Fermat's Library | Satellite Communication - An overview of the problems and programs annotated/explained version.

294 PROCEEDINGS

OF

THE IEEE,

VOL.

65, NO.

3,

MARCH 1977

Satellite Communication-An Overview

of

the

Problems and Programs

WILBUR L. PRITCHARD,

FELLOW,

IEEE

Abrmct-This

paper

introduces

the

subject

of

satellite

communica-

tion

in

its

broadest

aspects,

recounts

its

history

and

discusses

the

principal technical problems It

is

primprily communications-oriented

but

relevant

spacecraft

and

launch

considerations

are

summarized.

Tabular summaries

of

the

world’s

satellite

communication programs

are

given.

I. INTRODUCTION

OMMUNICATION satellites have several important

characteristics. One

is

certainly the availability

of

band-

widths exceeding anything previously available for inter-

continental communications. Although overland transmission

of

highquality TV pictures by microwave radio relays or cable

has been possible for some years, trans-Atlantic TV transmis-

sion took place for the first time only after the

first

active

communication satellite had been put in orbit. Interconti-

nental relaying of TV programs, now commonplace,

is

done

exclusively by satellites.

Another, perhaps the most important

of

all,

is

the unique

ability to cover the globe. In the future, it

is

likely that cables

will

use

much higher carrier frequencies, probably as high as

the optical region

of

the spectrum. If

so,

a multitude

of

TV

channels or their equivalent could be transmitted from one

continent

to

another without satellites. However, a cable still

has

two fixed ends and there must

be

a connection between

every pair of points to be in communication. Satellite systems

offer, in this respect, a flexibility that cannot now be dupli-

cated. Furthermore, this flexibility applies not only to fixed

points on earth, but also to moving terminals, such as ships at

sea, airplanes, and space vehicles.

With communication satellites, then, instant and reliable

contact can be established rapidly between any points on

earth, in addition to, and well beyond, the capabilities

of

avail-

able land lines, microwave line-of-sight relay systems, and

other techniques. Satellites are the elements

of

a communica-

tions revolution analogous to that in transportation resulting

from the airplane.

11.

HISTORY

Although the

origins

of the whole idea

of

satellite communi-

cation

are

obscure, there

is

no question that the synchronous,

or more accurately, geostationary, satellite was first proposed

by Arthur C. Clarke in an article in

Wireless World

entitled,

“Extraterrestrial Relays.” He recognized the potential for

rocket launches based on the German V2 work during the war

and also the conspicuous advantage

of

the geostationary orbit.

Prophetidy, his proposal was for the

use

of

these satellites

for FM voice broadcast rather than for telephone service.

Interestingly enough, Clarke also foresaw the

use

in space

of

Manuscript received April, 1976;revised September 7, 1976.

The author

is

with Satellite Systems Engineering, Inc., Washington,

DC 20014.

electric power generated by panels

of

solar cells. Implementa-

tion

of

his

idea still had to wait for the Space Age (Sputnik-

1957) and solid-state technology.

Thirty-one years have passed since

his

prophecy, and there

are now 22 satellite communication programs with satellites

in

orbit or under active construction. There are another score of

programs with earth stations only using the satellites

of

others.

A word

of

tribute to

his

exceptional vision

is

certainly in

order.

A.

The Early Years

Moon reflections for radar and communication purposes

were repeatedly demonstrated in the late forties and early

fifties. In July 1954, the

first

voice messages were transmitted

by the

US

Navy over the earth-moon path. In 1956, a

US

Navy moon relay service was established between Washington,

DC, and Hawaii. This circuit operated until 1962, offering

reliable longdistance communication limited only by the

“availability”

of

the moon at the transmitting and receiving

sites. Power used was 100 kW, with 26-m diameter antennas

at 430 MHz.

A metallized balloon

of

the correct size, launched by a

rocket and placed in orbit can be used as a scatterer of electro-

magnetic waves generated by an earth transmitter. Part of the

energy can be picked up by receiving stations at any point on

earth from which the balloon

is

visible, thus obtaining a pas-

sive communications satellite system.

Through the joint action

of

Bell Telephone Labs, NASA, and

JPL, the ECHO experiment was performed. Successful

communications across the

US

were first established in early

August 1960, between Goldstone, CA, and Holmdel, NJ, at

frequencies

of

960 MHz and 2290 MHz. The ECHO “balloon,”

in an inclined orbit at 1500-km altitude, was visible to the

unaided human eye.

Later in the same month, the first trans-Atlantic transmis-

sion occurred between Holmdel, NJ, and a French receiving

station [I]. This project alerted the entire world to

be

prospect of the new medium

of

communications although the

specific method was never exploited commercially.

Although passive satellites have infinite capability for

multiple-access communications, they are gravely handicapped

by the inefficient use

of

transmitter power. In the ECHO

experiment, for instance, only one part

in

10l8

of

the trans-

mitted power (10 kW)

is

returned to the receiving antenna.

Since the signal has

to

compete with the noise coming from

various sources, special low-noise receivers must be used.

Luckily, the invention of the maser in 1954 and its successive

development, permitted the construction of very low-noise

receivers (with temperatures in the neighborhood

of

10

K

which, used with horn-reflector receiving antennas having an

aperature

of

about 43 m2, made possible the transmission of

teletype, voice, and pictures).

PRITCHARD: WORLD’S SATELLITE COMMUNICATION

295

The advantage of passive satellites is that they do not require

sophisticated electronic equipment on board. A radio beacon

transmitter might be required for tracking, but in general,

neither elaborate electronics, nor, with spherical satellites,

attitude stabilization is needed. Such simplicity, plus the lack

of space-flyable electronics in the late fifties, made the passive

system attractive in the early years

of

satellite communications.

As soon as space-flyable electronics became available,

it

was

obvious that passive systems would be replaced by active

satellites. The mathematics

of

the inverse

square

law for active

satellites (versus the radar-like inverse fourth-power law appli-

cable to passive satellites) are overwhelmingly in favor

of

the

former.

The relative disadvantages

of

a passive system increase with

orbital altitude and the on-board power availability of the

active satellite. After the early experimental

trials,

all subse-

quent satellite communication experimental and operational

systems have been

of

the active type, and there

is

nothing to

indicate that the situation

is

likely to change.

