CA1278114C - Satellite communications system employing frequency reuse - Google Patents

Abstract

SATELLITE COMMUNICATIONS SYSTEM
EMPLOYING FREQUENCY REUSE

ABSTRACT OF THE DISCLOSURE

A satellite communications system employs separate subsystems for providing broadcast and point-to-point two-way communications using the same assigned frequency band. The broadcast end point-to-point subsystems employ an integrated satellite antenna system which uses a common reflector (12). The point-to-point subsystem achieves increased communication capacity through the reuse of the assigned frequency band over multiple, contiguous zones (32, 34, 36, 38) covering the area of the earth to be serviced. Small aperture terminals in the zones are serviced by a plurality of high gain downlink fan beams (29) steered in the east-west direction by frequency address. A special beam-forming network (98) provides in conjunction with an array antenna (20) the multiple zone frequency address function. The satellite (10) employs a filter interconnection matrix (90) for connecting earth terminals in different zones in a manner which permits multiple reuse of the entire band of assigned frequencies. A single pool of solid state transmitters allows rain disadvantaged users to be assigned higher than normal power at minimum cost. The intermodulation products of the transmitters are geographically dispersed.

Description

SATELLITE COMMUNICATIONS SYSTEM
EMPLOYING ~REQUENCY REUSE

TECHNICAL FIELD

The present invention broadly relates to satsllite communication systems especially of the type employing a satellite placed in geosynchronous orbit above the earth so as to forrn a communication link between m~ny sma11 aperture terrninals on the earth. More particularly, the invention involves a communication satellite having hybrid communication capability accomnodating both two-way and broadcast COrr3mUniCAtiOn systems. Two-way communications between small aperture earth terrrunals is achieved through multi-fold reuse OI a fixed frequency spectrum in contiguous zones of an area on the earth.

BACKGROUND ART
_ In domestic communication satellite systems, which interconnect large numbers of very small aperture earth terminals, the most important pararneters affecting the system capacity are the Effective Isotropic Radiated Power (EIRP) and the available bandwidth.
EIRP refers to a measure of the satellite's transmitter power which takes into consideration the ~ain of the antenna. EIRP is the power of a transmitter and isotroplc antenna that would achieve the same result as the transmitter and antenna which is actually err~ployed.

In the past, high antenna gain and multiple frequency reuse has been achieved by employing a plurality of up and down linlc beams covering the regions of a country or other areas of the earth to b~
served. Both frequency division and time division systems have been used or proposed to interconnect large numbers of signals from many geogr~phically separated earth stations. Time division systems permit the satellite transmitters to operate efficiently. lllis is because only one time division signal at a time is amplified in a transmitter, so it may be opera~ed at or close to signal channel saturation, the most efficient operating point. However, time division systems require high power ground transmitters and expensive signal processing and are therefore incompatible with low cost earth stetions. Frequency division systems are better suited to low cost earth stations, but have lower satellite transmitter efficiency because each transmitter handles multiple carriersO
Since multiple carrier amplifiers generate undesirable intermodulation products that increase in power as the transmitter efficiency is increased, the optimum compromise between transmitter efficiency and intermodulation product generation results in a relatively low transmitter efficiency.

In Ku band, the satellite communication band most suitable for two-way service between very srn~ll terminals, the attenuation of the signals by rain is an important consideration in the design of the system. In the previous systems, this attenuation is overcome on the downlink by using higher satellite transmitter power per channel than would be necessary for clear weather service, typically four times as much. This accommodation of rain attenuation therefore results in more expensive satellites having fewer available channels.

The available bandwidth of a satellite system is determined by the number of times the allocated frequency spectrum can be reused. Polarization and spatial isolation of beams have been employed to permit reuse of the frequency spectrum. As the number of isolated beams is increased, however, the problem of interconnecting all the users becomes very complicated and is one of the factors that limit the num~er of reuses of the frequency spectrum.

3L2~

The present invention is directed toward overcoming each of the deficiencies mentioned above.

or r~ r~ o~

The present invention provides a satellite communication system for interconnecting large numbers of very ~nall aperture earth terminals which maximizes satellite EIRP as well as the available bandwidth. The system employs highly directional, contiguous beams on the downlink or transmit signal which substantially increases the EIRP and a~lows multiple reuse of the assigned frequency spectrum. As a result, the number of communications that can be provided for point-to-point service is maximized. High multi-carrier transmitter efficiency is achieved as a result of the dispersion of intermodulation products and the deleterious effects of rain on the downlink channels are easily overcome by the use of pooled transmitter power. The interconnection of the meny users is achieved by a combination of a filter interconnection matrix and a highly directional addressable downlink beam.

According to the present invention, a system is provided for interconnecting any OI a plurality of earth terminals within an area on the earth for two~way communication using a communications satellite. A plurality of uplink beams are formed which respectively emanate from contiguous zones covering the area to be serviced by the satellite. The uplink beams carry a plurality of channels over a first preset range of uplink frequencies. Each uplink zone uses the same preset range of frequencies. The uplink frequencies are therefore reused by each zone, thereby effectively rnultiplying the number of communications chann01s that can be handled by the satellite. A
plurality of downlink beams destined for the downlink zones also carry a plurality of channels over a second preset range of frequencies. The beams for eaoh of the downlink zones RlSO use the same second preset range of frequencies to provide multiple reuse of these frequencies. The satellite employs a filter interconnection matrix for interconnecting the channels in the different zones.

~7~

A fan beam narrow in one direction, east-west for example, and broad in the orthogonal direction, is generated by a beam-forming network used in conjunction with the transmit arra~ antenna.
The east-west direction of the beam within the covered area is s deterrnined by the downlink frequency, which is related to the uplink frequency by a constant difference. The uplink frequency therefore determines the downlink frequency, which by action of the beam-forming network and the transmit array determine the direction and hence, the destination of the downlink beam. Such an arrangement is referred to as a frequency addressable beam. The side lobes of the beam are designed to be low enough to permit reuse of the frequency spectrum in the adjacent zones.

The transmitters, preferably solid state power amplifiers, are embedded in the transrnit array antenna, one transmitter being associated with each stave of the array. Each such ~mplifier arnplifies all of the tens of thousands of signals being served by the satellite. Since all of the downlink power is provided by this single pool of transrr~itters, it is easy to provide relatively high power to those relatively few signals directed to rain affected areas with only a very small reduction in power available to the much larger number of unimpaired signals.

f3ecause the transmit beam directions are related to the frequencies of their signals, and the frequencies of the intermodulation products generated in the power amplifiers differ from those of the signals which cause them, the intermodulation products go down in different directions than the si~nals. This process results in the spatial dispersion of the intermodulation products. This dispersion is enhanced by the use of multiple downlink zones, and is further enhanced by the use of zones of unequal widths. This results in a lower intermodulation product density at all ~round terminals in the frequency bands to which they are tuned. This reduced sensitivity to intermodulation products permits the power amplifiers in the satellite to be operated more efficiently.

