-A - -------- ---­

-------

T%/ vI.
(A~~ N E(HU dS

(CODE

dt-ctc9

(NSACROkTM
17Al
R D UM&
(CTGRY II
i
NAIOA TC
INFOMATOUSE
-C
 NOTICE

This document has been reproduced from the best copy

furnished us by the sponsoring agency. Although it is

rccognized that certain portions are illegible, it is

being released in the.interest of making available as

much hlflrcmoion as possible.

 CONTENTS

Page

rP.
FIODUCTION i.-i

PART I. BASELINE SYSTEM

ASELINE SYSTEM SUMMARY

I Introduction 2-1
General Configuration and Summary Characteristics 2-3
, System Operations 2-9
-- . ,RB Comrnunications System Interfaces -13
* System Compatibility 1-18
2.6 References 2-20

AN2ENNA SUBSYSTEM
1 Introduction 3-1
-. - K-Band Section 3-1
., S-Band Section 3-10
"VF Section 3-14

A '-IF SUBSYSTEM
.~ introduction 4-!
K-Band Section 4-1
3
.. S-Band Section 4-3
4.. VHF Section 4-3

iA"Z-;FCBAND SUBSYSTEM
* ; Introduction 5-i
K-Band Section
. 5-I
S-Band Section 5-9
VHF Section 5-11

PART I. PARAMETRIC ANALYSES

:-P-CATTONS
I' 13C-OM377 RY AND ANTENNA POIN TIN C

1 A r L ITIiti ioy
"-, s Isy t'n z xnd f]lcO],'].r Chit
a I-
'.U ­
i
 6. 3 -Mathematical Analysis of Space Station/DRS Geometry 6-6
6.4 Antenna Placement 6-Z!
6.5 DRS Handover Geometry 6-24
6. 6- 'Antenna Pointing Geometry 6-25
6.7 Analysis of Baseline Gimbal System 6-35

7. LINK ANALYSIS AND DESIGN TiADES

7. 1 introduction 7-1
7.2 Linear Link Design 7-I
7.3 Design of a Composite PSK Link Through a Hard-
Limiting Repeater 7-4
7.4 Design Trades 7-14
7. 5 References 7-15

8. ANTENNA ACQUISITION AND TRACKING STUDIES

8. 1 Introduction 8-1
8.2 Programmed Tracking Concept 8-3
8.3 Antenna Acquisition by Scanning 8-3
8.4 Theoretical Comparison of Autotrack Systems 8-Z7
8. 5 References 8-31

9. MODEM AND TIME DIVISION MULTIPLEX STUDIES

9. 1 Introduction 9-1
9.2 Demodulation of Convolutional Coded Biphase PSK
Signals 9-I
9.3 Design Study of the Space Station TDM System 9-14
9.4 References 9-23

10. EQUIPM\IENT SURVEYS

10. 1 Introduction 10-1
10. 2 TWTA's For Space Application (Z to 15 GHz) 10-1
10. 3 Microwave Preamplifiers 10-5
10.4 Viterbi Decoders and ZOP Ctmpressors 10-5
10. 5 References 10-12

APPENDICES
A. Link Power Budgets A-1
B. Electronically Scanned Arrays For Communication
During Artificial-G Operations B-1

iv.
 SPACESTATION

/ 'SFN

SPACESTATION
250 N.•i.ALTITUDE
55 INCL"ATION
ATTITUDE t.ODE
x-POP, SHOWN:
LOCALVERTICAL

Figure 1-t. Cornmunication Terminals

 1. INTRODUCTION

This document is the final report of the Parametric Analysis of RF

ionnimunication and Tracking Systems for Manned Space Stations, conducted
by the Space Sys'tems Division, Htighes Aircraft Company, for the Manned
Spa cocraft Center, National Aeronautics and Space Administration, under
Contract NAS 9-10409.

The 1-year (5 January 1970 to 5 January 1971) study was divided

in:n two major phases. Phase I provided analyses and assessment of Th:e
-:rnificant parameters and factors associated with IF communication sys­
tcms for manned space stations. Phase II used these results in the design
of ba scline e ternal communications system for a lZ--an Space Station.

Part I of this report desc-cibos the baseline system. The basic Space
Station m-ission profile is given in ra.hT rb 1-1. 'Wherever definitive require­
i.tcnts or constraints were require, during the design phase, the North
-inwrican Rockvell (NAR) Solar-Pocred Space Station was assumed. The
b;iric link geometry of the mission is illustrated in Figure 1-1. The low­
;-Ititude orbit Space Station utilizes a data relay satellite (DRS) for all normal
coimnunication to the ground. The stations of the Manned Space Flight Net­
work (MISFN) perform tracking of the spacecraft for orbit determination.
Commnunication links are established between the Space Station and the orbi-
t.-J, vehicle of the Advanced Logistic Systen (ALS) whenever the ALS approaches
cc-endezvous. Links are also provid&d to the Station's free-flying experi­
".ents nodule. Section Z provides a system level description of the all­
nutal system. The design outlinec" avoids the comrpatibilityproblems encoun­
rui in the Apollo Unified S-Band Fystein. The sybtem's antenna, RF-IF,
- 2 ba seband subsystems are detailed in Sections 3, 4, and 5, respectively.

Part ii of this report documaents the s.tuijey and np;arametric trades per­
-r'.er.r lany of the studies cai be ut ad in the t-veloprnent of conrnunica­
t+c,'~i~t s.em designs for application to rT-Jssiols other than the Space Station
.. ion (for exarnple, the ALS ni.nir.) but tle deta.iled design trades con­
only
.. the Spazce Station recuirk;neni en. trzimts. Suctio, C con
e..
t"""l geonmetcy of the various links, The pfn'ncohlcnls associatd
S.:. re of a ]I iJ'qr.in antlc:n')ini foi th.; d,-. z lieP
ink rt amn-)yzen1 in
cttiu 7 dcurib; thc li:Th .'lys.- p, .....tn.1 tht . ii- i .n

,ar', , ..... . ..
 TABLE 1-1. - MISSION PROFILE

o lZman Space Station

e NAR solar powered configuration

o Launch 1977

" Duration > 5 years

o Primary attitude mode: X-POP LV

o Orbital parameters: Z50 n. mi. altitude, 55 degree inclination

o Operational system interfaces:

MSFN

ALS

EM1

DRS

pointing technique. Section 9 includes analyses of a portion of the baseband

electronics for a digital communication system. The design of both a PSK
modem and a time division multiplex system are conside'red. Section 10
documents surveys made of the state-of-the-art in several different
technologies.
Appendix A contains the final link power budgets for the Space Station
communications system design. Appendix B summarizes a special task
performed as the request of MSC; namely, the use of electronically steered
arrays for communication during artificial gravity operations.

ktko
 PART I

BASELINE SYSTEM
 Z. BASELINE SYSTEM SUMMARY

- PFODUCTION
M

A brief outline of the baseline system design and the rationale behind
l-npent is helpful as an orientation to the description which follovis.
'1
, of the system. extends from digital and analog signals at baseband
, ziantenna and RIF transmission equip-nent. It provides color tele­
.-,)iCC, telemetry, and wideband digital data comm-nnunications services.
vtknm operates in three frequency bands (nominally 140 MIz, 2. Z GHz,

A data relay satellite system is employed as the prim-iary nnemil of

*-to-ground communications. The wideband service over this link uti-
t:,frUL. high gain tracking antennas at 14 0Hz. Onnidirectional coverage
1 c-prodd-ied at the other frequencies for voice and data links to
r'
, and the detached experiments module (EMl).

Except for the emergency voice se-vice, which is provided bya

- .t analog systen, digital modulation and associated source and chan­
J.tiing is the selected transmission mode for all the links and the
t 'lthey c.rry.

* 'in Design Objectives

I,',levtloping t-he baseline system, there wr-ere several obj ectives

crv'ed to guide the direct'ion and dcter-mine the scope of the design.
r,ne, which infclaenced several design deei& was
was,the intention
Sate the maxiarn comrnunication capabilitics --hat technology would
in the mission tirsj frame. This was to be done. not by extrapolating
Ali e]
device perforrnance levels to 7 years in th,. future, but by obsrr­
'"c,.bive growrh i different techra oloi; erd rranr'ing these in a sys­
which dered "Lnasimutn s""te t level prro-rr;ince. The det.ice
.cLulally I £'qwir"d aze- we- 7:llY &rtI c'-n.,I--:vativc a-,; aC'
1Ce
!ta of iOn. art. Stjci di 5LI 'tfl-2
o.'isU
- " ''ta relao link -, , ro CI,- (10 , O . .I .as
ni
lanultctc c's sOu* ,'t../dcAZ~ 'I['2Lf.
 . I.Ir MODEM- - -t

y>J I SELECT
r ~j
T-ANSELECT
__
CONFIG
._SWITCH L MODEM
sE c
TDM
SWI TC1­
F]
-sw'roH -B

Mm ,i [I]
! TRANSMIT/ MODEM
E_ +I"
J,, I RECEIVE
FROM IF
MSFN
,TRAACNGSYSTEM -13LS MODEM
INTERFACE TRANSMIT/

~7 CT' -. SLC EEC NEFC

L....SW"ITCI- RECEIVE
RECEIVE 1MODEM
SWITCH
." A ;
,LEMj MODEMJ zINERAC

1:.TRANSMITTER!/ F

iv7 LE
-IP,1
/-RECEIVER M0NI DI

.X - TRANSMITTER! CONFIG
SELECT INTERFACECOMMUNICATION
SWITCH '----OINTERNAL

C
_ELECT ,V RECEIVER
, u:=,.,._ _...

TRANSMITTER/ I COMMUNICATIONCONTROL

RCIE SME INTERFACE

Figur.e 2-1, System Block Diagram

 ,.")VI 1 ajor objective was to avoid the problcfl15 encountered in
.. e,±nunication system. The design philosophy provides a
re of independence between the functions of ranging, antenna
.. I., data transmission, and preliminary analysis indicates that
- :x i should prevent the incompatibilities found within the Apollo
*-Band System.

\r1cther objective of the baseline design was to determine technical

I or risk areas. For this reason, certain areas of the design were
.z !,td in more depth than others. For example, the time division
*-evr (TDIM) was investigated to the level of digital device block dia­
-n circuit board estimates. Similarly, achievable solid state power
i',r performance was carefully considered and the specified capabili­
-o based on devices and circuits currently employed in Hughes'

Two additional objectives in the system design were: 1) to assist

-mn planning by providing a representative set of performance capa­
versus Space Station resource demand, and 2) to provide a design
*t r requirements analysis and verification. In connection with the
tlhese, special emphasis was given to developing a conprehensive
mten of the system and operational interfaces.

.2 (:PNItAL CONFIGURATION AND SUMMARY CHARACTERISTICS

.Phe basic system concept is shown in Figure Z-1, where the sub­
are organized by equipment location and function, with the antenna
v.ten on the left, the RF-IF subsystem in the middle, and the baseband
tern to the right. Each of these, in turn, is divided into sections which
ivipi.,nd to the operating frequency band and communications circuits
in that band, The top (K-band) section handles all dhe conimunica­
taffic between the Station and the ground with the exception of the
.-- ncy voice link. The S-band section carries the telemetry and wide­
.±f'mnunications between the Space Station and the shuttle (ADS) and
!.ched experiments module (EM 1 ). The bottom- portion of Fig­
-1 9hows the dedicated, emergency voice facility.which provides com­
4ion: 1) between the Station and the ALS, and 2) to the ground, either
:!y or through the data relay satellite (DRS).

'in use of redundancy is indicated by the parallel equipment in each

the required level of redundancy is a, yet undetermined, but
t'I ,,nits are included here to assure that weight and power estimates
"...be. In general, a "hot" spare is provided where feasible which
* . ':ated via the configuration selection switcbes shown. This spare
in addition to the parallel units required during antenna hand­
*, tntennas dr not have t-i level or rrhii,,,a.cy bec:0use of their
. ".ight requirenients, altbougb t]here are bac]hu., provision" in

2-3,
 Three K-band antennas are provided "Which are the minimun required
to handle uninterrupted relay satellite handover with the spacecraft stabi­
lized in a standard mission aLtitude. In the event of failure of one (and under
some circumstances, two) of these antennas: the third could still provide the
full link capacity through the relay satellite. There would be an interruption
in service however, as the single antenna was repositioned for acquisition of
a second relay satellite.

The S-band and VHF antennas each consist of two independent arrays
(one circumferential and one axial) with their patterns orientdd for maximum
coverage in a given Station attitude mode. As these are not tracking anten­
nas, their reliability is quite high, If a failure should occur in the primary
pattern, the alternate array can be used. This generally would not provide
full capability (either in range or continuous visibility) but would give
,Minimum, essential service.

In the S-band section of the antenna subsystem, a tracking system

interface is shown. This is a pair of isolated transmit and receive ports
which hive full access to the S-band antenna for ground tracking or comr­
mand functions. The receive port follows a wideband preamplifier so that
the tracking system noise temperature would not be penalized by attenu htion
in waveguide runs, filters, and couplers associated with the communications
channels.

On the transmit side, it will be necessary to coordinate transmit

filter bandwidths and power levels so that appropriate isolation can be pro­
vided to prevent interference due to reflected signals in the common antenna
transmis sion line. I

in the K-band section of the baseband subsystem, parallel PCM and

single TV and voice sources are shown. This is because the TV and voice
signal processing is included within the communication system, whereas,
the PGM is provided from a system external to the communication system.
To provide continuous flow of data as rhay be required by the PCM users
during relay satellite handover, a parallel interface is provided. For the
TV and voice, the input and output circuits would be switched between satel­
lite circuits at the multiplexer (TDM) interface when the new link is estab­
lished. This woild cause a transient interruption of the analog signal, but
probably would not be noticed by the users.

The S-band section of the baseband subsystem contains separate

modulator/demodulators for the ALS and EM 1 circuits. This is required
because of the decision to employ quadriphase modulation for bandwidth
conservation in the EM 1 link. Rather than attempt to define a dual mode
moden which could be configured for either quadriphase or biphase, sepa­
rate modems were specified as these are relatively small, low power units.

The configuration selection switches, as already noted, are used to

switch in relundant units in the event of a failure. in addition, they are used
to connect the appropriate sntenna, RF, and baseband units together as
required by Sp!tce Station corninnications link geometries and service require­
nients. Thesc scitchcs are controlled by a coumrnunnicsions control ielity,

Z-'!
 . a1 to the system. It is assumed that this facility vil!
, 411 internal and external communication functions and preplanned
1! and handover sequences.
.iALocation s

£h.e mounting locations of the baseline conrnunictinns equipment

in Figure 2-Z. These locations are based on a North American
-11.l -ar-Powered Space Station configuration. (The solar cell panels
for clarity. ) The three K-band, high gain antennas are
i .m)nitted
....
, ,ut tie circumference at 120 degree intervals. Two are mounted
* dctZe of the conical section and the third at the rear. Each has a
~ t.r'dle and erection mechanism. The two near the conical section
7,daially against the cone surface while the third is folded across the
irface.

The S-band and VHF circumferential arrays form bands around the
, 2.ic31 section in the middle region. The exact location is not critical,
that the VHF array should be 10 feet or more from the ends of the
,--n so that a symmnetrical pattern is achieved. There is a single
")WI VIF element at each extremity of the longitudinal axis. One
Y wn in Figure 2-2 on the narrow end where the solar panels are
i'1; there are two more on the flat surface at the end of the cylindri-

At the base of each K-band antenna, there is an equipment section

1.)"es the RF pover amplifier, its power supply, and the servo
i,',,iics. These are unpressurized spaces with prime power and ther­
provisions. A compartment at the rear of the antenna reflector
lit.' tme preamplifier and the monopulse circuitry. Another equipment
It'i ", the external equipment bay, is located at a convenient point on the
.- :(:ia section. This contains the majority of the RF-IF equipment
siinilar power and thermal sink facilities. The bay is accessible
a"thatch in the surface for modular replacement of units.

Th,±re are two internal equipment bays; one in each of two pressur-
,Ctions of the Space Station. This is an agreement Vith the Station
philosophy which provides for two isolated areas for survival in an
hvey which threatens the habitable environment.

" ;e antenna locations were selected primarily by coverage require­

blockage problems, and by launch stowage considerations. The
the Llectronic equipment was governed by the desire to locate as
,.,f
t possible in a single external bay to mininize the following:
t'i?2tion management difficulties, Z) signal and power cable runs,
'w:,and thermF.. load demands on the xnanned C 3V5VOn
aent. It vns
i, locate the K-bzad pover anrplifiers; near their associetd anion­
to oid the e: c.ssive x-,avr niide lo- sos at tb-s frequ,-cy. Ibe
: 'er,: then inclgded .c'[ this Ioc'-'iol to 5inplify >Jr--r *.& a­
f" -.
a und 7
chcc.uu, -rol,0-n
 K-RAND ANTENNA

EXTERNAL
EQUIPMENT S
BAY

S-BAND AXIAL ELEMENT 'b

Y k INTERNAL

S-BAND CIRCUMFERENTIAL ARRAY

VHF AXIAL ELEMENT VHF CIRCUMFERENTIAL ARRAY

K-BAND EQUIPMENT
SHELF

Figure 2-2. Equipment Locations

z-6
 V
A
Wciglii, 11owe
'r, oItO1
5'tubys tenm Section pounds wttts cubic feet t.octin

ANTENNA K-band 299. 0 245.0 0.75 A, B, b C

S-band 423.0 18.0 0. Z4 A & C
VI-IF 94.0 0.3 0. 03 A &C

Total 816. 0 Z63.03 .0

RF-IF K-band 30, 0 20.0 1.28 C
S-band 96. 0 476.0 1.95 C
VHF 27.0 166.0 1.40 C

Total 153.0 496.03 4.60

BASEBAND K-band 410.0 Z17.0 6,40 D &E

S-band 102.0 59.0 2. 66 D &E
VHF 2.0 - 0,23 D &E

Total 514, 0 276.03 9.3

Total external
communications system 1,483. 0 1, 035.33 14.9

Electronic equipment only

Location code:
A = Antenna
B = K-band equipment shelf
C = External bay
D = Internal bay No. I
E = hIternal bay No. 2
3One K-band and two S-band links active
 The division between the RF-IF and baseband subsystems (locaited
t e--trna! and internal equipment bays, respectively) was determined
ipally by the objective of minimizing the cabling requirements. There
large number of command and monitoring functions associated with the
* in~nent grouped in the baseband subsystem which would lead to a severe
.... .]ty in vire harness weight if it was necessary to carry these to the
*'-N,?rnal bay from the internally located commnunication control facility.

-v:tenm Physical Characteristics

Table 2-1 summarizes the physical characteristics of the baseline

* wvtem. The weights indicated include all the equipment shown in the pre­
.q;us system block design, including redundant units. The power demand
if the average load with links established to a single DRS, ALS, and EM I .
"he volumes shown are for the electronic components only and do not include
the antenna apertures or gimbals.

Communications Performance

The communication capabilities of the baseline system are summarized

.iiTable 2-2. The channel capacities are'specified in terms of their digital
bit rates. The 12. 2880 'negabits/sec and 19. 2 kilobits/sec rates handle.the
color TV and engineering voice requirements, respectively, using appropri­
ate source encoding.

As stated above, the DRS system is employed as the primary means

of space to ground communications, although the emergency voice and S-band
links have more than adequate capability to work directly into a modest
ground terminal. For the EM I and the ALS, the achievable communications
range is given as an important communications performance parameter.

2. 3 SYSTEM OPERATIONS

The design of the baseline system assuimes autonomy of the Space

Station in establishing its half of the various communications links. In
addition, it is assumed that the Station ihitiates all acquisition sequences
according to a preplanned schedule. Appropriate monitoring and configura­
tl,n command test points, redundant units, and signal test patterns are pro­
'.'ided for circuit and equipment performance monitoring, fault isolation, and
r,toration.

These activities are under the control of a separate communications

z 'Tntrol facility. This facility coordinates all frequency and traffic assign­
Il'trts With its counterpart either on the ground or aboard the ALS or EM1
_iod establishes appropriate circuit interconnections with the internal con­
:V1 nIcation system. These functions are assumed to be automated with no
- ird Operator supervision during normal operations, but with certain
- n'l Override provisions.

Z-8
 TABLE1 Z-Z. SUMMARY COMMUNI CIATI(NS ,OuMA],(:

Channel Capacity* Range

Link Frequency Service (full duplex) Number n. mi.

DRS K-band Television 1Z. 2880 Megabits/sec 1 ­

Engineering voice 19. 2 Kilobits/sec 1z ­
Telemetry 307. 2 Kilobits/sec 7 ­
Priority telemetry 19. 2 iilobibs/sec 4

TDRS VHF Emergency voice 3 ldiz I ­

EM1 S-band Wideband data 7. 3728 Megabits/sec 1 450

ALS S-band Telemetry, voice 57.6 Kilobits/sec 1 303

ALS VHF Emergency voice 3 kl-z 1 3000

MSFN VHF Emergency voice 3 kI-z 1

kSFN S-band Ranging __MSFN standard ,

10 5
-'Digital channel BER =
 i
Table 2-3 summarizes the various communications system modes
.I.-de changes as a function of several -mission operations. For example,
it-. Station attitude changes,. the K-band link is interrupted during the
A nfuver and reestablished when the Station is stabilized in the new
nm.;,titon.

in the event of a mission emergency, power can be conserved and

-- ninin-al conmniunications capabilities sustained by powering down the two
,,deband links to the DRS and EM I . The emergency voice remains opera­
&,nai and the narrowband data link normally providing telemetry to the
" uan be used for a 57. 6 kilobits/sec link to the ground. The S-band
.,,cclli"g interface also remains operational.

A DRS link-handover is accomplished by establishing the new link

i: parallel with the old during the period of mutual visibility. (This assumes
t::±t the number and location of the relay satellites and the Space Station
.,rbit provide such overlap. ) There are three functions which must be per­
l rrred in sequence for each half of the duplex link: antenna acquisition,
. nal acquisition, and TDM synchronization.

The link is first established in the direction from the DRS to the
.ace Station. Prior to initiating the sequence, the data listed below must
be coordinated with the ground facility controlling the DRS. This informa­
1:11 may be transmitted to the Station via the DRS link at regular intervals.

1) DRS look angles in Space Station coordinates (The final data is

generated onboard the spacecraft. The fundamental calculation
is based on DIS and Space Station ephemerides.)

Z) Transmit and receive channel assignments

3) Beacon frequency doppler offset derived from item (1) data

4) Ground station TDM timing offset to compensate for path length

variation during the pass

The Station and DIS coordinate the pointing of their high antennas by
't- use of 1-band tracking beacons. The Space Station K-band antenna, which
tto carry the new link, is commanded to point at the DIS to within its beam­
:th (I. Z5 degree). The estimated doppler offset is inserted in the K-band
nt,-u local oscillator in the receiver and the beacon code tracking loop locks
PN-ceoded, DRS beacon signal. This beacon signal is transmitted over
Ie
.'arth coverage antenna of the DRS so it is not necessary to orient the high
-~n * ltnna on the relay satellite at this point. When code tracking synchro­
-*ti. acconplished, the Space Station antenna commences automatic
A the beacon signal.
ie i-n
n[UuSpace Station beacon is then transmitted to assist the DIS in
SiU high gain antenna. When this has been acconplished, a confir­
_.- sent over the DES b)eacon aind detected at the Space Stom.
S- v-t. tn ground and the Space Station that the conmmnunications link8
- k MabJysb]-fel.
2-I])
 TABLE Z-3. COMMUNICATION SYSTEM MODES

V ,,rrationls Response/ Configuration

.:,c change Programmed reconfiguration and re-acquisition

of antenna, signal,- and data

Emergency voice and MSFN telemetry operational

in emergency Powerdown EM 1 wideband data

Powerdown DRS wideband data

Emergency voice and MSFN telemetry operational

.andover Maintain communication on primary link

Programmed secondary antenna acquisition on
DRS beacon

Programmed signal and data sync acquisition On

communications channel

Accept data on new channel

A, : -ezvous/docking Programmed K-band antenna reconfiguration and

power control

Programmed S-band power control, automatic

antenna reconfiguration
?',re/maintain Select redundant baseband or RF-IF equipment

Select alternate antennas

Reduce capacity on all digital links

Manually assisted antenna pre-positioning

Replace major equipment m-odules

-i::cial g DRS coverage limited to two K-band antennas

Automatic switching of S-banC anrd VHF antennas

2-li
 At this point in the acquisition sequence,, the communications. signal
is transmitted by the ground station, relayed by the DRS, and received at
the Space Station. The design of the receiver is such that when the beacon
demodulator is tracking the beacon signal, there is automatic doppler com­
pensation provided in the comnunications channel. This communications
signal is therefore quickly acquired and processed by the communications
demodulator/decoder. The Space Station begins transnitting its cornmuni­
cationis carrier at about the same time as the ground station in order to
establish the other half of the full duplex link.

Initially, the signal carries no data, but has only the digital frame
structure with the sync word. The TDM equipment scans this sequence ahd
establishes frame and bit synchronization. As each half of the link achieves
TDM sync, it confirms this by transmitting an acknowledgement in the
reverse direction. This completes the link acquisition procedure and
traffic can now be transferred from the previous DRS link to the new.

During rendezvous and docking maneuvers, it is necessary to man­

age the antenna selection and transmitted power to provide continuous cover­
age and avoid interference with electronics systems on spacecraft maneuver­
ing in the near vicinity of the Station. S-band antenna selection is provided
automatically, while the K-band antenna selection is preplanned by considera­
tion of look angles and relative geometries.

in the event of circuit degradation or equipment failure, new circuits

or equipment can be configured by switching to alternative units or, for the
K-band link, reducing the capacity. Electronic modules may also be removed
for later repair or refurbishment. There are also provisions for tracking on
the DRS communications signal in the event of beacon failure and for manually
assisting the positioning of the high gain antennas for initial alignment and
acquisition.

During the artificial gravity experiment, the antenna look angle

considerations become complicated by virtue of the sun-oriented geometry
(See Figure 2-3). The two K-band antennas with collinear azimuth axes are
'despun' at the 4 rpm spin rate and used for DRS coverage. Depending on
thd number and longitude of the relay satellites which comprise the DRS
network, and the time of the year that the month-long experiment is conduc­
ted, there may be interruptions of this link.

While three high gain antennas at IZ0 degrees spacing can be located
to provide uninterrupted visibility of a given relay satellite on a single pass
for the normal attitude modes, this is not true for the artificial a mode. As
illustrated in Figure Z-3, the addition of the S-I stage as a counter-weight
limits any one antenna to hemispherical coverage on either side of the spin
plane. There are Station/DRS orbit configurations which cause the spin
plane to intersect the visible relay satellite, necessitating a handover during
a pass and the attendent interruption of service. This -could be avoided by
the addition of a fourth high gain antenna to provide tvo in each hemisphere,
but this was not considered justified for an experimental mode.

2-12
 ±b, "TIF and S-band antenn~.s also have unique handove considerations
•..-t'ucial g mode which impose some operational constraints involving,
Iv, Space Station-to-ground links. A link to a ground station at local
S .nkvould lie in the spin plane. The circumferential and axial array
then oriented such that switching from one array to the other would
- rel tk;wo tines per revolution. This could interfere v.ith both the
,-.c-vvoice link and a turnaround ranging link in the tracking system
Ly;. not properly designed. *- These are not insurmountable difficul­
;,u' should be considered in the design of this mission and the ground
t O;.Ipm~ent.

tP CO1MMUNICATIONS SYSTEM INTERFACES

The system interfaces, shown in Figure 2-4, have been separated

-xturnal and internal systems. The external systems are the relay
, (DRS), the shuttle (ALS), the detached experiments module (EIMI),
K Manned Space Flight Network (MSFN). The internal interfaces have
1run ped into baseband cormnmunication services, communication sys­
*;tt-ol, and ancillary systems.

Tatble Z-4 is a compilation of the communication characteristics of

- ccternal systems. This summarizes the RF characteristics of the
, StioL and the as sumed characteristics of the interfacing systems.
>S;-" characteristics are not included as they are not a lirmiting

Notice that in general, complementary, orthogonal circular polariza­

nit employed except for the VHF link where all right-hand circular polar­
.'t~i~,is specified. This permits both the ALS and the Space Station to con­
* ,.izt
to the ground through the DRS and directly to each other without
'?:, ,zto change antenna polarization. (The transnit and receive frequency
:'.-.!_ion is sufficient to use frequency diplexing for isolation.) Notice also
-1..1,e transmit and receive channels in the ALS and Space Station can be
V'rhiged as required to establish direct or relay links. The required
-, - .-ur per carrier (P.) is specified as are the antenna gains. Effective
'-,:,l pover (ERP) is the product of these parameters. The digital modu­
* i tes shown are the encoded rates which occur at the interface point
a
I'-fcarrier) and should not be interpreted as the information rates;
.''ere previously listed in Table Z-Z.

The internal interfaces are listed in Table Z-5. The number of

-' it Lre the number of discrete terminations either single-ended or
'.to ground. As m'entioned earlier in the discussion of the system
::ant, the voice and TV signas are proceaed within the R7 coin­
- >'= -5ysteun; therefore, only a single inte-FoCe is reuired for each

, •t:qx'te ,.iher ipcoherent dctdurlic-n tc ' (e. g. , Zn

:.' n*t,'-
i' l lr'.$eijCr
s
loor
V/O~
m oiio)-
,Le,,Eiicy
.), or e0011 lr-io
rL(, ~Uiat',
­
,v'.1itr
uj~
,]r o,,' ... :t -,: t ;,'L 'z­
 -+ Z

SOLAR-PANELS
E TENDED)

COREMODULE

ROTATION AT 4 REV/hAIN

Figure 2-3. Space Station Artificial Gravity Mode

EXTERNAL INTERNAL

ni-v <BNDy

I
-"-- V-.--J
2
I ENGINEERING
VOICE

VII
-D EMERGENCY
SPACE VOICE BASEBAND
STATION - COLOR TV COAMUNICATIONS
ill SERVICES
PCA WIDEBAND
DATA
ND hEM
i YIDERAND
RF DATA
COMM SFN TIM
S-BAN"I--- ALS TLM

S- ANDVHF •.---oEQUIPMENT STATUS

S-AND VHF
l/P4 CIRCUIT STATUS
EQUIPMENT
CONFIG CMD COMMUNICATIONS
SYSTEMCONTROL
ANTENNA ACQ CMtA

SIGNAL ACO CIAD

-- STATION-CLOCK
[ it' STATICN rOllEP, ANCIL" YST LMS

0- TACKING SYSTEM

Figure Z-4. RF Comrnm-nications Systern Interface

2-14
 A
iWEr2-4 EXTIERN-AL INTEI FACI i DCSCRIPTION

1forujatiotl Ant'lina G i.
rre ny o ra on d13

ra , tL Rcc-ive 'Tranismz Receive dBw 'I Bandwidth Type Chips/Sec Transmit aeceive
r2-b.in 3 14l.hS- 13. 70­
1(1.;b.5-
It 35 Gliz 14.36 M; i70? lLUGp 15.0 391 85 MU1. 2 PSK 40.4 x 106 42.5 41.9

K -b
aend
(beaconl) 15.35 GHz 14.26 GHz RHCP HGP -12.0 391 6.3 KHz 2s PSI- 1.5 . 101 42.3 41.7

Space S-band 2100- 2100- RHC? LHCP 19 4 438 15 MHz 44 PSK 7.38 c106 -3. 0 -3. 0
Slatl.nn (EMI ) 2300 2300

S-band 2100- 2100- RHCP LI-ICP 19.4 438 244 KHz ?