It is interesting

to

note that the first active

US

communica-

tion satellite was a broadcast satellite. SCORE, launched on

December 18, 1958, transmitted President Eisenhower’s

Christmas message to the world with a power

of

8

W

at a

frequency

of

122 MHz. SCORE was a delayed-repeater

satellite receiving signals from earth stations at 150 MHz; the

message was stored on tape and later retransmitted. The 68

kg payload was placed in rather low orbit (perigee 182 km,

apogee 1048 km).

The communications equipment was battery powered and

not intended to operate for a long time. After 12 days

of

operation, the batteries had fully discharged and transmission

stopped.

B.

The Experimental Years

Aside from early space probes like Sputnik, Explorer, and

Vanguard, as well as the SCORE and Courier projects, which

were early communication satellites of the record and re-

transmit type, the major experimental steps in active com-

munication satellite technology were the Telstar, Relay, and

Syncom projects.

Project Telstar is the best known

of

these probably because

it was the first one capable of relaying TV programs across the

Atlantic. This project was begun by AT&T and developed by

the Bell Laboratories, which had acquired considerable knowl-

edge from the early work

of

John R. Pierce and his associates,

and from the work with the ECHO passive satellite. The first

Telstar was launched from Cape Canaveral on July 10, 1962.

It was a sphere of approximately 87 cm diameter, weighing

80 kg. The launch vehicle was a Thor-Delta rocket which

placed the satellite into an elliptical orbit with an apogee

of

5600 km, giving it a period of 2-$ h.

Telstar

I1

was made more radiation resistant because

of

experience with Telstar

I,

but otherwise,

it

was identical to

its predecessor. It was successfully launched on May 7,1963.

The power of Telstars

I

and

I1

was 2.25

W

provided by a

TWT, with an RF bandwidth of 50 MHz at 6 and

4

GHz. Both

satellites were spin-stabilized. The overall communication

capability was 600 voice telephone channels, or one TV

channel. To overcome the low camer-to-noise ratio avail-

able in the down-link, receivers at the earth stations used

FM

feedback in order

to

obtain an extended threshold. Even

though the Telstar system was superbly engineered, it was

designed as an experiment and was not intended for commer-

cial operation. Among other things, the orbit used made it

only visible for brief periods.

A

project with similar objec-

tives, Project Relay, was developed by the Radio Corporation

of America under contract to NASA. It was similarly

successful.

In early 1962, the President sent proposed legislation to

Congress to start the commercial exploitation

of

these suc-

cesses. After extensive hearings on the Bill, the

US

Congress

passed the Communications Satellite Act of 1962, which led

to the establishment

of

the Communications Satellite Corpora-

tion in 1963.

On August 20, 1964, a significant event occurred when

agreements were signed by 11 sovereign nations which resulted

in the establishment of a unique organization-the Interna-

tional Telecommunications Satellite Consortium, known as

INTELSAT. This new organization was formed for the pur-

pose of designing, developing, constructing, establishing, and

maintaining the operation of the space segment of a global

commercial communications satellite system.

C. The Commercial Era

Commercial communications by satellite began officially in

April 1965, when the world’s first commercial communication

satellite, INTELSAT

I

(known as “Early Bird”), was launched

from Cape Kennedy. It was decommissioned

in

January of

1969 when coverage of both the Atlantic and Pacific was

accomplished by two series of satellites, INTELSAT’s

I1

and

111.

Interestingly enough, Early Bird was planned to operate

for only 18 months. Instead, it lasted four years with 100 per-

cent reliability.

The fully mature phase of satellite communications probably

is

best considered as having begun with the installation of the

INTELSAT IV into the global system starting in 1971. These

spacecraft weigh approximately 730 kg in orbit and provide

not only earth coverage but

also

two “pencil” beams about 4’

in diameter which can be used selectively to give spot coverage

to Europe and North and South America. INTELSAT IV is a

spinning satellite, as were its predecessors, but the entire

antenna assembly, consisting

of

13 different antennas,

is

stabilized to point continually toward the earth. Two large

parabolic dishes form the two spot beams. Each satellite pro-

vides about 6000 voice circuits, or more, depending upon how

the power in the satellite

is

split between the spot beams and

the earth coverage beam. INTELSAT IV can carry 12 color

TV channels at one time.

D.

Military Satellites

The first military satellites, the DSCS-I, were launched by

the

US

Air

Force in June

of

1966. These launches were

interesting because

8

satellites were launched simultaneously.

Finally, about 30 satellites

of

a very simple spinning type and

without station-keeping were placed in near synchronous

orbits. Some are still in operation today. The DSCS-I1 system

was initiated several years ago and constitutes the present

US

military system although it has had both spacecraft and launch

vehicle failures. DSCS-111

is

being planned.

111.

CATEGORIES

OF

SYSTEMS

There are some 42 satellite communication systems in the

world today, 22

of

which include both satellite and terrestrial

equipment (See Table

I).

By satellite system, we mean one

which is in active operation or one for which the equipment

is

being built under funded contract. There are literally dozens

296

PROCEEDINGS

OF

THE

IEEE,

MARCH

1977

TABLE

I

LIST

OF

PROGRAMS

1

-I

Cl...

Statu'

Coveram.

:6

-

x:

RODr-

?r.qwrSy

sada

Date

--

om)

IIRLull

5925

-

6425 up 1965

xx

X

2

US%

3

~I-lrud.

4

Ycston

hln

I

ICIutQLl

xxx

X

X

5725

-

6225

up

3400

-

3900.

ooo

-

lwo

dam

1965

xxx

3700

-

4200

dan

(C

Wd)

X

X

xx

xx

Snre

an

YESTAR

1976

5927

-

6403

3702

-

4178

1973

1974

5925

-

6425

3700

-

4200

6

kEFiXM

&TLLXTg

ccV.*

7

COnutIA~

(-)

YESrAR

Trnnapmder

1974

xx

X

X

X X

5925

-

6425 3700

-

42w)

1976

xx

14152.5

-

14192.5 11490

-

lU30

1977

X

8

i?sA

(on/ocr)

14242.5

-

14362.5

14455.0

-

1U60

1UM)

-

11700

11792.5

-

11797.5

9

ULL)u*

10

llmmzsu

X

xx

IhTEWT

Tranapoder

1975

X

X?