8~

It is therefore an object of an aspect o~ the invention to provide a communication satellite for interconnecting large numbers of very small aperture antenna terminals using high satellite transmit antenna gain and allowing multiple reuse of the assigned frequency spectrum, to substantially increase the number of communication channels that can be provided for point-to-point communication service.
An object of an aspect o~ the invention is to provide the downlink power from a single pool of transmitters, so that signals being attenuated by rain can be easily allocated more satellite transmiter power.
An object of an aspect of the invention is to provide a communication satellite which disperses intermodulation products in order to increase transmitter efficiency.
An object of an aspect of the invention is to provide a communication satellite as described above which provides both broadcast and point-to-point communications service.
An aspect of the invention is as follows:
A method of interconnecting any of a plurality of terminals within a geographic area on the earth for two-way communication using a single earth orbiting communications satellite, comprising the steps of:
(A) allocatin~ a set of uplink communication channels 2~ spanning a first preselected band of communication frequencies for use in each of a plurality of uplink zones within said area wherein said uplink zones are arranged seriatim and in contiguous relationship along an axis extending across said area in a manner to cover the entire said area, such that the same uplink communication channels are used by all of said uplink zones;
(B) transmitting from each of said uplink zones to said satellite a radio frequency uplink beam of electromagnetic radiation, the uplink beam from each of said uplink zones carrying a plurality of communication signals over said uplink communication channels 7~

5a (c) receiving said uplink beams at said sat~llite;
(D) allocating a set of downlink communication channels spanning a second preselected band of frequencies ~or use in each of a plurality of downlink zones within said area wherein said downlink zones are respectively essentially coextensive with said uplink zones, such that the same downlink channels are used by all of said downlink zones;
(E) associa~ed the downlink channels in each oE said downlin~ zones respectively with a plurality of geographic sections within the as50ciated downlink zone, said geographic sections being arranged seriatim along said axis across the associated downlink zone such that each of sections is serviced by a respectively associated one of said downlink channels;
(F) transmi~ting from said satellite to each of said down link zones a plurality of radio frequency downlink beams of electromagnetic radiation, said downlink beams carrying a plurality of communication signals originating from terminals in said uplink zones, said downlink beams being transmitted in a manner such that each of said geographic sections receives only those downlin~ beams having frequencies within the downlink channel associated with a geographic section in step (E).
The foregoing and further object and advantages of the invention will be made clear or will become apparent during the course of the following description of the invention.
B~IEF DESCRIPTION OF THE D~AWINGS
In the accompanying drawings:
Figure 1 is a perspective view of a communications satellite, showing the antenna subsystems;
Figure 2 is a top plan view of the antenna subsystems shown in Figure l;

Figure 3 is a sectional view taken along the line 3-3 in Figure 2;

Figure 4 is a sectional view taken along the line 4-4 in Figure 2;

Figure 5 is a view of the United States and depicts multiple, contiguous receive zones covered by the satellite of the present invention, the primary areas of coverage being indicated in cross-hatching and the areas of contention being indicated by a dimpled pattern;

Figure 6 is a block diagram of the comnunicution electronics for the comnunications satellite;

Figure 7 is a schematic diagram of a coupling network which interconnects the point-to-point receive feed horns with the inputs to the corrmunications electronics shown in Figure 6;

Figure 8 is a reference table of the interconnect channels employed to connect the receive and transmit zones for the point-to-point system;

Figure 9 is a diagrumnatic vlew of the United States depicting multiple contiguous transmit zones covered by the satellite and the geographic distribution of the interconnected chunnels for each zone, across the United States;

Figure 9A is a graph showing the vuriation in gain of the transmit antenna beam for each zone in the point-to-point system in relation to the distance from the center of the beam in the east-west direction;

Figure 9B is a graph similar to Figure 9A but showing the variation in gain in the north-south direction;

Figure 10 is a detailed schematic diagram of the filter interconnection matrix e~r~loyed in the point-to-point system;

Figure 11 is a detailed, plan view of the beam-forming netwo~k employed in the point-to-point systern;

Figure 12 is an enlarged, fragmentary view of a portion of the beam-forming network shown in Fig~ure 11;

Figure 13 is a front elevational view of the transmit array for the point-to-point system, the horizontal slots in each transmit element not being shown for sake of simplicity;

Figure 14 is a side view of the transmit element of the array shown in Figure 13 and depicting a corporate feed network for the element;

Figure 15 is a front, perspective view of one of the transmit elements employed in the transmit array of Figure 13;

Figure 16 is a front view of the receive feed horns for the point-to-point system; and Figure 17 is a diagramrnatic view showing the relationship between a transmitted wave and a portion of the transmit feed array for the point-to-point system.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Referring first to Figures 1-4, a communications satellite 10 is depicted which is placed in geosynchronous orbit above the earth's surface. The satellite's antenna system, which will be described S in more detail below, will typically be mounted on an earth-oriented platform so that the antenna system m~intains a constant orientation with respect to the earth.

The satellite 10 is of A hybrid communications-type satellite which provides two different types of communication services in a particular frequency band, for example, the fixed satellite service Ku band. One type of communication service, referred to hereinafter as point-to-point service, provides tWo-wQy communications between very small aperture antenna te~ninals of relatively narrow band voice and data signals. Through the use of frequency division multiple access (FDMA3 and reuse of the assigned frequency spectrum, tens of thousands of such communication channels are accomnodated simultaneously on a single linear pol~rization. The other type of cornmunication service provided by the satellite 10 is a broadcast service, and it is carried on the other linear polarization. The broadcast service is primarily used for one-way distribution of video and data throughout the geographic territory served by the satellite 10. As such, the transmit antenna bearn covers the entire geographic territory. For illustrative purposes throughout this description, it will be assumed that the geographic area to be serviced by both the point-to-point and broadcast services will be the United States.
Accordingly, the broadcast servlce will be referred to hereinafter as CONUS (Continental United States).