2300 Z300 PK 115.2 x 103 -3.0 -3.0
(ALS)

149.22 ­ 135.60­
VH-IF ------------
.------- RI-ICp RHCP 18.1 723 43 Ez M 3 KI- -3.0 -3.0
N15.- I149.22 - P=-.30
1 . 70 Mz .32 MHz
6
K-band 13. 70-' 14.85- LUG? RHO? 7.0 1200 85 MMd 24 PSK 40.4 : 10 41,9 42.5
(comrn) 14.2 G- 15.35 GHz

K-band 14.26 GHz 15.35 GHz LHCP RHCP -16.5 1200 6.3 KUz 24 PSK 1.5 x 103 15.0 15.0
DRS .(beacon)

'70 MIU;-
135.60 149.22 -
.3 2 MHz RHCP RH-C 12.7 C 800 43 5. 3KHz
M 5CM

S-hand 2100- 2100­

2300 2300 LHCp RHCP 18.4 800 244 K1Hz 2 PSK 115.2 x 103 -30 -3,0

A LS
135. 60 - 149.22 -
Vjq .70 MHz - .3Z MI z- RHP RHCP 20,1 800 43 KHz FM 3KHz -3.0 -3.0
1,19,22 135;60
32 N114 .70 MI-J.' i"

V1Sani 2100- 2100 LHCP RHCP 19,4 438 15Mgz 44 PSK 7,38x 106 20 20

T'anntled C/N a 15.0 dB at I-band

,',Tansmntted C/N a 12.5 dB at VHr
 TABLE 2-5. INTERNAL INTERFACE DESCRIPTION

Number
of Circuits
Circuit
Interface Group Descriptor Input Output Signal Description

Baseband Engineering 12 12 300 to 3000 Hz delta-mod voice,

comnmiunications voice inband signalling
services
Emergency 1 1 300 to 3000 Hz analog voice; dedi­
voice cated

Color TV 1 1 NTSC video

PCM wideband 2 2 Synchronous data in 2.4576 mega­
data bit increments to 17. 2032 mega­
bit; 2 increments maximum
simultaneous with TV

EM 1 wideband 1 I Synchronous data, 7. 3728 mega­

data bits maximum

MSFN telem-
etry I z
22 22
Two pairs each; four 19. 2
kilobits and 7, 307. Z kilobits
synchronous data channels
ALS telem­
etry

Communication Equipment A. . Continuous monitoring (AGC, RF

system status power out, etc. ) 0 to 5 volts,
control sZ = 1000 ohms; discrete (command
confirmation, etc. ) circuit clos­
ure to ground

A. R. : As Required
 Circuit
£.tc-t iLCC Group Descriptor Input Output Signal Description

Co n 'ication Circuit --- A. R. Digital channel sync word errors

sy'stern status (5 to 10 bit word); subframe
control assignments (8 bit word); frane
(continued) buffer register count

Equipment A. R. " .... Redundancy selection, parallel

configura- circuit closure to ground; antenna
tion corn- selection, parallel circuit closure
mand to ground

Antenna A. R. --- Parallel, axis 1.D., 10 to 20 bit

ACQ word
command

Signal acqui- A. R. --- Parallel, channel I. D., =L 1 volt

sition scaled
connand

Station 1 --- 19. 6608 rnegabits/sec clock pulse

clock

Ancillary Station A.R. --- 56 volts dc and 400 Hz per NAR

systrnms power

Tracking 2 2 S-band; compatible channel assign-,

sy'stem ment; omni coverage @-3 dB by two
overlapping patterns

A. R. As Required
channel. The POM data, however, is externally processed, so two interfaces
are provided for both transmit and receive to enable continuous handover
operations. The MSFN -and ALS telemetry circuits are similarly externally
generated, and so duplicate terminations are provided for each channel,

The detailed communication systen control interface cannot be

defined at this point as it is intimately related to the baseband and communi­
cation control equipment design. Typical signal descriptions are provided,
however, to give an understanding of the scope and nature of this interface.
The power system interface is straightforward, but as in the instance of
the control interface, the specific voltages required by each subsystem can
only be determined by considering detailed equipment designs. The track­
ing system interface has dual transmit and receive ports corresponding to
the circumferential and axial arrays.

2.5 SYSTEM COMPATIBILITY

The need was recognized during the design phase for maintaining
compatibility between the various portions of the baseline system. Incom­
patibility was prevented wherever possible simply by minimizing the number
of interconnections and interdependences. Design decisions were also
influenced by reports of certain problems which have been discovered dur­
ing operational use of the Apollo communications system. The Apollo track­
ing system is particularly susceptible to amplitude modulation at frequencies
of 50 Hz and 100 Hz, the error signal multiplexing frequencies. Incidental
amplitude modulation is caused by subcarrier oscillator imperfections,
carrier amplitude modulation, and intermodulation between the range code
and the subcarriers (References Z-1 and 2-2). And, in the wide-beam
acquisition mode, the Apollo command module high gain antenna occasionally
acquires a multipath signal reflected from the spacecraft surface. (Refer­
ence 2-3)

The baseline autotrack antenna system was especially designed to

avoid these problems and to minimize the possibilities of incompatibilities.
The system normally tracks on a beacon signal provided by the DRS.
This beacon originates at the DRS, and the communication signal cannot
interfere with either its generation or reception. No significant tracking
errors should occur due to "incidental amplitude modulation" of the beacon;
its biphase PSK signal structure is quite simple and easily implemented.
The antenna error signals generated by the autotrack system are demodu­
lated within the RF-IF subsystem by an (incoherent) envelope detector.
This type of detector was recommended for Apollo after studies of certain
autotrack system difficulties, due to up-link modulation. The Space Station
autotrack system will also not acquire signals reflected from the space­
craft surface. This is achieved by the use of precise initial pointing, a
narrow (I degree) acquisition bcamwidth, and careful antenna placement
on the vehicle.

Z-18
 ganging for orbit determination can cause no degradation in antenna
or comm-unications performance because it is accomplished by a
• -:,' svstem. Input/output ports to the Space Station S-band antenna
* '-'-l are provided at RF for a turnaround transponder compatible with

Because the ranging and antenna tracking functions are separated

:i the. data transmission function, the signal structures at both K-band
.,i s-band are optimized only for the communications function. The sig­
.rmnat used for the DRS link is biphase PSK directly on the carrier.
rj ivision multiplexing (TDMi) is used rather than frequency division
itinlexinlg with subcarriers. In Apollo, intermodulation between the base­
t .:. voice and the telemetry subcarrier may cause significant degradation.
St1e baseline system, the presence of nonlinearities will not generate
inificant intermodulation noise in the signal structure. The multiplexed
S,& types are generated by source encoders which produce highly random
1.1 sdreams. Furthermore, a convolutional encoder is used which will tend
., randomize the data stream. Therefore, the signal structure between the
WIvolutional encoder and decoder will appear to have few prominent spec­
tr, Uconponentso Any intermodulation products will be at very low

n the FDM structure of the Apollo Unified S-Band, a residual

K-Jm ]ponent is present at the carrier frequency for receiver tracking, how­
*-.er, the receivers have shbwn a tendency to lock on spurious components
ntar the carrier at frequencies related to the telemetry bit rate. The
t eaeetry rate on the 1. 0Z4 MHz subcarrier is 51. 2 kilobits/sec (Refer­
',eC 2-4). False acquisition by the'various digital receivers of the base­
line system presented here will not occur. The carrier phase tracking
l,.-ps track a replica of the carrier obtained by squaring the communica­
thions signal. (For the quadriphase link, a X4 loop is used. ) This straight­
' rvard implementation will not generate components that the tracking
loop might falsely acquire.

No incompatibilities are forseen between the comimunications

eiectronics and the Space Station power conditioning system. For example,
0he frame synchronization scheme was designed to accommodate the power
transients. Furthermore, the various electronics units are packaged for
protection against EvI and are separated physically from other systems.

As stated above, no attempt was made to design a unified communi­

-;dtlons system of the Apollo Unified S-Band type. The opposite approach
w..al- used in order that the individual functions could be accomplished by
'thit: optimized for that purpose. It reLnains the responsibility of the
tv~U~ufllications control facility to operate the system in a coordinated
',:nner either by automatic or manual control.

Z-19
Z.6 REFERENCES

2-1-: "Tracking Effect due to IAM on the Uplink Mode Containing

both S-Band Subcarriers, " TRW Systems Group, under
Contract NAS 9-4810, to the Manned Spacecraft Center,
18 March 1968.

Z-Z. '"Electronic System Test Program, CSM Block II Unified

S-Band System Compatibility and Performance Final Report
(Section 11), " EB 68-3119-(u) NASA MSC/ISD, February
1969.
2-3. "An Investigation of the Multipath Problems Associated
with the Apollo CSM and LM S-Band High Gain Communi­
cation Antenna Tracking Systems, " TRW Systems Group,
under Contract NAS 9-8166 to the Manned Spacecraft Center,
16 December 1968.

Z-4. "Electronic System Test Program, CSM Block II Unified

S-Band System Compatibility and Performance Final Report
(Section 10), " EB 68-3119 (u) NASA MSC/ISD, February
1969.

2-20
 3. ANTENNA SUBSYSTEM

1 INTRODUCTION
:k.

The antenna subsystem block diagram is shown in Figure 3- 1. The

K-band section contains the high gain antennas, the antenna masts and
t.rection mechanisms, the pedestal and servo systems, and the associated
RE gear, such as preamplifier and power amplifiers. The S-band section
consists of two broad-beam arrays and preamplifiers. The VHF section
also consists of two arrays and their preamplifiers. The S-band and VHF
power amplifiers are part of the RW-IF subsystem because they are
physically colocated with the IF equipment.

Figure 3-Z shows the NAR Space Station configuration and the loca­
ton of the various antennas, and Table 3- ' surmnarizes the physical
characteristics of the antenna subsystem. The weight figures are detailed
Ln Table 3-2. The power required assumes a typical operating condition
with only one K-band power amplifier active. The volume requirements
are specified for the electronics components only.

3.2 K-BAND SECTION

Three K-band antenna groups are employed. The antennas and

pedestals are mounted on masts spaced.120 degrees about the circum­
f:rence of the station. The two masts at the forward end of the station are
folded down to the conical section diring launch. These are arranged so
that their azimuth axes are collinear for despinning in the artificial g mode.
The third mast folds into the inner- stage area at the aft end. The masts
2rt! erected and locked after normal station operations are begun.

A 48-inch parabolic reflector is used with a five-horn Cassegr .in feed.

Table 3-3 summarizes the characteristics of the K-band antenna. The
h-tailed gain and loss budget is given in Table 3-4. Its physical dimensions
Tp.:ar in Figure 3-3. Note that an equipment compartment at the rear of
.:.t dish contains the monopulse circuitry and preanplifier.

3-1
 RECE TD I' Ad"'
PRE-A."P Momps POE.A
AND FILTE R CIRCUITRY AND FILTER

] 44 ,vTO
po
P RECEIVER

FOSVVO

SEEYO4tI w ANTENNA FROM

COFIURTINCONFRO SIGTION TiRA~lRS

CONFIGUTONRAT"IOANNITE
aN IFILAERFL 7
C OWANDS ELECTRONICS

COFGRAINSERVO TRP-AM AN---T FDRO A IE C~ AD

COMMANDS ~ ELECTRONICSGIBL
ICONFIGURATION COFGRTONCMAD

SSETORECEIVNOOUS -.- d ERS SD

BAND SECTION

ONFIURATION

7 ............ YyAPR-U
17 'TY
F r -1AtnONye T-2ND
FCONFIGURATION
Blo CMMANDD
SECTION
Ri.iE
fIL,.ME] PEM
EApp ' SWITCH FEEDN 0oR

(250=K))

SYS- TO REE
E-M ---- --­ V 2

- -- - -- -- - - C NFIGRATIN COVIID SSA DCTION

3-.7~
Fig~~~~~z'e L (r d3 S b s se lc z t

3-7
 S-BAND AXIAL ELEMENT VSF AXIAL

SOLR PANEVS K-BAND PARABOLIC S-BAND CIRCUMFERENTIAL

ANTENNA ARRAYT

33 7 FT

ELEMENT
|C5 F1

AXIAL
________ Vh CiRI..MWERENTIL ARRAY EILEMN
7 T(OCA T ?z'OFT FRO LMN

W FT

Figure 3-2. Space Station Configuration (North American Rockwell)

3-3
 ).'ONOPULSE FEED

48 an. DIA 6 1:

MONOPULSE HYPERBOLOID StJF-REFLECTOR

CIRCUIT AND
PREAMPLIFIER
FORSUBREFLECTOR
QUADRAPOD SUPPORT

PARABOLOID IAAtN REFLECTOR/D=0,4

Figure 3-3. Cassegrain Reflector Antenna

3-4
 Pounds 'Watts Cubic lnches <xt., ,i

K- 13and
Kefector and feed package 11. Z 6.0 --- A

Servo electronics 2.5 5.0 50 B

Pedestal 39.0 14.0 --- A
Mast and mechanism 22. 0 -- --- A
-SF power amplifier 24.0 220.0 300 B
Configuration switch 1. 0 --- 10 G

One Antenna Group 99.7 245. 0 360

K-Band Section" Z99. 1 Z45. 0 1, 080

S - Band
Circumferential array and feed 350. 0 --- --- ---
Axial array and feed 62.0 .........
Preanp and filters (3) 9.0 18.0 120 A
Configuration switch 2.0 --- 300 C
S-Band Section 423.0 18. 0 420

VHF

Circumferential array and feed 66. 0 --- --­

Axial array and feed 24.0 --- --­
Preamp and filters (3) 3.0 0.3 30 A
Configuration Switch 1.0 --- 30 C
VHF Section 94.0 0.3 60
TOTAL 816. 1 263.3 1560
tissurnes t.--a. an a group active
.'...E:lctronics only
 TABLE 3-2. ANTENNA WEIGHT ESTIMATES

Weight,
Pounds

Circumferential Array (1/16-inch aluminun)

Two radiating apertures (TX and RY) 88
Two corporate feeds (TX and PY) Z6Z

Total 350

Axial Array

Two radiating elements and feed 5

100 foot waveguide (20-mil aluminum) 32
100 foot cable (coax and power) 25

Total 6Z
S-Band Total 412
vHF -

Per radiating element

Dipole, balun, and hybrid 1

Pop-up mast and base 1

Element Total 2

Circumferential Array

Aperture (18 elements) 36

Corporate feed 30

Total 66

Axial Array

Aperture (two elements) 4

100 foot cable (coax and power) 20

Total 24

-690
 3-3. K-BAND ANTENNA SUMMARY CHARACTERISTICS

• -igCassegrain,
,"D-ign Monopulse 5 Horn Time Division

k ,tor0. Diameter: 48 Inches

Mbtnd Gain:

*raflsmi. 42. g aB @ 14. 825 GHz

Receive: 41.9 dB @ 13. 725 GHz

olArizatioAl (4 db Axial Ratio)

Transmit: RHCP

Receive: LHCP

Ey-lVnidth (-3 dB) 1.25 Degree

Uliibal Design: Three Axis, Hemispheric Coverage

Without Tracking Interruption
?v-rvo Design: Torque Motor Drive W..ith Dgtal
Position Feedback
Signal Inputs: Monopulse Tracking Errors,
Manual Slew and Digital Positioning
Commands

Figure 3-4 shows the three-axis antenna gimbal design. The

•-.t;u n of the cross-axis gimbal to the conventional elevation/azimuth
rpemits tracking without interruption over a viewing area greater
4 hemisphere. During auto-track operation, the cross-axis and
.'Lion axes are controlled by the monopulse error signals. The azimuth
slaved in such a manner as to maintain orthogonality between the
•Y. tt and cross-axis using the cross-axis angle transducer as a signal
- ,.~. As the line of sight to the DRS approaches the zenith region of the
.;ystrn
Yi (that is, along the azimuth axis), the azimuth gimbal
loop gradually becomes ineffective in maintaining this 6rthogonality.
.. . loop is eventually disabled by the zenith logic unit. The antenna
.. c X. through the zenith region as an X-Y mount. The highest
required of this system is 4. 5 deg/min on any axis. In the
.ate
node, the communications control facility provides a constant
- '-ocity command input to the azimuth axis at the Space Station

3-7
 TABLE 3-4. CASSEGRAIN R'EFLECTOR ANTENNA
GAIN AND LOSS BUDGET, dB.

Center Frequency, GHz 13. 725 14. 825

Area gain, dB 44.9 45.6

Losses, dB

Illumination efficiency 0. 5 0.7

Spillover efficiency 1. 0 0.8
Hyperboloid blockage 0. Z 0.3
Hyperboloid support blockage 0. 5 0. 5
Reflector surface error 0. Z 0. Z
(an = 0.015", as = 0.005")
Feed, ohmic , 0. 3 0. 3
Cross-polarized power 0.3 0..3

Total 3. 0 3. 1

Expected gain at center frequency 41.9 4Z. 5

Additional loss over 0. 950 GHz band 0. Z 0.2

Gain at band edge 41.7 4Z. 3

*Contributes to system noise temperature

Figure 3-5 illustrates the antenna servo system. The power amplifiers,
torque motors, tachometers, and resolvers are located on the gimbal structure.
The servo electronics shown is located in the equipment shelf. The rnonopulse
feed netwvork generates signals Al andAZ in proportion to the pointing error in
two orthogonal planes. The signals are amplitude modulated onto the sum chan­
nel after time division multiplexing. The resulting signal is passed through a
single preamplifier and then sent to the RF-IF subsysterr for further processing.
This subsystem recovers the tracking signals and forwards them to communications
control. Norinal operation entails tracking on the DlRS beacon; the design also
permits track-on-comm as a backup capability with negligible degradation.
The tirne nultiplexed Al and AZ signals are sent to the servo electronics
unit by comnunications control (at the bottom of Figure 3-5) and after
multiplexing, they are passed through shaping circuits and used to control

3-8
 c;oOS-A(IS 4 20 ,

INE-OF-SIGHT TO
I DATA RELAYSATELLITE

REFLECTOR

ELEVATION AXIS
130. FREEDOM
1. 20lBELONVHORIZON)
,,AVEGUIDE ROTARY
JOINT

5INCH DIA.

AZIMUTH-CONTINUOUS
360VFREEDOM

Figure 3-4. K-Band Antenna Gimbal System

K-BAND ANTENNA
WITH MONOPULSE
FEEDNETWORK 62

/
+ POWE RAMPLI FIER

TOQU MOTORI~

CLOI "I REDGITERS

C I TRQUE
+TE TACHG HOLD DI
4SUBTRADT E
LEAD

CULIPLEXER
CRO ML COD
ZENITH SPOSITyN*­

•'i,tug'e 3.5. }(-B3cuid Anten. ;s S.r'vto Blurt,. Diagr'am

3.-9
the cross-axis and elevation drive motors. The repositioning of the antenna
boresight by these motors completes the autotrack feedback loop.

The antenna servo system also functions in a slew mode for auto­
matic prepositioning of the antenna at handover or for manually assisted pre­
positioning. When the mode control logic selects the slew mode, the control
loop detailed in Figure 3-5 becomes active. The loop error signals are formed
by subtracting the contents of the command position register and the actual
positioi register. Slewing continues until the loop error is zero on all
axes. A rate limit is included to prevent excessive slew rates.

The sum channel of the monopulse feed network is initially proc­

essed in a unit placed directly behind the reflector. A receiver filter is
provided and a 250°K uncooled, parametric amplifier is used. The receive
K-band waveguide is separated at a coaxial rotary joint just below the
azimuth drive and then is routed within the antenna mast to the equipment
shelf at the base,

The transmitter includes a K-band traveling-wave tube amplifier

(TWTA) located within the equipment shelf. The TWTA output is 44 watts
with a 25 percent efficiency; the associated high voltage power supply is
80 percent efficient. Passive thermal control is used within the equipment
shelf containing the amplifier. A loss of 1. 4 dB is budgeted for the signal
path from the output of the TWTA to the antenna feed port. The transmit
signal is inserted through the coaxial rotary joint in the azimuth axis and
through a rotary joint in the gimbal opposite the receive channel at the
elevation axis.

3. 3 S-BAND SECTION

The S-band antenna system consists of two separate arrays. The

general coverage provided by these arrays is shown in Figure 3-6. The
circumferential array provides a toroidal pattern centered about the
Station's longitudinal axis. When the Station is in the X-POP (X-axis perpen­
dicular to the orbit plane) attitude mode, this antenna prdvides good coverage
to vehicles approaching the Station for rendezvous, and to ground tracking
stations. The axial array provides coverage in the +X and -X directions.
With the X-axis oriented along the velocity vector during Y-POP operations,
the axial array is utilized for the S-band links to the ALS and EM I .

The circumferential array has separate transmit and receive aper­

tures spaced 1 foot apart, center-to-center (see Figure 3-2). Figure 3-7
illustrates the power division network used for the feed. Each radiating
element is 6.47 feet long and has 18 slots and loops of the type shown in
Figure 3-8. The system transmits at S-band with left-hand circular polari­
zation (LHCP) and receives with right-hand circular polarization (RI4CP-).
A section of the transmit/receive patterns is shown in Figure 3-9.

The axial array uses one element at either end of the Station (see
F rigue 3-2). The a::ial element design is shovwn in Figure 3-10. This

3-10
 t.- MFERENIAL APRAY
.ZIRCU
-(OROI AL PATTERN)

AXIAL ARRAY AXIAL ARRAY

ORBIT PLANE
IN X-POP ATTITIDE MODE

Figure 3-6. Omnidirectional Antenna Placement and Coverage

S-band and VHTF

EACHRADIATING SEOMENT HAS I8

SLOTSAND LOOPS AND IS 6.47 FEET
LONG AT THlECIRCUMFERENCE

INF1JMT
PROBE.

Figure 3-7. S-Band Circumferential Array Feed Network

3-11
 SHUNT SLOT

4
Loop: 1IN..

VEHICLE SURFACE

Figure 3-8. S-Band ,

Circumferential Array IAI AE OT
Element SEG-ENT

SINLAYER~ED W G CORPORATE
FEED

0 . To- .

°
300 - 0

50 o50

40.

-14 Figure 3-9. Gain

2Patterns of S-Band
SECTION OF TOROIDAL Arrays
IV( APATTER1, OF CIRCUMFERENTIAL

S4-2---2--4--6--8-I01-1<

20 PATTERN2'
OF ONE
OF AXIAL
ELEM'ENT
30- ARRAY

40'

CIRCULAR TO RECTANGULAR S
l,'G (4 N 1.0) TPANSITION

FIVE FIN ORTHOMD ZA

2.01 POLARIZER TRANSDUCE R
POLARIZER FINS,

Figure 3-10.
S-Band Axial . -.

Element (K\ I0T

TRtANSMIT OR RECEP

CHO.E PINGS

3-1Z
* i,,nent provides both RHCP and LHCP for transmit/receive isolation and,
iz addition, the element design provides a high degree of isolation from
iurrounding structure. For this reason, the element may either be
nounted.flush with the skin surface or it may be permitted to protrude.
The pattern of one element is shown in Figure 3-9. Table 3-5 summarizes
c.eperformance of the two arrays.

Three 100°K uncooled parametric amplifiers and associated filters

are used in the receive channels. One is positioned at the output port of
the circumferential array and the other two are colocated with the two
axial array elements.

TABLE 3-5. PERFORMANCE OF S-BAND AR.RAYS

Circumferential Array-:* Axial Array":

Degrees
Off Axis G Ellipticity, dB G Ellipticity dB

0 -0.5 1 +4.5 1

10 -0.8 +4.3

20 -1.3 +3.5

30 -2.2 +2. 3

40 -3.5 +0.3

45 -4.3 5 -1.0 3

50 -5.Z -Z.5

60 -7.3 10 -6.5 6

70 -9.6 -10.5

80 -13.5 -14.5

U ,n-ic loss (per element) = 0. 3 dB

Rjppie t dB at 0 degree, ohmic loss =
-i . dB

3-13
".'I configuration switch shown within the S-band section perLits connection
Ct. desired antenna array and IF units and the input/output interface with
mS N tracking system.

t 4 VHF SECTION

The VHF antenna system consists of two separate arrays. The

-,overage these patterns provide is very similar to the &overage of the
4-bat'l arrays. (See Figure 3-6. ) The patterns are optimized to nrovide
Coverage in the directions defined by the principal axes of the spacecraft.
The circumferential array consists of 18 crossed-dipole elements. The
fvtd network is shown in Figure 3-11. Because the antenna must provide
ICP on both transmit and receive, transmit/receive isolation is achieved
by the use of a frequency diplexer. A typical element is shown in Fig­
nre 3-12. After launch, the mast telescopes out to the full 10-inch length,
.Thich the design requires. A section of the transmit/receive patterns is
town in Figure 3-13.

The axial array uses one of the crossed-dipole elements at each end
of the Station. The pattern of one element is shown in Figure 3-13.
.A~sunin-g a require=lent for only -3 dB gain, a useful beamwidth of 100
degrees was attained. Table 3-6 summarizes the performance of both VHF
arrays. Three %50 Xtunnel diode preamplifiers are used. One is located
ai the output port of the circumferential array. The other two are located
-each end of the Space Station near the axial array elerients.

3.-!4
 ItFL-. c;RCUI.WEREN1
ALAIRTA
~&Z
2IV 7 9 I0GWII 12 a3 14 i5 1617

Figure 3-11.
VTF Feed Network

Ar P.'r eVD

TX UItLEXER PER

POWFRDIVIDERCIRCUITISCOAXIAL, EQUIPhASE, AND EQUIAMPI.ISJDE

40 IN CROSSED DIPOLES

Figure 3-12. 2.0 IN DIAMETERDIPOLE

VHF Array Element

SUIOACE
10.0 IN OFF VEWI4CLE

TELESCOPING MAST

RECEPTACLE MOUNTING BOXFOR

INTERCHANG EASILI -CONTAII'lS
118 0ID TOR PIIASINO DIPOLES,
70D I
w 2INTERFACING COAX CONNeCTOr
AND BALUN
30. 30,

0,Fig -e 3-13.
°
30

PTTERN
x OF A EN T IAL Gain Patterns of VHF

EECBEL
GAIN, XArrays

3-15
 TABLE 3-6. PERFORMANCE OF VHF ARRAYS

V HF Antenna Performance

Circumferential Array'* Axial Array-

Degrees
Off Axis G Ellipticity, dB G Ellipticity dB

0 +0.5 1 +2.5 1

10 +0. 1 +Z.Z

20 -0.5 +1.5

30 -1.6 +0.5

40 -3.0 -1.1

45 -4.0 5 -2.0 5

50 -5.0 -3.0

60 -7.5 10 -5.5 10

70 -10.5 -8.5

80 -14.5 -Iz.5

'Olrnic Loss (per element) 0.4 dB

:'"Ripple = ±1 dB at 0 Degree,Ohmic Loss = 1. 5 dB

3-16
 4. R -IF SUBSYSTEMI

4. 1 INTRODUCTION

The RF-IF subsystem block diagram is shown in Figure 4-1, where

the units indicated actually occur in triplicate. The subsystem consists
basically of receive and transmit IF stages and the power amplifiers. The
K-band power amplifiers are not included because excessive line losses
forced their placement at the base of the antenna masts. The DRS beacon
demodulator is included in this subsystem becatse it interfaces only with
communications control -- it does not require a baseband interface with a
communication system user. Because of the simplicity of the VHF emer­
gency voice equipment, the moden is packaged with the RF-IF gear. The
RE-IF subsystem is actually defined (as shown in Figure 4- 1) as that hard­
ware located between the configuration switches in the antenna and base­
band subsystems. Figure 4-2 summarizes the physical characteristics of
the subsystem. Note that the power indicated is a typical load not a peak
load.

4.2 K-BAND SECTION

A single stage upconverter is used on the transmit side. The data,

modulated on a Z70 MHz source is up-converted to one of the satellite
communication channel frequencies (14.85 to 15. 35 GHz) by a channel
selection command to the frequency synthesizer. This stage also up­
converts the Space Station beacon to 15. 35 GHz.

The receiver downconverter uses two stages in the parallel con-.

rmsunication and beacon channels. The first stage of the conmunication
channel downconverter permits selection of the appropriate D-RS communi­
cations channel. The intermediate frequency of both the beacon and com­
.. nnication channel is 750 MHz. The second conversion generates a beacon
-,'tond IF at 30 MHz and a communication channel second IF at Z70 MHz.
* requency acquisition and tracking is accomplished at these second stages.
QLquisition is aided by a doppler compensating offset based on computer
•onerated estimates. Automatic frequency control (AFC) for both the
"1"tl-ounication and beacon channels is normally accomplished by tracking
'4-tin beacon. Communications signal AFC is used as a backup.

4-I
 To
.- C" w
DRIVER qT p
AND TOSESM,1N
BA D
UPC(NV qThP _p CONTROt SUBSYSTEM
TO AND EON CMS
ANTENNA
SUBSYSTEM

~
' CIANt.LL DOWN

PROGRAMMED DOPLER OFFSET

CHANNEL BEACON DRS

DOWN DEVODULATOP TELEMETRY

STDIS CODE
COMIOND
TRACKING
FREQUENCSIGNAL
ANTENNASIZR DVPROGRAMMED
DOPPLER
OFFSET

SELECTION COMNO A N

SYISST FIT R TRANSMIT *

FITERUI'COMMUNITE tp OWEP CONTROL Ch'D1
1- TIONCHANNEL

FILTES
SELECTOR
SWITCH

I OVERTE ALS OR NEM

SYASYSTEM FLE L =0 2 Ld UONM CHANNEL
TO TO
CONFICURAT0 BASEBAND
N
SUBSYSTEM
S TCOMMANDS
U
LCOL OE L~~ __L__- [ -VAT,,
CHANNEL
ALS CHTSNANEL

SYNTHESIZEiR'
E '-
I MH. LOCK
TRANSMIT CHANNEL
-SELECTION
SBAND SECTION

.MYER
DORIVER
AN D U ,4
ANTENNA TB
SUBSYSTEM'' " S g NTEB
A@' D

45CHANNEL
L SELECTION

OMN IT ETO
'IEACH, ALL OT14ERUNITS, 2 EACH

Figure 4-1. RF-IF Subsystem Block Diagram

4-2
 The beacon demodulator is detailed in Figure 4-3. The device is a
Z. , implementation of a delay lock PN code tracking loop and a biphase
g,, demodulator. The generation of the inphase (I) and quadrature (Q)
.. ,oflents occurs in mixers 1 and 2. The feedback connection of the data
1, ojcaition loop is seen at the bottom of the diagram. The code tracking
- rvor is generated by taking the difference of correlations of the incoming
.al with an advan'ced and retarded replica of the reference code. The
, .. of the desired DRS must be loaded into the PN reference generator by
:: zunications
.) control. The beacon demodulator outputs 150 bits/sec
' telemetry to communications control when beacon acquisition is
e 9tblished.

. S-BAND SECTION

A double stage upconverter is used in the transmitter. The signal

be transmitted enters the RF-IF subsystem at either 3 MHz (ALS data)
.
at 50 MHz (EMI data). After the first upconverters, these signals are
,,,itioned within parallel IF's centered at IZ0 MHz. The second mixer in
r-*h channel generates an S-band output which is then passed to one of the
.w:er amplifiers. The power amplifier is a medium scale integrated cir­
,t device; it generates 110 watts with 50 percent efficiency. The appro­
:r-ate transmit filter is selected by commands from communications
n trol.