14WO

-

14500 11700

-

12200

liu

had

1978

X X

X

5925

-

64425

3700

-

4200

1976

xx

c

hnd

2500

I'caat

1980-90

x

X

X

c

or

YU

2500 B'cast

11

XalUTnam

su

X X

X

12

ms

u

PnILIPPMLs*.

14

TmLI

xx

1975

INIELSAT

Transponder

X

X

X X

5925

-

6425

3700

-

4200 1976

,lS

uu

-

16

CQ5ATcemU@AUUT)

Satellite

-

Ship

Satellite

-

Sa*

1975

xx

X

300

-

312 1638.5

-

1662.5 up 6174.5

-

6425

up

248

-

260

1537

-

1541

dun

3945.4

-

4199.0

d-

1645

-

1660, 131.425

-

131.975

up

SOW

-

5125

up

1541.5

-

1,43 125.511

-

1S.Yl5

dull

3125

-

5210

do.0

1641.5

-

1644.5 up

1U90

-

10500 up

1540

-

1542.5

dm

11690

-

11700

dm

L

hmd

Corhhd

17

ESAlmaUCaerul-

Satallite

-

Ship Ltellitc

&re

198Ot

X X

X

19

==(mu==)

18

IsAo

Satellite

-

Planc

1979

X

X X

20

mopwl

IRWCASTDE

ma

Satellite

-

Crd

X

Satellit.

-

Ship

S.tellite

-

shore

1977

X

X

X

YU

b0J

m

1985

X

Xu

Band

198Ot

X X

21

FEXRM.

RemaLIc

o?

COIIuIll

U

AIS4

22

sw2mm.e

X

X

X

X

xx

5926

-

6424 up 3700

-

4200

dovn

1975

24

JMAI

-

carPlaruna

xx

Ylde

variety

of

axprianta

in

6

C.

S!

UlR,

ad

VR

&do

1974

X X

5925

-

6425

3100

-

4200

1917

xx

25

JAPAP

-

IQ-

14

OW

1978

xx

xx

zh

CIS

X

7975

-

8162 7250

-

7437

1970

X

x

xx

30

NATO

29

m

7900

-

8400

7250

-

7750

1966

xx

x

xx

28

Dscs

xx

X

27

SnIO

X

X

31

PLLTuK*w

xx

X

X

X

1977

Satallite

-

Ship

&toll1t*

-

She+.

32

LE1

SPRXES

xx

xx

x

.x

33

LKU

MIIW

34

ARRQKINA'

35

AU-LU

X

xx

X

1980'.

c

ib.

S

Ud

36

m4zu

X

?

X

31

MIPL.

?X

x

X

xx

X?

1978

5923

-

6425

3700

-

4200

30

IUS

X?

?X

x

x1

ImEWI

Tr.nopoldcr

39

MUYSIA+

X X

X

40

UXICED

KXm

x

1975

rWnWI

Trarupoder

61

sISTF2l.S

USm

I"KLSI*

xx

X

30

WO

12

OW

1975

7976

-

8005

7257

-

7281

1969

17009

-

17782

1977

l4

Oo0

11

OOO

20

ooo

1U31

-

11862

I

xx

X X

225

-

400

7900.bLoo

1965

wide

-rlety

of

exprLnto

at

IAW

.od

K

sad.

X

X

IHnl5AI

Irarupondor

X?

xx

X??

IrnLSAI

Tr.a.por*lcr

X?

--

--

62

ASW

IUIIOFS

XIX

X

X

X X

IWlZlST

Iranapoadcr

Varioua

X

7

.?-r~h

:t.:lsn.

0n:y

of other systems under study in various phases. They range, in

likelihood

of

implementation, from remote possibilities

to

being on the verge

of

realization.

We can characterize and categorize satellite systems in a

variety of manners, either by their technical characteristics or

by their operational

use.

The latter seems more natural and,

indeed,

is

instructive in the sense that after having categorized

the satellites operationally, we can examine the diversity

of

technical methods used to achieve similar operational results.

The first category of system is certainly international civil

telecommunications,

of

which we have only

2

examples

today: the highly successful INTELSAT and the inchoate

STATSIONAR

of

the Soviet Union.

The second category

.is

that

of

regional and .domestic satel-

lites, principally for civil telecommunications but occasionally

for the distribution of television programs. In this class are:

An&, Westar, Satcom, Comstar, Palapa (Indonesia), and

Molniya. For military communications (either to fixed or

mobile terminals), we have

4

systems: NATO, Skynet, DSCS,

and FLTSATCOM.

There are as yet no operational direct broadcast systems

although there are several experimental ones in that category.

It

is

likely that even operational ones will be joined with tele-

communications services.

And, finally, still in the operational rather than the experi-

mental category, we have those systems planned specifically

for communication with mobile terminals: MAROTS,

MARISAT, and AEROSAT.

The experimental programs cover a wide range

of

purposes,

from experimenting with operational problems and proving

spacecraft technology to acquiring propagation data

in

fre-

quency bands of potential interest, especially above 12 GHz.

PRITCHARD:

WORLD’S

SATELLITE

COMMUNICATION

291

In this category, we have:

ATS-6,

Canadian Technology Satel-

lite, the Japanese Communication Satellite, Japanese Broad.

cast Satellite, OTS, Symphonie, SIRIO, and the Lincoln-

Laboratory Series (LES). The

gross

technical features of these

systems are summarized in Table

11.

IV.

MAIN

TECHNICAL CONSIDERATIONS

In order to discuss the various ways in which the different

systems have chosen to cope with the particular technical

problems, we

must

examine, at least briefly, the technical and

operational problems of satellite communications. These

problems cover virtually the entire spectrum

of

physics and

engineering. It is our intent here to look particularly at those

that are special and even unique to satellite communication.

Clearly, a whole range of problems results from the necessity

to insert the satellite into the desired orbit, normally geo-

stationary, to control it remotely, and to maintain it in good

operating condition. Even if satellite communications were

to

consist of passive reflectors, the later problems alone would be

obviously nor negbble. The wide bandwidths now easily

available in the use of satellites in order to be exploited

properly lead to another group

of

problems; many

of

which,

of

course, are familiar in connection with terrestrial microwave

relay.