The antenna system of the satellite 10 includes a conventional omni antenna 13 and two antenna subsystems for respectively servicing the point-to-point Qnd CON US systems. The point-to-point antenna subsystem provides a two-way communicQtion link to interconnect earth stations for two-way communications. The CONUS antenna system functions as a transponder to broadcast, over n wide pattern covering the , .
' entire United States, signals received by one or more particular locations on earth. The point-to-point transmit signal and the CONUS receive signal are vertically polarized. The CONUS transmit and point-to-point receive signals are horizontally polarized. Theantenna system includes a large reflector assembly 12 comprising two reflectors 12a, 12b. The two reflectors 12a, 12b are rotated relative to each other about a common axis and intersect at their midpoints. The reflector 12a is horizontally polarized and operates with horizontally polarized signals, while the reflector 12b is vertically polarized and therefore operates with vertically polarized signals. Conseguently, each of the reflectors 12a, 12b reflects signals which the other reflector 12a, 12b transmits.
A frequency selective screen 18 is provided which includes two halves or sections 18a, 18b and is mounted on a support 30 such that the screen halves 18a, 18b are disposed on opposite sides of a centerline passing diametrically through the satellits 10, as best seen in Figure 2. The frequency selective screen 18 functions as a diplexer for separating dif~erent bands of frequencies and may comprise an array of discrete, electrically conductive elements formed of any suitable material, such as copper. Any of various types of known frequency selective screens may be employed in this antenna system. However, one suitable frequencv selective screen, exhibiting sharp transition characteristics and capable of separating two frequency bands which are relatively close to each other, is described in Canadian Patent Application Serial No.
543,179, filed July 28, 1987, and assigned to Hughes Aircraft Company. The frequency selective screen 18 effectively separates the transmitted and received signal~ for both the CONUS and point-to-point subsystems. It may be appreciated that the two halve 9a 18a, 18b of the screen 1~ are respectively adapted to separate individual signals which are horizontally and vertically polarized.
The CONUS subsystem, which serves the entire country with a single beam, has, in this example, eight conventional transponder~ each having a high power traveling wave kube ampliPier as its transmitter 82 (see ~igure 6). The CONUS receive antenna uses vertical ~z~

polari~ation, sharing the vertically polarized reflector 12b with the point-to-point transmission system. CONUS receive signals pass through the frequency selective screen half 18b and are focused on the receive feed horns 14 located at the foc~l plane 28 of reflector 12b. The antenna pattern so formed is shaped to cover CON US. The CON US
transrn~t antenna errploys horizontal polarization, and shares reflector 12a with the point-to-point receive system. Sign~ls radiating from the transmit feeds 24 are reflected by the horizontally polarized frequency selective screen 18a to reflector 12a whose secondary pattern is shaped to cover CON US.

The point-to-point subsystem broadly inclufles a transmit array 20, a subreflector 22, and receive feed horns 16. The transmit array 20, which will be described later in more detail, is mounted on the support 30, immediately beneath the screen 18. The subreflector 22 is mounted forward of the tr~nsrnit ~rray 20 and slightly below the screen 18. The signal emanAting frorn the transmit array 20 is reflected by the subreflector 22 onto one half 18b of the screen 18. The subreflector 22 in conjunction with the main reflector 12 functions to effectively magnify and enlarge the pattern of the signal ernanating from the transn~it array 20. The signal reflected from the subreflector 22 is, in turn, reflected by one half 18b of the screen 18 onto the large reflector 12b, which in turn renects the point-to-point signal to the earth. Through this arrangement, the performance of a large aperture phase array is achieved. The receive feed horns 16 are positioned in the focal plane 26 of the reflector 12a. It consists of four main horns S0, 54, 58, 62 and three auxiliary horns 52, 56, 60 as shown in Figure 16.

lleferring now also to Figures 13-15, the transmit array 20 comprises a plurality, for example forty, transmit waveguide elements 106 disposed in side-by-side relationship to form an array, as shown in Figure 13. Each of the transrnit waveguide elements 106 includes a plurality, for example twenty-six, of horizontal, vertically spaced slots 108 therein which result in the generation of a vertically polarized signal. As shown in Figure 14, the transmit array 20 is fed with a ~;~78~

transmit signal by means of a corporate feed network, generally indicated by the nurneral 110 which excites the array element in four places 114.
The purpose of the corporate feed network 110 is to provide a broadband match to the transmit waveguide element 106. Signals input to the waveguide opening 112 excite the array slots 108 so that the slot excitation is designed to give a flat pattern in the north-south direction.

Attention is now directed to Figure 5 which depicts a generally rec tangular beam coverage provided by the horizontally polarized point-to-point receive system. In this particular example, the area serviced by the point to-point system is the continental United States. The point-to-point receive system cornprises four beams Rl, R2, R3, R4 respectively emanating from the four uplink zones 32, 34, 36, 38 to the satellite, wherein each of the beams Rl-R4 consists of a plurality of individual uplink bea~ originating from individual sites in each zone 32, 34, 36, 38 and carries an individual signal from that site. The uplink beam signals from the individual sites are arranged into a plurality of chaMels for each zone. For example, zone 32 may include a plurality, e.g. six~een 27 MHz channels with each of such channels c~rrying hundreds of individual bearn signals from corresponding uplink sites in zone 32.

The signal strength for each of the four beEun pattern contours, respectively designated by numerals 32, 34, 36 and 38, are approximately 3 dB down from peaks of their respective beams. The antenna beams have been designed to achieve sufficient isolation between them to make feasible in the cross-hatched regions 39, 41, 43, 45 reuse of the frequency spectrum four times. In the dotted regions 40, 42, and 44, the isolation is insufficient to distinguish between signals of the same frequency originating in adjacent zones. Each signal originating in these regions will generate two downlink signals, one intended and one extraneous. The generation of extraneous signals in these areas will be discussed later in more detail.

It rr~y be readily appreciated from Figure 5 that the four zones covered by bearrls 32, 34, 36, 38 are unequal in width. The East Coast zone covered by beam 32 extends approximately 1.2 degrees;
the Central zone covered by beam 34 extends approximately 1.2 degrees;
S the Midwest zon0 covered by beam pattern 36 extends approximately 200 degrees, and; the West Coast zone covered by beam pattern 38 extends approx~rnately 2.0 degrees. The width of each of the four receive zones 32, 34, 36 and 38 is determined by the number of terminals and thus the population density in the various regions of the country. Thus, beam pattern 32 is relatively narrow to accorrrnodate the relatively high population density in the Eastern part of the United States while beam pattern 36 is relatively wide due to the relatively low population density in the Mountain states. Since each zone utilizes the entire frequency spectrurn, zone widths are narrower in regions where the population density is high, to accommodate the greater demand for channel usage.