The first mixers of the receiver are followed by parallel IF's for
tX. ALS and EM 1 signals. Each is centered at IZ0 MHz; the bandwidth of
ALS amplifier is I MHz and that of the EM 1 is 16. Z MHz. The second
*ca.;e in each of the IF's performs the AFC function. The second IF ire­
u-tces are 3 MHz for the ALS and 50 MHz for the EV 1I .

',.4 VT-IF SECTION

The 3 kHz voice signal is frequency modulated on a 30 MHz carrier.

1T;is is up-converted to either 149 or 136 MHz. The power amplifier is a
f,
tercent efficient solid-state device, capable of 80 watts output at VHF.
.c receive side uses two stages of conversion (either 136 or 149 MHz to
S,,o 3 MHz), Automatic gain control is performed at both IF stages; the
b 'sn'idthof each is 43 kHz. The demodulator is a threshold extension

4-3
 i I
3
l:EW,' -WEIGHT (LaS) POWER(,1 VOLUME (IN ) LOC.AICz ,­

i, .. h(; NCS 10.0 23.0 70 C

'C IP('30.0 2.0 211.0 C-,tC

'. .REAP,'IIER 10.0 2200 20 C

C. , ".ATTOTSWITCHES&FILTERS 2.0 --- 21 C
•
ESiEN~.2.0 18.0 no c
C'.. C ON-Up 32.0 2.8.0 1,124 C
.IAN SECPON ** (3) 96.0 476-'0 3,372 C

40 160 0 130 C
foRA'.PIIFIER 5.0 6.0 700 C
.rI1 ELECTRONICS
ONE GROUP 9.0 166.0 80 C
VHF SECTION' (3) 27 0 W4..0 2,430 C

TOTAL 153.0 496... 7,872

A• -SUMESONE GROUP ACTIVE

ASSUMESIWO GROUPS ACTIVE
... ASW'.LMESK-BAND AND S-BAND ACTIVE

Figure 4-2. RF-IF Subsystem Physical Characteristics

IIt- I~~,A. TRACKIN

FiemtoR
Fgr 4-. D SBncnDDATA

AT"0
[[ AGC .

goo MIER-2 P04 r

j : LOW PASS ANALOG-TO- I __X=r=

•,•"
rf PSEUDO-NOISE

GEERTO

Delay lock P -AoTTE Rat.biW loop

 5. BASEBAND SUBSYSTEM

5. 1 INTRODUCTION

The baseband subsystem block diagram is shown in Figure 5-1. As

before, the units shown actually occur in triplicate for purposes of link conti­
nuity during handover and for redundancy. The subsystem consists basically
of modulator/demodulators, digital channel encoder/decoders, and time divi­
sion, multiplex equipment. In the K-band section, digital source encoder/
decoders are alsoa provided which process the analog video and voice signals.
Only digital data streams are passed across the S-band section interface;
source encoder/decoders were not required. The VHF section contains little
baseband equipment, as the FM modem is contained within the RF-IF subsys­
tem. The baseband subsystem is delineated with respect to the other, subsys­
tems in Figure 2-1. Figure 5-2 summarizes the physical characteristics of
the subsystem. Note that the power indicated is a typical load, not a peak load.

5. 2 K-BAND SECTION

Modem

The modem consists of digital and analog portions. The convolu­

tional encoder (K = 5, V = 2) accepts a data stream at approximately Z0. 2
megabitt/sec and generates 40.4 x 106 subbits per second. The biphase
modulation is accomplished at Z70 MHz; the data is then passed to the config­
uration switch for routing to the proper RF-IF units. The biphase demodu­
lator recovers a coherent carrier reference and subbit tining information.
A filter and sample approach is used in the detection operation, but rather
than a "hard decision" on the subbit, an analog to digital (A/D) converter
quantizes the detector output and generates a 3-bit word. This digital
output is processed by a-parallel Viterbi decoder and the result is a 20. 2
rnegabits/sec data stream. All the I-band gear can operate at 307. 2
4ilobits/sec in a fall-back node.

-l)1l Systerm

The transmit portion! of the tir1 ,s divi-don nultiplex unit is shovwn in

' t-ure 3-3. The vnit accepts standardizcd Station-clucic-synchronono c%.ta
i, .nis. The rates of Qither the TV or PCM[ daua can vary from zero to
17. "i)-2megabits/sec in inccemcntu 0! ."776 mg.bits/sec
11, vith the

5-I
 - ESELE~i- STATIO

STATUS VOICE*
MONI RIND'G MOD

To
FF
CONFIG
SELECT CONFIGC
SELECT CONFIG" MOD
Rs 'SISIt. SWITCH SWITCH SELECT
SWITCH1SWITCH

OTLM

MODE SELECT _ VOICE

COMMACDS DEMOD

20 CONVOLUTI!ONAL DATAO
TO NTRNAL
DECODER

COMMUNICATION COMUICIK-BOND SECTION

TRACKING SIGNAL CMD

---- SELECT

MOSDAN SELaIONM

TO
COOVOUTION

DEMDUDEODE

RFS-BAN SELECTON
SYTO SWTCHN STTU STATON[INTRNA

SWITCH INTER,
FACE CV

I A04, ALL OTER UNITS 3 EACH VHFSECTON

Figure 5-1. Baseband Subsystem Block iagram

5-2
 3)
N C
ITEM WEIGHT (LBS) POWER (W) VOLUME(N) LOCATION c,
10

MODEM 17.0 44.0 800 D&E it

62.0 100.0 2,000 D&E W
TDM
CONFIGURATION SWITCHES 3.0 --- 1,000 D8E
16.0 12.0 800 D&E
K-BAND VOICE MODEM 54.0 61.0 800 D&E
VIDEO MODEM
ONE GROUP 152.0 217.0 5,400
K-BAND SECTION .410.0 217.0 11,000 D&E

'ALS MODEM 12.0 19.0 400 D8

IEM0 MODEM 21.0 40.0 800 D&E
'S--AND
S-N
ICONFIGURATION SWITCHES
J ONE GROUP
3.0 --- 1,000 D&E
, 36.0 59.0 1,840 D&E
[ S-BAND SECTION 102.0 59.0 4,600 D&F

VHF CONFIGURATION SWITCH 2.0 --- 400 D&E

TOTAL 514.0 276.0 16,000

* ASSUMES ONE BASEBAND GROUP ACTIVE.

Figure 5-2. Baseband Subsystem Physical Characteristics

 TV DE TV AND

PCM
2/7 DAA SUMADATA,-

PRA.t U CCLOK FPNtEOVERHAD

AAULOC SUBFRAME

VOICE 32 SUB-GROUP 8t

t DENOTES STATION CLOCK

I.- STMT ~AJAE SA TADA

Figure 5-3. Transmit Block Diagram

~jTRANSMIT

FRENOE
SYNTATTRN
CLOC

MI
IZ T-
,
I

I -

Figure 5-4. Frame Timing Diagram

5-4
 * : ctioxi that the total TV and POM rate can be no greater than 17. 2032
.4nt3/!sec. The twelve voice channels are each 19. Z kilobits/sec and
)i the four priority telemetry channels are 19, 2 kilobits/sec. Seven
S- o'bits/sec telemetry channels are also provided. The total avail­
dat capacity without overhead is 19. 6608 megabits per second. Fig­
,..shows ... the frame timing diagram for the nultiplexing process, includ­
,:Ie addition of the channel ID words and the frame synchronization word.

Figure 5-5 details the more complex of the subframe multiplexers.

,:-:nnications control commands the configuration logic device to gener­
t'zw appropriate shift register logic (SRL). These coimands indicate
:*.zLta rates being received in the PCM and TV channels. In the PCM
t-.rtvL logic unit, the pulse stream is gated successively into subframe
i -:;rough 7 storage according to the clock rate. The TV channel logic unit
.. ultaneously gates data into these same storage devices beginning with
.. 7 and progressing upward toward No. 1. If the PCM channel is oper­
at 12. 2880 megabits/sec, subframe 1 is used for PCM data and sub­
t :2Zs 7, 6, 5, 4, and 3 are used for TV data. The SRL configures the
' storage unit into a linear-feedback shift register and outputs a
' ,ido-randpm code. The signals from the parallel frame multiplex
/::: to the storage units are generated by the frame mux device. These
:. .als cause the storage devices to output their data in appropriate time

The frame mux diagram is given in Figure 5-6. Driven by the

.!r it clock at 20. 1984 MHz, this unit clocks out the data from the sub­
.:Oec registers and inserts the synchronization pattern and the channel
'.. at the appropriate times (according to the timing sequence given in
:.'r 5-4).

The receive portion of the TDM gear is diagrammed in Figure 5-7.

'cause the K-band link is full-duplex, the overall structure of the receive
.ir as very similar to the transmit unit shown in Figure 5-3. However,
.. ;- data is first received, the "start frame" time is unknown and further­
,rt-, during operation, the received clock frequency may vary (doppler
'. or clock drift). Frame synchronization and receive clock buffering
c,woinplishedby the frame dernux diagrammed in Figure 5-8. The
Vved data stream and received clock is accepted from the modem. The
-bit counter then clocks out successive 20-bit words which the sync word
4-hirer cross-correlate-s with a reference pattern and also with the
rn's complement. A search for the complementary pattern must be
.' rined because of the 180 degree ambiguity in the recovered carrier.
-c .e;cary,the data stream is re-inverted in the data channel logic unit.
5-9 describes the scanning operation and nanner in which the main­
S.-w xiode is attained. Once synchronization is achieved, the remaining
. is en-abled and the deinuitiplexing operation can be initiated.
*, rt.cc-% ed data and the associated clock can be buffered by the frame
.,t:& up to one full frame time of 0. 149 rns. It is assumed that the
.l'.tion
adjusts its transmission rate such tha-3t the rate seen by the
-ithlin this buffering capability.

5-5
 KM . NO.1 DATA
DATA 4 PC"
CANNEL"U 2 STORAGE
-
PCM LOGIC * UL rRM'E
C OCK SEL AMUCIOCK

-- I NO.2
PARALLEL CONFIG jA-
CONFIGURATION - T]OIN SRL
COMMANDS LOGIC

SRL

TV
DATA VT
I
2
I NO.6
CHANNEL Lr- -
TV LOGIC
CLOCK
SUBERAME

NO,.7 DATA

FRAME
PARALLEL
SrL MUX CLOCK

Figure 5-5. Z/7 Subframe Multiplexer Diagram

rS

DATA FROM o DATA

SUS-PFAME GATE

PARALLEL -AND SYNCBI

CONFIGURkATION I.D. AND - f DLA

LOGIC COUNTETO

REGISTER 0 0 I ^ R

L6,

Z'igure 5-6. Frame Multiplexer Diakram

5-6
 SUB A -ES S T NVAL
-7i L ek
DAATA

FAATD DEMUX DATFAVE

DELOUX CLOCKC
CLOCK

fMA

DATA D)EMUX • LM DE 'UX TLM

S-DENOTES STATION CLOCK

Figure 5-7. Receive Block Diagram

DATA LOGI
C5 AND
LOGIC DATA
DELAYPLUS
LOGICE

f..
; WORD a 6
IR ECOGNIZER

BIT F
COUNTERS

SYIIC

AND PARALLELy

11UEDL Q I; RE 15E
COVERLO

Figure 5-FIam euli xR DEAgrY

CLOCK ELAY7
 SCANNIkG MECOGNIZER SEARCHES
FORTHE 20 BITSYNC PATTERN
ANDITS COMPLEMIENT.NO ERRORS
PERMITTED.)
AFTERTHEINITIATION OF SCANNING, A PATTERNRECOGNITiONS
WILL BEMA.DEWITHIN ONE FJME TIME WITH A Pk:OEABILITY
GREATER THAN 0.999a

THEPROSABILITYTHAT THISSYNC IS FALSEIS LESSTHAN

6X IRIS
ANOTHER FRAMEWORD ISEXAMINED BEFORESYNC IS CON
SIDEREDESTABLISHED.THEFROBABILITY THATFALSESYNC IS
NOT DISCOVEREDAFTERTHISSECOND FRAMEIS GIVEN BY
"7
9.5 X 10 . THEPROBABILITY THAT CORRECTSYNC IS REJECTED
IS 2 X 1D-4

MAINTFNANCE (I;O OR MORE ERROPS &UST OCCUR IN TWO SUCCESSIVE

SYNC WORDS BEFORERETURNTO SCANNING.)