Most important of all, we have that category of problem

unique to satellite communication that arises from the neces-

sity and desirability

of

exploiting the geometric availability of

a geostationary satellite to any point over almost a third of the

earth’s surface. Before this convenience can be realized, it

is

necessary to choose a system

of

multiple access. In a very real

sense, we can call this

the

problem of satellite communications.

The problems in the three categories are not independent of

each other. We

will

discuss, at least briefly, the most impor-

tant interrelations among them.

A.

The

Communication Link

The exploitation of the wide bandwidth and the multiple

access problem, and the related question of choosing appro-

priate modulation systems, are

all

best examined by starting

with the elementary equation for the performance

of

a

communication

link.

The carrier power received at the earth station is given by

where

PT

satellite transmitter power

GT

transmitter antenna

gain

(over isotropic)

AR

earth station antenna effective area

=

CRh2 /4n

h

wavelength

R

distance from satellite to earth station

Li

incidental loss.

The noise density

NO

is equal to

kT,

where

k

is

Boltzmann’s

constant and

T,

is

the equivalent system temperatures defined

so

as to include antenna noise and thermal noise generated at

the receiver. Let

us

further define effective isotropic radiated

power (EIRP) as equal to

PT

GT

and a dimensionless spaceloss

equal to

(4nRIh)’.

It

is

routine

to

show

C EIRP

GR

1

No L,Li

T,

k

-

=-

--

dBHz.

Communications engineers usually make these calculations

in dB by taking 10 times the loglo of both sides of

(2).

Care

should be taken since some terms are dimensionless and some

others are-not. (e.g.,

(C/No)

is in dBHz,

Li

will be in dB, and

EIRP in dBw.)

Even more interestingly, the transmitter antenna gain

GT

is

inversely proportional to the solid coverage angles which in

turn depends on the terrestrial area to be covered

A,,,

Equation

(3)

has been written

so

as to hghlight the most

basic aspects of space communication. On the left side of the

equation are the desired parameters

C/No

and the area to be

covered on the earth. On the right side, by using the reference

simple relations, we have succeeded in eliminating all the terms

other than the transmitted power in the satellite, the effective

physical size of the receiving antenna, the dissipative losses,

and the system temperature.

The camer-to-noise density

(C/No)

can

be

written as

(C/N)B

where

B

is the noise bandwidth and

(C/N)

is the desired

carrier-to-noise ratio.

(C/N)

typically has a threshold value,

the achievement of which will permit substantial improve-

ments in demodulated signal-to-noise ratio. It

will

also be a

function of bit-error rate in digital systems and in general,

depends on the modulation used. It has values between

8

and

13

dB for most space communication systems. We now can

writ e

Equations

(3)

and

(4)

are both fundamental to appreciating

the essential problems of space communications. Although

they have been written for a down-link, the same equations

apply to the up-link from earth station to satellite if appropri-

ate parameters are substituted.

Either one or both links may determine the overall perfor-

mance. It

is

usually assumed that the noise from all sources

(including intermodulation to be discussed later) is additive

and it

is

then easy to show that

(KT

=

($1;

+

(y

+

(y

(5)

No

T

No

D

No

I

where the subscripts

U,

D,

I,

and

T

apply to the carrier-to-noise

density ratios as calculated for the up-link, down-link, equiv-

alent intermodulation noise, and total, respectively.

Note that in both

(3)

and

(4)

certain terms do

not

appear,

e.g., satellite altitude, carrier frequency, and earth station

(GIT).

The carrier frequency has an important secondader effect

since clearly the available transmitter power, receiver temper-

atures, etc., does depend on it’s state of the art at this fre-

quency, but the dependence is not a function of the com-

munication performance equation. The commonly used

figures of merit EIRP and

G/T

are not basic and should be

used with caution. EIRP has an implied coverage and

G/T

an

implied carrier frequency and when used as figures of merit,

this must be remembered.

The implications

of

the term

C/No

are fundamental and

account for its frequent use in communications systems

engineering. If one examines Shannon’s expression for the

information capacity

of

a channel and allows the bandwidth

298

PROCEEDINGS OF THE IEEE, MARCH 1977

TABLE

I1

CHARACTERISTICS

SATELLITE

WLSS

-

__

-

SYSTM

LAUNCH

VEHICLE STABILIZATION MULTIPLE ACCFSS

MODtiLATION

AND

EARTH

STATION

ANTENNA

DIAUEI'ER

CAPACITY PER SATELLITE

1.

IWTELSAT

IV

IKTEISAT

w-a

700

KG.

after

apogee

rotor

firing

apogee

=tar

790

KG.

after

firing

S

ATLW-CEWTAUUR FHPideo

FDU/Fn/FDEvL

29.5

M.

for Std.

A

13

M.

for Std.

B

29.5

n.

7500 channels

+

SPADE

+T,

12,000 channels

+

SPADE+

W

I I

~~ ~

(elliptical orbit)

A-2-e and SL-12

Fn

12

n.

25

n.

1

TJ

channel

+

unspecified telephones

Single

curler

Fn

Uulti carrier FM

Single channel/

carrier.

Delta

Mu-

latlon PSK

TDUA

Heavy

bute

-

301.

Northern

10.1

1.

Network

IV

-

10.1

M.

Tele-

oxmmicationr

Resate

n

8.U./

Thin

mute

S.lU./

4.714.

4.711.

12 transponders of

36

Wz

bandwidth

3.

TELESAT

-

Canada

272.2's. DELTA 1914

I

I

SSB/FM/FDU

Single

c

MIltiple

Access

Vidco.

SCPC:

PSK/TDCI/TOW

24 Video channels w/34

9000

channeldtransponder

WP

bandwldth

-

FLMA:FDX/Fn

6

PQVPSK

for voice data, 4 PSK

for digltal data,

m

color

n

for monochrome

or

data

-

TDHA:

POI/PSK for

voice/

Ion.

1-/

7.

COPISAT/ATT-

(CCMSTAR

)

28,ROO

one-way

tek-

ikony

ch-cls

or

'1,000

data

750

K;.

ATLAS--ZMTAtiR

1.

ESA IOTS/ECS)

I

324

KG.

I

DELTA 3914

1-120

HHz

Transponder

1-40

MHz

Transponder

1-5

MHz

hasponder

FH

Vldeo

4-Phase PSK Eulobem

A

131.