As shown in Figure 9, the point-to-point transmit system comprises four beams Tl, T2, T3, T4 respectively covering the four transmit zones 31, 33, 35, 37, wherein each of the bean~ Tl-T4 consists of a plurality of individual downlink beams destined for the individual downlink sites in each zone 31, 33, 35, 37 and carries an individual signal to that site. The downlink beam signals, destined to be received at the individual downlink sites, are arranged into a plurality of channels for each zone. For example, zone 31 may include a plurality, e.g. sixteen 27 MHz channels with each Oe such channels carrying hundreds of individual beam signals to corre-sponding downlink sites in zone 32.

The use of multiple downlink zones and downllnk zones of unequal widths assist in causing the intermodulation products, generated by the later-discussed solid state power arnplifiers, to be geographically dispersed in a manner which prevents most of these products from being received at the ground terminals. The net effect is that the amplifiers may be operated more efficiently because the system can tolerate re intermodulation products. Although the widths of the transmit zones 31, 33, 35, 37 are nearly the same as those of the receive zones R1, R2, R3, R4, small differences between the two sets have been found to optirnize the capacity of the system.

The half power beam width of the individual transmit bean~ 29 is substantially narrower than that of the transmit zones 31, 33, 35, 37. This results in the desirable high gain, and avoids the zones of contention 40, ~2, 44 characteristic of the receive zone arrangement.
l~ese individual beams 29 must be steered within the zones in order to maxirnize the downlink EIRP in the directions of the individual destin&tion terrninals. The transrnit point-to-point frequency addressable narrow beams 29 are generated by an array 20 whose apparent size is magnified by two confocal parabolas comprising a main reflector 12b and a subreflector 22. The east-west direction of each bearn 29 is deterrruned by the phase progression of its signal along the array 106 of transnit elements 20 (Figures 13 and 15). This phase progression is established by a later-discussed bealr-forming network 98 and is a function of the signal frequency. Each of the transrnit array elements 20 is driven by a later discussed solid state power amplifier. The power delivered to the array elements 106 is not uniforrn but is instead tapered with the edge elernents being more than 10 dB down. Tapering of the beams 29 is achieved by adjusting the transmit gain according to the position of the tran~nit array elements 20. The excitation pattern determines the characteristics of the transrnit secondary pattern, shown in Figure 9A.
Referring to Figure 9, the closest spacing between transrnit zones 31, 33, 35, 37 occurs between zones 31 and 33 and is approxin~tely 1.2 degrees.
This means that a signal addressed to zone 33 uslng a particular frequency would interfere with a signal using the s~me frequency in zone 31 with its side lobe 1.2 degrees from its beam center. However, the individual transmit gains have been adjusted to provide low side lobe levels, thereby perrnitting freguency reuse in adjacent zones. Referring to Figure ~A, it is seen that the side lobe level at this angle off beam center is more than 30 dB down, so that such interference will be negligibly ~anall. The same frequency uses in zones 35 and 37 ~re ~urther removed in angle, hence the side lobe interference in those zones is even srnaller.

Figure 9B is an illustration of the transmit beam pattern in the north-south direction. The twenty six slots 108 in each of the transmit waYeguide elements 106 are excited in a manner which creates a nearly ~lat north-south pattern, extending over the covered range of plus and minus 1.4 degrees from the north-south boresight direction.

Both the point-to-point and CONUS systerns rn~y utilize the same uplink and downlink frequency bands, with the point-to-point system using horizontal polarization for its uplink polarization, and the CONI~S system using vertical polarization, as previously mentioned. For example, both services may, simultaneously, utilize the entire 500 MH~
uplink frequency band between 1~ and 14.5 GHz, as well as the entire 500 MHz downlink frequency band between 11.7 and 12.2 GHz. Each of the receive zones 32, 34, 36, 38 and tran~nit zones 31, 33, 35, 37, employing the point-to-point service utilizes the entire frequency spectrum (i.e. 500 MHz). Furtherrnore, this total frequency spectrum is divided into a plurality of channels, Ior example, sixteen channels each having a usable bandwidth of 27 MHz and a spacing of 30 MHz. In turn, each of the sixteen channels may accorrmodate approximately 800 subchannels. Hence, within each zone, approximately 12,500 (16 channels x 800 subchannels) 32 kilobit per second channels rnay be accommoclated, at any given moment. As will be discussed below, the communication ~rchitecture of the point-to-point system allows any terminal to communicate directly with any other terminal. Thus, within a single polarization, a total of 50,000 subchannels may be accomnodated nationwide.

Referring now particularly to Figures 1, 2, 6, 7 and 16, the point-to-point receive feed array 16 employs seven receive horns 50-62. Horns 50, 54, 58 and 62 respectively receive signals from zones 32, 34, 36 and 38. Horns 52, 56 and 60 receive signals from the zones of ~7~

contention 40, 42 and 44. Using a series of hybrid couplers or power dividers C1-Cg, the signals received by horns 50-62 are combined into four outputs 64-70. For example, a signal originating from the area of contention 44 and received by horn 60 is divided by coupler C2 and portions of the divided signal are respectively delivered to couplers C1 and coupler C~ whereby the split signal is combined with the incoming signals received by horns 58, 62 respectively. Similarly, signals originating from the area of contention 42 and received by horn 56 are split by coupler Cs. A portion of the split signal is combined, by coupler C3, with the signal output of coupler C4, while the re~ining portion of the split signal is combined, by coupler C7, with the signal received by horn 54.

Attention is now particularly directed to Figure 6 which depicts, in block diagram form, the electronics for receiving and transmitting signals for both the CONUS and point-to-point systems. The point-to-point receive signals 64-70 (see also Figure 7) are derived from the point-to-point receive feed network in Figure 7, whereas the CONUS
receive signal 72 derives from the CONUS receive feed horns 14, (Figures 1 and 3). Both the point-to-point and CONUS receive signal are input to a switching network 76 which selectively connects input lines 64-72 with five corresponding receivers, eight of which receivers are generally indicated at 74. The receivers 74 are of conventional design, three of which are provided for redundancy and are not norm~lly used unless a mslfunction in one of the receivers is experienced. In the event of a malfunction, switching network 76 reconnects the appropriate incoming line 64-72 with a back-up receiver 74. Receivers 74 function to drive the filters in a filter interconnection mEItrix 90. The outputs of the receivers 74, which are connected with lines 64-70, are coupled by a second switching network 78 through four receive lines R1-R4 to a fllter interconnection m~trix 90. As will be discussed later below, the filter interconnection matrix (FIM) provides interconnections between the receive zones 32, 34, 36, 38, and the transmit zones 31, 337 35, 37.
Operating in the above-mentioned 500 MHz assigned frequency spectrurn, separated into sixteen 27 MHz channels, four sets of sixteen filters are err~loyed. F ach set of the sixteen filters utilizes the entire 500 MHz frequency spectrum and each filter has a 27 MHz bandwidth. As wi~l be discussed later, the filter outputs Tl-T4 are arranged in four groups, each group destined for one of the four transmit zones 31, 33, 35, 37.