O THEPROBABILITY OF SYNC LOSS DUETO CHANNEL NOISE =

15
1.4 X 16-

Figure 5-9. Data Synchronizer

Characteristics

PARALLEL
FRAMEDEMU
CLOCK l I
SUFFRAA E. E
NO, 1 PCM
STORAGE CHANNEL PCM DATA

~~~~PARALLCNi
[
IU
1
-LOGIC
PARALLEL
M

- NO. 27LC

DATA-G LOGICLOI
Figure 5-0 7f em lipeerDaga
PAPA ME
DP ,RUX
FRAME -
COC
CLOCKj

Figur 5i0. 2/ CHANLtTeVe DArA

NO5-8
 The 7/2 demux diagram in Figure 5-10 shows that this unit is very
,!..dlr to the Z/7 mux unit of Figure 5-5. The 7/2 demux accepts seven
-:-tnbrames and outputs the data in two channels. The parallel frame
-~.--x signals gate the data into subframe storage registers 1 through 7.
.a a is then unloaded at synchronous rates into either the PCM channel
ol TV channel according to the configuration commands. Registers con­
1,:ijiig PN bits are ignored during this process.

ource Encoder/Decoder

At the time of DRS link handover, two TDM units will be operating.
The propagation delay difference in the links will cause the data streams
from these units to have a significant differential delay. Since the PCM
channel user may require uninterrupted data flow, he is provided the out­
ijuts of both operating units. The user may then buffer as required.
tfu:nentary data dropout will occur in the voice and television channels
during handover. The twelve 19. 2 kilobits/sec outputs from a single TDMI
are switched into a bank of delta modulators which generate the analog
voice waveforms. These units also encode the 3 kHz voice signals for
transmission. The IZ. 2880 megabits/sec television data is switched from
one TDM into a data expander which regenerates the commercial quality
analog waveform. A data compressor encodes the TV originating at the
Space Station and generates a 12. 2880 megabits/sec bit stream.

5.3 S-BAND SECTION

The S-band section of the baseband subsystem provides modems for

the EMI and ALS links. It also provides a means of automatic antenna
selection for transmissions to these vehicles. Figure 5-1 shows the
arrangement of the S-band section. The equipment shown is provided in
triplicate. Two ALS demodulators and two EM 1 demodulators are active,
wvith another set as standbys. Each active pair is continually receiving
data from both of the two antennas. Within each modem a measurement is
nade on the power received via a particular antenna. The antenna logic
unit determines through which antenna the received ALS signal power is
g-reater and through which antenna the received EM 1 signal power is
i-reater. The communication control then configures the system to receive
arid transmit both ALS and EM 1 data through the appropriate antennas. A
particular antenna may be used for both ALS and EM I data simultaneously.
A particular power amplifier may be used for either link, but not simul­
t*'neously.

Figure 5- 11 shows the more complex of the two moderns. The EM 1

.. Odenl accepts data at a 7. 3728 megabits/sec rate synchronous with the
:=tion clock. The convolutional K = 5, V = 2 encoder consists of a 5-bit
*nft register with tap connections to two adders. For each bit into the
.iHr, one subbit is generated at each adder -- that is, each adder out­
' 7. 3728 x 106 subbits per second. The data stream from adder No. 2
- ,h:,Le modulated on a 50 MHz source. The data stream from adder
I i; biphased modulated on the sane sourcV in phase quadrature.

5-9
 pHASE CLOCK
GENERATOR STATION CLOCK ,

MODULAfOR

FoM IF (FROM ADDIF NUMF6R 2) TONALU.

3,ja8 MYV16.2 Mft AP(RMADRNME )P EulCO)DE 7.38 mr, s

.LT
I MH

CLC [ INTEGRATE
CHAIN

RF-OF IFPHTASE

Figure 5-11. ThM1 Modem~

5-10
rlw sunmation of the two biphaso chanqIls generates a quadriphase signal
With 7. 3728 x 10 6 chips per second. Th6 receive side of the modem sees the
same signal format that the transmit side generates. The modem recovers
a coherent reference from the signal itself by a X4 phase tracking loop.
This reference and its 90 degree phase shift is used to translate the two
biphase channels down to baseband. The integrate and dump operation is
accomplished by the use of two units in parallel (for each channel) which are
used alternately. The chip time synchronizer determines the zero crossings
in the received data. The chip time clock determines the integrator sampling
times and the timing of the multiplexing of the eight-level digital samples.
The samples from the two channels are processed by a parallel implementa­
tion of the Viterbi decoding algorithm. This decoder outputs data which is
nominally 7. 3728 megabits/sec. Note that the modem also provides an AGG
signal and a signal to the antenna selection logic. Furthermore, multiplying
the recovered carrier with the local clock generates an error signal for the
AFC loop which controls the VCO at the second stage of the downconverter.

The ALS modem accepts data at 57.6 kilobits/sec, uses a K = 5,

V = 2 cdnvolutional code, and performs a biphase modulation of a 3 MHz
source. The demodulator employs optimum integrate and dump detection.
An eight-level quantization of the integrator's output is processed by the
serial Viterbi decpder. This device then outputs a data stream at the
received clock rate at a nominal 57. 6 kilobits/sec. Buffering of the EM 1
and ALS received data permits the interface data rates to be synchronous
with the Station clock.

5.4 VHF SEGTION

The configuration select switch permits the user to select a particu­

lar transmit/receiver unit. The interface with the internal communications
system consists of two wire pairs for the full-duplex 3-kHz voice.

5-I1
 PART II
PARAMETRIC ANALYSIS

Part II of this report documents the ana'lyses and parametric

trades performed during this study. Many of the sections
discuss design studies performed in support of the Space
Station communication system design. Several sections
provide data of a parametric nature which can be used in the
development of communication system designs employing
alternate implementation techniques. Where applicable, the
results of these analyses were used to evaluate alternate
designs and inplementations prior to the baseline selections.
The analyses included in this report will aid.in evaluating
modifications to the baseline system and will provide a basis
for the design of communication systems with requirements
differing from those of the baseline.
 6. COMMUNICATIONS GEOMETRY AND ANTENNA
POINTING ANALYSIS

. INTRODUCTION

An iri-depth analysis of the geometry associated with the Space Station

links was begun early in the study. It was concluded that three high gain
antennas were required on the Station for uninterrupted DRSS communication.
This effort also developed the need for a three-axis antenna gimbal system.
Computer analysis allowed a quantitative description of the required gimbal
-,ystem performance. A second computer program was developed to analyze
the performance of the selected cross-axis/azimuth design. Various types
of parametric daia were generated during the communications geometry
study and are summarized below.

Subsection 6. Z consists of a collection of curves giving range and

range-rate for. the various links. These results are of sufficient accuracy
that they can be used for preliminary design studies. Subsection 6.3 provides
a mathematical analysis for determination of the line of sight angles between
tvo earth-orbiting spacecraft. Although the results are applicable to Station/
ALS communication, the application of special interest is-the Space Station
to DRS link. The analysis permits determination of the required motions of
a Space Station antenna which must track the DRS. Subsection 6.4 describes
the geometrical considerations involved in placement of an antenna on the
Space Station which is to be used for DRS communication. Subsection 6. 5
considers the problem of predicting DRS occultation and of anticipating the
need for handover. Subsection 6. 6 considers the basic Space Station high
gain antenna pointing, problem. . Both two and three gimbal approaches are
discussed. The final subsection gives an analysis olf the baseline three-
Pimlbal antenna positioning system.

1. 2 PRELIMINARY RANGE AND DOPPLER SHIFT DATA

Figures 6-1 through 6-6 present the maximum range and doppler
"lift data for the various communication link geometries that may be
1"1,cott-ntered by low earth orbit manned spacecraft. The data is based on
;rcular orbits and simple geometrical relationships, but is useful for pre­
"i n' ry Link analyses. The equations used to produce these curves are
: flflarized briefly below.

6-1
 25.0

. 24.5

24.0

22.5

S23.50_____

0 100 200 300 400 500 600 700 800 900 1000

CIRCULAR ORBIT ALTITUDE, NAUTICAL MILES

Figure 6-1. Maximum Range Between Space Station and

Synchronous Data Relay

2.68

2.60

I2.52 ____ ____

2.44

2.36 _+--._

2.28

2.20
0 100 200 3D0 400 500 600 730 830 900 I0o

CIRCULAR ORBIT ALTITUDE, NAUTICAL MILES

Figure 6-Z. Maximum Percentage Doppler Shift Between Space

Station and Synchronous Data Relay

6-z
 ELEVATION ANGLE =0­
z
44
Z 1500

0 100 2000300 400 5_, 60 00 F,0 0 e_

5W

2.3

9" 20 __2___
0 0 100 200 300 400 500 600 700 &30 900 1000

CIRCULARORBIT ALITUDE, NAUTICAL MILES

C
Figure 6-3. Maximum Range Between Space Station and
S0 Ground Station
2.46

2.3 __________

2.1 _ _ _ _- _ _ _

.0

ELEVATION
NANGLE =10- 0

0 100 200 300 Soo50 600 700 800 M0 1010

CIP.CUL/R Q'SIT ALTITUDE, NAUITICAL VILES

Figure 6-4. M,4aximum Percentage Doppler Shift Between Space

Station and Ground Station (Equatorial Orbit)

6-3
 2.6

2.5

2.4

2.3 _ _ _

ELEVATION ANGLE O

2.0

1.9
0 100 200 300 40 5"o 600 700 800 900 1000

CIRCULAR
ORBIT ALTITUDE,NAUTICAL MILES

Figure 6-5. Maximum Percentage Doppler Sh4If Between Space

Station and Ground Station (Polar Orbit)

3000 1 R
EPACTSTATION =
OCAOT
ALTITUDE, .C 300

2250

: 25O

O 2000

:2
x 1500

50 ISO 150 00 250

ALSCIKCULAP ORBIT AITITUDE, NAUTICAL MILES

Figure 6-6. Maximum Range Betveen Space Station and ALS

6-4
* -tce Station Data Relay Satellite Link

dr RR S - R + ZREh + h) Figure 6-1)

.-here
d = maximum range between Space Station and DRS
r
h circular orbit attitude of Space Station

RE = radius of earth (3440 n. rni.)

RDRS geocentric range of data relay

Af
tRE~
=-s ) - (See Figure 6-Z)

where

Af = maximum fractional doppler shift between Space Station and DRS

v c= velocity of body in circular orbit at earth's surface

(2. 595 x 104 fps)

c = velocity of light (9.85 x 108 fps)

Space Station - Ground Station Link

g RE+ h+ R ZEE h + RE) sin cos 6 + cos e sin 8

sin 6 = h ± RE cos 6 (See Figure 6-3)

cosb = (1 - sin z6)/2

d1:, = maximum range between Space Stotion and ground station

h = circular orbit altitudle of Space Station

6-­
 S=ground station elevation anile
RE radius of earth (3440 n. mi.')

3/Z
- RE e 1 (See Figure 6-4)
e c R~E/ CCose0

where
Af e = maximum fractional doppler shift between Space Station and
ground station for equatorial orbits.

v = tangential velocity of body at earth's equator (I. 5Z x 103 fps)

Af = C ( /RE\
_ 31 cos e (See Figure 6-5)
p C \ h tE/

where

Af = maximum fractional doppler shift between Space Station and

P ground station for polar orbits.

Space Station . ALS Link

2 /
d ZRh +h2 I/Z h
ds = (E hAL S + tLS)
+ E SS + hss)

(See Figure 6-6)

where

d = maximum range between Space Station and ALS

h SS = circular orbit altitude of Space Station

hALS = circular orbit altitude of ALS

RE = radius of earth (3440 n. mi.

6.3 MATHEMATICAL ANALYSIS OF SPACE STATION/DRS GEOMETRY

Quantitative data concerning antenna pointing geometry for the Space

Station/DRS link requires that the line of sight (LOS) between the two
vehicles be related geonmetrically to the Space Station and the antenna pqsi­
lioning mechanisms. Of interest are the look-angles with respect to coor­
dinates referenced to the Station and the requirements of an antenna position­
n, mechanism, Specifically, gimbal freedonm, total angular movement and

6-6
a!Wular rates are quantities to be determ-ined. Look-angle data is required
in making mounting decisions, and the antenna positioner requirements may
bd determined for several implementations to provide a basis for a choice.

This subsection develops mathematical expressions for look-angles,

ginbal angles, and gimbal rates. The development is general and is orien­
ted toward machine computadon. The matrices defined and equations devel­
oped may be readily implemented in a computer program, allowing variation
of attitude mode and gimbal system mechanization. The following subsection
defines coordinate systems useful in the analysis and defines transformations
between these systems. Later, notation is discussed under that heading,
and Definitions defines several vector quantities essential to the analysis.
The mathematical expressions for look-angles, gimbal angles, and gimbal
rates are developed under the subheading Analysis.

Coordinate Systems

Before analysis of antenna pointing can be initiated, coordinate sys­

tems and associated notation must be defined. There are five sets of coor­
dinates which prove useful in developing mathematical expressions for
quantities of interest.

inertial Coordinates (I 1 £2 13)

This system is shown in Figure 6-7. The origin lies at the earth's
center. I 1 coincides with the vernal equinox vector0, 13 coincides with the
earth's polar axis, and I Z lies in the equatorial plane, completing the right­
handed system.

Space Station Orbit Coordinates (SI S Z $3)

This system is also shown in Figure 6-7. S3 is perpendicular to the

orbit plane, S 1 coincides with the line of nodes and lies in the direction of
the ascending node, and S2 lies in the orbit plane, completing the right­
handed system.

A vector, V, expressed in Si S Z S3 coordinates, is denoted by V S

and can be expressed in li 1? 13 coordinates when operated on by the linear
transformation, AIS, given'by

CosX -sin X 0\ /1 0 a0
= in0 cos i -sini

0 0 1/ \ sin i cosi
(6-1)
/,Cos X -Sin X Cos i sin k. sin i A
A in ks S cos i S
cos % -cos ks sin i s
0 sin i cos if

6-7
 S3 L3

l'sl

S
T

Figure 6-7. Iihertial, Space Station Orbil

and Local Space Station
Coordinates

13

03

12
D D2

Figure 6-8. Inertia] and Data Relay

Orbit Coordinates

t) ­
Thus, denoting V in iI I z 13 coordinates by V,:

VI =Ais VS (6-2)

Data Relay Orbit Coordinates (D 1 D 2 D 3 )

This system is shown in Figure 6-8, and is' conceptually the same
as the Sj S Z S3 system. D 3 is perpendicular to the orbit plane, DI coincides
with the ascending node direction, and D 2 completes the system. The trans­
formation from D 1 D 2 D 3 coordinates to I] U
13 is denoted by AID and is the
same as Ais, but with the subscripts changes, i. e.,

cos XD -sin XD cos iD sin D siniD)

AID= (sin XD ccs)Cos i~jD Cos X D sin iD (6-3)

\
0o sin i D cos iD /

Denoting V in D 1 D 2 D 3 coordinates by VD'

VI = AID VD (6-4)

Local Space Station Coordinates (L I1 L- L3)

This system is shown in Figure 6-7, where L 3 coincides with S 3 ,

the orbit normal. L Z lies along the velocity vector, and Lj coincides with
the vector from -the earth's center toward the space station. The transforma­
tion from Li L z L 3 coordinates to Sj S Z S3 coordinates is denoted by ALS
which is expressed as the matrix

/cos0 -sine 0
A = sin 0 cos e s 0) (6-5)
0 0 i

Thus,

VS = ASL VL (6-6)

6-9
 Space Station Coordinates (X Y Z)

These coordinates have been adopted from the North American

RockveLl Space Station Phase B Definition Program and are illustrated in
Figure 6-9. Different attitude modes for the Station result in different
transforimations from the XYZ system to the L 1 L Z L 3 system, but the gen­
eral transformation will be denoted by A LX Four candidate attitude modes
are treated below.

Earth Pointing, X-POP. The X coordinate is perpendicular to the

orbit plane, Y lies along the velocity vector, (LZ), and Z coincides with the
radius vector, (L 1 ). Thus,

ALx 1 0 (6-7)
LXI 0 0

and

V L = ALX 1 V x (6-8)

where V denotes a vector in the XYZ coordinates.

Earth Pointing. Y-POP. X lies along the velocity vector, Y is per­

pendicular to the orbit plane, and Z coincides with the radius vector.

LXZ 0 (6-9)

Inertial, X-POP. The Space Station's attitude is held constant in

inertial space with the X-axis perpendicular to the orbit plane.

/0 sin - 0s) cos ( - 0s)\

ALX 3 - ( -cos (0, Ose) sin (60 - elS 1 (6-10)

o o /

6-10
T.'here (60 is the angle between Z and 'I at e - 0 (where L1 = Si) in the
direction of orbital angular motion. Note that the transformation from
xrZ coordinates to S1 S Z S 3 coordinates is given by

/0 sin cos Eo
10 01
ASX3 =A SL A LX3 = 0 -Cos 6 sin 00(6-1-1)

Inertial, Y-POP. The Space Station attitude is held constant in

inertial space with the Y-axis perpendicular to the orbit plane.

A-sin (E - as ) 0 cos (80-

AkLX 4 Cos (E° - Es ) 0 sin (e0 s). (6-1Z)

0 .1 0

and

/-Sin a 0 cos

ASX 4 ASL ALX 4 cos ° 0 sinO (6-13

0 i 0!

where 60 is the angle between L1 and Z at 6 = 0 in the direction of orbital

angular motion.

Gimbal Reference Coordinates .(GI G 2 G3). In a two-gimbal system,

one axis is fixed rigidly to the vehicle. This axis is calle'd the primary
axis. A second rotating mechanism is attached to it, usually in an orthogonal
arrangement." This second axis of rotation, called the secondary axis, is
positioned by the prim-ary axis. The gimbal reference system is a coordin­
ate system representing the "zero" position of the two rotating axes. This
system may be referenced go the Space Station coordinate system (XYZ) or
the orbital coordinate system (Sj S? S3). That is, a vector V in the G G
03 system is expressed in XYZ or Si SZ S3 coordinates by transformations
AXG or ASG, respectively. These matrices will be further defined in the
subsequent analysis.

The angular orientation of the antenna boresight can be easily refer­

enced to the G 1 G2 G3 system. Referring to Figure 6-10, the primary axis

6-11
(TOWARDEARTH)

Figure 6-9. Space Station Coordinates

G3

(ANTENNA 92
BORESIGHT) a

SECONDARY

/ 02

Figure 6-10. Gimbal System Coordinates

6-12
 I

coincides with G1 , and in the zero position, the secondary axis and antenna
boresight coincide with G? and G3, respectively. Denoting angular rotation
about the primary and secondary axes by gl and g 2 , respectively, the
antenna boresight unit vector, B, is expressed in G1 02G 3 coordinates by

B - ( sing Cos g (6-14)

cos g 1 cos g 2 /

Notation

As may already be noted, letter subscripts refer to coordinate sys­

tems. Matrix transformations have two subscripts, the first referring to
the coordinate system to which the vector is being transformed and the sec­
ond referring to the coordinate system from which the vector is being trans­
formed. Thus, ASG represents a transformation from G G G 3 coordinates
to S1 S S3 coordinates.

I is the inverse of A 3 0 , and it can be shown that

AG AT =A (6-15)

where AT is the transpose of the matrix ASG. This property is true of all
the transformations defined earlier because they are orthonormal
transformations.

Components of a vector are denoted by a number in addition to the

coordinate system letter subscript. Thus, VSZ is the component of the
vector V along the S2 coordinate axis. Since the magnitude of a vector is
simply a number and does not depend on any coordinate system, it will be
denoted by the letter used for the vector but with no subscript. Thus,
IVs I = IVI = V. Also, a vector not referred to a coordinate system will
have a "bar" above it, e.g., V.

Definitions

The vectors to the Space Station and DRS are denoted byS and D,
respectively, and are defined in SI S2 S 3 coordinates by

6-13
 D3

2C

SPACESTATION

D
2

DRSt' !.D
PLAN9

I
SPACE '
ORBITPLtN ;
N
I/
oD-

DRS

Figu z z-11. Range Vector and Space

Station/DRS Geometry

6-14
 S =A (6-16)
D0 /
/cos eD)

0/

These two vectors are shown in Figures 6-7 and 6-8, respectively. Their
derivatives are given by
t

SS Ws S Cos 0 s (6-18)

0 e)
/ -s
n a1D \

DD = DD (cos D) (6-19)

do de
where s -dt an dt

The range vector is defined to be directed from the Space Station to

the iRS and is denoted by R. The geometry is illustrated in Figure 6-11.
Thus,

R fD - S (6-Z0)

Adalysis

The analysis of antenna pointing amounts to expressing the range

vector in appropriate coordinate systems for determining look angles, gim­
bal angles, etc. In particular, the range vector must be expressed in Space
Station coordinates (XYZ) and gimbal reference coordinates. This is accom­
plished using the transformations defined under Coordinate Systems and the
definitions under Definitions, specifically Equations 6-16, 6-17, and 6-20.

6-15
 The range vector in inertial coordinates is given by

R I = DI -S I AID DD - AIS SS (6-2)

Then, using property (Equation 6-I5),

RS = SI 1 -TI=i
AI I

RS DS
is AIDDD-
R S =AIS A(6-2Z)

The magnitude of R is defined by

RZ R =TR T TT

D T ID ATT
ID D +S T
AisAis S TA IS SS
- DDAID
T T A(6-Z 3)
S iS ID D

I2 ID +Sz - 2 [iS S) T (AID DD)]

The angle between D and S is denoted by V and is shown in Figure 6-11. By

the law of cosines, it can be shown that

(AsSs)T (ADD )

cosp = 1iA S S) ID D (6-24)

Space Station Look Angles

The range vector can be expressed in spacecraft coordinates by

transforming RS:

i X = AXL ALS RS

Wli~=e A=T andA A

SL LX LS SL.

6-16
 Substituting from Equation 6-ZZ

RX= A T TD
LX A SL is ID -( ST) (6-Z5)

Define the unit vector in the R direction, expressed in XYZ coordin­

\t
,by

R ux = R (6-26)

"cin the X component of RUX is the cosine of the angle, aX, between the
. axis and R. Denoting the direction cosines by aX, cy, and aZ

cos aX = RUxi

Cos a - RUXz (6-27)

cos a = RUX 3

Calculation of these angles for different attitude modes will pr6vide useful

tor making mounting choices and studying structural interference.

Gimbal Angles

For proper communication, the antenna boresight axis, B, must

nominally coincide with the range vector. Transforming H to G
coordinates

R -AT RX
G XG

=A T AT AT A A D -Rfl (6-28)
R DD
G XG LX -SL IS ID
HG
R
=AXG
TAASX
TAT A
isIDDD D -R)
o':nc B is a unit vector, correct antenna pointing requires that

6-17
 I
fl-a i" sin g.z )

(cs
BG- -(sin g1 Cos gz (6-29)

g 1 cos gz

RGZ
(6-30)
Rg'G 3

RGl
tan -
IG3
R cos g (6-31)
5a

'sr equivalently

g= tan-1 -R G 2
aG31
R (6-32)

2= tansta (6-33)
RYG3 Cos y
11

iT gZ is expected to lie between +90 and -90 degrees, then from Equation 6-24

gz=sin- RI (6-34)

I he rate of change of the gimbal angles may be derived from Equations 6-32
6-33
t above.

1l GZ 'G3 - ' G 2 (6-35)

dt -g 1 = 2 6
RGz + RG 3

6-18
 dg z (RG Rb 3 - R G34 cos - g, RGI R sin
=t g = R z-G+ (RG Cos gl ) Z (6-36)

xnere the derivative of is given by

A T AT (ATA D
,G XG SXD - S
+T "T (6-37)
T
AXG AsX (AT AID DD -SS)

where i)D and are given by Equations 6-18 and 6-19.

ASX = ASL ALX

Thus

rp T T
ASX A LX 'SL

and so

XI AIx A T+ TX ATL (6-38)

For earth-pointing modes, ALX is constant; thus,

ASXLX
z&= SL
AT AT (earth pointing modes) (6-39)

and from Equation 6-5

(-sin e -cos e s 0)

ASL = Uo
s , cos 8s -sine 0 (6-40)
0 0 0/

6-19
 x

A) INERTIAL MODES" EACH-ANTENNA COVERS A HEMISPHERE 2

(X-POP, Y-POP, ARTIFICIAL G)

ANTENNAS ANTENNAS

SOLAR PANELS

-h~~~
--- - -- -- -- -- -- -- ­

I SOLAR PANELS
z
POINTS TOWARD EARTH
B) LOCAL VERTICAL MODES- BOTH ANTENNAS COVER SAME HEMISPHERE
(X-POP, Y-POP)

Figure 6-12. Mounting of Two DRS Link Antennas

6-20
 For inertial modes, ASx is constant, and so

G -- AuT G AASx (is AID D

T (D SS (inertial nodes) (6-41)

6.4 ANTENNA PLACEMENT

Introduction

The discussion below is directed toward the factors determining the

proper placement of the DRS high gain antennas on the Space Station. Many
of the results also will aid in locating omni antennas used for DRS communi­
cation or for links direct to ground:

The relay Link antennas should be placed so as to accomplish the

following:

1) Permit stowage at the fore or aft end of the Station during

launch,

2) Mininize obstruction of the LOS by the Station structure

3) Minimize the number of handovers and time required for each

handover (and therefore, minimize data loss).

Antenna Placement Schemes

One Antenna

If only one relay link antenna were to be used, it would be placed so

as to look in the -Z direction relative to the Station. In the local vertical
attitude modes, this direction is along the local vertical away from the
earth, thus providing good coverage of the DRS System (DRSS). In the
artificial-g mode, this single antenna could provide coverage to the DRSS on
one side of the spin plane. For the inertial modes, there is no preferred
LOS direction to the DRSS relative to the spacecraft. Therefore, there is
also no preferred location for the antennas.

Two Antennas

Figure 6-12 shows two different schemes for mounting two antennas,
One is designed for the inertial attitude modes and the other is intended to
, used with an earth-orien d spacecraft.

In the inertial modes, complete spherical coverage must be pro­

A4:t i, consequently, the antennas are placed 180 degrees apart. They are
out along the i-Z and -Z axes to perinit their azn-uth gimbals to
o.trl

6-Z1
 In STATION' SIM .- ! 55'
X-POP, LOCAL VEOTICAL

D0SS ONE EQUAIOIAL DRS

I \ I

'C ...... -. 4----.. -- ­

­
"ToPc To x
0 . PEC EC TEOPEC <-/ TO - AXI

060 20 IT ESP0C 120 140 160 IS

DIRECTION COSINE ANGLE, DEGREES

Figure 6-13. Distribution of Direction Angles

X LOS

Y
°
STAfION 270nr , 55 C
X-POP, LOCAL VERTICAL
DRSS _ ONE EQUATORIAL DRS / .

0 MEASURED
IN Y-Z PLANE FROM -Z
O MEASURED
FaOM X

135 .

45

0 45 10 13 IS 225 270 315 360

PHI, 0), DEGREES

Figure 6-14. Line of Sight Angles Gand _Required

for Uninterrupted Communication with a Single DRS.

G-22
 .. tne antennas when the Station operates in the artificial gravity mode.
pin
'.-: that with this two-antenna configuration, a handover is required when­
-,r the LOS to the visible DRS passes between the hemispheres seen by the
nWo antehnas. Furthermore, if the antennas are separated by 180 degrees,
!A C olar panels may block both at the same time because the two solar
;.rincls are co-planar and are gimbaled about two axes.
If the Station were only to be used in the local vertical modes, the
antennas should be mounted out away from the earth. In this configuration,
;n antenna tracks a particular DRS from h6rizon to horizon. Handover is
only required when a DRS becomes occulted by the earth. Assuming that
the deployment of the DRSS is such that periods of dual coverage will exist,
the two antennas can provide essentially uninterrupted communication.
While one antenna is tracking a DRS, the other antenna is automatically
positioned toward the DRS just appearing above the horizon. Note that the
antenna masts are not oriented exactly in the -Z direction and that they are
on opposite ends of the Station. In the X-POP local vertical mode, the
antennas would interfere with one another if placed on the same end of the
Station. If they were both out in the -Z direction, problems might also
arise in the Y-POP LV mode when the LOS is generally along the length of
the spacecraft. Mounting the antennas on opposite ends of the space station
also gives the system the capability for "looking" past rendezvousing
vehicles such as the EM I or ALS.

Baseline Design: Three Antennas

The Space Station may be used in any of the various attitude modes
mentioned above according to the latest mission plans of NASA. The place­
ment of three antennas in the manner suggested in the baseline design com­
bines the desirable characteristics of the two schemes described just above.
As shown in Figure 3-2 the three antennas are placed 120 degrees apart
about the spacecraft. The two antennas on the -Z side of the spacecraft are
used during the local vertical modes. The two antennas on the forward end
of the station are used during artificia gravity operations. They have their
azimuth axes oriented parallel to the Z,-direction so that the gimbal system
can remove the artificial-g spin by a counter-rotation about this axis. All
three antennas can be used during one of the other inertial modes when the
spacecraft is not spinning.

Computer Results

In the local vertical attitude modes, the LOS to the DRS is generally
v'aay from the earth. " To quantitatively describe the possible pointing
tarections of an antenna mounted on the Space Station, a computer program
developed based on the analysis of Subsection 6. 3. The program com-
LOS angles to a DRS relative to Space Station coordinates. The pro­
.tm requires inputs on the station's attitude mode and its orbital
• ranleiers. Orbital parameters of the DRSS are also required. Fig­
(.-13 and 6-14 are outputs from the "relative frenuency" version of the

6-23
 ;e.,,,ram. This version calculates the aImount of time gimbal angles and
.r rates can be expected to occur during a long duration mission. The
, ivC frequency of a quantity is the ratio of this occurrence time to the
mission time.
m.-

Figure 6-13 shows a plot of the relative occurrence of a LOS direc­

tin defined by direction angles relative to the Station's X, Y, and Z axes.
t•e curves correspond to a cumulative probability function -- that is, the
,rdinate gives the relative frequency of occurrence of a value less than or
• iual to the abscissa. For the Station's attitude shown, - the earth is always
'dcvwn" in the +Z direction and the Station's XY plane is always at 90 degrees
to this reference. Note that the plot indicates that the LOS can approach to
within 68 degrees of the +Z direction or 22 degrees below the XY plane. An
anena on the Stationis never required to look closer than Z6 degrees to
cither the +X or -X directions.

Figure 6-14 contains the same basic data as is shown in the previous
'aure. The data here is in the form of a "scatter plot" on a 0 and grid.
'ihe areas delineated are those portions of space, relative to the Station, in
which a DRS may be found. For a more complete analysis, this plot could
bu: used in conjunction with a program for determination of blockage by the
:;pacecraft structure. Figure 6-14 shows the LOS angles required for
uninterrupted communication when the DRS's are above the horizon. The
additional program would determine those LOS angles which result in an
'uiobstructed view by the antenna. In Figures 6-13 and 6-14, only one DRS
,,asassumed because this -constitutes a "worst case" situation. The required
coverage is significantly less when utilizing a DRSS with many satellites.

o.5 DRS HANDOVER GEOMETRY

introduction

Communication with a particular DRS must be discontinued whenever

the LOS is occulted by the earth (or effectively occulted by the earth's
ittnosphere). DRS handovers may also be desired for operational reasons
( .uchas to permit DRSS access by multiple users).

Dotermination of Effective Occultation

Figure 6-15 shows the algorithm included in the Space Station/DRSS

(;-O'-netry Program for deterinination of effective occultation. The DRS is
.t_:,itned to be effectively occulted when the LOS vector passes within an
Itdltude, ha, of the earth. This parameter is determined by the atmospheric
,'fl,:cts and CCIR requirements on RFI that exist at the link frequency. At
1-1fncies above S-band, the signal will be severely attenuated if it passes
Ili rain storms in the lower atmosphere. At frequencies below K-band,
COIR h1as placed restrictions on the power density which may graze the
- ,.rh at various angles. (See Figure 6-16. ) At X-band, for example, the
,.' r censity is limited because terrestrial micro\vave relay syste-ms util-
S :" ,: frequencies and could be jammed if thu satellite's antc:noa beam
....4 U: ~afrth. This condition can be prevented by stipulating thit the

6-2!
-OS (and consequently, the main beam) 'should pass no closer than h a miles
frt'n the surface of the earth (see Figure 6-17).

At the 14-GHz band chosen for the Space Station baseline design, no
coJla requirement currently exists. An h a = 10 n. ni.'was chosen in order
that the beam will pass above the most severe weather. This value also
allows for anomalies of the nonspherical earth.

Determination of Handover Times

DRSS deployment may provide for dual coverage of certain portions

of the user's orbit. Handover from DRS to DRS can occur at any time dur-
Ing this dual coverage period. Handover initiation near the beginning of this
period permits the most time -for antenna and receiver acquisition. How­
ever, during the middle of the dual coverage period, the LOS geometry to
the two satellites is approximately the same and the LOS is well above the
horizon. In a "late" handover, communication with the "setting" DRS must
be maintained down very close to the effective atmosphere limit. Determin­
ation of handover times will ultimately be fixed by operational needs. A
particular DRSS user may require uninterrupted service during certain
periods when performing important experiments.

The first step in developing an operational DRSS user schedule is a

computer analysis of the geometry for the time period of interest (perhaps
24 hours). The Space Station/DRSS Communications Geometry Program
developed during the study' performs these basic calculations for space­