TD(u

Eurobea

B

31.

+

Spot

Beam

I

LO.

1mIA

300

KG.

DELTA 2914

FDWFU

Hultiple

carriers

9.m.

per

transponder 7.311.

I I

Voice:

SCPC-Fn

1.221., =bile term-

lnals

PSK

inals

Data: 2-Phsse coherent 12.81.. shore term-

-

16.

XIUSAT GEWEIW

326.6

KG.

on

S

I~ISlrT)

DELTA 2914

station

t

5-80

kHz

c-.

ch.

for

ground-to-air

and

surveillance

15-40

kHz

corm.

ch. for

air-to-ground

2-80

kHz

C-.

ch. for

ground-to-ground

1-400

kHz

or

10

MHz

experimental channels

~~

.7. ESYtOnsAT GENEIUL

470

KG.

IMROSAT)

B

1

DELTA 3914

.8.

ESA

1-S)

DELTA 3914

466

KG.

B

Voice:

NBFM, PDU and/or

VSDM

La

gain airborne

antennas

-

Data:

PSWFSK

FLM

m

F@nA

TDW

Atnut

lu.

Shore-to-ship: tip to

50

voicehigh-speed data

Ship-to-shore: Up to

60

channels

volcehigh-speed

data

channels

Shore-to-shore: tip to

3

voicehigh-speed

data channels

to

approach infinity as a limit,

it

turns out that the inforrna-

tion that can be transmitted in a channel is proportional

to

If the

RF

power budget,

as

determined from the above

questions, yields a particular

C/No

and if a particular band-

width

B

is

available from a frequency allocation point

of

view,

then

C/(Vo

is

quite

simply divided by the bandwidth to

determine the carrier-to-noise ratio

(C/N)

available for detect-

ability with the particular modulation system. For a fixed

(C/N)

and modulation system,

B

is

proportional to the number

C/NO

f

of

channels.

Thus

the number

of

channels

in

a given coverage

can depend only on the transmitter power, receiver antenna

size, and system temperature as before. Multiple-beam

antennas reduce the coverage and hence increase the number

of

channels for a constant

total

power. If the power

is

divided

among the beams proportionally, then

B

remains constant for

each beam and the

total

bandwidth available 'per satellite

increases by the number

of

beams.

Increasing capacity by increasing

B

through frequency reuse

so

as

to avoid the limitations

of

(4)

costs considerably in

PRITCHARD: WORLD'S SATELLITE COMMUNICATION

TABLE

I1

(con?)

299

22.

SYMPHONIE

'

230+5KG.

B

I

DELTA

2914

!

i

1

23.

as-6

1356

KG.

B

TITAN 111-C

'I

!I.

8M.

M.

1200

one-way

telephaly

~

2

color

TV

channels

or

I

,.9H.

I.

IVariousI

12

Video channels

at

2.6

GHz

11

Video channel at UHF

(860

MHz

I

C-band transponder has

40

NHz

I

bandwidth

11.5

GHz

transponder has

or

12

MHz

bandwidth

I

-

0.911.

1

lV

Channel

I

-

2.43M.

-

3.0%.

11

sound broadcast

channel

-

9.14M.

I10

duplex

vmce

channels

:X,

Video

!

I

26.

CTS

350

KG.

B

I

DELTA

2914

27.

SIR10

188

KG.

THOR-DELTA

t.

,.

.M

Video

>M

Sound

LO

ch.

(FDMI

of

broadcast

FM

duplex

YOLC~

t

CY-PSK,

2-phase

far

voice

Inarrow

band

'H

or

Dmital

for

lv

comunlcations)

cations)

(videband cmuni-

12

-

100

kHz

telephone channel:

total

1.5

MHz

band-

width

or

1

-

4

MHz

baseband for

TJ

I

500

KG.

S

1

TITAN 111-C

jtaqe

la

Of

proqram:

FDMA

6

cma

<t*'.b

of

program:

FDKA

6

CEHA

stage

IC

Of program:

FDNA

&

CDKA,

phasing

into

TDWPM

rn

itage

2

Of program:

temnals (Fixed1

or

transportable

Total

of

410

MHz

of

232

KG.

S

1

DELTA

2313

(SKYNET

2A

6

281

I

-

'.EM.

12.2M.

11

-

2

MHz

channel

11

;l

-

20

MHz

channel

j24

12400

bps1 data channelS

Iff

4M.

6.4M.

/280

mice

channels

IV

or

1

t;

340

KG.

S

j

DELTA

2914

30. NArO

+

!

'lo

UHF

domlin*

9

UHF and

1

SHE

upllnL

Each

UHF

has

25

kHz

bandwidth

+

IPSK

dovnllnk

2PSK

conferenclnq

link

3-ary

FSK

forward

upllnk

3-ary

MFSK, happed at

200;5ec.

2M.

for

ABNCP

temlnal

IL~~~~~~

jms,

DPSK, to other

!36-38

GHz:

Labs1

!

04.

for airborne

I

20

kb/s,

DPSK.

to

ABNCP

!

LES

satellites

450

KG.

TITAN 111-C

32.

LE5

SERIES

II

,

M.

for

Navy

t&m-

tenn.31,

AN/asc-22

inal

8

ary

FSK

forward

upllnk;

QPSK

conferenclnq

lplink

50

K-ary

symbolshec

from

ABNCf

terminal

75

b/s to

Navy

(s5ipboardl

terminal

ternma1

spacecraft antenna size and complexity. A channel may be

available only in a particular beam and must be switched

if

it

is

desired in another. The switching can be done slowly,

mechanically, as

is

now done in INTELSAT IV and

will

be

done in INTELSAT V. Increased capacity is obtained at the

expense

of

increased transponder complexity-there

will

be

hundreds of switches in INTELSAT V-and a

loss

of flexibility

that is acceptable because

of

the traffic patterns for a world-

wide fixed system. The switching can also be electronic and

rapid, such as would

be

the case in a timedivision switched

satellite in which multiple access was achieved by burst trans-

missions from each earth station that would be switched in the

satellite to the appropriate beam.

The multiple-beam configuration

is

suitable to high-traffic

fixed systems, such as INTELSAT and large-area regional

systems. It is less appropriate

in

domestic systems, such as

Canada, Indonesia, and the United States.