The transmit signals Tl-T4 are respectively connected, ViQ sw~tching network 34, to four of six driving amplifiers 92, two of such amplifiers 92 being provided for back-up in the event of failure. In the event of the failure of one of the ~mplifiers 92, one of the back-up amplifiers 92 will be reconnected to the corresponding transmit signal T1-T4 by the switching network 94. A similar switching network 96 couples the amplified output of the amp]ifiers 92 to a bearn~forming network 98.
As will be discussed later in more detail, the bearn-forming network 98 consists of a plurality of transrnission delay lines connected at equal intervals along the four delay lines. These intervals and the width of the del~y lines are chosen to provide the desired centerband be~ squint and the beam scan rate with frequency for the corresponding transmit zones 31, 33, 35, 37 to be serviced. The transmit sign~ls, coupled from the four delay lines, are summed in the beam-forming network 98 as shown in Figures 11 and 12S to provide inputs to solid state power arnplifiers 100, which may be embedded in the point-to-point system's transmit array 20. In the illustrated embodiment discussed below, forty solid state power amplifiers (SSPAs) 100 are provided. Each of the SSPAs 100 arr~lifies a corresponding one of the forty signals formed by the beam-forrning network 98. The SSPAs 100 possess different power cap~cities to provide the tapered array excitation previously mentioned. The output of the SSPA 100 is connected to the input 112 (~igure 14) at one of the elements of the trensmit array 20.

The receive signal for CONUS on line 72 is connected to an appropriate receiver 74 by switching networks 76, 78. The output of the receiver connected with the CONUS signQl is delivered to an input multiplexer 80 which provides for eight channels, as mentioned above.
The purpose of the input multiplexers 80 is to divide the one low level CONUS signal into subslgnals so that the subsignals can be amp]ified on ~.Z7~

an individual basis. The CON US receive signaLs are highly amplified so that the CONUS transmit signal may be distributed to very ~11 earth terminals. The outputs of the input multiplexer 80 are connected through a switching network 84 to eight of twelve high power traveling wave tube amplifiers (TWTAs) 82, four of which TWTAs 82 are employed for back-up in the event of faiiure. The outputs of the eight TWTAs 82 are connected through another switching network 86 to an output mutliplexer 88 which recombines the eight arnplified signals to form one CONUS
transrnit signal. The output of the multiplexer 88 is deliversd via waveguide to the transrnit horns of the CONUS transmitter 24 (Figures 2 and 3).

Attention is now directed to Figure 10 which depicts the details of the FIM 90 (Figure 6). As previously discussed, the FIM
90 effectively interconnects any terminal in any of the receive zones 32, 34, 36, 38 (Figures 5) with any terminal in any of the transmit zones 31, 337 35, 37. The FIM 90 includes four waveguide inputs 120, 122, 124 and 126 for respectively receiving the receive signals Rl, R2, R3 and R4.
As previously mentioned, receive signals R1-R4, which originate from a corresponding receive zone 32, 34, 36, 38 (Figure 5), each contain the entire assigned frequency spectrum, (e.g. 500 MHz), and are separated into a plurality of channels, (e.g~ sixteen 27 MHz channels). The channels are further separated into a plurality of subchannels, where each of the subchannels carries a signal from a corresponding uplink site.
The FIM 90 includes 64 filters, one of which is indicated by the numeral 102. Each of the ~ilters 102 has a passband corresponding to one of the channels (e.g. 1403 1430 MHz). The filters 102 are arranged in four groups, one for each receive zone 32, 3~, 36, 38, with each group including two banks or subgroups of eight filters per subgroup. One subgroup of filters 102 contains those filters for the even numbered channels and the other subgroup in each group contains eight filters for the odd-nurnbered channels. Thus, for exarnple, the filter group for receive signal R1 corrlprises subgroup 104 of filters 102 for odd chaMels, and subgroup 106 of filters 102 for even channels. The following table relates the receive signals and zones to their filter subgroups:

Filter Subgroues Receive Zone R0ceive Signal Odd Channels Even ~hannels 32 Rt 104 106 34 1~2 108 ll0 The filters are grouped in a unique manner such that when the receive signals Rl-R4 are filtered, the filtered outputs are cornbined to form the transmit signals. The transmit signals T1-T4 also utilize the entire assigned frequency spectrurn, (e.g. 500 MHz) . In the illustrated ernbodirrJent, each of the transmit signals Tl-T4 possesses sixteen 27 MHz wide channels, and comprises four channels from each of the four receive zones 32-38 (Figure 5).

The incorning receive signals Rl-R4 are divided into the corresponding subgroups by respectively associated hybrid couplers 128-134 which effectively divert 50% of the signal power to each subgroup. Hence, for example, one-half of the Rl signal input at waveguide 120 is diverted to transmission line 136 which services the subgroup 104 of filters 102, and the remaining half of the Rl signal is diverted to transmission line 138 which services subgroup 106 of filters 102. In a sirnilar manner, each of the subgroups 104-118 of ~ilters 102 is served by a corresponding distribution line, similar to lines 136 and 138.

The construction of subgroup 104 will now be described in rnore detail, it being understood that the rernaining subgroups 106-118 are identical in architecture to subgroup 104. At intervals along the transmission line 136, there Qre eight ferrite circulators 140, one associated with each of the odd-numbered channel filters 1û2. The function of the circulators 140 is to connect the transrrussion line 136 to each of the odd chaMel filters 102 in a lossless rnanner. Thus, for exarnple, the R1 signal enters the first ciraulator 140a and circulates it counterclockwise whereby the 27 MHz band of signals corresponding to channel 1 passes through it to circulator 142. All other frequencies are reflected. These reflected signals propagate via the circulator toward the next filter where the process is repeated. Through this process, the 5 Rl receive signal is filtered into sixteen channels by the sixteen filters 104-108 corresponding to the R1 signals. Hence, the R1 signal with frequencies in the range of channel 1 will pass through the first ferrite circulator 140a and it will be filtered by filter 1 of group 1û4.