craft in circular orbits. Two mathematical algorithms for handover sched­
uling were written for this program. The first algorithm causes communi­
cation to continue until just before the DRS is occulted; the next DRS utilized
is the first vehicle to the east which is visible. This algorithm is shown
pictorially in Figure 6-18. The second algorithm causes communication to
be passed to whatever DRS is closest to the user spacecraft. Figure 6-19
shows the use of the algorithm for an equatorial space station orbit. For
polar orbits, the "shortest-range" algorithm is a poor choice. Figure 6-20
shows that this scheme can cause two unnecessary handovers during that
period in which the user passes through the triple coverage region at the
poles.

6.6 ANTENNA POINTING GEOMETRY

Introduction

Articulation of an antenna for DRS communication requires a

lntChanical system with at least two degrees of freedom. The discussion
tJlQuw centers on the possible use of a two-gimbal system. This discussion
hn basis of the rationale for selection of a three-gimbal design. The
lilu three-gimbal system is discussed in Subsection 6. 7.

6-z5
 DATARELA
ORI
SATELLITE
RANGE MAX

OREIT

STATIN ORBIT

REFF HEARTH hATMOSPHERE

R

h = FUNCTION OF ATMOSPHERIC EFFECTS

ON RFI
AND CCIR REQUIREMENTS

1I =REART hD
H

h ORBIT
= ALTITUDE OF RELAYSATELLITE
1Is1 REARTH+ S

I S ALTITUDE OF STATION'S ORBIT

1
MAX EF - CONSTANT

WHEN THERANGE > RANGEMAX THEDRSIS OCCULTED.

Figure 6-15. Determination of Effective

Occultation by Earth and the
Effective Atmosphere
Assumes circular orbits and
spherical earth

LOS TO DRS
SPACESTATION ORBE ATION
LOS TO DS

MICROWAVE LINK
TER3ESIPIAL

CCIRREQUIREMENTS AT 4 GHz
MAXFIMUMINCIDENT POWErR PER
SQUAREMETERPER4 kH
-5IS Lo~dB,
/ j
15/

rB..re 6-16. CCIR Requirements Figure 6-17. Effect of CCIR Requirements

on RFI on Parameter h a

6-Z6
 TIME SEQUENCE
Al INITIATE
COMMUNICATION
WITHIFIRST':;I
,STO THE AI Figure 6-18. DRS land.j.,r
IINTIL 0~g ' I5 Nigie . I - Sae t,.*.
':LATLY CC 'L o7I-bac tD!'~
.%

HNNDO ER TO FI-SI Equa to ia] Orbit

VISIBLEDS TO THE
EAST
C' HANOOVER TO NE\T
D'S TO THEEST
m)S 2 D) HAM-OOVERTO N!/T
DRS10 THEEAST

TIME SEQUENCE
A) INITIATE
COMMUNICATION
VITH THENEAPEST Figure 6-19. 2 -DRS l-fandover Alcorxt
VISIBLEDRS INo. Space Station in
NEAREST
DHS Equatorial Orbit
C)HANDOVER TO
NEARESTDS
D) HANDOVER TO

DRS2 NEARESTDRS

TIMESEQUENCE
S - COI MUIICATION
WITH LTHENIIFAREST
"ISIP E Dr'S
H'B
SHA4F0 EU Figure 6-20. DRS I-andover A.
No. ? - SDace Statio i:.
' c%.C''.
*-C- HA',DAro~ TOPolar
HA,TDO/[p.TO Orbit

1,7
 SECOND __Y __ SECON APY S

-PRIMAR&Y PRIM&ARYlir

//y//2#//2/ ,// /4,/.. ///

II TWO-GIMBAL AXES b) THREE-GIMEAL AXES

Figure 6-2l. Gimbal System Schemes

LOS
PLANE NORMAL TO gl

SPL&NI NORMAL
/S
TO LOS,
NORMAAL
TO tOS

NORMALTO0V,

LOS

II- -PARALLEL TO LOS

"/-P TO LOS
i A LEL

Figure 6-zZ. Geometric Determination Figure 6-23. Primary Gimbal Axis

of Gimbal Rates Rate Increase When LOS and
Primary Axis are in
Close Proximity

6-28
1pwo-Gimbai Versus Three-Gimbal Designs

To mechanically orient an antenna with respect to the spacecraft,

there must be one axis of rotation fixed rigidly to the spacecraft and one
a.xis fixed rigidly to the antenna. For a positioner with only two rotational
axes, the axis fixed to the spacecraft will be called the primary axis, and
the axis fixed to the antenna will be called the secondary axis. If there are
three axes of rotation, the axis attached to the antenna is again called the
secondary.axis, and the axis that orients the secondary axis is again called
the primary. The axis that is attached to the spacecraft and orients the
primary axis will be denoted here as the third axis. Thus, for the three­
gimbal system, the primary axis is the middle axis that connects the two
axes which are attached to the spacecraft and the antenna. These two
schemes are illustrated in Figure 6-21. Wherever possible, the two­
gimbal scheme is used because of its simplicity.

Theoretically, an antenna can be pointed in any direction if it has

two connected axes of rotation -- one fixed in inertial space (the primary)
and the other fixed to the antenna (the secondary). One heuristic argument
supporting this statement is based on the fact that rotational motion about the
axis defined by the LOS is unnecessary and/or unimportant; hence, only two
other degrees of rotational motion are required. The major problem with a
tvo-gimbal system is that the angular rate about the primary axis can be
very large. To see how and why this occurs, consider the following heuris­
tic discussion.

Denote the primary axis of an orthogonal two-gimbal system by gl

and the secondary axis by . Orthogonal means that gl is perpendicular to
gZ. Let the antenna pointing direction be perpendicular to gZ. Such a con­
figuration is shown in Figure 6-Zla. Since the pointing direction must coin­
cide with the LOS to the target, gZ lies in the plane normal to the LOS, but
also must lie in the plane normal to g, because of the orthogonal design.
Thus, the rotation of gz about gl must be such as to make gZ coincide with
the intersection of these two planes. But, in addition to this positioning
requirement about gl the relative ang&lar rate of the LOS with respect to
the body to which gl is fixed must be c6mpensated.

It was mentioned previously that rotational motion about the LOS is

unimportant. Thus, the angular rates, gl and mZ,
must compensate for the
component of the LOS angular rate, £ , which is normal to the LOS. As just
vxplained, g 2 is defined by gl and the LOS. And consistent with the above
argument, l and are defined by the projection of the Q vector along the
LOS onto the g, - gZ plane. This geometric interpretation is illustrated in
Figure 6-2Z. The gimballing problem occurs when the LOS and gl coincide
nr nearly coincide. Figure 6-Z3 indicates geometrically when g, becomes
vury large; namely, the projection of the LOS angular rate, 0, along the
O onto the g, - gz plane results in a large vajue of gl" Theoretically,
'.lvn the LOS and gl coincide, gl must be infinite in order to point the
.. u'n:'a at the target.

6-Z9
 AZ. P.AZ

EL

Figure 6-24. Two-Gimbal System on Mast of X-POP Station

63O
Use of Two-Gimbal Designs

Two-gimbal antenna pointing systems can be used successfully under

one or more of the following conditions:
1) The gimbal system can be oriented such that the LOS does not
pass near the zenith -- that is, the primary or azimuth axis.

2) The system can be designed to track at rates great enough to

maintain the target within the beamwidth of the antenna when
encountering a zenith condition.

3) Short communications outages can be tolerated when antenna

system pointing lags behind the motion of the LOS.

For the Space Station mission, it was assumed that communication

outages (condition 3) would not be tolerated. It was also assumed that con­
dition 2 did not apply to the Space Station DRS antenna. During the design
phase, narrowing the beamwidth would necessitate servo system redesign
to achieve higher tracking rates. Furthermore, the excessive motions of
a two-gimbal system during a zenith pass could shorten gimbal life; the
system must opeiate for more than 5 years.

It was discovered that for certain phases of the mission, a two-gimbal

system could be oriented such that the LOS to the DRSS would never be near
the system zenith. Ther.efore, condition 1 could be met. Figure 6-24 shows
how an AZ/EL system might be mounted on a Station used in an X-POP
(X-axis perpendicular to the orbit plane) mode, Figure 6-25 shows pictori­
ally (that with certain inclinations) that this arrangement prevents coincidence
of the LOS and the primary axis. Figure 6-26 shows the maximum declina­
tion oi the celestial sphere of the LOS to the DRSS and (in dashed lines) the
declination of the orientation of the primary gimbal axis. This is a worst
case geometrical condition. Note that for 80 degrees s<i < 100 degrees, the
LOS can pass through the gimbal system zenith. However, the baseline
inclination of the Space Station's orbit has been set at 55 degrees. There­
fore, a two-gimbal system with the primary axis perpendicular to the orbit
plane can be used with the Station in its baseline orbit.

When the station is in the local vertical mode, the principal motion
of the azimuth axis is a 1 rev/orbit rotation to remove the earth-pointing
rotation of the spacecraft. The antenna motions required to "track-out"
Station and DRS orbital motions are very slow. The required motions were
dttermined quantitatively by the use of the computer program discussed pre­
V",°usly. Figure 6-27 shows a histogram of the required azimuth and eleva­
ion angular rates. (The curves correspond to probability density functions.
The computer plot shows that the azimuth rate is centered about 1 rev/orbit
3(0 deg/90 min. = 4 deg/min. The azinuth raLe is always between -4. 7
and -2. 8 deg/nin. The elevation rate is between -0. 6 and +0. 6 deg/min.

6-31
 ORIENTATION

Figure 6-Z5. LOS/Primary Axis Geometry - Primary

Axis Perpendicular to the Orbit Plane

20r

o If _______

z DECLINATION OF PRIMARY AXIS

VMICIH IS ORIENTED POP
C// \ (PERPENDICULAR TO ORBIT PLANE)

.MAXIMUM DECLINATION OF LOS I MAXIMUM DECLINATION OF LOS,

/(FOR 80-S- 5S'0 THE LOS CAN

EXACTLY COINCI DE V.TK WE AXIS)
0 0 60 90 120 150 I8O

INCLINATION OF 300 N MI CIRCULAR O'BIT, DEGREES

TI.u rec 6-26. rorst Case LOS/Primary Axis Geometry - Primary

Axis Perpendicular to the Orbit Plane

6-32
 z'''
'L~AZIMUTH ­
I -,

-- -- - ELEVATION
-.STATION- 270 . i 55
X-POP, LOCAL VERTICAL
GIMBAL SYSTEM PRIMARY ALONG X
0.2 - D5Ss S
ONE EGUATORIAL DR5
VISIBILITY" h. IOn , H ONLY PEALIZABLE
INCLUDED
QEO ILTRIES
- -- I i -_'_I _

0.05 _

p 0 02 ...... j1

0.01

0.002 ­

0.01 -.
-10 -8 -6 -4 -2 0 2 4 6

PERMINUTE
GIMBAL ANGLE RATE, DEGREES

Figure 6-Z7. Gimbal Rate Data

0.8

t//

<0.6 ___

o 0.4

,- -- ELEVATION
'C

0.2 -Y-POP, STATION: 270n .m,55

LOCAL VERTICAL
GIMBALSYSTEM, PRIAy ALONG X
CUSS: ONE EQUATORIAL IMSE
=
VISIBILITY. h. 10 n -, WITHON LY REALIZABLE
IGR?mfTRICS$ NCUDED
0 2 4 6 8 10

PE? MTI*fl:
MAGNITUDE OF GIMBAL ANGILR RATE,DEGPRES

Figure 6-28. Required Gimbal Rates

6-33
 L/L
$TAIION

DRS

Figure 6-Z9. Girnbl Coordinates

G3

Figure 6-30. Definition of Angles aand

6-. 3t
 It has been shown that a two-gimbal system can be mounted on an
A-'POP Station with an orbit inclination of 55 degrees such that a zenith con­
,;tion never-occurs. However, if the Station can maneuver into other atti­
tude mode's, this preferred geometry may no longer exist. Figure 6-28
ehovs a computer-generated "cumulative probability density" of the required
atrnular rates assuming a change to the Y-POP local vertical mode.
Decause the antenna must track through its zenith, the azimuth rate is no
longer Limited to 4. 6 deg/min. Approximately 5 percent of the time, the
azimuth rate must be greater than 10 deg/min in order for the boresight to
be kept coincident with the LOS.

Selection of a Three-Gimbal Design

The two-gimbal system discusEed above was considered undesirable

because itcould only be used ove+ a limited range of the possible mission
attitude modes (and it could also be unsatisfactory on a Station placed in a
polar orbit). The three-gimbal system discussed in the following section,
can be used with any Space Station orbit or attitude mode. This particular
design was selected for the baseline system.

6. 7 ANALYSIS OF BASELINE GIMBAL SYSTEM

For the three-gimbal system shown in Figure 3-2, a history of the

gimbal motion is useful for determining: 1) the limitations of the system,
Z) the logic requiredl for three-axis control, and 3) understanding handover.
The following analysis does not consider the dynamics of the control system,
but only the basic control logic and the Starion/DRSS geometry.

Gimbal Coordinate System

Figure 6-Z9 illustrates a coordinate system for a positioning system

mounted on an earth-oriented vehicle. The coordinate unit vector, G I , is
parallel to the local vertical and points away from the earth's center. G 2 is
perpendicular to the orbit plane in the direction of the orbit angular velocity;
G is parallel to, but in the opposite direction of, the velocity vector.

The time-varying vector, R, from the Station to the DRS, can be

expressed in these coordinates as a function of orbital parameters as is
shown in Subsection 6. 3. In Figure 6-30, the angles a and P define the
direction of R. with respect to the G 1 G? G 3 coordinates,

The angular position of the gimbals is defined by angles gl g2, and

P.,, where "I" refers to the azimuth axis (A), "Z" refers to the elevation
.."'s (E), and "3" refers to the cross-axis (C). Note that gl' g 2 , and g3 are
tflt necessarily simply related to the angles a and P. The zero points of the
U.rce angles are defined such that gl = 0 when the elevation axis coincides
ti h%G Z (the orbit normal), g? = 0 when the antenna boresight is parallel to
C;! - G 3 plane (local horizontal), and g 3 = 0 when the boresight is per­
1: ; arQ to
- cular the elevation axis. The baseline maechanical limits on these
as follows­

-180-5 gl < 180 degrees (unlimited motion)

6-35
 1
-20 :s gz - 110 degreds

-20 : g 3 < 20 degrees

Positioning Logic

Gimbal angle motion depends on the time history of the vector, R,

with respect to the G 1 G Z G 3 coordinate system and the control system
logic, which splits the required angular motion into the three gimbal angles.
There are three positioner modes with the logic proposed: 1) azimuth­
elevation mode, 2) X-Y mode, and 3) transient mode.

Azimuth-Elevation Mode

As an azimuth-elevation mount (g 3 = 0) approaches zenith, the

azimuth angle rate of change required for perfect tracking approaches
infinity. To prevent tracking errors at zenith, the proposed LAS gimbal
operates as an az-el mount for low elevation angles and as an X-Y mount
for high elevation angles, If a satellite orientation and altitude is assumed,
an elevation angle, E, can be chosen below which the maximum required
azimuth rate will be within the slewing capability of the azimuth drive.
Then, referringto Figure 6-30, when -20 degrees : <E 1 , the p'ositioner
behaves as a typical az-el mount with

gl =a, = I and 03
0

It was in Subsection 6. 3 that

9, = tan-1I - G2) =

g- tan- I RG; cos g) = sin- (R= 1 )

93 = 0

G2 G3 -2 G3

GZ G3

6-36
 kG1 I G 3 - RGI nG 3) cos gl - 'I GI IG 3 sin g,
= z 2
G3 + \R 1 cos g)

3= 0

For the proposed system, E 1 = 80 degrees.

X-Y Mode

In the X-Y mode, the elevation axis and cross-axis are used to
track the DRS. When the elevation angle exceeds E, the azimuth angle, gl,
ceases tracking, and remains at the value, ao, which it attained at g = El .
And, it remains at this position for as long as the gimbals are in theh-Y 1
mode. To prevent the elevation and cross-axis gimbals from hitting their
stops, thresholds are set on the angles gZ and g 3 ' which are sufficiently far
below the mechanical limits that a return to az-el mode can be effected
without reachingthe limits. The proposed thresholds are 95 and 10 degrees,
respectively. Thus, for

800< g 2 < 950 and -J00< g 3 <100

the positioner behaves as an X-Y mount. It can be shown that the gimbal
angles and gimbal angle rates are given by the following equations:

g1 =" 'o (6-42)

9?= tan- 1 H (6-43)

sn= I (RH 2 (6-44)

= 0

6-37
 fi3 F. - R F.
= H
- R-13 H i 1-13 (6-46,)
R23 + R 2
HZ Hi

(N11 3 RH - RIH3 RH 2 ) g 2 + Z RHZ FH

Cos 3 sing 2 (6-47)
3 2 2 &
Cos7co ­
113 + R HZ s

RH = AHG R G (6-48)

AH to cos g sing, (6-49)

-sin g1

where a is the value of g, when g2 reached El.

Transient Mode

When the gimbal system is in the X-Y mode, and the thresholds for
92 or g 3 are reached, the azimuth axis is reactivated. Typically, g, will
be different from e, and this error must be nulled while g 2 and g 3 adjust
to maintain the boresight unit vector, B, equal to Il/i.. When gl = a, the
gimbal system resumes the az-el mode. Figure 6-31 illustrates a simpli­
fied control system error response that is useful for this preliminary anal­
ysis. This response amounts to a constant rate slew of the azimuth axis in
the proper direction to minimize the error nulling time. The baseline value
for r, the azimuth slewing rate, is 36 deg/min. For this mode, the gimbal
t-ngles and rates can be found from Equations 6-42 through 6-49 with the
following azimuth gimbal rate

r - gl ) < 0 or >180'

g1= t0r (a- gl)1=

(Ce- ) 0
gl > 0 or < -180'
(6-50)

.. rIormning the computation, g, is found from j, and the initial value,

*i - --a0, where t = 0 vhen starting the transient rode.

6-38
 o
-]B

Figure 6-31. Simplified Error Response

for Nulling Azimuth Error
During Transient Mode

Computer-Generated Time History

Using the logic and equations described above, a computer

,irogram was developed to compute the gimbal angles and gimbal rates as a
ixwction of time. For an illustiative example, the Station is assumed to be
in a 55 degree inclined orbit at Z50 n. mi. Three equatorial synchronous
DRS are also assumed. The Station's initial position is such that at 24 min­
,'tes after t = 0, the spacecraft passes directly under one of the DRS's.
Ie particular K-band system analyzed here is that with its azimuth axis
.winting out along the +Z axis with the station in a local vertical mode.
-t.te that only the tracking behavior of the antenna system is generated by
f." program. Slewing and acquisition by the antenna is not included.
'Urtantaneous handover" between DRS's is shown in Figures 6-32 and
- 3. Note in the first figure that as the Station passes under the DRS (at
r ,,,t t = 0. 5 hours), the antenna must track through its zenith and the
"-axis become active. Figure 6-33 shows the rates which occur on
'"rte axes during this transient mode and during normal AZ-EL mode.

6-39
 AZI"USH

ELEVATION
CROSS AXIS

• 120

060~ b/ 4,

.160

0 0.25 0.5 0.75 2.0 1.25 1.5 1.75 2.0

TIME,HOUR

Figure 6-3Z. Gimbal Angle Time History - Baseline Three-Gimbal System

6 , jj TRANSIENT MODE 91 34de 0 men AZIhUTh

h4AXIMUMATES - 3.21 dog ., --- ELIEVAT[ON

03 5.51do0 m n CPOSSAXIS
_________ I --------
DRS02 DRS 2 - RS
D.--- 3....-..-DRS I­
4 -

. I

II , ",

, !I - ______________

2
o.0 0I 5 075 lO 1.25 5 1.75 2.0

HOURS
TIMvE,

Figure 6-33. Gimbal Angle Rate Time History --razeline

Thr'ee-Gimbal System

6-40
 7. LINK ANALYSIS AND DESIGN TRADES

7. 1 INTRODUCTION

Throughout the study, link calculations were performed to update

the Space Station communication system design in view of changing con­
traints and requirements. The final link powir budgets are given in
Appendix A as supportive material for the system design in Part I.

Subsection 7. Z reviews conventional linear link analysis and gives

a sample calculation. Subsection 7.3 outlines an analysis approach for a
composite PSK link through a hard-limiting repeater when it is used in
frequency division multiple access. Subsection 7.4 summarizes the con­
clusions reached as a result of the various link analyses performed in
support of the Space Station communications system design.

7. Z LINEAR LINK DESIGN

introduction

The equations employed in the linear link analyses are given below.
The performance data required to evaluate these equations can be found
elsewhere:

I) For geometrical data, see Subsection 6. Z.

Z) For achievable TWTA powers, see Subsection 10. 2.

3) For achievable system temperatures, see Subsection 10. 3.

4) For results on convolutional coding, see Reference 7-1.

kange Equation

Consider a space-to-space communications link. The received

' ,rrier-to-noise density ratio can be written as

C Pt Gt Gr
No LL sN O
0 t S

7-1
 I CH-IPVSURT POV[
PERSUBBIT BRITIL
BIHASE PERBIT
MODULATOR ENCODER BITSA

RIF
CHANNEL

l ~I
5UEiBI
1 B'DECION

81HASS DECISINP
1C! I1' VITE:: V SU8EITS

CHIPS LOFMODULATO]
+NOISE EOE RJ
;

NOTE- A QUADRIPHASE MODE GENERATES I CHIP

FOR EVERYTWOSUBBITS.

Figure 7-I. Definition of Bit, Subbit, and Chip

7-2
-,,hnere

Pt Gt
Lt the effective radiated power,

22Z
Ls the space loss = 1 Z

Gr the receive antenna gain,

and

N0 the receive systEm temperature.

Also, N kT s, where k = Boltzmann's constant and T = Tantenn a

+ 290 (L -1) + T
r receiver'

C/N 0 Required for Biphase PSK

The performance of digital links is described in terms of the bit

rate, RB , and the bit error probability, PB. Modulation theory determines
the required bit energy-to-noise density, EB/No, for a given PB. For
example, if biphase PSK is coherently detected, an EB/No = 9. 6 dE at
the demodulator results in a PB = 10- 5 in the output data stream.
C/B
Given C/N o and RB, the EB/No at the demodulator is simply N0

C/No Required for Convolutional Coded Links

Simulations of convolutional encoder/decoder performance deter­

mine the required received ED/No for a given PB" These results can
therefore be applied directly to link design as shown above for uncoded
P3K. It should be noted, however, that a code of rate = I/V generates
• - subbits for every information bit. Therefore, the biphase modulator
,flperates at a clock rate of VRB since one biphase chip represents one subbit.
(Stce Figure 7-1. ) The link summaries are given in terms of the RF chip rates
AL-d the required ECHIP/N o . This clarifies the RF structure of the data links
.ard identifies the SNR seen by the phase tracking loop and the subbit time
.),nci}ronizer.

"4-Dresentative Calculation

TabJe 7-1 shows a representative calculation for a convolutional

-',v, biphase PSK link. The first r,rtion of link sumnmctry determines

7-3
 :.v. sienal power received at the ALS. Given the ALS receive system noise
tn.Iperature, the received C/N o is then calculated. The chip rate in the
,-hannel is 2 x 57. 6 chils/second, since a V = 2 code is used. Therefore,
t e' received Ec/N o is C/RcIjIp/No = 50.6 dB-Hz. The Z dB "degradation
n the detector" includes the effects of intersynbol interference, carrier
:Aase tracking error, and chip synchronization error. A 1. 0 dB margin
n allocated to the detection process independeit of the 1. 9 dB margin in
the RF system. Since a rate 1/2 code is used, the bit energy available
Lo the Viterbi decoder is two times the received chip energy. This is
indicated as a factor of EBIT/ECHIP = 3. 0 dB. The EB/No received by
the decoder is then 5. Z dB. Simulations at MSC have shown that at this
EjIBNo, the Viterbi decoder outputs on the average only one erroreous bit
ctcision in every 10 5 bits (Reference 7-1).

7.3 DESIGN OF A COMPOSITE PSK LINK THROUGH A HARD-LIMITING

REPEATER
Although a communications satellite is conceptually simply a
"repeater, " the characteristics of typical relay satellite designs should
be considered in a composite link design. The use of a hard-limiting
repeater precludes the use of the conventional linear link design tech­
niques described above. The analysis approach below could be used in
the preliminary design of a composite PSK link through a hard-limiting
repeater.
Typical Repeater Designs
The DRS repeater may be used by several low altitude vehicles
on a random basis. Assuming that a frequency translation design is used,
the repeater bandwidth can be used by multiple users with orthogonal sig­
nals. Orthogonal signaling schemes can be achieved by access in time or
frequency division or by the use of pseudo-random codes. Frequency
division multiple access (FDMA) has been the most widely used approach
and will be assumed in this study.

The signals and noise entering a typical repeater are sometimes

simply translated and linearly amplified. However, if the input signal
levels can vary significantly, the input may be limited or automatic gain
control may be employed. Furthermore, the output power amplifier of
a communications satellite is often driven to saturation because of the
resultant increased efficiency. The limited input and saturated output
amplifier repeater is often called simply a "hard-limiting repeater,

Inter-m-odulation Noise in a Hard-Limiting Repeater

Although a link through a linear repeater can be analy :cd ]argely

1%tie basis of retransmitted noise and downlink noise, the hard-iirn .iug
t'ter produces a nonlinear distortion of the signal and an intermodu­
"noise'' when n-ultiple userzs are present. Int, rnlodulation noise (TM).
& systems sets an upper bound on the quality of the relayed d&:c.

7-4
 TABLE 7-1. ALS DATA LINK SPACE STATION
TO-ALS (f = 2100 - 2300 MHz)*

V
-PERFORMANCE
- SUMMARY

Bit rate 57. 6 kilobits/sec

Bit error rate 10- 5

Range 303 n. ni.

Channel Coding at Baseband at RF

Encoding (Z, 5) Convolutional Code Biphase PSK
Decoding Viterbi decoder (q = 8) Coherent integrate and
dump

LINK SUMMARY

Transmitter power (including output circuit losses) 20.4 dBw

Line losses -1.0 dB
Transmit antenna gain -3. 0 dB
Free space loss -154.Z dB
Ellipticity loss -1. 0 dB
Receive antenna gain -3. 0 dB
Link margin - 1. 9 dB
Power received at ALS C = -143.7 dBw
Noise power density (T. = 800 0 K) No = 199.5 dBw/Hz
C/N = 55.8 dB'Hz
Chip rate (115. 2 x 10 /see) 50.6 dB'Hz
Received E c /N 0 5. 2 dB
Degradation in detector -2. 0 dB
Detector margin -1.0 dB
E,BIT /E CHIP +3. 0 dB
EBN to Viterbi decoder 5. Z dB

i3 : requires Z44 KHz within this band

7-5
 16

-2- - C­
z

00
-

0 -2 -4-6 -8

OUTPUITBACI<OFFBELOWSINGLE CARRIER,05

Figure 7-2. Output Backoff Versus

Average Signal to Average
Distortion Bower

7-1,g

7-6
indcpendent of ther-mal noise considerations. Studies have shown that the
,'utut signal to Iv noise power ratio approaches about 10 dB as the num­
be. of equal-power and equally-spaced FDMA users grows very large (and
is at least dix) (References 7-2 and 7-3). Figure 7-Z shows how this sig­
nal to Ivi ratio varies with the amplifier (TWT) backoff below the saturated
(.single carrier) case. Given the maxinum signal to IM ratio, the maxi­
num attainable EB/No in a PSK channel can also be calculated. if the TM
noise is spread in the ZRB bandwidth of a PSK signal of bit rate B ,
then the

EB IS\ TB 1

) 1 where TB RB
ZR
B

and with u.)=10

__ (10)x Z -13dB.
0

Design of the Composite Link

The design of the "composite" data link should consider (as a minimum)
uplink noise, IM noise, power sharing, TWT backoff, and downlink noise. A
ruLe-of-thumb analysis which includes all these factors is given below. The
resulting equation (at the bottom of page 7-13) was used to obtain Figure 7-3.
The curves illustrate the penalties which IM noise imposes on the design
of PSK links to and from the satellite. The links must be of very high
quality if very low bit error rates are to be obtained; very little noise in
addition to the IM noise can be tolerated. Note that in the case of
(E B/No)uP = 13. 0 dB, the downlink must be essentially noiseless to
obtaina < 10- 5 . In order to obtain an improvement in P B from !0
to 10- 5 with (EB/No)Up = 13. 0 dB, the downlink quality must improve
by 11 dB,
Use of Coding Techniques

The use of error correcting codes on the DRS link permits a reduction
in SNR to the point where TM noise in the repeater is not a major factor. Rate
1/2 convolutional codes can be considered for this application, but research
hl)uld be done on their performance in an TIM noise environment. The K = 5,
V I code with Viterbi decoding requires an EB/No of only about 5. 2 dB
.tt 10-5 BER. Since two biphase PSK channel symbols (chips) are
; for each information bit, the ECHpiN o = (5. 2 - 3. 0) = 2. 2 dB.
-Itt
2dB1 is allowed for degradation in the detector and about 3 dB

7-7
 23.5

HARD-LIMITING (.5
1PEATEP
dB BACKOFF)
MUtTIPLE PSKUSERSIN FDMA

24.5

13 0 16.0 19.0 (UPLINK E83 dB)

0o.5

16.5

12.5

8.5
- -
10-2 104 10 5 10 8 10-10

BIT ERRORPROBABILITY AT GIOUND STATION

Figure 7-3. Composite Link Bit Error Performance

7-8
 r a margin, the RF system must'deliver EC/N o = 7 dB to the PSK
M1
te .rate and dump. At Ec/N o = 6. 8, the corresponding chip error rate
. ib-3. From Figure 7-3, it-is seen that this sort of performance is
casily achievable in a relay system utilizing a hard-limiting repeater.
taecall that each RF chip corresponds to a bit in a conventional PSK svs­
tm.n- Therefore, in Figure 7-3, read "bit error probability" as "chip
error probability. 'I)
Figure 7-4 shows the various tradeoffs in designing a relay system
- 3 (or a bit error rate of 10 - 5
which will achieve a chip error rate of 10
when using convolutional codes and allowing reasonable margins). The
ordinate of this curve is actually the required downlink E C /N O minus the
TVT backoff. This gives a "cost function" which takes into account the
reduced EM and reduced power when a TWT is "backed-off. " The ordi­
nate can be interpreted as far more than just DRS antenna gain. The
relative link performance required can be achieved by varying the DRS
or ground station antenna, the receiver system temperature, the weather
losses (by varying the frequency), or by choosing a TWT of different
saturated-single-carrier rating. The curves indicate that with a high
quality uplink, the W is tolerable and to minimize the requirements on
the downlink system, the TWT should not be backed-off. When the uplink
is severely constrained and say (Ec/No)-p < 8.5 dB, the downlink system
becomes severely burdened when the TN T is driven to saturation. It
appears that choosing an(ECI/No)up = 10.0 dB would be a reasonable
choice for the Space Station ' - data relay link. This does put a severe
burden on the downlink when using a hard-limiting repeater. The required
DRS antenna gain is a-iso shown to change very little as the TWT is
backed-off into the linear region. This permits more flexible repeater
operation.

Composite Link Equation for Multiple Biphase PSK Signals Through a

Hard- Limting Repeater

Let the total available RF power of the relay satellite be fixed

at bC, where

C = single carrier TWT power

b = backoff.
This power will be shared by the signals and noise inputted to the repeater.

'1 = fraction of output constituting relay noise

6i = fraction of output diverted to the i th signal

7-9
 75- - PLIrNK §

8.0

2 26f

7 BIPHASEP5K
-
0 CHIP ERRORRATE10
E 6.SdB
IN RECEIVEDAT THEGROUND
o STATION

O 9.5
00.

16.0
19.0
10
-4.5 -3 5 -2.5 -1.5 -0.5 0
TWTBACKOFF, DI1

Figure 7-4. Downlink Capacity Versus TWT Operating Point

7-10
 i =1

r V definition, output power = + 6 bC.

Relayed Noise Power

The noise input to the TWT is due to thermal noise, IM noise, and
various cross-terms (which'we neglect here as second order effects).

noise in = thermal noise power + IM noise power

NoB + a (n Pi)

where

N = density of thermal noise

B = repeater bandwidth

P = power in i t h signal.

nP. = total input signal power resulting from n signals.

a (nPi) level of noise generated by intermodulation of the

n equal power equally-spaced signals (We
assume it is approximately uniform across the
band).

noise in
noise in + signal in

(N B + anP)
(NoB + an P) + (nP)

7-11
A"I the relayed noise power

(N)a = 1 bC.

'delayed Power per User

signal in
noise in + signal in

P
(N B + a itP) + (riP)

(P)R 6 bC.

No at Ground Station

= KIN +(T)
(N)ground = 'N 'relayed ( downlink

K = Sum of all system gains and losses.

If we assume the repeater bandwidth is just wide enough to pass the spectra
of the n PSK signals each of data rate R, its bandwidth is n(ZR).

Therefore,

(N) o K [(N)]
G LZnR JoD + (NoD

Znergy at Ground Station Per User

(E)G =K [ c]

7-12
 at Ground Station Per User

i 6 KbC

llu~t bC 1 E in the downlinkl per user

oaD a

() =(No)(D ) Down
G. - + 1
2 $IT} Down

Now rewriting the expressions for q and 6 using P = ER, B: = n(ZR),

an -- PIPR
aI- P, we have
-0 2 + a~
2 + No
11= zRQ ) up
UP UP

and

n UT

UP up

Therefore,

2 )UP____P
:-
l-:' UPr
0+ al F+
L'+ (& +a) D .,­

7-13
y: z.v"be remembered that a definite relationship exists between a and
- ;,)Do\' since both are a function of the TWT backoff. Reference 7-I
the expression

"
a = .19 (b 8

for - 1.5 > b > -3.75 dB.

.4 DESIGN TRADES

All links except the emergency voice link are convolutional coded
... ,tal systems. This approach is the most efficient transmission scheme
-'ich could be implemented in the early 19701 s. Frequency modulation is
e,,snnloyed in the emergency voice system for simplicity.

All the digital links use biphase PSK modulation except the EM1
t.nn. This nodulation (with "integrate and dump" detection) is the
,..Wihnum technique among the class of bistate modulations. Biphase PSK
L z.tlsp very easily implemented. To conserve bandwidth at S-band, the
I,,S links use quadriphase modulation. This scheme can be analyzed
inplemented) as the union of tvo biphase systems in phase qua dra­
lsA
t-;rc. The RF bandwidth it requires is I/Z that of a comparable biphase
,y,,:tem, but its EB/No requirembnt is theoretically the same.

The ALS data links and the EM 1 links employ S-band frequencies
'i-c.use of NASA's allocations at those frequencies and because of the
vtailability of space-qualified hardware. It was concluded that frequen­
* ..above S-band were desirable for the Z0 mnegabits/sec DRS data link.
*-hiher frequency band would provide greater available bandwidth and
r,Ild give a relative link improvement, assuming fixed apertures at the