When very high carrier frequencies are used because

of

increased crowding in the more desirable bands, the dissipative

and scattering losses and lower available powers will probably

300

PROCEEDINGS

OF

THE IEEE, MARCH

1977

force multiple-tteam operation even where

it

might otherwise

be undesirable.

Broadcast satellites, especially those for areas covering

several time zones,

will

use

them since the same programs do

not necessarily have

to

go simultaneously to different areas.

The European Broadcasting Union plans, when brought to

fruition, envision multiple beams for Europe.

Satellite connections require up-links and down-links.

Normally, since

it

is relatively easy

to

supply high transmitter

powers and antenna gains at earth stations, the performance

is

determined by the down-link. The limitations

of

satellite

power and the necessity for covering the appropriate terrestrial

area

limit

the overall performance. The problem

is

compli-

cated by the necessity for multiple access and the resulting

possibilities for intermodulation noise in nonlinear tran-

sponders. In the case

of

small terminals, such as in mobile and

data gathering systems, the limitation

is

often in the up-link.

As

we see from

(3),

the ultimate limit in the down-link

performance

is

the transmitted power in the satellite, and this

limitation cannot be avoided for an assigned bandwidth and

the requirement for terrestrial coverage.

Satellite power

is

directly translatable into weight in the

spacecraft and, even more pointedly, into cost. Although the

carrier frequency

is

not a first-order problem in satellite

system planning,

it

has many very important second-order

effects. External sources

of

noise, such

as

the galaxy and

propagation through the ionosphere and atmosphere, are

generally frequency dependent.

B.

Multiple Access

To exploit the unique geometric properties

of

wide-area

visibility and multiple connectivity that go with satellites, the

various communications links using it must be separated from

each other. This can be accomplished in several ways.

I)

Space-Division Multiple Access (SDMA):

One

is

to use

different antenna beams and separate amplifiers within the

satellite. This

is

SDMA. Flexibility

is

only possible at the

expense of complications within the satellite, increased weight,

and occasional operational difficulties.

2)

Frequency-Division Multiple Access (FDMA):

A second

basic way, and the one in most common use,

is

that

of

using

different carrier frequencies for each transmitting station.

This

is

FDMA and permits many stations to use the same

transponder amplifier until finally the overall noise level limits

the capacity

of

that amplifler. Multiple carriers in any non-

linear amplifier produce intermodulation products which raises

the apparent noise level. This requires a "back-off"

of

drive

on the amplifier in order to reduce this intermodulation noise.

The carrier level received

is

less and thus the effect

of

thermal

noise generated in the earth station receiver

is

increased. This

reduction in drive must thus be optimized. Even optimized,

the effect

is

not trivial and the reduction in capability

of

a

transponder over that it would have if all the available infor-

mation was multiplexed on a single carrier frequency can be

as much as

6

dB. Nevertheless, FDMA is the most popular

technique for commercial communication satellites. It

is

efficient

if

one is not power limited, and it

is

the natural

expansion of terrestrial communication methods.

FDMA can be implemented

in

two ways. One

is

to

multi-

plex,

in

the conventional terrestrial manner, many channels

on each carrier that

is

transmitted through the satellite.

Another

is

to use a separate carrier frequency for each tele-

phone or baseband channel within the satellite. If many

carriers are used, the intermodulation problem

is

still more

serious. On the other hand,

it

does approach, asymptotically,

a limiting level that

is

usually acceptable. This single-channel-

per-carrier approach has particular advantages

.in

systems

where there are many

links

to

be made, each one having only a

few circuits

to

be handled at any one time. Normal multi-

plexing

is

very convenient terrestrially but may be economical

only

if

each carrier has traffic, for example, in a group

of

12

channels or more.

Both systems are in extensive

use

today. INTELSAT

uses

both systems, the SPADE system being a singlechannel-per-

carrier multiple-access system. Canada, Indonesia, Algeria,

to

mention a few, use

singlechannel-per-carrier

systems. The

modulation for

singlechannel-per-carrier

systems

is

a separate

decision and in

use

today, we have PCM, Delta modulation,

and narrow-band

FM.

The arguments as

to

which is best are

rather complicated and are discussed elsewhere

in

this issue.

3)

Time-Division Multiple Access (TDMA):

The next basic

method

is

TDMA,

in

w-

'

:h each earth station

is

assigned a

time

slot

for its transmismon, and

all

the earth stations

use

the

same carrier frequency within a particular transponder. In

terms

of

total satellite performance, this

is

the superior

method because the intermodulation noise

is

eliminated and

there

is

an increase

in

capacity. The required back-off

is

much

less,

just

that required

to

achieve acceptable spectrum

spreading.

The

price paid is a considerable increase in com-

plexity

of

the ground equipment. It does seem as

if

the long-

term trend

will

be toward more and more TDMA since

it

fits

naturally with the digital communications systems that are

so

rapidly proliferating terrestrially, not only for data transmis-

sion but more and more for digitized voice.

Various experimental TDMA systems in the

6

Mbit/s

to 60

Mbit/s range have been built and tested by INTELSAT and

others. Their efficiency advantage over FDMA can be

illus-

trated by comparing the approximate channel capacities

of

an INTELSAT IV global beam transponder operating with

standard INTELSAT

30

m earth stations, using TDMA and

FDMA, respectively. Assuming

10

accesses, the typical

capacity using FM/FDMA

is

about

450

one-way voice channels

[

21.

With TDMA,

using

standard

64

kb/s voice frequency

PCM encoding, the capacity

of

the same transponder

is

approximately

900

channels. If Digital Speech Interpolation

(DSI)

is

used to process the PCM bit streams, the capacity is

further increased to about 1800 channels.

A TDMA system went into commercial operation on Telesat,

Canada's system, starting in May 1976. Numerous other

TDMA systems

are

planned for regional and domestic satellite

systems throughout the world.

This trend

to

digital systems both terrestrially and via satel-

lite is reinforced by the ease with which the TDMA methods

can

be combined with SDMA by switching transmission bursts

from one antenna beam

to

another depending on their ulti-

mate destination. This notion

of

time-division switching,

although not yet exploited

in

any satellite, seems inevitable for

the reasons stated in connection with the discussion

on

the

link equation. It

is

efficient in its exploitation

of

both the

satellite power and the frequency spectrum, and both these

resources are in short supply. The price paid

is

increasing com-

plexity. That seems less and less

of

a price considering the

awesome technology

of

large-scale integration and micro-

computers.