The outputs from a fiiter subgroup 104-118 are selectively coupled by a second set of ferrite circulators 142 which swns, in a criss-cross pattern, the outputs from an adjacent group of filters 102. For example, the outputs OI channel filters 1, 5, 9, and 13 of group 104 are swrrned with the outputs of channel filters 3, 7, 11 and 15 of filter group 112. This sum appears at the output terrninal for T1 144.
Referring to Figure 8, these signals correspond to the connecffons between receive zones R1 and R3 and to transmit zone Tl.

Attention is now directed to Figures 8 and 9 which depict how the transrnit and receive signals are interconnected by the FIM 90 to allow two-way cornnunication between any terminals.
Specifically, Pigure 8 provides a table showing how the receive and transmit zones are connected together by the interconnect channels while Figure 9 depicts how these interconnect channels are distributed geographicaliy across the transmit zones 31, 33, 35, 37. In Figure 8, the receive signals Rl-R4 are read across by rows of interconnect channels and the transmit signals T1-T4 are read by columns of interconnect channels. It can be readiiy appreciated from Figure 8 that each of the transmit signals T1-T4 is made up of sixteen channels arranged in four groups respectively, where each group is associated with one of the receive signals R1-R4. The satellite comnunications system of the present invention is intended to be used in conjunction with a ground station referred to as a satellite network control center which coordinates communications between the ground terrninals ViA packet switched signais. The network control center assigns an uplink user with an uplink frequency based on the location of the desired downlink, assigning the available frequency whose downlink longitude is closest to that of the destination. The frequency addressable downlink transmit beams 29 are ~hus addressable by the frequencies of the uplink signals.
S This strategy maximizes the gain of the downlink signal.

As shown in Figure 9, the continental United States is divided into four primary zones 31, 33, 35, 37. Zone 31 may be referred to as the East Coast zone, zone 33 is the Central zone, zone 35 is the Mountain zone, and zone 37 is the West Coast zone. As previously mentioned, each of the zones 31, 33, 35, 37 utilizes the entire assigned frequency spectrwn (e.g. 500 MHz). Thus, in the case of a 500 MHz assigned frequency band, there exists sixteen 27 MHz channels plus guard bands in each of the zones 31, 33, 35, 37.

The numbers 1-16 repeated four tirr~s above the bear~
29 in Figure 9 indicate the longitude of the beam; corresponding to the center frequencies of the channels so numoered. Because of the frequency sensitivity of the beam;, the longitude span between the lowest and highest frequency narrow band signal in A channel is approximately one channel width. Each beam is 0~6 degrees wide between its half power point, about half the zone width in the East Coast and Central zones and nearly one-third the zone width in the Mountain and West Coast zones.
The antenna beam; 29 overlap each other to ensure a high signal density;
the more that the beams overlap, the greater channel capacity in a giv0n area. Hence, in the East Coast zone 31, there is a greater overlap than in the MountQin zone 35 because the signal traffic in the East Const zone 3L is considerably greater than that in the Mountain zone 35.

The interconnect scheme described above wi~l now be explained by way of a typical comnunication between terminals in different zones. In this exarnple, it will be assum~d that a caller in Detroit, Michigan wishes to place a call to a terrninal in Los Angeles, C~lifornia~ Thus, Detroit, Michigan, which is located in the Central zone 34, is the uplink site, and Los Angeles, California, which is located in the West Coast zone 37, is the down~ink destination. As shown in Pigure ~, each geographic location in the continental United States can be associated with a specific channel in a specific zone. Thus, Los Angeles is positioned between channels 14 and 15 in trQnsmit zone 37.

5Referring now concurrently to Pigures 5, 8 and 9 particularly, receive and transmit zones R1 and Tl lie within the East Coast zone 32 and 31, R2 and T2 lie within the Central zone 34 and 33, R3 and T3 lie within the Mountain zone 36 and 35, and R4 and T4 lie within the West Coast zone 38 and 37. Since Detroit lies in the Central 10or R2 zone 34, it can be seen that the only channels over which signals can be transmitted to the West Co~st or T4 zone 37 are channels 1, 5, 9 and 13. This is determined in the table of Figure 8 by the intersection of row R2 and column T4. Therefore, from Detroit, the uplink user would uplink on either channel 1, 5, 9 or 13, whichever of these channels is 15closest to the downlink destination. Since Los Angeles is located between chennels 14 and 15, the network control center would uplink the signal on channel 13 because channel 13 is the closest to channel 14.
The downlink beam width is broad enough to provide high gain at Los Angeles.
20Conversely, if the uplink site is in Los Angeles and the downlink destination is in Detroit, the intersection of row R4 and column T2 in Figure 8 must be consulted. This intersection reveals that the signQl can be transmitted on channels 1, S, 9 or 13 depending upon which channel is closest to the downlink destination. The network control 25center would uplink the signal from Los Angeles on channel 9 since channel 9 is closest to channel 11 which, In turn, is closest to Detroit.

Returning now to Figure lû, the conversion of a receive signal to Q transmit signal will be described in cormection with the example mentloned above in which the uplink site is in Detroit and 30the downlink site is in L,os Angeles. The uplink signal transmitted from Detroit would be transmitted on ch~nnel 13 carried by receive signal R2.
Thus, the R2 receive signal is input to tran~nission Une 122 and a portion of such input signal is diverted by the hybrid coupler 130 to the input line of subgroup 108 of filters 102. Subgroup 108 includes a bank OI
eight filters for the odd channels, including channel 13. Thus, the incoming signal is filtered through by filter 13 and is output on a line 164 along with other signals from subgroups 108 and 116. The channel 13 sign~l present on line 164, is combined by the hybrid coupler 158, with signals emanating from subgroup 106 and 114, and forms the T4 signal on output line 150. The transmit signal T4 is then downlinked to Los Angeles.
It is to be understood that the above example is somewhat simplified inasmuch as the network control center would assign a more specific channel than a 27 MHz wide band channel, since the 27 MHz wide channel may actually comprise a multiplicity of sm~ller channels, for exalTple, 800 subchannels of 32 KHz bandwidth.
Referring now again to Figures 5, 8 and 9, in the event that an uplink signal originates from one of the areas of contention, 40, 42, 44 (Figure 5), such signal wil1 not only be transmitted to its desired downlink destination, but a non-neglible signal will be transmitted to another geographic area. For example, assurne that the uplink signal originates from Chicago, I~linois which is in the area of contention 42 and that the signal is destined for Los Angeles, Californi~. The area of contention 42 is produced by the overlap of the bea~ forming zones 34 and 36. Hence, the uplink signal can be transrnitted as receive signals R2 or R3. The network control center determines whether the uplink comrunication is carried by receive signals R2 or R3. In the present example, since Chicago is closer to zone 36, the uplinlc comnunication is carriec~ on receive signal R3.