Ation and at the DRS. Frequencies at K-band (13. 7 to 15. 4 GHz) were
en because this frequency is currently within the state-of-the-art
•dcrowave technology. At frequencies above K-band, the beamwidths
".v:eloped by antenna apertures of typical size (3 to 4 feet in diameter)
;* yUl,e so narrow that poirting problems become severe, Because the
I-ergency voice system requires only 200 kHz, a VHF frequency band
chosen. This permits full communications service with relatively
- 2 ter levels and omnidirectional antennas. The power levels required
>tl:i be about 250 times greater if S-band were used.

Since all the links are full-duplex, the communications terminals

remte vehicles (either the ALS, EM I , or DRS) are approximately
/:'nt to the corresponding terminal on the Station itself. The charac­
.. " :a.x;sumed for these remote terminals are commensurate with their
The ALS uses only flush-mounted ornni antennas for communi­
-: 1the
( Station, because it must re-enter the earth's atmosphere.

7-14
RIetractable high gain antennas would add significant complexity. The EM 1
1! 0s utilize a high gain antenna. This vehicle is only used in a space
.nironment and furthermor, its principal mission is to collect data and
return it to the Space Station. The DRS is assumed to have earth coverage
antennas at VHF and K-band and one or more high gain antennas at K-band.
its transmitted power per carrier has been assumed to be modest (5 watts
aS 14 GHz and 18 watts at 140 MHz), since it may serve many users.

7.5 REFERENCES

7-1 "Program Definition for Computer Simulation of the Viter'bi

Decoding Algorithm, " Lockheed Electronics Company, under
Contract NAS 9-5191 to the Manned Spacecraft Center, 1 May
1970.

7-Z "A Study of Channelized Repeaters for Tactical Satellite Com­

munications, " (Classified Confidential) Hughes Aircraft Go,
IR and D Program, 9 December 1968.

7-3 "Applied Research Study on Satellite Communications, Report

No. 5, " Contract No. 00442(5), ITT Federal Laboratories,
for the U. S. Army Satellite Communications Agency, 30 June
1966.

7-15
 8. ANTENNA ACQUISITION AND TRACKIING STUDIES

81 INTRODUCTION

Analysis early in the sfud? indicated that DRS data link constraints
would require the use of a high gain antenna on the Space Station. A study
was begun which examined problems of initial antenna pointing, antenna
acquisition, and antenna tracking.

Consideration was given first to the programmed tracking concept

outlined in Subsection 8. Z, in which slewing, acquisition, and tracking
become essentially the same operation. It was determined that computer
control of the antenna was indeed feasible for the Space Station mission.
This approach was not chosen for the baseline design because of its open­
loop nature and because it requires a continuous interface with either a
computer or a mass storage device. The programmed tracking study did
indicate that the Space Station K-band antenna could be initially pointed
toward the DRS within its 3-dB beamwidth.

Antenna beam broadening was not given extensive consideration

simply because the DRS is normally expected to be found within the beam­
width required for communication. Beam broadening requires additional
RI complexity. Acquisition by scanning entails only additional use of the
slew or manual positioning mode. Scanning can sometimes cause excessive
gimbal wear, but it should be required Very seldom during the Space Station
nission. Subsection 8. 3 provides a sta'tistical model of antenna acquisition
by scanning. The analysis develops expressions for the probability of acqui­
sition and false acquisition given the basic system parameters. It also
develops estimates of the mean time to acquisition.

Subsection 8.4 gives the basic expressions for the theoretical per­
formance of various autotrack teclniques. A comparison of the techniques
is also made assuming similar system parameters.

The Space Station high gain antenna design is an outgrowth of the

.rious studies. The antenna is initially pointed toward the DRS on the
tis of conputer-generated pointing commands. If the DRS beacon sig­
-'1 is not present within the beamwidth, communication control generates
t *a n pattern using the slew mode. If the DRS signal is still not located,

8-1
 -:Zj DRS C

RESOLVERS

MOTORSACTIJ L

HEEGISTE
POSITION

G[ERAL
[ AEOR

PURPOS
... OSOITION

COA UTENTENRPOLATION

EHEM'ERIS

POINTING DATACOULD BE TR.ANSMITTED

TO THE
SPACECRArT
OR COMPUTED ON-BOARD

Figure 8-1. Programmed

Tracking Concept

Pl-z
a neW position estimate is taken from the computer data and another sequence
is begun. If acquisition is not attained after several attempts, the operator
is alerted. Antenna tracking is accomplished by the use of time-shared mono­
pulse. Its theoretical performance is at least equal to other schemes and
it has been used successfully in many hardware designs.

8.2 PROGRAMMED TRACKING CONCEPT

Although monopulse techniques have been used for many years, open­
loop, programmed tracking provides an attractive alternative. This approach
has been used at ground stations when ephemeris data permits the target
vehicle track to be accurately predicted.

The suggested programmed tracking concept is diagrammed in Fig­

ure 8-1. Note that while the tracking is open-loop, the antenna positioning
is closed-loop, since the actual position of the antenna is monitored by the
angle resolvers. The computer interface required by the programmed
tracking approach appears to be a negative factor. However, computer
generated pointing commands will be a necessity for initial antenna point­
ing in any mission utilizing a DRSS, and even when a monopulse technique
is used for tracking, a computer interface will probably be required for
acquisition. The general purpose computer itself need not be located
on-board the spacecraft when used to generate either initial pointing or
tracking commands. Programmed tracking does require a continuous inter­
face with either the computer or some mass storage device.

The success of an open-loop design requires that all possible error

sources can be determined and that their total effect is less than the required
error tolerance. Table 8-1 gives a pointing error budget for the Space Sta­
tion antenna pointing problem. Assuming that the structural deflection is
negligible, the pointing error is acceptable even without use of inputs from
the spacecraft attitude control system.

Table 8-2 summarizes the advantages and disadvantages of program­

med tracking. The references referred to document contain problems that
have been encountered with the Apollo monopulse systems.

8.3 ANTENNA ACQUISITION BY SCANNING

The Scanning Problem

To establish the communication link between a spacecraft employing

a high gain antenna and a DRS, some method must be used to grossly point
the antenna toward the DRS. This section treats the method of link accuisi­
ti-n c lled scanning, in which the antenna system then searches that angular
rc'ion of space in which the DRS is expected. When a thresholding device
iurlcatez the presence of the DRS signal, the antenna system switches to
th. autotrack mode and discontinues scanning.

8-3
 TABLE 8-1. POINTING ERROR BUDGET

Initial Extensive
Acquisition, Calibration,
degrees degrees

Station attitude stability* 0. Z5 0. Z5

Station structural deflection - Unknown

Alignment of boom on station 0. 05 (N)

Boom deflection 0.02 (N)

Orthogonality of axes 0. 02 (N)

Antenna boresight collimation 0. 05 (N)

Antenna boresight deflection 0. 0Z (I)

Antenna positioner accuracy 0. 006 0. 006

nntenna gimbal encoder accuracy 0.003 0o 003

Dynamic overshoot 0.01 0.01

Accuracy of ephemeris and local clock 0. 006 0. 006

Computational-step size 0. 005 0. 005

Maximum sum, error 0.440 0.280

Expected RSS error 0. z63 0. Z51

(N) - Negligible
:This error dominates. It can be tracked out given the actual spacecraft
attitude.

8.-i
 TABLE 8-Z. ADVANTAGES AND DISADVANTAGES
OF PROGRAMMED TRACKING CONCEPT

Advantages

a. Pointing is maintained under low signal-to-noise conditions or

even during signal drop-out.
S

o Tracking performance is unimpaired by the presence of low

frequency amplitude modulation on the communications signal.
(In Apollo, an incidental amplitude modulation on the uplink
signal has caused the appearance of an unwanted signal within
the tracking system bandwidth.)(Reference 8-1).

o Tracking performance is unimpaired by the presence of multi­

path signals. (In testing, it was noted that in the wide-beam
mode, the CSM antenna may falsely acquire a signal reflected
from the spacecraft surface) (Reference 8-Z).

a The approach is easily adapted to anticipate shadowing by the

vehicle itself or DRS occultation by the earth.

o Corrections for-axes misalignments, calibratable structure

flexures, or encoder errors may he conveniently introduced.

a The programmed tracking approach greatly simplifies the

antenna feed design and eliminates the usual error signal
processing electronics.

a The use of digital techniques greatly reduces the system's

sensitivity to component drift.

Disadvantages

o The technique -- like all open-loop designs -- is not adapt­

able to systems in which all contingencies cannot be antici­
pated in advance. An error source which goes unrecognized
during the design phase could prevent proper system
operation.

o A near-real time interface is required with a computer or a

mass storage device.

a To achieve the required tracking accuracy an interface may

be required with the spacecraft attitude control system. This
is especially true when the vehicle (such as the ALS) is
capable of e.tensive rnaneuvering,

8-5
 I )C-2.

C
(
___ ___ :_____ )
__ -_------- =--)- ~ _______
C =),C. cc: _____ C)
( ::) c .t. 2)

A) SQUARE AREA EAR SCAN B) CIRCULAR AREA EAR SCAN

C) SPIRAL SCAN

Figure 8-2. Scan Patterns

8-6
7f 6 <10 degrees and e < Z0 degrees,

1- 44Z (8-4)

An approximation relating antenna 3 dB beamwidth and gain, valid for beam­

widths less than 10 degrees, is given by

EZ = Z.7x 4 (degrees) 2 (8-5)

G

Substituting equation 8-5 into 8-4

k = (1.48 x 10 - 4 ) GI 7 (qsin degrees) (8-6)

If k- 1, then A s a, i. e., the antenna view area is as large or larger

than the search area, so no scanning is required.

If T s is the total allowable search time, the view period Tk, for a
.given angular position in the search area is given "by

Tk = T s/k (8-7)

Detector Operations

The basic detector assumed in the analysis below is the square law
device. Both linear and square law devices have been analyzed by Marcurm
(Reference 8-3) and shown to have nearly identical detection- characteristics.
.The detector, itself, is preceded by a bandpass filter and followed by a
sampler. The sampler is, in turn, followed by an adder which sums the
last in outputs from the sampler. A threshold switch is used to indicate a
signal detection when the adder output voltage exceeds a bias level, b.
This combination of functions is shown in Figure 8-4.

If TI denotes the time that a DRS is within the receiving pattern of

the antenna and t s is the time between samples, then m is given by TK/t s .
The samples may be considered independent if t s ->I/B, where B is the
input filter bandwidth, For acquisition, the bandwidth must be large enough

8-9
to accommodate the effects of dopplei shift and oscillator variation on the
transmitted signal. The maximum doppler shift is given by

VS
S + VD
D cos i (8-8 )
(AfD)

where C is the speed of light, VS and VD are the velocities of the station
and DRS, i is the orbit inclination, and f is the nominal transmission fre­
quency. For a Space Station altitude of ZOO n. mi. andIan inclination of
55 degrees,

(AfD) Max = (3 x 10 - 5 ) f (8-9)

Oscillator stability generally results in an order of magnitude less varia­

tion. Assuming a stability of one part in 10-6,

(Af) = (3. 1 x 10) f (8-10)

The bandwidth must be twice this value, i. e.,

B (6. Z x 10 - 5 ) f (8-11)

and so

- 5 (8-IZ)
rn S (6. Zx 10 ) f TK

The noise will be assumed to be white gaussian noise due to thermal

motion of electrons in the receiver. The average noise power, N, is given
by

N = KTB (8-13)

w;,here K is Boltzmann's constant and T is the effective noise temperature (K).

8-10
Probability of Error
When no signal is received, the noise alone may cause the adder
output to exceed the threshold bias level. The output of the detectot will
then have a probability density of:

exp - V :>0
p(v) (8-14)
0 v < 0

where:

N = average IF noise power

v = detector output voltage

This density is derived in Reference 8-4, pages 253 - 254.

Normalizing the noise power

(eY y e 0
P(Y) =. (8-15)
0 y < -0

where y = v/N is the normalized detector output. The characteristic

function of this density is given by

C(p) p (8-16)

The characteristic function of the sum of in independent samples is

= (p = Lpl(p) (8-17)

8-11
Taldng the inverse transform, the density function for the normalized adder
output voltage is given by

•( Sr'- i e -y
y>0

J0 y< 0

The probability of error, Re' is the probability that the adder output due to
noise alone will exceed the threshold level, b. The normalized threshold
level is denoted by bn' where

bn = b/N (8-19)

Then

P- = pr(y) dy (8-Z0)
^n

In Reference 8-5, b n is computed as a function of P and m.

Probability of Detection

Using characteristic functions as above, it can be shown (see

Reference 8-3) that the probability density of the adder output with signal
plus noise at the input is given by

rn-i
( e (v+mR) I
(Z.R 'rn-I (Z \mrRv) v : 0
qm(v) = (8-z2) I
0 v<0 "r

where R is the signal-to-no 3 power ratio and r-i is the modified Bessel
function of the first kind and of order rn - 1.

P,-I2
*,:.nthe probability of detection, Pd, is given by

Pd f m(V) (8-ZZ)

This integral may be expressed as

P 1- Tb- (Zm - 1, m - 1, Tia--) (8-23)

n

where

b = b/N = normalized bias level

n

b = threshold bias level

N = rms noise power

R = signal-to-noise power ratio

T = incomplete Toronto function (see Reference 8-3)

P is the probability that at a given time the detector will indicate a detec­
t-on "whenthe DiS signal is not present. Pd is the probability that a detec­
tion will be indicated when the DRS signal is present.

Probability of False Alarm

A false alarm occurs when the detector registers a signal detection

and there is no signal present at the time. This happens if the output of
the adder exceeds the threshold bias level due to noise alone. The proba­
bility of false alarm is given by

Pfa = l - - e ))k-l (8-24)

If G, jJZ and Pr, are fixed, then P is determined by Equations 8-24 and 8-6,
ich in turn Aetermines b n by using Equation 8-Z0.

8-13
 Note the follolving relationships and conditions:

Pe =PLa k< Z

(8-25)
iP = f k L- Z, P k << !
e -Ie

-Lfa.
i rnay be notedsofrom Equation Z4 that Pgs Thus, if kPfa<< I then
."PC
<< , and

P
P - k - 2 (8-26)
e k-I

Probability of Acquisition

The probability that the DRS signal will be properly detected during
the scanning process is given by:

k
Pacq = PA Pd' (I - Pe) (8-Z7)
n= 1

where PA is the probability that the DRS lies in the search area and Pi is
the probability that the DRS lies in nth angular position of the search area.
The above summation expression is the sarm of probabilities that an error
was not rmde in the previous (n - 1) positions and that a correct decision is
made in n h position times the probability that the DRS is in the nth position.

A useful and simplifying assumption is thatP n = I/k. This assump­

tion is not generally correct, but is" sufficiently accurate to alloVI a good
estimate of the probability of acquisition. Using it allows Equation 8-Z7
to be further reduced

P - P [I ]
pI- (8-28)
acq kPge

8-14
1; p is small << 1, then

k P k(k -I) (8-29)

[1 -(1 Pe)k]

Thus for P e << 1,

pac 2 A Pd jp1 e (8-30)

and ifkP << 1,

e

pacq pA
Pd (8-31)

< 1, then
It should be noted that ifk

Pfa pe

and

acq Ad

where A is now the area defined by the antenna beamwidth which is larger
than the required search area. Figures 8-5, 8-6, and 8-7 show P /PA
as a function of Pd with Pe and k as param,eters. The first two in cq
cate the validity of Equation 8-31 for kPe' I.

Parametric Results

From Equation 8-ZO, b n is a function of P and rn, and from Equa­

tion 8-23, Pd is a function of n , m, and the signal-to-noise power ratio, R.
Thus, Pd is a function of P., m, and R or inversely R is a function of Pd,
Pe, and m.

if it is assumed that kPe << I (i.e., kP e < 0. 1), then from Equa­
tion 8-31, acq = PAPd • If A is taken to be the area defined by letting q)

8-15
 =0.0 -0.01

--- -0 0.0
0.8 0.6O

I- 1

tFxgure 3-5. Probability of Acquisition j I

When k = 10 0.4 12 T i

0.2 ___ _ _

0 02 0.4 0.6 0.8 10 1.2

Id

I.(

p 0.0l

0.0, Figure 8-6. Probability of Acquisition

"0.4 When k = 100

0.2 ____

0.2 0.4 0.6 0.8 1.0 1.2.

I'd

1.0 1

I 0C01

rq.,re 8-7. Probability of Acquisition

0.6

.0t- - ... .

When k = 10.

0 0.2 0.4 0.. 0. I 0 1.2

'd

8-16
in Equation 8-I equal the 3 T value of antena boresight deviation from the
noninal position, then /
PA = 0.9974 (8-32)

and

Pd = P aq/0.9974 (8-33)

And then R is a function of Pac , P. and m. The value of rn will probably

depend more on hardware impl~mentation considerations than on the theore­
tical limitations of Equation 8-23.

For instance, it is expected that 25 dB <- G < 60 dB and 0<W5<3'

(for the Space Station mission) thus, from Equation 8-6, k < 1. 33 x 103.
It was shown above that due to doppler shift and oscillator uncertainty,
B - (6. Z x 10- 5 )f. Furthermore, it is expected that T s > 30 seconds. Thus,
for f = Z GI-Iz, BT s - 3. 72 x 106. Then, from Equation 8-23, in > 21, 000.
It is not expected that the detector could add this many samples.

In Reference 8-5, bn is tabulated as a function of in and Pe for values

of m between I and 150 and Pe between 10 - 1 and 10-12. In Reference 8-5,
Pd is plotted as function of bn , m, and mR for in = 1, 3, 10, 30, 100,
300, 1000 and 3000.
- From Equation 8-33, the following list is valid:

p p
acq Pi

0. 90 0. 9023
0.95 0.95Z5
0.99 0.9926
0. 995 0. 9976
Combining this list with the data from the above mentioned References, the
required signal-to-noise ratio is listed in Table 8-3 as a function of lacq,
PC and n. For Pacq = 0. 99, 4curves of R versus m with Pe as a param­
eter is shown in Figure 8-8.

I Using Equations 8-6 and 8-25, the signal-to-noise ratio may be deter-
Tined as a fun6tion of G 2 with m, Pacq, Pfa as parameters. Using the
& ta of Table 8-3, curves of R versus GtP 2 with n as a parameter were
lAtted for P = 0.99 and Pfa = 0-5. These curves are shown in
:re 8-9. acq

8-17
Figure 8-8, Required Signal-to-Noise 0- '0_8

Ratio Versus N~umber of 5 ­

05
0.2

0.1
o
1
I 0 20 50 o
Rati Vesu Nube-o----
INUMUfRor SAMPLES,M,
zz__ ­

Poc 0.99 I.,0I0-3­

-5
Pf 10

200

10, 2 5 0 2 5 105 2 1 06

2
G 2 DEGS UIMBTEDo, MPE

Figure 8-9. Required Signal1-to-Noise Ratio Versus Gq'?

8-_18
 TABLE 8-3. SIGNAL-TO-NOISE POWER RATIO

Pa 0.90 0.95 0.99 0. 995

p m bac
e n pd 0. 9023 0. 9525 0.9926 0.9976

1 11.51 18.0 Z0.5 25.3 27.7

3 16.55 7.3 8.17 10.0 11.z

101 10 29.52 3.0 3.20 4.0 4.30

-30 59.29 1.38 1.57 1.88 2.0

100 148.5 0.68 0.74 0.86 0.94

1 13.82 21.Z 23.5 28.5 32.0

3 19.13 8.5 9.5 11.4 12.5

101 0 32.71 3.3 3. Z 4.4 4.8

30 63.55 1.59 1.74 2.02 2.22

100 154.9 0.74 0.84 0.94 1.01

1 16.12 23.5 25.8 31.7 35.2

3 21.67 9.53 10.5 12.7 13.8

10-7 10 35.79 3.70 4.0 4.8 5.2

30 67.60 1.75 1.88 2.22 2.37

100 160.9 0.81 0.87 1.01 1.08

1 18.42 26.5 29.4 36.0 39.0

3 24.18 10.5 11.4 13.6 15.0

10- 8 10 39.80 4.0 4.3 5.2 5.7

30 71.48 1.87 Z.03 2.32 2.53

100 166.6 0.87 0.95 1.08 1.16

8-19
 Conclusions on Detection

It has been shown that the number of samples used in the detector's
adder has a dramatic influence on the required signal-to-noise ratio, whereas
the effect of the gain-search area product (proportional to GIj 2 ) is much less.
In Figure 8-9, where P a = 10-, the change due to the maximum variation
of G Z causes the signal-to-noise batio to change by a maximum factor
of 1. 44 or approximately 1. 6 dB, when m = 1. The change due to 047 is
much less when m > 1.

The change due to a variation of rni is much larger; in Figure 8-9,

for GO Z = 10 5 , increasing m from 1 to 3 reduces the required SNR from
29 (14. 6 dB) to 11. 5 (10. 6 dB). If m is further increased to 10, the required
SNR is reduced to 4. 5 or 6.55 dB.
Thus, -the number of samples added in the detector prior to the
threshold switch is the major factor in determining the required signal-to­
noise ratio for given probabilities of acquisition and false alarm.

Acquisition Time
It has been shown above that the parameters of interest are related
by

T B
m (8-34)
k

where k - 4AI/EZ, the number of view areas contained in the search area.

For a reasonable relationship between search area and antenna

beamwidth, k <- 100. Then, if Pe is limited to a maximum of 10 - 7 , the
probability of false acquisition will be less than (k - 1) P e = 10 - 5 . Refer­
ring to Figure 8-8, if a probability of acquisition of 0. 99 is desired with
a signal-to-noise ratio of Z(3 dB), then in - 40. The IF bandwidth should
be as small as is compatible with the expected doppler shift and oscillator 5
.instability. This requirement is shown to place a lower limit of (6. 2 x 10- )f
on the IF bandwidth, B, where f is the transmission frequency.
Substituting the above numerical quantities into Equation 8-34:

T - (Z.58 x 106) - seconds (8-35)

This inequality seems to indicate that the search time may be decreased by
increasing the transmission frequency, f. This, of course, is not neces­
sarily true because the resultant larger IF bandwidth will reduce the

8-20
 *..-to-noise ratio. The same effect is possible by increasing the
xwidth at any frequency, f. Inequality (Equation 8-35) relates the search
-~t
d the other parameters for the minimum IF bandwidth. hidependent
:. analysis will insure the required signal-to-noise ratio, taking into
?fl-lnt antenna sizes and transmitter power.

Figure 3-10 is a result of using Equation 8-35 for transmission

,requencics of 2. GHz and 14 C-Hz. From this figure, it may be seen that
:or small values of /O, the search time may be quite small; thus, the
actual search time may be determined by mechanical considerations rather
t.tn by the operation of the signal detector.

Figures 8-11 and 8-12 show the search time as a function of P/'I
awli signal-to-noise ratio (SNR) for transmission frequencies of Z. 2 GHz
ann 14 GHz, respectively. The relationship between m and SNR of Fig­
ure 8-Z was combined with inequality (Equation 8-34) to produce these
figures.
Mean Time to Acquisition
The minimum time which must be allowed to scan the search area
has been discussed above. The mean time to acquisition, Ta, is related
to the search time and is given by the following expression:
k
Ta 1 n dn T + PnPe (n T± TL+ T a (8-36)
n= l

where

Pn z probability of target lying in the nth antenna view

area within the search, area

PD = probability of detection when target is within the

antenna view
Pee = probability of error when the target is within the
antenna view

TL = time lost due to an error

= Ts/k

8-Zi
 *, _=0
20
10722G 5,=3dg

0~q
. 99
F= 14GHz/

1O4 5 10-3 -1
2 5 102 2 5 j0 2 5 I 2 5 10

MINIMUM SEARCHTIME, TS, SECONDS

Figure 8-10. Minimum Search Time Versus P/e

8-22
 5­
f 2.2 GHr
P .99
"
p =l0
_______
2

SIGNAL-To-NOISE
RTIO, dB 01

02

10,
0.1 0.2 0.5 1.0 2 5 10 20 50 IO

Figure 8-11. Minimum Search Time Versus 44e at S-Band

8-23
 SIGNAL-TO-NOISE
RATIO, d3
$
2 4 6 a 10 12

lO
f = 14 H
=0.99
=

10

29

0.1 0.2 0.5 1 10 20 50 10

FiT

2-2
i, it is assu-med that the target -maylie at any point within the search area
vith equal probability,, then Equation 8-36 reduces to

a I e PTa + + T (8-3-)

for
k =L 2 > 1(8-37)

The uniform probability assumption is not correct but in many cases leads
to results very nearly the same as with a more accurate and more complex
probability distribution.

If P <- 10 - 6 , as will usually be the case, then Equation 8-37 becomes

e

a Zdi + ±R + e s-L (8-38)

And if TL/T s - 100, then

Ta P41+ : (8-39)

Figure 8-13 shows Ta/Pd T s as a function of q/e.

Example

Let 4 be the 3o- deviation of the boresight from the nominal. Then,
as shown previously, for P e << 1

Pd = Pacq/0. 9974 (8-40)

8-25
 h'O 1 I
o.9 i I t 2

T 0 a

T Pd

0.7

0.6 _

0.5
01
1
02 05 I 2 5 0 20 50 to

Figure 8-13. Ta/TsPd Versus p,/e

8-Z6
 - P acq is the probability of acquisition. And Let the following system
•.- eteTs be given!

P = 0.99
acq

y = 3 degrees

6 = 1 degree
- 7
P = 10
e
rn = 40

F'hen Pd= 0. 9926, P/0 = 3, and there are

rintenna view areas in the search area. From Figure'8-10, the total search
time which must be allowed is 1. 7 ms for f = 14 C-Hz and 11 ms for
f = 2.2 0Hz. From Figure 8-11, T /T PA = 0.514, and soT = 0. 866 ins
for f = 14 GHz, and T a z 5.61 ms %r 2. GHz.

8.4 THEORETICAL COMPARISON OF AUTOTRACE SYSTEMS

This section compares the theoretical tracking accuracy due to

thermal noise of three types of autotrack systems: 1) true monopulse,
2) conical scan, and 3) time-shar-ed monopulse.

The derivation of the tracking accuracy for the true nonopulse sys­
temls has been presented in various references, one of which is Refer­
ence 8-6. In this reference, the rms tracking error for a single axis has
been derived for both correlation detection and coherent detection of the
e:rror signals. These equations are shown below

True Monopulse -- Correlation Detection

e
rnis = /S
+ ()j
i (8-41)

r-n7

8-27
 True ivionopulse -- Coherent Detection

=
tmSs ~ )(-4Z)
where

8 s = rms tracking error along single axis in fractional

rms beamwidths

2IBN = two-sided servo noise bandwidth

BIF = IF bandwidth

S.
N.I Sum channel IF signal-to-noise ratio at the receiver
= input

km = error slope factor of the monopulse difference

pattern normalized to the antenna beamwidth

The derivation of the tracking accuracy for the conical scan system
has also been presented in various references, including References 8-6
and 8-7. Here the tracking error is shown to be,

Conical Scan - Square-Law Detection

N
SLi
8rms­

8k-
 Conical Scan -- Coherent Detection

B
r
0 rms kcI s BIF
S (8-44)

where all parameters are defined as before except

k cs = normalized
error slope to
factor of the conical scan system
the antenna bearnwidth.

The theoretical tracking accuracy for the time-shared rnonopulse

system has been derived in Reference 8-8. The implementation of the sys­
tem is also described in this reference. The equation for the tracking
accuracy is given below.

Time-Shared Monopulse -- Square-law Detection

BIN F S. \1

rms S k ( 8- 45)

where all parameters are defined as before except

k loss of sum channel signal power due to coupler

(throughput loss)

k2 loss of error signal power due to coupler

(coupler loss)

In order to compare the systems on a comparable basis, the follow­

ing assumptions were made:

1) No line losses between antenna output and receiver input.

Z) XF band-width/servo noise bandwidth = 10 6 .

8-29
 0.012 ­

z
0.010

0008---
-~
ii --
TRUEMONOPULSE - CORRELATION
DETECTION '0 dB FEEDLOMs

CON
.... I
CALSA - SQUARF-LAW
(I 0 dB CROSSOVER LOSSI

- SOUA'E.LAW
PA -ISHAREDA'ONOPULSE
i
DETECTION
I
SUM CHANNEL tOSS DUETO COUPLER)

0.004

0 2

0-10 -5 0
15 20

RECEIVED
IFSIGNAL TO NOISE RATIO. 08

Figure 8-14. RMS Tracking Error Due to Thermal Noise

3) True rnonopulse system incurs 0.5 dB loss of signal power due

to the use of multiple feeds.

4) Conical scan system operates with 1. 0 dB crossover loss.

5) Time-shared monopulse incurs "0.5 dB loss of signal power

due to multiple feeds and another 0.5 dB loss due to coupler.

6) The error slope factor for the monopulse feeds, kin, is assu­
med to be 1. 57/beamwidth error. This is based on the dis­
cussion in Reference 8-9, pp. 27Z-Z75.

7) The error slope factor for the conical scan system, kcs, is
assumed to be 0. 8/bearnwidth error. This is based on the
discussion in Reference 8-9, pp. Z69, for a 1. 0 dB crossover
loss.

The tracking accuracy due to thermal noise for the three systems
is shown in Figure 8-14. The curve for the true monopulse system is baseq:
on correlation detection, and the curves for the conical scan and time-shared
n-onopulse are based on square-law detection.' The abscissa of Figure 8-14
is based on the IF signal-to-noise ratio before reduction of the signal power;
by assumptions 3), 4), and 5) above.

8-30
 It can be seen from Figure 8-14 that for the parameters chosen, the
cnnical scan system provides 4 to 6 dB poorer performance than the true
nmonopulse, and the time-shared monopulse provides from 6 to 10 dB poorer
pjeformance.1At is of interest to note that if the coupler loss were 3 dB for
tje time-shared monopulse, i.e. , k, = kZ = 1/2, the rms tracking error
for the time-shared system with square-la1v detection would be exactly
twvic'e as great as for the true monopuise system vith correlation detection.

8.5 REFERENCES
8-1 "Tracking Effect due to AMk on the Uplink Mode Containing
both S-Band Subcarriers, " TRW Systems Group, under Con­
tract NAS 9-4810, to the Manned Spacecraft Center, 19 March
1968.

8-2 "An Investigation of the Multipath Problems Associated with

the Apollo CSM and LM S-Band High.Gain Communication
Antenna Tracking Systems, " TRW Systems Group, under
Contract NAS 9-8166 to the Mahned Spacecraft Center,
16 December 1968.

8-3 5. I. Marcum, "A Statistical Theory of Target Detection by

Pulsed Radar, " with mathematical index, RAND Research
Memo RM-754, 1 December 1947.

8-4 W. B. Davenport and W. L. Root, Random Signals and Noise,

MtGraw-Hill, 1958.

8-5 J. Pachares, "A Table of Bias Levels Useful in Radar Detec­

tion Problems, " IRE Trans. on Information Theory, IT-4,
No. 1 (1958) pp. 38-45.

8-6 J. M. Stephenson, et al: "Design Criteria for a Large Multi-

Purpose Tracking Antenna," ASTIA Document No. AD Z67147,
January 1961.

8-7 G. M. Pelchat, "The Effects of Receiver and Antenna Noise

on the Performance of a Conical-Scan Tracking System,
The Microwave Journal, February 1965, pp. 37-40, 61.

8-8 E. T. Hazelton, "Thermal Noise Errors in Time-Shared

Monopulse, " Hughes IDC No. Z794/61, Z2 January 1970.
8-9 D. K. Barton, Radar System Analysis, Prentice-Hall, Inc.,
Englewood Cliffs, Ne.w Jersey, Chapter 9, 1964.

8-31
 9. MODEM AND TIME DIVISION MULTIPLEX STUDIES

9. 1 INTRODUCTION

Figure 9-I shows the conceptual arrangement of a digital

communications system for the transmission of several analog signals.
The source encoder/decoder devices accomplish the mapping of the data
source signal from the analog domain to the digital domain and then back
into the analog. The digital portion of the channel coding device generates
one or more subbits for every information bit. These subbits are repre­
sented in the RF communications channel by the analog modulator. Its
output symbols are called chips.

After consultation with the Manned Spacecraft Center (MSC), the

selected channel encoding utilizes biphase or quadriphase PSK in conjunc­
tion with convolutional coding. The performance of a representative biphase
demodulator is discussed in Section 9. Z. This device recovers both subbit
timing and a coherent carrier reference from the PSK signal itself. The
performance of various convolutional codes with Viterbi decoding has been
documented by MSC and has not been part of this study. The physical char­
aZteristics of Viterbi decoders were investigated and are summarized in
Section 10. 4. Section 9. 3 summarizes the design study of the Space Station
time division multiplex (TDM) system. Thie statistical analysis of the frame
sync technique is given special emphasis. The source encoding problem
was not addressed during the study. The physical characteristics of a
particular source encoder/decoder suggested by MSC vere estimated.
These results appear in Section 10.4.

9. Z DEMODULATION OF CONVOLUTIONAL CODED BIPHASE PSK

SIGNALS
This section considers the technique shown in Figure 9-2 for the
demodulation of convolutional coded biphase PSK signals.
Subbit Synchronizer

The subbit tim-e synchronizer phase detects the data, locates the
"t:Uliit 'transitiontimes, and outputs a sampling time signal to the integrate
Au: dunp circuit, A synchronizer has been built at the Jet Propulsion

9-1
 [SOURCE
SOCE ANALOG DATA C

CHANNEL ENCODER E

CHIPS RE SUBBITS BASEBAND BITS MUTPEXRSURCE

[ MODUL~~~rOR ENCODER6ECDR NLGDT

C[ANNEL NOISE SOURCEi

SOURC ANALOG DATA
DATA

DECODER (+ NOISE)
CHANNEL DECODER
N SUBBIT BIT
RF BASEBAND INSSOURCE
D__DECISIONS ANALOG DATA

ET SOURCE > ANALOG DATA

DECODER (+ NOISE)

REPRESENTATIVE DESIGNS
SOURCE ENCODER/DECODER; DELTA MODULATION
(VOICE)
RF CODING: BIPHASE PSK WITH COHERENT DETECTION
BASEBAND CODING: CONVOLUTIONAL CODING
WITH VITERBI DECODING

Figure 9-1. Conceptual Arrangement of Digital Communications System

L.,i-bratory (JPL) which provides negligible system degradation down to-
Esubbi INo = -5.0 dfl at data rates up to 250 i:lz (Reference 9-1). The