Timedivision switching will be a major factor

in

communi-

cation satellite technology.

A

satelliteswitched TDMA system

(SS-TDMA) using a microwave switch matrix

of

redundant

design shows an increase

of

over

30

percent in available

30

1

rApacit~ :over FDMA/TDMA [3] (separate frequency bands,

each carrying TDMA). The satellite-switched TDMA concept

uses a single

400

MHz channel, as distinguished from the

FDMAITDMA system, which

uses

5

channels

of

80

MHz. Its

keying rate

is

300 MBd/s, rather than

60

MBd/s. Four-phase

PSK

is

used, as with FDMA/TDMA. A total channel capacity

of 39

700

is

achieved by SS-TDMA, compared with 29

870

for

TDMADDMA. Note that this time-division switching must be

done in nanoseconds

so

as

to connect successive bursts

to

different spot beams. Diodes

of

the p-i-n type and similar

solid-state switches

will

be necessary and are under develop-

ment along with extensive ancillary logic circuitry.

4)

Code-Division Multiple Access (CDMA):

The final basic

method of multiple access

is

that of CDMA, called occasion-

ally “spread-spectrum multiple access.” In either case, the

idea

is

the same. The transmission from each earth station

is

combined with a pseudo-random code

so

as to cause the

transmission to occupy the entire bandwidth of the tran-

sponder. The station for whom the transmission

is

intended

has a duplicate

of

this pseudo-random code and by cross-

correlating techniques

can

extract it from the “noise level”

created by the simultaneous

use

of

many other stations.

It has considerable advantage in military systems because the

spread-spectrum technique must be used anyway to harden the

satellite receiver against possible jamming and the pseudo-

random sequences are necessary to provide cryptographic

security.

The use of such crypto and anti-jam systems provides

automatic multiple access. In a sense, it

is

free. The difficulty

is

that it

is

not nearly

so

efficient an exploitation

of

the

resources of power and frequency spectrum as is even the

FDMA system, not

to

mention TDMA. In addition, it requires

extra equipment at both ends

of

the link.

Nevertheless, it is used and will continue to

be

used for

military systems. The possibility of its limited

use

in commer-

cial systems may appear as satellite

users

become increasingly

concerned with the possibilities of both malicious interference

and unauthorized listening. Users

of

satellite systems for

commercial data transmission

of

the kind envisioned in

domestic US systems may well be the

first

to consider at least

the crypto secure aspects of these methods.

C. Multiplexing

Multiplexing is the process

of

combining a number

of

infor-

mation-bearing

signals

into a single transmission band. This is

either a terrestrial or satellite problem and is not to be con-

fused with the related multiple-access question. Theoretically

almost any sequence of terrestrial modulation-terrestrial

multiplexing, carrier modulation to the satellite, multiple-

access system-can be used. For instance, the standard

INTELSAT, Telesat, DSCS-1, and Molniya systems use

single-sideband AM and frequencydivision multiplexing on the

ground, FM to the satellite, and separate carrier frequencies

for each earth station. In abbreviation, this system

is

SSB/

FDM/FM/FDMA. The proposed TDMA referred to earlier

would be described as PCM/TDM/QPSK/TDMA. The SPADE

singlechannel-per-carrier system

is

written

as

PCM/QPSK/

FDMA. The most common terrestrial multiplex method in

use

is

frequencydivision multiple (FDM), which

is

used

throughout the world. Frequency-division systems include:

a) single-sideband suppressed carrier (SSC or

SSB);

b) single-sideband transmitted camer (SSTC);

c) double-sideband suppressed carrier (DSSC);

d) double-sideband transmitted carrier (DSTC).

Most terrestrial and space systems use SSB, although some

short-to-medium-haul systems

use

other techniques.

Timedivision multiplexing (TDM) is becoming

of

increasing

interest in satellite communications. Timedivision systems

can use many modulation systems, such

as

pulse-amplitude

modulation (PAM) and pulse-duration modulation (PDM). By

far the most important for satellite communication are pulse-

code modulation (PCM) and delta modulation (DM). Within

these headings there are variations, such as differential PCM

and variable-slope delta modulation. The tradeoffs are

complex.

Although FDM goes naturally with FDMA, and TDM with

TDMA, nevertheless hybrid systems are entirely conceivable

and

will

be used; e.g., a FDM-Master Group Codec (coder-

decoder) has recently been designed for use in the Telesat

TDMA system

[

41

.

A low-loss multiplexer for satellite earth terminals has been

developed to eliminate the broadband high-power transmitter

and thereby improve satellite earth station reliability and

efficiency

[5].

The 5925- to 6425-MHz frequency band is

divided into 12 contiguous channels, each 36 MHz wide. Each

channel

is

amplified with a separate air-cooled TWTA.

Channels can be added by using modular units consisting

of

two 3-dB quadrature filters. Time delay and amplitude

responses are connected with waveguide equalizers placed

before the TWT, thereby avoiding the equalizer loss in the

high-power TWT output.

These units are expected to find wide application in small,

unmanned earth terminals. Successful implementation of

multiplexer and equalization circuits has demonstrated the

practicality of the modular transmitter as an alternative to

single, large, high-power transmitters currently used in satellite

earth stations.

D. Demand Assignment (DA)

Earth stations having continuous traffic over a given number

of

channels

use

preassigned channels. However, many channel

requirements, as in any communications plant, are of a short-

term nature,

so

a channel and terminal equipment economy

technique known as demand assignment is used.

Increased space segment efficiency in a fully variable DA

network arises from the fact that all channels are pooled and

may be used by any station, according to its instantaneous

traffic load. This may be contrasted with a system

using

pre-

assignment in which all channels are dedicated, i.e., both ends

of

the channel are fixed. With this system, when traffic to a

particular destination

is

light, the utilization

is

poor. Also, for

a given system traffic load, the blocking probability for a

system employing preassignment

is

higher than for a system

employing DA. This occurs because some number of channels

are “locked

in”

to a particular link. In a system employing

DA, unused channels may be made available to other users.