As previously discussed, the downlink destination, Los Angeles, is located in zone 37 ~nd lies between channels 14 and 15. As shown in Figure 8, the intersection of R3 with colwmn T4 yields the possible channels over which the comn~mication can be routed. Thus, the Chicago uplink signal will be transmitted over one of channels 2, 6, 10 or 14. Since Los Angeles is closest to channel 14, chann~l 14 is selected by the network control center as the uplink channel. Note, however, that an unclesired signal is also transmitted from zone 34 on channel 14.
To determine where the undesired signal will be downlinked, the table of Figure 8 is consulted. The table of Figure 8 reveals that uplink signals carried on channel 14 in the R2 zone 34 are downlinked to the Tl transmit zone 31. The desired signal is transmitted to Los Angeles and the undesired signal (i.e. an extraneous signal) is transmitted to the East Coast (i.e. zone 31). The network control center keeps track of these extraneous signals when making frequency assignments. The effect of these extraneous signals is to reduce slightly the capacity of the system.

Referring now again to Figure 6, the beam-forming network 98 receives the transmit signals T1-T4 and functions to couple all of the individual com~mnication signals in these transmit signals together so that a transmit antenna beam for each signal is ~ormed. In the example discussed above in r~hich the assigned frequency spectrum is 500 MHz, a total of approximately 50,000 overlapping antenna beams are formed by the beam-forming network 98 when the system is fully loaded with narrow band signals. Each antenna bearn is formed in a manner so that it can be pointed in a direction which optimizes the performance of the system. The incremental phase shift between adjacent elements determines the direction of the antenna beam. Since this phase shift is determined by the signal frequency, the system is referred to as frequency adcire~sed.
Attention is now directed to Figures 11 and 12 which depict the details of the beam-forming network 98. The beam-forming network, generally indicated by the numeral 98 in Figure 11, is arranged in the general forrn of an arc and may be conveniently mounted on the communication shelf (not Yhown) of the satellite. The arc shape of the beam-forming network 98 facilitates an arrangement which assures that the paths of the signals passing therethrough are of correct length.

1 The beam-forming network 98 includes a first set of circurnferentia~y extending transmission delay lines 168, 170, ~ second set of transmission delay lines 172, 174 which are radia~y spaced fron de~y lines 168 and 170, and a plurality of radia~y extending waveguide assem~lies 176. In the i~ustrated embodim~nt, forty waveguide assem~lies 176 are provided, one for each of the elements 106 of the transrnit array 20 (Figure 13). The waveguide assemblies 176 intersect each of the delay lines 168-174 and are equA~y spaced in angle.

Each of the waveguide assemblies 176 defines a radial line summer and intersects each of the delay ~nes 168-174. As shown in Figure 12, at the points of intersection, between the radial line sunnTers 176 and the transmission delay lines 168-174, a crossguide coupler 180 is provided. The crossguide coupler 180 connects the delay ~nes 168-174 with the radial line surnners 176. The function of the crossguide couplers 180 wi~ be discussed later in n-ore detail.

Four delay lines 168-174 are provided respectively for the four transrnit zones T1-T4 (Figure 9). Hence, transn~t signal T1 is provided to the input of delay line 170, T2 is provided to input of delay line 168, T3 is provided to the input of delay line 174, and T4 is provided to the input of delay line 172. As shown in Figure 12, the distance between the radial line sunTners is indicated by the letter "l" and the width of each of the radial delay lines is designated by the letter "w".
Although the r~dial line suunners 176 are spaced at equal angular intervals along the delay lines 168-174, the distance between them varies from delay ~ne to delay line due to the fact that the delay lines 168-174 are radia~y spaced fr~n each other. Thus, the further from the center of the arc, which is formed by the radial line summers 176, the greater the distance between the radial line sunnners 176, at the point where they intersect with the delay lines 168-174. In other w~rds, the spacing "l"
between radial line sumners 176 for delay line 168 is less than the spacing between adjacent radial line summers 176 than for delay line 174.
Typical values for the dimensions "l" and "w" are as fo~ows:

1 Delay Line ~ignal l, inches w, inches 168 T2 1.66 0.64 170 Tl 1.72 0.66 172 T4 2.45 0.74 174 T3 2.55 0.76 The width Or the delay lines 168-174, "w", and the distance 1~1!1 between adjacent radial line summers are chosen to provide the desired center beam squint and beam scan rate so that the beam pointing is correct for each channel. This results in the desired start and stop points for each of the tran~nit zones T1-T4.

Referring particul~rly to Figure 12, the transmit signal T2 propagates down the delay line 168 for a precise distance, at which point it reaches the first radial line sunnner 176. A portion of the T2 signal passes through the crossguide coupler 180, which ~ay, for exainple, lS be a 20 dB coupler, such that one percent of the transmitted power of tran~nit signal T2 is diverted down the radial line sun2ner 176. This diverted energy then propagates down the waveguide 176 towards a corresponding solid state power a ~ lifier 100 (Figures 6 and 11). This process is repeated for signal Tl which propag&tes down delay line 170.
The portions of signals Tl and T2 which are diverted by the crossguide couplers 180 (i.e. 0.01 Tl and 0.01 T2) are summed together in the radial line summer 176 and the combined signal 0.01 (T1 + T2) propagates radia~y outwardly toward the next set of delay ~nes 172, 174. This same coupling process is repeated for signa~ T3 and T4 in delay ~nes 174 and 172 respectively. That is, 0.01 of signals T3 and T4 are coupled via crossguide couplers 180 to the radial line surnner 176. The resulting co ~ ined signal 0.01 (T1 + T2 + T3 + T4) propagates radia~y outwardly to an associated solid state power aJnp~fier 100 where it is aTplified in preparation for transmission.

After encountering the first radial line surnner 176, the remaining 0.99 of signals T1-T4 propagate to the second radial line sunTner where an additional one percent of the signals is diverted to the summer 176. This process of diverting one percent of the signals T1-T4 is repeated for each of the radial line summers 176.

The signals, propagating through the radial line summers 176 towards the SSPAs 100, are a rnixture of all four point-to-point transmit signals T1-T4. However, each of the transmit signals T1-T4 may comprise 12,500 subsignals. Consequently, the forty signals propagating through the radial line suJTmers 176 rr~y be a mixture of all 50,000 signals in the case of the errbodiment mentioned above where the assigned frequency spectrum is 500 MHz wide. Therefore, each of the SSPAs 100 amplifies all 50,000 signals which emanate from each of the plurality of wave guide assemblies 176.