design's performance is attributable to the use of a digital loop filter and a
specially designed phase detector. Although implementation of these tech­
niques -nay not be possible at 40 MHz rates because of the Limitations
imposed by circuit speeds, it will be assumed here that suboit timifg is"
perfect. (Subbit timing information can also be recovered at baseband ­
after the colherent reference is reinserted'and before the integrate and
dump. )--

Squaring Loop

The carrier phase tracking loop design is shown in Figure 9-3. This
"1squaring loop" takes the modulated carrier and recovers the original car­
rier. (It has been shown that this sort of system design is more efficient
in terms of total power than that iii which a separate reference signal is
used (Reference 9-Z). Another loop design might be used, but the squaring
loop is representative of the possible configurations. ) Now, an expression
for the overall system error should include the effects of both the PLL and
the integrate and dump. It has been shown that the probability of a detection
error for P5K given a phase error, t , is

P (Error/l)= I exp " - d,

2E2

where t N Cost
0

p (Error/4)= Erfc Lts cost]

Viterbi (Reference 9-3) has shown that for a first order PLL the steady­
state phase error (mod ZTr) probability density is given by

exp (a cos )lII

0

Wflere a is the SNR in the loop noise bandwidth and I. is the imaginary
Pettssel function of order zero. Lindsay has shown experiirently (Refer­
*lllcc V-4) that- this formh is also approt.imately correct for the second order

9-3
 *DERIVES SUBBITTIMING

- RECOVERS
CARRIERPHASE

INTERGRATE
AND DUMP

SUBBITSYNC
TRACKING LOOP ANLGSAUEI
ANALOG VALUEIS
GENERATED EVERYI/R S so
S(') 7PCOS (vWt 0) BY SAMPLINGTHE INTEGRATOR
+ n(t)

S(t) 1 ATRS SUBBITS/SEC CARRIERPHASE

nt) =VHITE GAUSSIAN NOISE TRACKING LOOP

Figure 9-Z. Demodulator for Biphase PSK

S(tI -2P¢o (Vtt-0)CARMtODULATED °

() " CARRIEP

REFERENCE
SQUARE
LAWYDEVICE

Figure 9-3. Carrier Phase Tracking Loop (Squaring Loop)

9-4
loop when a is large enough to be of practical interest. The overall P (the "
subbit error probability) can now be written:

P S = P (Error/) P (4)

T2E
- eo (t (a)os Erfc oF~Cos

this expression has been evaluated numerically by both Lindsay and

Viterbi. We shall first determine the range of a and Es/N o which are of
interest. Next, 'or a loop giving good tracking, the acquisition character­
istics will be investigated.

Bandpass Filter and Squarer

To determine the loop parameters which will-give good tracking, the

effects of the filter and square law device must be evaluated.

The received signal plus noise "seen" by the filter is

y(t) = x(t) + n(t)

:where

x(t) = 47j5 sin (wct + 4 + s(t) j) = s(t) <ZP cos (Wct + 4)

s(t) = ±1 (the random data)

P = carrier power

0 = phase random process

n(t) = white Gaussian noise of one-sided density N0o

The bandpass filter is matched to the spectrum of the signal - which is
roughly 2lRs MIhIz wide where Rts is the subbit rate. The total carrier fre­
quency uncertainties are given approximately by Table 9-1, assuming no
programmed Doppler offset. Since they are an order of magnitude less than
ZR2 , a sharp cutoff filter of width W ZR s will pass the signal without sig­
nificant distortion.

9-5
 TABLE 9-1. TOTAL CARRIER FREQUENCY UNCERTAINTY

fc = 2. 2 GHz MHz

Doppler shift 0. 112

Oscillator shift 0. 004

Total 0. 116

fc = 14 GHz

Doppler shift 0. 714

Oscillator shift 0. 028

Total 0. 74Z

Now, since the noise bandwidth BL of the PLL following the squarer
need not even be as large as the frequency uncertainty which must be
"tracked out, " it holds that V>>BL. The signal X noise and noise X noise
terms out of the square law device can then be approximated as Gaussian
noise with a new density

N' : 4 N P NZ W

o O2 0

This assumption is used by Viterbi (R eference 9-3 , p. 291), and by

Lindsay (Reference 9-5), and was examined in detail by Marcum (Refer­
ence 9-6). The signal term out of the square law device is of the form

2 Cos Z((0c t + 8) + 1.
[s(t)] 2 (2p) z

and :. [1] (P) [cos (2cct + ZIP)] after filtering. The squaring operation maps
the s(t) = -I modulation into (:l1)2 = 1. (Note ,that the carrier phase refer­
ence may be in error by 180 degrees. The subbit characters may become
"inverted". They can be re-inverted by examining the frame sync word.)

9-6
Phase Locked Loop
The baseband PLL model used by Viterbi to derive the phase error
probability density (given above) is shown in Figure 9-4. It can be applied
to the system of interest if we replace our VCO and doubler with a single
VCO which generates twice the first VCO's frequency (for the same input
voltage). From the linear theory, the phase error variance of the reference
at 2W is
c

N, BL
P2 /O

where N -.BL is the noise power and P 2 /Z is the carrier power. However,
since the loop output is taken to be at wc the phase error variance is
1/4 2 orc

N' B L 1
.2pz

. NBL ( N W)

Since

P =ER
S S s

1 NoB (
0~ kl±E+
5 ZRJ

Constraints on B L for near theoretical PB

In the expression for Ps given above

Ps -xp (acos4) Erfc .._.2 cos ') d4

r
T (a) N
Jo 0 0

9-7
 n'('0
0
Io

(tA s in ( ) L EA

T
, 2(0 )
0

n'(t)

Figure 9-4. PLL Baseband Model

the error probability is a function of a and ES/N O. Lindsay's integration of
the expression is graphed as a finction of E N with 6 = a/(E /Nolas a
parameter.

The upper set of curves in Figure 9-5 give Lindsay's results. The
two sets taken together can be used to determine the Ps degradation due to
inperfe.ct tracking. The curves also show the required Eb/N o increase
needed to meet a specified iB given constant 6. Figure 9-6 illustrates the
use of Figure 9-5. (We assume here that the analog signal which constitutes
the integrator's level at the sampling time has the samte characteristics
whenever the P. is the same. Actually, a hard-decision on which subbit was
sent is never made.
Now it is desirable to make a (and 6) as large as possible so that the
PE degradation is negligible. This does not require increased power but
only a decrease in B .

N0 NEBL 0~2

R N
s 0
L
- B. N

This curve is plotted in Figure 9-7. Note that for 6 10 (0.4 dB degrada-"
tion) the BL = 2 MHz. it seems likely that 6 = 100 represents less than
0. 1 dB degradation. At 6 = 100 the BL = 0. 2 MHz.

Loop Acquisition Performance

Now consider the acquisition properties of a PLL with 0. Z MHz <

BL < 2. 0 MHz. (Since the available literature deals with this problem only
for the noiseless case, it can be assumed that the results below are an
upper limit on the performance.) When operating linearly with an input
signal of voltage A and loop gain K, a second order PLL has a closed loop
function

H-I(S) = AK (S +,a)
S2 + AKS + AKa

9-9
 I.
i
t I __ __ _ I -
PL IE PE FEC T
1RACING
~(FROM/
Q
LIN'DSAY[DI
I fKG
0 -2

g0
-4.0 -32 -2.4 -46 -I0IR0 E-0.8
12"3 IM,
No

I_______

M.
IRATECES,t
Co * -- _ __ _
NN
-,
40
2
1
-3, . 2.4 3-26
I
.0
(FO\E~E
.8 0
I .
80 3 0.
' \0
f/No, DO REQUIRE ED ORF 1­

5 i t i Ea

0.
3K4 %K=6 2.4
2 rrr robability

0-1
The two parameters A and K can be written in terms of the damping
coefficient S and natural frequency cn"

6= 1 2
6 Z fal--I['M
I W2 = ai
in aAK

Also,

BL 1T H (i.)I dio

and

B _ AK + a for the H(S) above.

L 4

It is common practice to fix one of the two loop parameters to give a

"good" transient response. With 6 = 1/ 4-2 the PLL is said to have a good
balance between stability and speed of response. This leaves one param­
eter undeterminea; this parameter is B L through its relationship to wU

1 1
but =

a/AK = L/Z

B AK+a 3
BL 4 ; AK.
12 32

=aAK = (AK) z =-BL

nO 9 Ls

Although the second-order loop can theoretically "track-out" any

constant frequency off-set (doppler shift), the pull-in range is usually
limited to several times B L . When the off-sct is within this range, the
time to pull-in is given by Gardner (RIeference 9-7) as

9-11
 100

oC
200
{ 2. Gr1z
14GHz
0 12 MF1t
O.Y4Mri

7
S0.5 -i501 .T
0 MAXMU A_ 2f,-

n I ­

0 A1

to
50I

o
.
0.0,
.

0.02- poo
-.0
V 3 e3 V
0Es
L. 61
•o v
.0_._
n-
-
VALUEFOR.
T EECODES

AND PERFECT
z

20
05I4

I!
lIo

.0I 2 i
03

I 0 2.0

_______
30 02. 0.4

_______ _____
1

Figure 9-7.
Versus
-20, EL/N
Loop Noise Bandwidth (BL)

5 is proportional to loop SNR

5
ot
N* RI______
ene6 For
0.5
0.2

0.1
Fl-T

_____ I_____
1.6
N0

L4 _ _ _ 002
0_ _ _

0.21 0.2 0.4 0.6 0.8 1.0 1.2

i Figure 9-8. Loop Pull-In Time Versus

Loop Bandwidth For Expected Frequency
tOffsets (Without Noise)
2

S0.6 _ __

0.4

0.2

0­
0.2 0.4 0.4 0.0 1.0 1.2

Y"IAHZ
?igure 9-9. Ma-xinaurn Sweep Rate
Versus Loop Noise Bandwidth (BL)
(Without Noise ),
9-12
 2 1
4. Z (Af)
sec for 42
BL

The second-order loop cannot perfectly track a frequency ramp

e'tuppler rate). The steady-state error is given by

Phase er-ror - a
n

9 )
32BL

The PLL cannot attain or maintain lock if the nL, is much greater than
I/Z wZ . This characteristic determines the rate at which a VCO can be
swent in order to enlhance loop acquisition.
Figures 9-8 and 9-9 are plots of the above equations. At fc = 14 GHz,
the time to lock becomes greater than 10 seconds when B L < 0. 6. If it is
determined that B L mu~t be less than about 0.6 MHz, sweeping of the VCO
should be considered. At fc = Z. Z GHz, the acquisition time will be only a
few seconds for any BL of interest. , Since the doppler rate is very small,
the steady state error is negligible (that is, the principal contribution to
tracking error is the channel noise). If VGO sweeping is desired at
fc = 14 Hz, the permissible rate is 0.05 MHz/sec at B L = 0. Z MHz. The
entire uncertainty of 1.48 MHz can be swept in 30 seconds.
Conclusions

A squaring loop can provide a near-perfect carrier reference in a

high rate convolutional coded biphase PSK system in which the data is
directly modulated on the carrier (such as the Station/DRS data link). A
second order loop with BL I MHz gives less than 0.4 dB degradation in
the station/DRS data. Such a loop can also acquire and track the maximum
frequency offset within a 3-se6ond lockup time. Since the carrier compo­
nent tracked by the PLL is very "clean, " false lock is not considered a
difficulty.
The use of programmed doppler offset during acquisition would per­
"li the lLoop bandwidth to be reduced by an order of magnitude. This would
truatly improve the loop's tracking capability.

9-13
9. 3 DESIGN STUDY OF THE SPACE STATION TDM SYSTEM

The structure of the Space Station Data Relay Link TDM System is
determined to a very large extent by the format of the main data frame.
The various factors influencing the TDM frame design are listed in
Table 9-2 and are discussed in the following paragraphs. The design
resulting from this study is detailed in the discussion of the baseband
subsystem.

System Constraints

At this early period in the design of Space Station's communications

system, it is desirable to maintain a high degree of simplicity and sym­
metry within the system structure. This yields a system which can be
readily modified at the detail level during future design iterations and still
retain the desired system characteristics. The future need for asymmetry
or special services should receive special scrutiny in order that a "patch-i
work" design does not result. To assist configuration management, and
simplify the interfaces, the data rates at interface points with other sys-
tems have been set at the IRIG standard of 75 x ZN bits/sec, where
N = 0, 1, 2, . . ., . Table 9-3 shows the rates to which the communications
system is constrained.

TABLE 9-Z. FACTORS INFLUENCING THE TDM FRAME DESIGN

Consider overall system problems.

e IMaintain system symmetry and simplicity

o Maintain standard interfaces
o Design for high information content by minimizing
the percentage of bits used for sync and control

Provide requiied information capacity

o Consider the error quality which the various data

types require
G Consider the need for estimates of the channel
quality for the various data types
o Minimize buffer complexity

Provide control words for system flexibility

Provide a frame synchronization word

a Consider both scan and maintenance mode requirements

(i Consider channel noise, internal noise, and expected
data drop-out times

9-14
 TABLE 9-3. STANDARD BIT RATES (75 x Z)

Bits/Sec Rilobits/Sec. Megabits/Sec

0 75

150

2 300

3 600
4 1. z

5 Z.4

6 4.8
7 .9.6

8 19.2

9. 38.4

10 76.8

11 153.6

12 307. Z

13 614.4

14 1.2288

15 2.4576

16 4,9152

17 9.8304

18 19. 6608

19 39.3216

20 78. 643Z

9-15
 In formatting the TDM frame, the number of bits devoted to the sync
word and control functions should always be minimized to use the RF link
c:fficiently. Since the link performance can rarely be designed (or even
mjeasured) to better than 0. 1 dB (or 1. 0Z3 numeric), any loss in information
capacity 4ue to framing and control bits can be ignored if these bits consti­
tute only about 2. 3 percent of the data stream.

Data Requirements

Representative data types and their nominal rates are listed in

Table 9-4. It is apparent that except for the delta modulated voice, these
nominal rates are not commensurate with the standard rates.

No requirements have been placed on the need to monitor the quality

(bit error rate) of the relay link. It will be assumed that link quality
measurements are an ancillary function of the frame sync word recognizer.

The multiplexing of the various data types of Table 9-4 involves the
buffering and interleaving of several high rate data streams. The complexity
of these buffers increases with the ratio of the input/output rates and the
number of bits within the buffer at any bne time. As the frame rate is
increased, fewer bits are required per frame, and the buffer size is
reduced. However, as the number of information bits per frame is made
small, the framing bits may become an unacceptably large portion of the
data stream. For a preliminary design, let the bits constituting the "over­
head" be fixed at around Z percent of the total bits per frame. The buffering
required will be assumed to be acceptable.

TABLE 9-4, REPRESENTATIVE DATA SERVICE REQUIREMENTS

FOR RELAY LINK (FULL-DUPLEX)

Type Nominal Rate Bit Error Requirements*

3
12 voice channels 19. 2 kilobits/sec 10
(delta modulation) (per channel)

Digital television 12.0 megabits/sec 10­ 4

Experiments data 3. 0 megabits/sec*- 10 -4

-Mnoden bit error rate has been fixed at 10- 5 .

S5j.C|taneous with 12. 0 megabits/sec TV.

9-16
 .- -! Requirements

Table 9-4 summarizes the data services of the RF communication

v Ltenl.
-. It is assumed that it is necessary also to provide bits which indi­
c.te which of the data channels is indeed active. (When a channel is inac­
tive, bits-from a pseudo-random generator will be inserted into the data
ctream. ) The signaling function for the voice circuits is assumed to be
performed in-band. Bits should be provided to indicate whether a high rate
channel is devoted to television, is assigned to experiments data, or is not
being used at all. These three states require that at least two bits be
appended to the data per channel. Although these bits themselves may not
be used to reconfigure the TDM system receiving the data, they do provide
a positive indication of the data present in the channel at any given time.

Synchronization Word Constraints

In a TDM system, the various data types arrive at the demultiplexer

in time sequence. In order to separate the TDM frame into the appropriate
channels, the demultiplexer must recognize wheil a new frame begins. This
can be accomplished by prefixing a frame synchronization word to each
frame. The frame time synchronizer that uses the sync word operates in
two modes: A scan (or acquisition) mode and a maintenance (or tracking)
mode. A typical synchronizer and its analysis is presented below. (The
final design arrived at is very similar.

Scanning

During scanning, the "recognizer" correlates each successive n

data bitswith an n-bit frame sync word reference. When the recognizer is
examining random ones and zeros, the correlation will be near zero. When
both framing bits and information bits are present, the use of a proper sync
code will result in a negative correlation with the reference word. (This is
the "overlap" region. Note that to the recognizer, all bits which are not
framing bits are "information" bits. ) If the sync word itself is correlated
with the reference word, the correlation will be at a maximum. Since chan­
nel errors may corrupt the received word, there is always a porssibility that
even when the sync word is actually encountered, the measured correlation
may not reach the theoretical maximum.

While scanning, there is a certain probability that n successive

random data bits will occur in the frame sync word pattern. This probability
can be bounded above (see Reference 9-9) when it is assumed that:

1) The true sync word has just been missed.

2) The sync word is a random bit pattern (not an optimum sync

pattern).

9-17
 2
0o
0
z
Figure 9-10. Average Number of Frames
in Scanning Mode Versus Sync Word
Length, Given R = r/r + m
z6

z
Z2
0.2 ­ _.0o_
(M Versus m, Given R) 0,
0_
ASSUMPTIONS
5
0 R 99. - BER XU
SYNC WORDJUST MISSED

• RANDOM BITPATTERNUSED
FORFRAMESYNC WORD

0 02

0.01
0 IH 20 30 40 50 60

LENGTH OF FRAMESYNC WORD, BITS

10-2

10I

0
"
to 2 S -N MSE fO CJS

- 4
10

-~ BE -9.90
FRW
YNrWA
FOR~~

09-1
0 -1

10-16I -- WhleSanin erusFam Sn

LEOT SYNC WORD

OFRAME , IS(tVru ,GvnR

9-18
 probability
c of false sync on the f$rst frame is

+n - I
P = I (I - Pi ) r

n = number of bits in sync word

rl-n = total frame length

n
P. = =1
(i) since each bit has a probability of 1/Z of occurring
as the correct framing bit.
, on the average, vi frames will actually be scanned, where

M= (1-
P
+ P.C+ (I1 P r+n ]
P. (r--n)
P fs -fs )
i

P =

Pc = (I- Pee-) n
Pe = bit error probability

e assume no errors are permitted in word recognition. Figure 9-10

S vr + n Once scanning is completed, the
2 .hability that the established sync is false is given by

P 2ifs
Ptfs f PPffs + PPlP:
c (I - Pfs )

;2 "igure 9-11, the probability Ptfs < 10-3 for R = 99 percent and any
I ( his means that the system will attain true sync'within only
a very
Irnns,) When a false sync condition is discovered, the system returns
:,In mode and again searches for Lrue sync. The average number
Ss - examined before false sync is rejected is given by

1
J - I-P!

9- 19
 3.0

2.5 -_

2.2
Figure 9-12. Frames Until False Sync
2.0is Rejected-Versus Sync Word Length,
0 k= 3isRjce
Given Number of Errors Permitted

1. (5 Versus rn, Givcn k)

o > 5 ASSUMPTIONS
-5
'BER 10
k 2
k-2
1.2

1.0
0 10 20 30 40 60 1i6

=LENGTH OF FRAMESYNC WORD, BIS

1014

, 1010

108 -
___.
-. K=I

10 2

Figure 9-13. Maintenance Mode Rejection L ASSU' TIONS,

Versus Word Length Given That k <S=I

Errors are Permitted ,0 o 40 50

(K Versus m, Given k) LENGTH OF FAJME SYNC WORD, BITS

- 0
v.here

k In
P= 1 N '

and k = the number of errors permitted in the pattern. The quantity J is

plotted as a function of n in Figure 9-1Z.

The Maintenance Mode

When a frame synchronization system concludes that it has found the

beginning of the frame, it passes from the scan to the maintenance mode. In
the latter mode, the system "threshold" is reduced by permitting one or more,
errors in the sync pattern so that occasional channel noise will not disrupt
synchronization. The system also sets a "clock" so that it only examines
the n bits which arrive at the predicted beginning of a new frame. Fig­
ure 9-13 shows the average number of frames which will pass through a
system before sync is rejected due to channel noise. Although it appears
that one should lower the threshold significantly, i. e. , increase the number
of permitted errors, it should be remembered that this increases the number
of frames required to reject a false sync condition.

EMI Considerations

The above analysis on synchronization performance assumes that the

only perturbing influence is channel noise. Problems of internal electromag­
netic interference and data drop-out are also quite important. The design
suggested in the next section considers the effects of the Space Station power
system in terms of transients, noise at the power distribution frequency,
and noise due to ac ripple terms. it is undesirable that only occasional short
data drop-outs or error bursts due to power transients should cause the
frame synchronizer to return to the scan mode and cause further data loss.
A design should be considered which fly-wheels' through the expected power
transient times, The frame period should not be synchronous with the
period of expected EMI. The interference could cause frequent loss of sync
(if it occurred during the sync word) while the data would be unaffected. Con­
versely, the data c.ould be degraded periodically without the sync word recog­
nizer giving any indication of data quality degradation,

S,;lction of the TDM Frame Time

Figure 9-14 summarizes some of the.considerations involved in

'"lt'cting a TDM frame time. The curves are plots of the frame period
rsus n given r + n" Since r includes information at 19.66 megabits/sec
- fed
. number (64) of format bits, the frame period is given by

9-21
 POWER DISThIBUTION

\4oz)FREQUENCY
=
R fI\ 9. .FREQUENCY AFTER
I _
__ _- z), RECTIFICATION

z
0
S0.5 _ _ _ _ _ _
9.u 0",- PERIODOF AC RIPPLE(2.4 kiz)

0
0.2 ----- - DESIGN POINT _ HARMONIC
FIRST
RD-PLE

0. - ,-SECONDHARMONIC-- - >­

0,05'
0p5
-TI

THIRDHARMONIC
HAhO I -

I
"I -1 ­
o

-FOURTI HARMONIC HO
0.02

0.01 ___ ____ __ _

1 2 5 10 20 50 100 200 Sco woo

- LENGTH OF FRAMESyNC WORD, BITS

Fignre 9-14. Constraints on Main Frame Period

9-22
 I
P = (r-64) 6
19. 66 x 10 bits/sec

(n 1 RR - 64) x (0. 051 x seconds).

llhe power system constraints on the design are shown along the right hand
side. The dc regulator frequency is in the range of I to 10 Kl-Iz. As a rule­
of-thunb, the period of the switching transients will last between 0. 01 and
o.10Ims. The primary power is distributed as three-phase, 400 Hz ac;
single-phase, full wave rectification may cause EMI at 800 Hz. Three­
:)hase ac ripple on the dc supply will be seen at Z.4 kl-iz and its harmonics.
From Figure 9-14, it appears reasonable to choose a frame sync time of
0. mns, and a sync word length of Z0. The information percentage is high
and the system should be relatively immune to EMil.

The synchronization scheme chosen .for the Space Station TDM system
is a simple extension of the typical approach discussed above. During scan,
two successive pattern recognitions (with no errors) must occur before
:ync is established. During the maintenance mode, sync is not assumed
lost unless at least two errors are discovered in two successive pattern
4xainatons. At the frame rate selected above, this approach permits loss
of one frame during power transients without a return to the scan mode.

9.4 REFERENCES

9,-i. W. J. Hurd and J, 0. Anderson, "Digital Transition Tracking

Symbol Synchronizer for Low SNR Coded Systems, " IEEE Trans.
on Com. Tech., Vol. 18; No. 2, April 1970, pp 141-147.

1-2. J. J. Stiffler, "On the Allocation of Power in Synchronous Binary

PSK Communication Systems, " Prpc. 1964, Nat'l Telemetering
Conf., June 2-4.
"-3. A. J. Viterbi, Principles of Coherent Communication, McGraw-Hill
Book Co., New York, 1966.
W. C. Lindsay, "A Theory for the Design of One-Way and Two-Way
Phase-Coherent Comnunication Systems, " TR 3Z-986, Jet Propul­
sion Laboratory, Pasadena, California, July 15, 1969.

'W.C. Lindsay, "Phase-Shift-Keyed Signal Detection with Noisy

Reference Signals, " IEEE Trans. on A and ES, Vol. 2, No. 4,
July 1966, pp 393-401.

J. I. Marcum, "Statistical Theory of Target Detection by Pulse

Radar, " IRE Trans- on I. T., p 174, April 1960.

9-23
 /
C,rdnr, Phaselock Techniques, John Viley and St,"

*A.\. Huller, "Sequential Decoding: Improved Perforn:anc,- *':

9 :tstr-int Length Convolutional
Codes, " JPL Space Progra:i
-tifu nary 37-56, Vol. II, pp 8Z-83.

j._ . Milstein, "Summary of Frame Synchronization Technicj-u,-,

HAC IDC ZZgz/655, March 10, 1969.

9-24
 10. EQUIPMENT SURVEYS

1 INTRODUCTION
Iw

To determine the level of current technology, several equipment

.surveys were performed during the study. The communications system
design described in Part I utilizes this state of the art technology.
Scction 10. 2 summarizes the power output and power efficiencies achievable
in current traveling-wave tube amplifiers (TWTA). TWTA physical charac­
t(.ristics are also included. Section 10.3 gives noise temperature data for
two types of microwave preamplifiers. System temperature estimates are
given, assuming two types of background noise. Section 10. 4 gives physical
,i,tta on hardware for source encoding and digital channel encoding. These
dt!signs do not require the development of digital devices beyond the current
state of the art.'

10.2 TWTA'S FOR SPACE APPLICATION (2 TO 15 GHZ)

. I
For medium power space applications above S-band, the TWT
remains the most desirable transmitting device. A survey of available
"IWTA's is summarized below.

Output Power, Gain and Efficiency

Currently available continuous wave spaceborne TWTA's typically

:enerate 10 watts RF with a 45 dB gain and 33 percent efficiency. &-s
3
Piture 10-1 shows, the characteristics of a given unit can be modified by
increasing the beam voltage, but increased output power is achieved at the
--xpense of decreased gain.

nandwidth, Gain Flatness

Bandwidths of 10 percent or more are attainable depending upon the

.'.:1n-f1atness specification. For example, at S-band carrier frequencies,
9Ts
make an RF bandwidth of greater than ZOO MHz a possibility. The
VLtdband nature of these devices is also a liability in that they require RF

10-1
 0 . OUTPUT -- I­

-4 2MA2B
BEA EFFECENCT
38

.z

0 32

OVERALL EFFICIENCY AO
SSSTATE-OF-THE-ART TOTL(970)
FrASSUMES TWO TUBESIN PARALLEL

0 -­ 40% INCLUDES POWE R SUPPLY

459O R EFFICIENCY OF 0%

200 / 2001

r m2 2W­

25 2 ~ a m THIHR ARIRFREQUENCIES225
- 50 _____ n 100 ______

o 0.

SBAND

0 .0 7-F =
020 40 60 SO I0 120 5 10 20 50 1
POWER OUTPUT,
IE WATTS
CAERIEN
FPEQUENCY, GGZ

'i,Pre 10-Z. Variable Power Operation Figure 10-3, TWgT Transmitter,

of the WJ-395 Efficiency and Power Output
Versus Frequency

3.0-2
 i, t, Weight, and Prime Power

The TWT units themselves typically weigh 18 ounces, and occupy a

space 2 by 2 by 10 inches. Each tube, however, must be accompanied by a
power supply which delivers up to 5, 000 volts do. These units are typically
b0 to 85 percent efficient, weight 3 pounds, and occupy a space 4 by 6 by
Z inches. -Their weight and size increase with the power level handled, the
voltage, required (which generally increases with frequency and efficiency),
and the number of telemetry points taken within the unit.

Reliability

TWTA's have demonstrated excellent reliability in both actual space

missions and laboratory life tests. The failure of a spacecraft communica­
tion system has yet to be traced directly to a Hughes tube. Life tests of
space-qualified TWTs have yet to produce a failure. After more than
8 years, Syncom 2.5 watt L-band tubes are still operating at the Hughes
Electron Dynamics Division lifetest facility. Their only degradation appears
to be a 0. 1 dB drop in power level every 3 years.

Available TWTs

Table 1 0-1 is a partial list of the units now available from Hughes and
is representative of the current technology. The list is ordered according
to frequency, but as the missions of the tubes indicate, this is also the
approximate order of tube deirelopment. The movement to higher frequencies
has prompted TWT development. Up through 1 0 0Hz, TWTs are available
which provide 20 watts output with 33 percent efficiency. Although tubes at
higher frequencies have been operated in space, their output levels and
efficiencies are notably lower. At 14 Gl-Iz, the only Hughes TWT which has
been flown delivers 1 watt with an efficiency of 20 percent.

Of the tubes developed by the many other companies in the TWT field,
two tubes built by Watkins-Johnson for JPL should be singled out. (Refer­
ence 10-1) The tubes are shown in Table 10-Z to have been designed for
high output levels and high efficiencies (at the expense of bandwidth and
gain). Furthermore, their output levels can be adjusted over a 10 dB range
by simply programming the power supply voltages. Figure 10-Z shows how
the efficiency of the 100 watt varies with output power and can be used to
determine the true prime power savings when operating at reduced output
levels. Although these tube'designs have sone remaining problems, (Refer­
ence 10-Z) they may be very useful for Space Station S-band applications.

The current state of the art in TWTs in terms of efficiency and power
output is summarized in Figure 10-3 as a function of frequency. It can be
assumed that this technology will be improved significantly in the next
seve-ral years due to research and development for commercial and military
a t.plications.

l0-3
 TABLE 10-1. MEDIUM POWER CW TRAVELING-WAVE TUBES
(PARTIAL LIST OF HUGHES TWT'S)

Powe Approximate* Overall

Frequency Power Prime Tube Saturated
Tube Range, Out, Power, Efficiency, Gain,
Band No. (GHz) watts watts percent db Status Mission

S 314H 1.5- 2.5 2.5 16. 20 33 Operational Syncom

Z14H 2.5- Z.4 8.0 30 33 Z7 Operational Pioneer
394H 1.8- 2.6 Z0.0 76 33 26 Operational (a) Apollo
C Z35H 3.0- 5. 0 10.0 37 34 43 Operational Intelsat ITT
261H 3.0- 5.0 6.0 24 31 50 .Operational Intelsat IV
X Z40H 7.0- 7.7 Z0.0 76 33 43 Operational (b) TACSAT
219H 7.5- 9.0 Z0...0-.- 7Z 35 30 Operational ---
K 837H 10.0-15.5 1.0 6 20 '45 Operational
274H 10.0-15.5 5.0 30 20 46 Experimental
818- 12.5-15.0 18.0 107 Z1 45' Operational(c) Lab TWTA
267 17.5-21.5 4.0 31 16 50 Developmental ComsatCorp.
Z68H 18.0-22.0 Z.0 17 15 42 Developmental ATS F and G
Z54H Z7.0-33.0 Z.0 Z3 11 4Z Developmental ATS F and G

*Prime power estimate assumes 80 percent efficient power supply.

(a) Also delivers 5 watts RF with 25 percent efficiency.

(b) Two tubes are used in parallel to deliver 40 watts.
(c) Although now used in a laboratory amplifier, this unit could be space-qualified.
 TABLE 10-2. WATKINS-JOHNSON S-BAND
50 AND 100 WATT TWT'S

:rApproximate Overall
Power Prime Tube Saturated
Frequency, Out Power, Efficiency, Gain,
Tube GHz watts watts percent dB

WJ-448 Z.3 50 105. 43 30

WJ-395 2.3 103 263 45 38

'Prime power estimate assumes 80 percent efficient power supply

10.3 MICROWAVE PREAMPLIFIERS

The noise temperature ranges of various state of the art low noise
microwave amplifiers are shown in Figure 10-4. "Figure 10-5 shows the
system temperature of a receiver with a "worst case" tunnel diode preampli­
fier and galactic background noise. Figure 10-6 shows the system tempera­
ture when a TDA is used and a Z90°K earth is "seen" by the receive antenna
(as in the case of a DRS).

10.4 VITERBI DECODERS AND ZOP DATA COMPRESSORS

Introduction

The digital units which are discussed in this section are located within
the overall communication system as described in Section 9. 1. Since the
Space Station links have been specified as full-duplex, their counter.parts will
also be required at the remote communications terminal.

Compressor (Source Encoder)

Table 10-3 shows the weight and power required for the data compres­
sor. The zero order predictor algorithn was assumed. This technique has
been found to have a good balance between performance and equipment con­
plexity (Reference 10-3). The algorithm compares a "new" sample value
with a prediction (which is the value of the last nonredundant sanrple). If
the difference between the new sample and the prediction is within some
tolerance, the sample is deleted from the data stream. If the new sample
is not redundant, it is "tagged" and sent to the buffer for transmission.

Encoder (Digital Channel Encoder)

The encoder schematic for a convolutional code of rate 1/h- and con­
straint length K is shown in Figure 10-7. When implemented, such a device

!0-.5
 I -,

5 2__ WIN

a -N

2 -UNNEL DIODE AMPLIFIERS

(REFERENCE 41 .

tREFEPENCE 41

, jLi

.1 I 02 O. 5 1 2 5 1 20 50 I N

FREQUENCY, GPZ

Figure System Noise Temperature Versus Frequency

A0-4.
Eeceiver

10-6
 5

2
EFFECTIVE
SYSlrTEM
NOtSE

NOISE

2\.

101 2
0.1 0 2 0 5 1 2 5 10 20 50 Too

FREQUENCY, GHZ

Figure 10-5. Effective System Noise Temperature Versus

Frequency (Data Relay to Space Station Link)

10-7
 5
/s­

' J
SYSTEM NOISE
EFFECTIVE
/
TEMPERATUREDUETO
TEMPERATURE_
1,EARTH •.
i .-
°3

z
_ 0

2
_ __ - AMPLIFIER TEMPERATURE

NOISE TEM'EAAIURE
EARTHEFFECTIVE
-- _

> 2
10

_____ _____
101
0.1 0.2 0.5 1 2 5 10 20 50 100

FREQUENCY, GHZ

Figure 10-6. Effective System Noise Temperature Versus

Frequency (Space Station to Data Relay Link)

10-8
 TABLE 10-3. ZOP DATA COMPRESSOR
PHYSICAL CHARACTERISTICS

Bit Rate Weight, Power,

(Beforei4:1 Compression) pounds watts

50 kilobits/sec 7. 9 11.4

3 megabits/sec 7. 9 11.4

50 megabits/sec 1Z. 0 45.0

Includes power supply (for each) and supply efficiency.

requires negligible weight and power compared to the rest of the system.