Conversely, for a given blocking probability, the number of

channels required

to

pass a given amount of traffic in a pre-

assigned system is greater than in a DA system. The hghter the

traffic per destination, the greater the advantage of the DA

system.

DA offers two main advantages when compared to pre-

assigned systems: 1) more efficient utilization

of

the space

segment;

2)

more efficient utilization of terrestrial intercon-

nect facilities. Corollary advantages are more direct service

302

(the need for “via” or transit routing is eliminated) and a mitted. By making the prediction threshold adaptive, the

consequent possible slight increase

in

communications quality system

can

be made able to respond to rapid changes in traffic

on such

links

because of the elimination of tandem loading.

connections. In application, both SPEC and digital TASI (the digital

There

are

many possible forms which a DA communications counterpart of the well-known Bell system for intersyllabic

system may take, and there are various ways of categorizing channel sharing developed for trans-Atlantic cable

use)

exhibit

their makeup. If, for example, both ends of all channels in the approximately the same advantage but SPEC is superior in its

system are undedicated

so

that any station may

use

any freedom from speech clipping, simple algorithms, lower com-

channel, then the system

is

termed “fully variable.” On the plexity, and lower cost.

other hand,

if

blocks of channels are reserved to an originating Using this approach, a 2:l reduction may

be

achieved in the

station or a destination station, but are

still

used only on bit rate required to transmit a group of voice channels. It

is

demand, then

it

is

a “semi variable” system. important to note that such economies may be achieved only

Another way of categorizing DA systems

is

as follows: when when at least

30

voice channels are processed as a group. One

carriers

(in

FDMA systems) or bursts (in TDMA systems) are

of

the advantages of SPEC

is

its graceful and sut degradation

assigned on demand, the approach may be termed DA multiple in the face of an overload condition.

access (DAMA); when channels on existing carriers (FDMA) or

E.

&,-Board processing

time slots in existing bursts (TDMA) are assigned on demand,

Increases

in

circuit

reliability

that

have

accompanied

then the approach may

be

referred to

as

baseband DA (BDA).

advances

in

solid-state

allow

significant

increases

in

Various

Of

these

approaches may

be

created,

the complexity

of

satellite onboard

circuitry;

con~quently,

the choice

On

traffic

and

On

user

designers now can give serious consideration to advantages of

requirements. For example,

if

a network has a multiplicity of

on-board

processing

previously

considered

too

unreliable.

users but only a few large earth stations, then a BDA approach

Although

on-board

processing

is

being

done

effectively

in

may

be

most

If

there

are

a

great many

earth

the

me

of arth-resourcs and weather involved

in

in

a

each

with

low traffic

a

kge+cale data gathering missions, it has not been attempted

variable DAMA approach would seem

best.

In an application

with

commercial communications

satellites

since

their

chief

where priority control of access

is

vital, the greater restriction

purpose

has

been

the

provision

of

a

link

between

points

on

the

enable a certain number

of

channels incoming to each station

television.

or each of several stations to be reserved for priority traffic.

It

is

also

quite possible and perhaps desirable to

mix

ap-

In military satellites, up-link

signals

accompanied by

proaches within the same system. The choice depends solely

unwanted interference can be converted to baseband, pro-

on the user’s requirements.

cessed for interference removal, and then remodulated on a

down-link carrier. This at least prevents the repeating of the

an example, are characterized by a per-user access technique.

In the SPADE system, for example, each user accesses the

signal in the transponder. The increased noise level must be

satellite with a singlechannel-per-carrier (SCPC). Other

dealt with by other methods.

possible approaches to DAMA are singlechannel-per-bust

sent

using

TDMA

will

be

sorted

on-board

the

Experiments are now being designed in which packets of

(SCPB) TDMA and CDMA. With these approaches, the carrier

(FDMA or CDMA) or burst (TDMA)

is

not transmitted until

This

is

the

previously

discussed

the-division switching.

On-

satellite and transmitted via one of the several spot beams.

it

is

required. The amval at the ground station via a terrestrial

line of a call request results in the establishment of the carrier

board DA, the “switchboard in the sky,” also

is

expected to

or the burst for the duration of the

call.

As stated earlier,

since

satellites

historically

have

been

power

limited,

while

become important toward the end of the coming decade.

these techniques are attractive in systems in which there are a

earth

stations

tend

to

be

bandwidth

limited,

it

is

conceivable

large number

of

users

whose individual traffic requirements are

small.

that a future system may

be

designed

in

which the uplink

The INTELSAT network uses a common TDM signalling

channels are transmitted using single sideband (amplitude

channel on which all stations

call

each other and keep track of

modulation) to minimize bandwidth, while the down-link

the available channels which

are

seized in

turn

on a firstame-

required

from

the

satellite.

Other

combinations

of

up-link

channels are transmitted using PCM/FM to minimize power

first-served basis. This avoids the politically awkward problem

of central controlled systems and the choice of country in

and down-link modulation and multiple-access systems can

which to locate such control-Canada and Algeria, for domes-

arrangements

would

also

be

accomplished

with

demodulation

also

be

conceived for various optimization plans. Such

tic service only, use central common control and avoid the ex-

pense of increased equipment complexity necessary to avoid it.

and remodulation on board the satellite.

(DSI) or speech-predictive encoding (SPEC). In the SPEC

With the launch of CTS, Sirio, and other satellites in 1976,

type system, each voice channel

is

monitored, and the present

the 14/12 and 18/12

GHz

bands

will

enter active

use

for

sample value for each channel

is

compared with the previous

satellite communication. The Japanese also plan

use

of these

sample valfie for the same channel.

If

the samples differ by an

frequencies

in

their BS and CS satellites. Still higher

fre-

amount which exceeds some threshold, (i.e., the present

quencies are under active consideration, not only because of

sample

is

not predictable from the previous sample) the

spectrum crowding in the lower frequency bands, but because

present sample

is

transmitted.

If

the difference

is

less than or

of the desire for broader bandwidths to accommodate higher

equal to the prediction threshold, the sample

is

not trans-

data rates than are now being sent commercially.

of a semi-variable system may be advantageous, since it would ground

for

the unaltered

of

voice,

data, and

DAMA

systems,

Of

which

the

INTELSAT

SPADE

system

is

interference and the nonlinear effect

of

a

BDA may be implemented

using

digital speech interpolation

F. Higher Frequencies