Since each of the SSPAs 100 arTplifies all 50,000 signals which are destined for all regions OI the country, it can be appreciated that all of the narrow, high gain downlink beams are formed from a comnon pool of transrnitters, i.e. all of the SSPAs 100. This arrangement may be $hought of as effectively providing a nationwide pool of power since each of the downlink beams covering the entire country is produced using all of the SSPAs 100. Consequently, it is possible to divert a portion of this nationwide pool of power to accommodate specific, disadvantaged downlink users on an individual basis without materially reducing the signal power of the other bearr~. For example, a downlink user rnay be "disadvantaged" by rain in the downlink destination which attenuates the signal strength of the beam. Such a rain disadvantaged user rnay be individually accolrmodated by increasing the signal strength of the corresponding uplink beam. This is accomplished by diverting to the disadvantaged downlink beam, a small portion of the power from the pool of nationwide transmitter power (i.e. a fraction of the power supplied by all o~ the SSP~s 100). The power of an individual uplink bearn is proportional to that of the corresponding downlink beam.

~;~7~

Consequently, in order to increase the power of the downlink bearn it is merely necessary to increase the power of the uplink bearlL

In practice, the previously rnentioned network control center keeps track of all of those regions of the country irl which it is raining nnd deterrnines which of the uplink users are placing comrnunications to downlink destinations in rain affected areas. The network control center then instructs each of these uplink users, using packet switched signals, to increase its uplink power for those signals destined for a rain Rffected area. The increase in power of the uplink user's signals results in greater collective arnplification of these signals by the SSPAs 100, to produce corresponding downlink bearns to the rain affected areas, which have power levels increased sufficiently to cornpensate for rain attenuation. Typically, the nurnber of signals destined for rain affected areas is small re1ative to the total number of signals being handled by the total pool OI SSPAs 100. Accordingly, other downlink users not in the rain affected zones do not suffer substantial signal loss since the small loss that may occur in their signals is spread out over the m~ny thousand users.

The SSPAs 100 (Figures 8 and 11) rnay be mounted, for exarnple, on the rim of the commlmication shelf (not shown) of the satellite. The signals arnplified by the SSPAs 100 are fed into the corresponding elements 106 of the transrnit array 20 (Figure 13 and 14).

As previously discussed, an incrernental phase shift is aehieved between the signals that are coupled in the forty radial line sumners 176. Elence, the bearn-forrning network 9~ permits the antenna bearns emanating from the transrr~t array 20 (Figures 1, 2, and 13) to be steered by ~requency assignment. The incrernental phase shift is related to the time delay between the waveguides 176 as well as frequency.
Attention is now directed to Figure 17 which is a diagrarr~natic view of four of the forty transmit array elerr~nts 106 (Figure 13), showing the wavefront em~nating therefrom, wherein 'td" is equal to the spacing between transmit array elements 106. The resulting antenna beam has an ~27~

--2~--~ngular tilt of ~ , where ~ is defined QS the bearn scan angle. This means that ~ is the angle from norrn~l of the trQnsmit beam center.
The incrernental phase shift produced ~y the delay line Qrr~ngement is The relHtionship between ~ ~ ~nd ~ is given by Q~=2 d sin9 where:
A = slgnal wavelength 3 = beam scan angle d = spacing between array elements EIence, the east-west direction of the Hntenna beam is determsned by the incremental phase shift which is different for the four delay lines 168-174 of the beam-forming network 98, resulting in the four transrnit zones T1-T4 previously noted.

Having thus described the invention, it is recognized that those skilled in the art mEIy nEIke various modifications or additions to the preferred embodiment chosen to illustrate the invention without departing from the spirit und scope of the present contribution to the art. Accordingly, it is to be understood that the protection sought and to be afforded hereby should be deemed to extend to the subject ~r~tter claimed and Qll equivalents thereof fairly within the scope of the invention.

Claims (6)

1. A method of interconnecting any of a plurality of terminals within a geographic area on the earth for two-way communication using a single earth orbiting communications satellite, comprising the steps of:
(A) allocating a set of uplink communication channels spanning a first preselected band of communication frequencies for use in each of a plurality of uplink zones within said area wherein said uplink zones are arranged seriatim and in contiguous relationship along an axis extending across said area in a manner to cover the entire said area, such that the same uplink communication channels are used by all of said uplink zones;
(B) transmitting from each of said uplink zones to said satellite a radio frequency uplink beam of electromagnetic radiation, the uplink beam from each of said uplink zones carrying a plurality of communication signals over said uplink communication channels;
(C) receiving said uplink beams at said satellite;
(D) allocating a set of downlink communication channels spanning a second preselected band of frequencies for use in each of a plurality of downlink zones within said area wherein said downlink zones are respectively essentially coextensive with said uplink zones, such that the same downlink channels are used by all of said downlink zones;
(E) associated the downlink channels in each of said downlink zones respectively with a plurality of geographic sections within the associated downlink zone, said geographic sections being arranged seriatim along said axis across the associated downlink zone such that each of sections is serviced by a respectively associated one of said downlink channels;

(F) transmitting from said satellite to each of said down link zones a plurality of radio frequency downlink beams of electromagnetic radiation, said downlink beams carrying a plurality of communication signals originating from terminals in said uplink zones, said downlink beams being transmitted in a manner such that each of said geographic sections receives only those downlink beams having frequencies within the downlink channel associated with a geographic section in step (E).

2. The method of claim 1 including the step of:
(G) associating the frequencies in said first band with the frequencies in said second band such that the frequency in said first band selected to carry a communication from a terminal in an uplink zone is related to the frequency in said second band which is contained within a downlink channel associated with a location in a geographic section containing a terminal intended to receive said communications; and (H) selecting a frequency in said first band based on the location of the geographic section containing the terminal intended to receive said communication.

3. The method of claim 1, including the step of selecting the geographic area covered by said uplink zones such that said uplink zones are of different widths measured along said axis, and the widths of said uplink zones are respectively functions of the density of terminals within said uplink zones.

4. The method of claim 1, wherein step (D) is performed in a manner such that two adjacent sections in two respectively associated downlink zones are serviced by channels which are different in frequencies to prevent communication interference between terminals in said adjacent sections.

5. The method of claim 1, including the step of selecting the geographic areas covered by said uplink and downlink zones, said selecting step being performed in a manner such that the geographic areas covered by said uplink zones are slightly different in size from the geographic area covered by said downlink zones.

6. The method of claim 1, including the step of selecting the widths of said uplink and downlink zones measured along said axis, said selecting step being performed in a manner such that at least some variation exists in the widths of said uplink and downlink zones.