It would probably be advantageous to integrate the encoder with other
digital units and save packaging weight and volume. (The simplicity of the
encoder makes coding especially attractive for deep space missions where
a decoder is needed only at the ground terminal).

Decoder (Digital Channel Decoder)

Table 10-4 shows the physical characteristics of Viterbi decoders

for three bit rates and two codes. The highest bit rate of 12 megabits/sec
corresponds to a 4:1 compression of commercial quality digital television.
The codes assumed were suggested by MSC. Note that for rates above
several hundred kilobits/sec, .a parallel implementation is required. For low
rates, a serial design permits using fewer packages with corresponding
weight savings. Now, since the number of operations required by the
Viterbi algorithm grows exponently with K, the parts count and power con­
sumption also grows rapidly with increasing K. Operation at higher data
rates with the same code also requires greater power. (The ultimate decod­
ing speed is dependent on the speed of the available logic. The Linkabit
Corporation pow has a 40 megabits/sec system operational with K = 4, V = 2.
(Reference 10-4).

Expander (Source Decoder)

Table 10-5 shows the weight and power required for the data expander.
This device takes the data format generated by a zero order predictor and
outputs a conventional PCVi data stream.

Conclusions

The data compression and coding system discussed (4:1 compression

of 50 megabits/sec data and K = 8, V = 3 code) requires hardware which
t ghs approximately 50 pounds and consumes 21 0 watts.
Q

30-9
 K-STAGE SHIFT REGISTER
INFORATION

Ii . . vMOD-2 ADDER

CODES SYMBOLS GENERATED FOR EACH

INFORAATION BITSHIFTEDINTO CODER

Figure 10-7. Binary Convlutional Codes

10-10
 TABLE 10-4. DECODER PHYSICAL CHARACTERISTICS

Bit Rate = 50Kilobits/sec (Serial Decoder)

Code: K=6, V=3 K=8, V=3

Weight, pounds: 6.0 8.0
Power, watts: 17.0 30.0

Volume, cubic inches: 80. 0 110.0

Bit.Rate = 0.7 Megabits/sec (Parallel Decode'r)

Code: K = 6, V = 3 K = 8, V 3

Weight, pounds: 10.0 35.0

Power, watts: 30.0 100.0

Volume, cubic inches: Z10.0 720.0

Bit Rate = 1Z.5 Megabits/sec (Parallel Decoder)

Code: K = 6, V 3 K = 8, V 3

Weight, pounds: 10.0 35.0

Power, watts: 40. 0 150. 0

Volume, cubic inches: zi0.0 720. 0

Includes power supply (for each) and supply efficiency. Assumes available
high speed units.

10-ii
 TABLE 10-5. ZOP DATA EXPANDER PHYSICAL
CHARACTERISTICS

3it Rate. Weight, Power,

(After 1:4 Expansion) pounds watts
50 kilobits/sec Z. 5 1.7

3 rnegabits/sec Z. 5 1.7
50 rnegabits/sec 5.6 15.5

Includes power supply (for each) and supply efficiency.

For future tradeoff analyses, it should be noted that data compression

is especially desirable, since it lowers the data rate requirement of both the
encoder/decoder and the channel. The weight of the decoder is roughly
independent of the data rate for megabit rates, but the power it requires
increases with increasing data rates. The decoder weight increases greatly
with increasing constraint length. While the K = 8 code requires a decoder
which weighs over three times the decoder for K = 6, the link performance
is improved by only about 0. 8 dB (based on the results due to Heller
(Reference 10-5)).

10.5 REFERENCES

10-1. E. R. Dornseif and L.A. Roberts, "TModern Traveling-Wave Tubes

for Space Application, "Microwave Journal, Vol. 1Z, No. 4,
April 1969, pp. 55-59.

10-2. Robert S. Hughes, "Spacecraft S-Band 10-100 W R Power Amplifier

Tubes, " paper No. 70-506, AIAA 3rd Comnmunications Satellite
Systems Conference, Los Angeles, California, April 6 - 8, 1970.

10-3. "Potenial Applications of Digital Techniques to Apollo Unified S-Band

Communication System, " Martin Marietta Corporation, Denver
Division, February 1970, under Contract NAS 9-985Z.

10-4. Private Communication with R. Martin of Linkabit Corporation,

July 21, 1970.

1 1-5. J. A. Heller, "Sequential Decoding: Improved Performance of

Short Constraint Length Convolutional Codes, " JPL Space Programs
Summary 37-56, Vol. III, pp 8Z-83.

10o-12
 APPENDIX A. LINK POWER BUDGETS

The following tables present the final power budget sunmaries for
the baseline design of Part I. The link analysis approach used is discussed
in detail in Section 7. 2.

A-I
 TABLE A-i. TRACKING BEACON, DATA RELAY SATELLITE
TO STATION (f. = 14. Z6 GHz)

PERFORMANCE SUMMARY

Bit orate 150 bits/sec

lo- 3
..Bit error rate
3T tracking error 0.16 degrees

Channel coding at Baseband at RIE

Encoding PN code of length 10 Biphase PSK

Decoding Ref. PN generator Coherent integrate and dump

LINK SUMMARY

Radiated power - 16.5 dBw

Transmit antenna gain (beam edge) +15. 0 dB
Free space loss (Z3, 800 n.mi.) -208.5 dB
Ellipticity loss -1. 0 dB
Receive antenna gain (4 foot diameter) +41.7 dB
Receive pointing error loss (±0. 25 degrees) -0.5 dB
Link margin - 1. 3 dB
Received power at Station C = -171. 1 dBw
Noise power density (T= 3910 K) N = -202. 7 dBw/Hz
C/N = 31.6 dB- Hz
Chip rate (1500/sec)*0 -31.8 dB-Hz
Degradation in detector -Z. 0 dB
Detector margin - 1. 0 dB
EBIT/ECHIP +10. 0 dB
SI/N
BIT o
+6.8 dB
SNR in 3kHz IF* -3.2 dB
0K
System temperature calculation,
Antenna temperature = 75.4
(L-1) Z900, L = 0.5 dB = 35.4
Normalized preamp temperature = 280. 0
... System temperature = 390.8
'ItLermines tracking performance
- PN code of length 10 is transmitted every bit time.

A-2
 TABLE A-2. TRACKING BEACON,STATION TO
DATA RELAY SATELLITE (f = 15.35 GHz)

PERFORMANCE SUMMARY

Bit rate 150 bits/sec

Bit error rate 10 - 3

Channel Coding at Baseband at RF

Encoding PN code of length 10 Biphase PSK
Decoding, Ref. PN generator Coherent integrate and dump

LINK SUMMARY
Transmitter power (including output circuit losses) -10. 6 dBw
Line losses -1.4 dB
Transmit antenna gain (4 foot diameter) +42. 3 dB
Transmit pointing error loss (±0.2.5 degrees) -0.5 dB
Free space loss (Z3, 800 n.mi.) "-209. 1 dB
Ellipticity lOss -1. 0 dB
Receive antenna gain (beam edge) +15.0 dB
Link margin -0.9 dB
Received power at DRS C = - 166. 2 dBw
Noise power density (T = 1200 0 K) N = -197. 8 dB/HZ

C/N = 31.6 dB-Hz

Chip rate (1500/sec) -31.8 dB-Hz

Degradation in detector -2. 0 dB
Detector margin -1. 0 dB
EBT/ECCHIP +10. 0 dB

EBIT/N +6.8 dB

A-3
 TABLE A-3. DRS DATA LINK,DAXA RELAY
SATELLITE TO STATION
(fc = 13. 70 - 14.2-0 GHz)§
r­
r PERFORMANCE SUMMARY
Bit rate 20. Z megabits/sec
Bit error rate l0-

Channel Coding at Baseband at RF

Encoding (2, 5) convolutional code Biphase PSK
Decoding Viterbi decoder (q = 8) Filter and sample

LINK SUMMARY

Radiated Power per carrier +7. 0 dBw

Transmit antenna gain -+41.9 dB
Transmit pointing error loss (±0. 25 degrees) -0. 5 dB
Free space loss (23, 800 n. mi.) -208.3 dB
Ellipticity loss -1. 0 dB
Receive pointing error loss (:E0. Z5 degrees) -0.5 dB
Receive antenna gain (4 foot diameter) +41.9 dB
Link margin -1. 5 dB
Received power at station -C = -121. 0 dBw
Noise power density (T 390. 8°K)* N = -20Z. 7 dBw/Hz
CIN = 81.7 dB-Hz
Chip rate ( 4 0.4 x 10 6/sec) 76.0 dB-Hz
E/1N in downlink to station +5. 7 dB

Degradation due to uplink-,-, -0, 5 dB

Detector degradation -2, 0 dB

Detector margin -1. 0 dB

EBIT /ECHIP +3. 0 dB
EBIT IN to Viterbi decoder at station +5.2 dB

4Bandwidth of 80 AHz is required within 500 M1z band

'Sec Table A-I
Assumes a linear repeater with (Ec/!o)p _ 15. 0 dB

A-4
 TABLIE A-4. DRS DATA LINKSTATION
TO DATA RELAY SATELLITE
(fc = 14.85 - 15.35 GHz)§

PERFORMANCE S UlvMARY

Bit rate 20.2 negabitslsec

Bit error rate 10- 5

Channel Coding at Baseband at RF

Encoding (Z, 5) convolutional code Biphase PSK
Decoding Viterbi decoder (q = 8) Filter and Sample

LINK SUMMARY

Transmitter power (including output circuit losses) +16.4 dBw

Line losses -1.4 dB
Transmit antenna gain (4 foot diameter) 4Z. 5 dB
Transmit pointing error loss (-0. 25 degrees) - -0. 5 dB
Free space loss (23, 800 n. mi.) -208.9 dB
Ellipticity loss -1. 0 dB
Receive pointing error loss (+0. 25 degree) -0. 5 dB
Receive antenna gain +4Z. 5 dB
Link margin -3. 0 rIB
Power received at DRS C = -113.9 dBw
Noise power density T .=(1200°K) N = -197.8 dBw/Hz

C/N 0 = 83. 9 dB
6
Chip rate (40.4 x 10 /sec) 76.0 dB-Hz
E C IN 0 in uplink to DRS 7.9 dB
Degradation due to downlink** 2. 7 dB
Detector degradation -Z: 0 dB
Detector margin -1. 0,dB
EBIT /ECHIP +3. 0 dB
EBIT/N to Viterbi decoder at ground +5. 2 dB

§Bandtwidth of 80 MHz required within 500 IMHz band

:;'-*Assumes a linear repeater with (Ee/N 0 ) down ;- 8.6 dB

A-5
 - I
TABLE A-5. EMERGENCY VOICE LINK,STATION TO
DATA RELAY SATELLITE
(fc = 149. ZZ - 149. 3Z)§

PERFORMANCE SUMMARY

Voice bandwidth 3 kHz

Output SNR 32 6 dB

Channel Coding at Baseband at RF

Encoding Pre-emphasis FM

Decoding De-emphasis Threshold extension demodulator

LINK SUMMARY

Transmitter power (including output circuit losses) 19.0 dBw

Line losses' -0.9 dB
Transmit antenna gain -3.0 dB
Free space loss (Z3, 800 n. mi.) -168.9 dB
Ellipticity loss -1. 0 dB
Receive antenna gain (beam edge) +15.0 dB

Link margin -1.0 dB

Power received at DRS C = -140. 8 dB
Noise power density (T s = 800 0 K) N = -199.5 dBw/Hz
G/N O = +58.7 dB-Hz
Receive bandwidth (36 kHz + 7 kHz doppler) -46.3 dB.Hz
Received C/N at DRS 12. 4 dB
Degradation due to downlink* -5.4 dB
Demodulator margin -1. 0 dB
C/N to threshold extension demodulator +6. 0 dB

T
' wo 50 kHz channels available for uplink
"'Assunes a linear repeater and (C/N) down 8. 5 dB

A-6
 TABLE A-6. EMERGENCY VOftE LINK,DATA RELAY
SATELLITE TO STATION (fc '135.60 - 135.70 MHz)§

PERFORM\ANCE SUMMARY
Voice bandwidth 3 kHz
output SNR 3Z. 6 dB

Channel Coding at Baseband at RF

Encoding Pre-emphasis FM
Decoding De-emphasis Threshold extension demodulator

LINK SUMMARY
Radiated power (per carrier) +12. 7 dBw
Transmit antenna gain (beam edge) +15. 0 dB
Free space loss (Z3, 800 n.mi.) -168.0 dB
Ellipticity loss -1. 0 dB
Receive antenna gain -3. 0 dB
Link margin -1. 0 dB
Power received at Station C = -145.3 dBw
Noise power density (T. - 723°K) No = ZOO. 0 dBw/Hz
I C/N0 54.7 dB'Hz
Receive bandwidth (36 kHz + 7 kHz doppler) -46.3 dB-Hz
C/N in downlink to station +8.4 dB
Degradation due to uplink* - 1. 4 dB
Demodulator margin -1. 0 dB
C/N to threshold extension demodulator +6.0 dB

System temperature calculation, °K

Antenna temperature = 250
(L-) 290 ° , L = 1.,5 dB = 119
Normalized preamp temperature = 354
System temperature = 723

§Two 50 kHz channels available for downlink

tt4ssumes a linear repeater and (C/N) up 2 12. 6 dB

A-7
 TABLE A-7. EMERGENCY VOICE LINK,SPACE STATION TO
ALS (f = 135. 60 to 135. 70 MHz or 149. 22 to 149. 32 MHz)§
C

PERFORMANCE SUMMARY
-,Voice bandwidth 3 kHz

Output SNR 32. 6 dB

Range 3000 n. mi.

Channel Coding at Baseband at RF

Encoding Pre-emphasis FM
Decoding De-emphasis Threshold extension demodulator

LINK SUMMARY

Transmitter power (including output circuit losses) 19.0 dBw

Line losses -0.9 dB
Transmit antenna gain -3.0 dB
Free space loss (3000 n.mi.) -150.3 dB
Ellipticity loss - 1.0 dB
Receive antenna gain -3.0 dB
Link margin -7.0 dB
Power received at ALS C = - 146.2 dBw
Noise power density (T s = 800 0 K) N = -199.5 dBw/Hz
C/N = 53.3 dB Hz
Receive bandwidth (43 kHz) -46.3 dB Hz
Received C/N at ALS 7:0 dB
Demodulator margin -1. 0 dB
C/N to threshold extension demodulator +6. 0 dB

' Both the Station and the ALS can either transmit or receive within these
bands. Each band containing two 50 klz channels.

A-8
 A8LE A-8. EMERGENxCY VOICE LINK,ALS TO SPACE STATION'
(fc = 135.60 - 135.70 MHz or 149.22 - 149.3Z MHZ)§

PERFORMANCE SUMMARY

v" "- bamndwidth 3 kHz

0- Ut SNR 3Z. 6 dB
Rtange 3000 n. mi.

Channel Coding at Baseband at RF

Encoding Pre-emphasis FM
Decoding De- emphasis Threshold Extension Demodulator-

LINK SUMMARY

Radiated power (per carrier) 20. 1 dBw

Transmit antenna gain -3.0 dB
Free space loss -150.3 dB
Ellipticity loss -1.0 dB
Receive antenna gain -3.0 dB
Link margin -9.5 dB
Power received at Station C = - 146. 7 dBw
N4oise power density (T s = 723 0 K) N = -200.0 dBw/Hz
C/N = 53.3 dB.Hz
Receive bandwidth (43 kHz) -46.3 dB-Hz
Received C/N at Station 7. 0 dB
Demodulator margin -1. 0 dB
C/N to threshold extension demodulator +6. 0 dB

"Both the Station and the ALS can either transmit or receive within these
,1;.n4s. Each band contains two 50 kHz channels.

A-9
 TABLE A-9. ALS DATA LINK,SPACE STATION TO ALS
(fc = 2100 - 2300 MHz)§

--
PERFORMANCE SUMMARY

Bit rate 57.6 kilobits/sec

Bit error rate 10 - 5

Range 303 n. ni.

Channel Coding at Baseband at RF

Encoding (Z, 5) convolutional code Biphase PSK
Decoding Viterbi ddcoder (q = 8). Coherent integrate and
dump

* LINK SUMMARY
Transmitter power (including output circuit losses) 20.4 dBw
Line losses -1. 0 dB
Transmit antenna gain -3.0 dB
Free space loss -154.2 dB
Ellipticity loss -1. 0 dB
'Receive antenna gain -3.0 dB
Link maigin -1. 9 dB
Power received at ALS C = -143.7 dBw
Noise power density (T "800-K) N = -199.5 dBw/Hz

CINO = 55.8 dB-Hz

Chip rate (115. 2 x 10/sec) 50.6 dB-Hz
Received E - IN 5.2 dB
c 0
Degradation in detector -2.0 dB
Detector margin -1. 0 dB
EBIT /ECHIP . +3. 0 dB
EB/N to Viterbi decoder 5. Z dB

Kink requires 244 kHz within this band.

A-10
 TABLE A-I0. ALS DATA LINK,ALS TO SPACE STATION
(fc = Z100 - 2300 MHz)5

- PERFORMANCE SUMMARY

Bit rate 57.6 kilobits/sec

Fit error rate 10 - 5

Range 303 n. mi.

Channel Coding at Baseband at RE

Encoding (Z, 5 convolutional code Biphase PSK
Decoding Viterbi decoder (q = 8) Coherent integrate and
dump

LINK SUMMARY
Radiated power (per carrier) 18.4 dBv
Transmit antenna gain -3. 0 dB
Free space loss -154. Z dB
Ellipticity loss -1. 0 dB*
Receive antenna gain -3z 0 dB
Link margin -3.6 dB
Power received at station C = -146.4 dBw
Noise power density (T s 437.8-K) N = -202.2 dBw/Hz
C/N = 55.8 dB'Hz
Chip rate (115.2 x' 10/sec) 50, 6 dB-Hz
Received E c /N o 5.2 dB
Degradation in detector -2. 0 dB
Detector margin -1. 0 dB
EBIT /ECHIP +3. 0 dB
EBIN to Viterbi decoder 5. Z dB

System temperature calculation, °K

Antenna temperature = ZOO. 0
(L-l) 2900, L - 1.3 dB, = 102.0
Normalized preamp temperature = 135.8
sten temperature = 437.8
• 'q es 244 IHz within this band.

A-i1
 TABLE A-II. EXPERIMEIT MODULE DATA LINK,
STATION TO EM 1 (fc = Z100 - 2300 MHz)V

PERFORMANCE SUMMARY

Bit rate 7. 38 riegabits/sec

Bit error rate -10 ­ 5

Range 4"50 n. mi.

Channel Coding at Baseband at RF

Encoding (5, 2) convolutional code Quadriphase PSK
Decoding Viterbi decoder (q = 8) Coherent integrate and
dump

LINK SUMMARY
Transmitter power (including output circuit losses) 20.4 dBw
Line losses , k -1. 0 dB
Transmit antenna gain - 3., 0 dB
Free space loss -157.9 dB
Ellipticity lo ss -1. 0 dB
Receive antenna gain +20.0 dB
Transmit pointing error loss -0.5 dB
Link margin -2.3 dB
Power received at EM 1 C -125.3 dB
Noise power density (T s = 437. 80 K) N O = -Z02.Z dBw/Hz
C/NO = 76.9 dB'Hz
Chip rate (7.38 x 10 /sec) -68.7 dB-Hz
Received Ec IN 8. Z dB
Degradation in detector -2.0 dB
Detector margin - 1.0 dB
BIT/ECHIP.. 0. 0 dB
EB IN to Viterbi decoder 5.Z dB

9 Link required 15 MiHz within this band.

*The V = 2 encoder generates two subbits per bit; the quadriphase modu­
lator outputs one chip per twio subbits. See Sec 7.2.

A%-12
 TABLE A-IZ. EXPEREVIENT MODULE DATA LINK,
EM TO STATION (f = Z100 - Z300 MHz)Y

- PERFORMANCE SUMMARY

Bit rate 7.38

- 5 megabits/sec
Bit error rate 10

Range 450 n. rni.

Channel Coding at Baseband at RF

Encoding (5, 2) convolutional code Quadriphase PSK
Decoding Viterbidecoder (q = 8) Coherent integrate and
dump

LINK SUMMARY

Radiated power (per carrier) 19.4 dBw

Transmit antenna gain 20. 0 dB
Transmit pointing error -0. 5 dB
Free space loss -157.9 dB
Ellipticity loss -i. 0 dB
Receive antenna gain -3. 0 dB
Link margin -Z. 3 dB
Power received at station C = -1Z5.3 dBw
Noise power density (T. = 437. 8°K)* N = -Z02. Z dBw/Hz
C/N = 76.9 dB Hz
Chip rate (7.38 x 10 /sec) -68.7 dB Hz
Received E IN 8.2 dB
C 0
Degradation in detector -2. 0 dB
Detector margin - 1, 0 dB
EBIT /E CHIP 0.0 dB
E BIN to Viterbi decoder 5. Z dB

5 Linkrequires 15 MHz within this band.

.Ste Table A-10

A-13
 APPENDIX B. ELECTRONICALLY SCANNED ARRAYS
FOR COMMUNICATION DURING
ARTIFICIAL-G OPERATIONS

1:22 ROD UCTION

Space Station operation in the artificial-g mode imposes the

reouirement that the K-band antenna system must provide spherical coverage
frun a platform rotating at 4 rpm. It appears that this spherical coverage
r.,quirement can be met by the use of multiple mechanically steered antennas
t.s used in the baseline design). " At the request of the Manned Spacecraft
Center, a special study was undertaken to consider the possible use of multi­
ple electronically steered arrays. It was concluded that the technique was
ii.deed feasible, but that the approach requires far greater weight and volume
than the mechanical system. The ESA system weighs Z563 pounds and utili­
.0es SO square feet of spacecraft surface to a depth of up to 2. 5 feet.

The study summarized below considers problems of coverage, weight,

g;,ze, power, and the state of the current technology. Recommendations for
fiture work are also inclfded.

SPHERICAL COVERAGE WITH ESA

To obtain spherical coverage with electronically steered arrays

requires an assembly of arrays that are a compromise among many fac­
tors. The number of apertures and the orientation with respect to each
other is determined principally by the design and configuration of the
Station structure. Consequently, it is not feasible to consider all possible
, onfigurations to obtain the configuration that is "best" in some sense.
What has been done for this study was to consider that coverage which is
obtained by using a number of planar arrays that would each cover a dif­
I,'.-,nt but overlapping angular region. As the pointing direction moves
irn). one angular region to another, the appropriate array would be
witched on, all others would be off. Such an arrangement would then
•juire only sufficient electronics for a single array, plus additional logic
4--i control circuitry to perform the switching operation.

The determination of the number of arrays and the number of ele­

:t to .. perform the scanning electronically nust take into account a
.Lwer of factors. The number of radiating elements required depends on
1,:. t"1 coverage region of each array and on the required array gain. To
tz:tanid this dependence, consider the followving argument. The beam­
,tjf each element in an array must be large enough to encompass the

B-1
 (A) BACK-TO-BACK PYRAMIDAL FRUSTUMSNA=IO

1­

Figure B-1. Array Configurations That

Provide Spherical Coverage by Mveans
of Electronic Scanning

T3-2
 -age region of that array. if t e coverage region is large, then
.i-tnaxst be electrically and physically small. A large number of
:,.r rray is then required to obtain the desired gain. If the array
. cLgion is small, the element beamwidth can also be small. Con­
.e, th clement gain is higher and fewer elements are required.
.Yr tys are required, however, to cover 4r sr so that the total number
-nts
. does not vary rapidly. There is, however, some optimum hum­
*rravs that result in a minimum total nunber of elelents. In addition,
: ,iectronics
v is to be shared among arrays, it is desirable to attempt
-u:;.ize the number of controllable elements per array. This approach
lylcannot be pursued to its limit, however, since there would result
...:, array for each beam pointing direction. Some compromise is
,,re required.

In a previous study related to hemispherical coverage, it was found

t tive arrays whose faces lay on the surfaces of a frustum of a pyramid
af vquate coverage. Two such configurations back-to-back - that is,
i-rays - would give spherical coverage (see FigureB-i). With this
t .::njent the maximum scan angle required of an array is 47. 1 degrees.
. -lcnients in the array are assumed to have a gain beamwidth product

GEP e 2 8, (B-I)

GcE is the element gain and Pe is the 3-dB beanwidth, then the
rgain is
it

GE z / = 8/(94. Z/57.3) z = Z.96

G- 4.72 dB

.tI.)IL (B-i) assumes a radiation efficiency of about 80 percent. The

of elenents in each array is then given by

2
e
N GA
NE= =GA - = 0. 3 3 8 GA (B-2)

the array gain. The number of elements are listed in

- lafor gains ranging fromn 20 dB to 49 dB. The table relates peak
!% na direction normal to the array to the total number of array
-: -However, if the gain is to be achieved at the extreme limit of
: gion, the number of elenents must be doubled, since at the
overMle region, the array gain is down 3 dB from its value at

3-3
TABLE B-1. NUMBERS OF ELEMENTS VERSUS GAIN FOR A
TEN-ARRAY CONFIGURATION (BACK-
TO-BACK PYRAMIDAL FRUSTUMS)

-- a. Peak Array Gain GA

Numerical NdB
NOE NNT = E

20 100 34 340
23 Zoo 68 680
26 400 135 1350
30 1000 338 3380
33 2000 676 6760
36 4000 1352 13520
40 10000 3380 33800
43 20000 6760 67600
46 40000 13520 135200
49 80000 27040 270400

b. Peak Array Gain GA

dB Numerical NE NT 10 N E

20 100 54 540
23 200 107 1070
26 400 214 2140
30 1000 535 5350
33 2000 1070 10700
36 4000 2140 21400
40 10000 5350 53500
43 20000 10700 107000
46 40000 Z1400 214000
49 80000 42800 428000

BA 41
 phase scanned arrays.
An additional source of loss is present in
"jzis loss occurs in the phase shifters and the feeding structure and may
.'ry' from 1 to Z. 5 dB or more, depending on frequency. If a value of
2 .ub is selected for illustrative purposes, the number of elements in each
Array would be given by

NE 0. 535 G A

The numbers of element required with this loss is given in Table B-lb.

The half-power beamwidth versus gain of a uniformly illuminated

circular array is shown in Figure B-Z for dissipative losses of 0, 1, 2 and
3 dB. The curve for 1 dB loss corresponds approximately to an element
gain beamwidth product of 8 and no phase shifter loss. The curve for
3 dB loss corresponds to an element gain beamwidth product of 8, with an
additional Z dB of phase shifter loss.

In the above section, an array of ten planar arrays arranged on the

faccs of the two pyramidal frustums back-to-back was considered as a
reasonable design approach to the problem of providing spherical coverage
with a minimum number of elements. Other arrangements include NA
arrays on the faces of a prism. When N4 < 10, this arrangenent results
in smraller scan requirements than those of the pyramidal frustun con­
figurations. For NA = 5 to 10, the maximurn scan angle, number of ele­
ments per array, total number of elements and element gain are listed in
Table B-2.

Because mutual coupling at the aperture varies with beam scan

position, an array cannot be impedance matched over the full range of
coverage. The resulting reflection losses at the aperture are minimized
by the selection of elements with a cos 0 power pattern (120 degrees beam­
width) and the reduction of the maximum required scan angle (one-half the
clement beamwidth). From Table B-Z, it is evident that to minimize scan
loss and the number of elements and to meet a gain requirement over the
range of coverage, there is some optimum value of NA. This value most
likely lies in the range 7 to 10. Probably the most important loss factor
ill selecting NA is the variation in ellipticity ratio as a function of scan
angle as ellipticity increases rapidly beyond 45 degrees.
/

..- 5
 TABLE B-Z. CHARACTERISTICS OF SEVERAL MULTI-ARRAY CONFIGURATIONS
THAT PROVIDE SPHERICAL COVERAGE

Array
Maximum Total Gain,
Number of Scan Number of Number Element GA
Arrays, Angle, Elements of Gain (I dB loss

Configuration NA Degrees Per Array Elements dB assumed)

Two pyramidal 10 47. 1 NE lONE = NT 4.72 2 . 96 NE

frustums

Prism 9 48.3 1. 052 NE 0.948 NT 4.49 Z.81 NE

Prism 8 49.1 l. 089 NE 0.86 N T 4.34 2.72 N E

Prism 7 51,0 1. 17Z N E 0.82 N T 4.0Z Z.53 NE

Prism 6 54.8 1:35 NE 0.81 N T 3.40 2. 19 NE

Prism 5 69.2 side 2. 16 N E side 0.83 N T 1.37 side 1.37 NE side

45.0 end 0. 9 13 NE side 5.10 end 3.24 N end
 1o

50

. 0 __ _ _ __ _ _

20

.5
02

0I 1
22 26 30 34 38 42 46 50

ARRAY GAIN, DB

Figure B-Z. Halfpower Beamwidth

Versus Gain of Uniformly
Illuminated Circular Aperture

B-7
 LOOK ANGLES
ARRAY
-
REFLECTARRAY
A FERATRE
RCOMWMAND BA
o PROCESSOR BEAM
STEERING
CONTROLLER

ORTIIOMODE
FEED
ARRAY 2

ROTATION
PATE
~COLIMi ND
PROCESSOR

TO RRYELECTION
< TO RECEIVEWTPANSMIT
RECEIVER E 0 SELECTSWITCHES
FROM
TRANSMITTER PA
APRAY
SELECTION B
0

NA

PROCESO

Figure B-3. Electronic Steering Antenna System

B-8
 ,tE$SENTATIVE ESA DESIGN

Coverage: Spherical

Gain: 36. 5 dB (minimum)

Frequency: 14GHz

Polarization: Dual sense circular transmit

and receive.
rn
-n- Technique

The method of electronic steering considered most suitable for the

-+,,c-Station application is termed command controlled phase scanning.
jN this technique, the antenna pattern is formed by electronically con­
trolling the relative phase of an incident signal incremently across the
.Lp:crture. Other techniques which accomplish the same phase steering by
,:sing fixed beam-forming matrices, were considered significantly less
ittractive because of their attendent weight, bulk and thermo-mechanical
zttortion problems.

To further reduce weight and volume, in consideration of the large

±unber of elements required, it was decided to use a reflectarray con­
!'huration with an drthogonal feed so that a complex corporate feed network
not be required. Figure B-3 is a block diagram of the recommended
I SA system.

As shown, each array aperture is fed from an orthomode feed and

i lectronically configured by instructions from a beam stee-ring con­
troller. The transmit and receive signals are automatically switched
:rqxn feed to feed as the desired target passes from the visibility of one
.rray to the next. While the target is within the scan angle of a single
.,rray, appropriate phase comamands are sent to that array's command
':pt cessor to point its beam accordingly. The beam steering controller
',-Quires data on look angles to the desired DRS in Space Station coordi­
- tts. (Alternatively, a tracking feed could be used with the reflectarray
:end tracking signals processed in the normal manner. However, the com­
I:ter facility required to generate the beam pattern should be able to
:t.zily perform look angle calculations.)
:-v-4tent Description

The ESA system that will best meet these requirements consists of
of electronics that is shared among an array of circular, planar
t'-.t
. oriented on the faces of a symmetrical prism or two back-to-back
.nis
of pyramids, The number of arrays will be between 7 and 10.

- -. . the angular relationship required between arrays; their actual

-.ti11g locatiuns can be distributed conveniently about the Space Station
ttr, as long as this relationship is maintained.

.A- 9
 PHASESHIFTEP-HYBRID SECTION

POLARIZERS
ORTHOS-MODVVRANSDUCERS
RECEIVE
RECEIVE AND TRANSIT FEED
NEA'MP'S

HORN FEED

TETNSDUCE STPRCUAL
SECTIONSUPPORT

//0.
DIAMETER
D
A) SIDE VUEW W}END VIEW

Figure B-4. Reflecarray Design For Command

Controlled Phase Scanning Array

FIN
RECEIVE" l I |

TRANSMIT TASTO

ORTHOGONAL
MO1DE
TRANSDUCER

Figure B-5. Circularly Polarized Fin-Loaded I-Torn"

B-10
 , the configurations -vere diagrammed in Figure B-I to show the rel­
...- ivntations of the arrays in either the prism or pyramidal frustum
. The arrays are connected to the electronics section, as
zn l'igure B-3. Eadh array is the same and consists of a reflector
:, n containing the array elements and reflective phase shifters and a
:.ens C? f6ed. This design referred to as the reflectarray design is
in Figure B-4. Eahh horn contains a polarizer and orthogonal node
...:cr. The polarizer converts a linear polarized wave to a circular
.o zcdwave, and the transducer separates the orthogonal transmit and
*.. ie signals (see Figure B-5). A typical array element, as shown in
.'c B-6, includes a crossed waveguide as the radiator, hybrid coupler,
.Mt) phase shifters (one for transrmit, and the other for receive). The
or
i separate phase shifters solves the beam-pointing error due to the
-,ificant frequency separation of the receive and transmit bands. The
z.Mscd waveguide radiator is uniquely suited for the close packing of
n-:icnt radiators required for wide angle scanning of a circularly.polar­
,! beaam. Figure B-7 depicts the best possible packing arrangement for
dn
v-. tvcguide b to a ratio of 0. 216.

ry Size, Radiating Efficiency, and Beamwidth

The gain requirement of 36. 5 dB will be net by providing enough

A .rture area to compensate for the losses itemized in Table B-3 and for
!:- 3 dB fall-off in gain \when scanning the beam from broadside to the
I hando-er angles.
3rAQst

TABLE B-3. ITEMIZED LOSSES IN 14 GHz REFLECTARRAY

Source of Loss Loss in dB

Phase shifter, 3-bit PIN diode*t 1. 6 minimun

hybrid couplert 0. 1
Spillover and tapered illumination 1. 3 minimum
reed blockage 0.2
Polarizert 0. 1
Phase quantization* 0. 2
I,'pedance mismatch at aperture due to scanning 0. 5 maxim-uln
50 degree scan angle)

Total loss 4. b dB
-!-it device is selected as a tradeoff between decreasing loss and in­
-:ing conputer storage capacity as the number of bits increases. A
Stvi~cZ will lead to about 0. 9 dB phase quantization loss, a 4-bit
abuut 0. 1 dB.

B-i
 HORIZONTAL 8
POLARIZATION

IEIVrR T(WIt4IALS 13OP

FCEI'E AND |IANSMIIT
COAXIAL PhASE I5IF]l RCOMMAI IDS
,
PORTS

PLAPIATIo ACKAGE FOR Tw

EHPHASE StIFTERS,
POLARIATIONDRIVERS, AND HYBRID
RADIATING COUPLER
APERTURE RADIATING FLEMENT

A)RADIATING ELEMENT FOR CIRCULAR POLARIZED El SIDEVIEW OF ELEMENT MODULE (LENGTHs

RADIATION 4INCHES AT 14 GHzI

TPANSWITTER
CONTROL HYBRID

DIGITAL PHASE RADIATING

SHIFTER. ELEMENT

RECEIVER TrANSMIT
CONTROL TRAPFILTER
C) ELECTRICAL CIRCUIT

Figure B-6. Array Element Module

22'

Figure F,-7. Closest Packing of Crossed Waveguide Elements

(Occurs When Tilted at Angle of Approximately 22 Degroes)

1.3,-1]
 That is, a lossless array with a peak gain of 43. 5 dB satisfies the
.,.!n-coverage requirement. From Figure B-Z, it is seen that the half­
,r'. bearxvidth will be larger than 1. 2 degrees; fron Table B-2, the
:.;n,,ber of elements will be

NE = GA(I dB loss)/Z. 7Z 17, 800/2.72 = 6550

1or an eight-array configuration. Assuming inter-element spacings of 0. 56

wavcength (X) for a square lattice arrangement, the array diameter will be

D = 0.56 ( E0.56.(0. 0705) T= 3.6 feet

and the surface area

w.D2
A = '_? = 10.Z square feet

Assuming a 2-foot spacing between the primary feed and the array
surface, the f/D ratio will be about 1/2. The depth of the array aperture
section and phase-shifter hybroid section -will be approxLmately 4 inches
(see Figure 6b).
i The basic antenna system components and their estimated numbers
and weights are listed in Table B-4.

A summary of array performance and mechanical characteristics is

given in Table B-5. The choice of a circular shape for the array leads to
a 17 dB first sidelobe level if the illumination of the aperture is uniform.
Since a tapered illunination is obtained from the basic reflector design,
sidelobe levels greater than Z0 dB should -be achievable.

RECOMMENDATIONS FOR FURTHER DEVELOPMENT

Percentage
Problem Area Remarks Effort
C Phase shifter Insertion loss must be 3D
reduced from present 2. 5
dB to about 1.5 dB.
Develop mass production
technique.

B-13
Th BLE B-4. PARTS, NUMBERS,!AND WEIGHTS OF ESA SYSTEM
Weight,
Component Number Pounds
Radiating element 52, 400 (eight Arrays) 157
Phase shifter 104, 800 Z, 180
Driver
Hybrid coupler 5Z, 400
F ramne 8 6
Back plate 8 84
Horn, polarizer, and ortho- 3z 38
mode transducer (four per
array)
Receiver preamplifier, feed 8 sets 48
supports, and miscellaneous.
Time division multiplex device 16
RF waveguide (5 feet) 16 8

Total Antenna System Weight Eight Arrays Z, 5241

Single-pole, 8-throw RF Z
switch
Logic circuits 1
Beam steering controller 1
Look angle computer 1

42

System Total 2, 563

B -14
 ". ]LE B-5. SUMMARY OF REFLECTARRAY CHARACTEPISTICS

r,,,]ized gain at handover 36. 5 dB

axirnu-fn scan angle 49. 1 degrees

-fl.adiating efficiency 39.8 percent
hlalf-power beamwidth at broadside' >1. 2 degrees
Sidelobe level >17. 0 dB
Nunber of elements 6, 550
NTunber of phase shifters 13, 100
Diauneter 3. 6 feet
Surface area 10.Z square feet
Overall depth -Z. 5 feet
Veight -315 pounds

Percentage
Problem Area Remarks Effort

Radiating element Obtain patterns with <1. 5 dB 40

ellipticity to 50 degree scan
angle. Obtain >25 dB isola­
tion between horizontal and
vertical polarized waves.
Minimize length. M\,atch
within array and determine
mutual coupling behavior.
Packaging of two Utilize stripline and micro- 30
phase shifters, strin technology. Rleduce
hybrid trap filter, weight;
and two driver
circuits.

About 50 percent of the phase shifter research and development work

.:-ctcd to be performed for other ESA applications at X-band and below.
r.flectarray design approach proposed in this study is technically
r',iqv bccause itis an extension of ESAIRA technology and because the
- .h:,:, element and phase shifter have been built at X-barld. However,
0-71 state of the art technology, a radiation efficiency of only 20
:rrc-ent would be realized.

-- 5