telephone network that ITT
was installing at the time.

The timeline in Figure 1.2 indicates that 1975 was pivotal for mobile satellite communications, since this was the year Marisat 1 was launched.
This L-band GEO satellite was operated by COMSAT as a means to improve ship-to-shore communications, which at the time was still dependent on ionospheric reflection at high frequency. The first shipboard Marisat terminals still needed a dish type of antenna to provide adequate link performance for SCPC telephone and telex service. At the other end of the link is a land station acting as a gateway to the telephone network in the respective country. This service was so successful that COMSAT created Inmarsat, another alliance to promote the proliferation of earth stations in the global mobile ground segment. We will discuss again how Mobile Satellite Servicesgrew to extend to the air and land, providing communications to a wide varieter terminals.

First Satellite Earth Stations

Satellite earth stations are different from previous ground-based installations in that they are intended to transmit to and receive from spacecraft. The first of these were used to track the early vehicles that were launched into orbit and deep space. The function is called tracking, telemetry, and command (TT&C) and often includes a requirement to receive various types of sensor information. Such stations could also be equipped for communication with manned spacecraft and orbiting repeaters (e.g., communication satellites).

Both the United States and the former Soviet Union introduced these capabilities as part of their respective space programs, which began in the mid-1950s. The first TT&C ground stations were, in fact, radio telescopes that had been modified for bidirectional transmission. An example is the 90-m Goldstone, California, tracking station antenna, which was installed in the late 1950s to track Explorer 1 by Jet Propulsion Laboratories (JPL), which at the time was under contract to the U.S. Army. In 1959, JPL was transferred to the National Aeronautics and Space Administration (NASA) as part of the still-operating Deep Space Network (DSN).

Also installed at Goldstone was the 30.5-m earth station for use with the Project Echo passive balloon reflector satellite. Bell Labs in Holmdale, New Jersey, provided the other end of the link with their reflector horn antenna. Horn antennas of this type have the added feature of very low side and back lobe radiation and reception, something that helps reduce noise pickup from extraneous sources. Later, this antenna was the instrument used by Bell Labs’ scientists A. A. Penzias and R. W. Wilson to make the discovery of the cosmic background noise level, produced by the Big Bang.

While experimenting with noise measurements, they were trying to find where about three degrees of excess noise was coming from.
The first active repeater satellite was Bell Laboratories Telstar 1, which allowed the United States and the United Kingdom to communicate realtime TV and voice. As the first true earth stations, these facilities were as large and elaborate as the TT&C stations and large radio telescopes that they emulated. Size did matter because of the low power and antenna gain provided by the little Telstar satellite. Because Telstar 1 was in low earth orbit (LEO), the earth stations required tracking systems; also, the service was interrupted whenever the satellite was not in simultaneous view of the end points.

All of these earth stations were highly customized and experimental in nature. They were not constructed to provide a service to subscribers and certainly were not operated as a business. In the next section, we consider how earth stations evolved into commercial ground segments installed to provide profitable communications services.

Microwave and Radar Development

Earth stations use microwave frequencies, which lie between approximately 1 and 30 GHz, and therefore owe much to the development of Radio Detecting and Ranging (RADAR) systems. Many readers are aware that a radar antenna in the Hawaiian Islands detected imperial Japanese aircraft that bombed Pearl Harbor on December 7, 1941 [3]. However, much earlier, in 1929, G. Ross Kilgore, an engineer at Westinghouse, generated 18 GHz of microwave energy with an experimental split-anode magnetronvacuum tube. This particular device could measure Doppler reflections from moving automobiles and railroad cars. Radar improved in RF power output and sophistication during World War II and shortly thereafter, providing a technology base for terrestrial and satellite microwave communications.

Professor Wilmer Barrow of the pioneering Radiation Laboratory at MIT experimented with electromagnetic horn antennas for static-less ultrahigh-frequency wave transmission. The system consisted of a conducting tube, a transmitting terminal device, and either a receiving terminal unit or the radiating horn. Other developments at the Rad Lab include a multitude of microwave components like the klystron, numerous waveguide devices like diplexers, and antenna systems for radar and communication applications. An early horn reflector antenna was developed at Bell Laboratories in 1942, a precursor to antennas used in terrestrial and satellite microwave communications. After the war, Raytheon used microwave technology in an innovative (and noncommunication) manner with their invention of the microwave oven. Not surprisingly, the first food item to be cooked was popcorn.

AT&T recognized that microwave technology could increase the capacity and reliability of long-haul communication lines. Line-of-sight microwave links were established across the developed regions of the world during the 1950s and 1960s. The terminal ends and intermediate connection points were very much like earth stations in their design and use, namely to interface the long-distance link with local users. The overall microwave network offers much that a modern ground segment can, although it is tied to the specific routing and associated real estate.
Radio astronomy, while not able to command the investment and revenues of commercial telephone and television services, still benefited from the availability of microwave technology. Here, the challenge is to receive very weak signals (effectively noise within the noise). The parabolic dish antennas grew in size to provide greater ability to discriminate distant radio emitters.

The principle behind this is that the width of the narrow antenna beam is inversely proportional to the diameter of the reflector. Thus, radio telescopes in the 30-to-100-m range soon appeared, topped by the giant 305-m Arecibo dish antenna in Puerto Rico (this antenna is actually constructed in a small lake basin and was featured in the James Bond movie Golden Eye). Constructed in 1960, and operated by Cornell University under a cooperative agreement with the National Science Foundation (NSF), the Arecibo radio telescope not only receives celestial noise, it can also transmit radar signals to map the planets of our solar system.

Radio telescopes more along the lines of earth stations were constructed at several locations in the United States and around the world. The 91-m (300-ft) Green Bank Telescope was constructed in the 1960s but experienced mechanical failure in 1988. At the time of this writing, another 100-m radio telescope was under construction, this time using the offset parabolic reflector design popular for home DBS installations (and many spacecraft reflector antennas as well). Another type of terrestrial radio communications system is the tropospheric scatter communication link, which employs microwave signals that can be propagated over the horizon (OTH).

The tropo installation shown in Figure 1.4 was installed in 1967 by the U.S. Army Signal Corps at Pleiku, Vietnam, where it provided a single-hop link of about 230 km to Nha Trang (four times line-of-sight range). As can be seen, this 15-m antenna points nearly at the horizon to acquire the relatively weak but reasonably stable) signals that have been dispersed by the troposphere.