zotero/storage/DIX6DY3F/.zotero-ft-cache

This electronic version (PDF) was scanned by the International Telecommunication Union (ITU) Library & Archives Service from an original paper document in the ITU Library & Archives collections. La présente version électronique (PDF) a été numérisée par le Service de la bibliothèque et des archives de l'Union internationale des télécommunications (UIT) à partir d'un document papier original des collections de ce service. Esta versión electrónica (PDF) ha sido escaneada por el Servicio de Biblioteca y Archivos de la Unión Internacional de Telecomunicaciones (UIT) a partir de un documento impreso original de las colecciones del Servicio de Biblioteca y Archivos de la UIT. ‫( ﺍﻹﻟﻜﺘﺮﻭﻧﻴﺔ ﺍﻟﻨﺴﺨﺔ ﻫﺬﻩ‬PDF) ‫( ﻟﻼﺗﺼﺎﻻﺕ ﺍﻟﺪﻭﻟﻲ ﺍﻻﺗﺤﺎﺩ ﻓﻲ ﻭﺍﻟﻤﺤﻔﻮﻇﺎﺕ ﺍﻟﻤﻜﺘﺒﺔ ﻗﺴﻢ ﺃﺟﺮﺍﻩ ﺍﻟﻀﻮﺋﻲ ﺑﺎﻟﻤﺴﺢ ﺗﺼﻮﻳﺮ ﻧﺘﺎﺝ‬ITU) ◌ً ‫ﻭﺍﻟﻤﺤﻔﻮﻇﺎﺕ ﺍﻟﻤﻜﺘﺒﺔ ﻗﺴﻢ ﻓﻲ ﺍﻟﻤﺘﻮﻓﺮﺓ ﺍﻟﻮﺛﺎﺋﻖ ﺿﻤﻦ ﺃﺻﻠﻴﺔ ﻭﺭﻗﻴﺔ ﻭﺛﻴﻘﺔ ﻣﻦ ﻧﻘﻼ‬. 此电子版（PDF版本）由国际电信联盟（ITU）图书馆和档案室利用存于该处的纸质文件扫描提供。 Настоящий электронный вариант (PDF) был подготовлен в библиотечно-архивной службе Международного союза электросвязи путем сканирования исходного документа в бумажной форме из библиотечно-архивной службы МСЭ.
© International Telecommunication Union

XVII A.R
DUSS6LD0RF Si.5-1.6 1990
CCIR
XVIIth PLENARY ASSEMBLY DUSSELDORF, 1990
NTERNATIONAL TELECOMMUNICATION UNION
REPORTS OF THE CCIR, 1990
(ALSO DECISIONS)
ANNEX TO VOLUME IV - PART 1
FIXED-SATELLITE SERVICE
CCIR INTERNATIONAL RADIO CONSULTATIVE CO M M ITTEE
Geneva, 1990

XVII A.R
DUSSGLDORF 21.5-1.6 1990
CCIR
XVIIth PLENARY ASSEMBLY DUSSELDORF, 1990
INTERNATIONAL TELECOMMUNICATION UNION
REPORTS OF THE CCIR, 1990
(ALSO DECISIONS)
ANNEX TO VOLUME IV - PART 1
FIXED-SATELLITE SERVICE

CCIR INTERNATIONAL RADIO CONSULTATIVE C O M M ITTE E

92- 61- 04191-4

Geneva, 1990

© I.T.U.

I
ANNEX TO VOLUME IV FIXED-SATELLITE SERVICE
(Study Group 4)

TABLE OF CONTENTS

Page

Plan of Volumes I to XV, XVIIth Plenary Assembly of the CCIR (See Volume IV - Recommendations) ...................................

Distribution of texts of the XVIIth Plenary Assembly of the CCIR in Volumes I to XV (See Volume IV- Recommendations) ..................

Table of contents

. . .. .......................

I

Numerical index of texts ................................................. VII

Index of texts deleted ....................................

IX

Section 4A - Definitions.........................

1

There are no Reports in this Section.

Section 4B - Systems aspects - Performanceand availability - Susceptibility to interference

4B1 - Systems aspects

Report 552-4 Use of frequency bands above 10 GHz in the fixed-

satellite service..........................................

3

Report 1139 General system and performance aspects of digital

transmission in the fixed-satellite se r v i c e .............

37

Report 1134 Digital satellite dedicated n e t works...................... 56

Report 451-3 Factors affecting the system design and the selection of frequencies for inter-satellite links of the fixed-satellite service.................................... 76

Report 1237 Satellite news gathering (See Annex to Vol. X I I ) ........

89

Section 4B2 - Performance and availability

Report 208-7 Form of the hypothetical reference circuit and

allowable noise standards for frequency-division

multiplex telephony and television in the fixed-

satellite service..........................................

91

II

Page

Report 997-1 Characteristics of a fixed-satellite service hypothetical reference digital path forming part of an integrated services digital network..................................................... 97

Report 706-2 Availability of circuits in the fixed-satellite service..................................................... 130

Report 214-4

The effects of Doppler frequency-shifts and switching discontinuities in the fixed-satellite service..................................................... 135

Section 4C - Earth station and baseband characteristics - Earth station antennas - Maintenance of earth stations

Report 391-6 Radiation diagrams of antennas for earth stations

in the fixed-satellite service for use in interference

studies and for the determination of a design

o b j e c t i v e ..................................

143

Report 390-6 Earth-station antennas for the fixed-satellite service..................................................... 159

Report 998-1 Performance of small earth-station antennas for the fixed-satellite service............................... 182

Report 868-1 Contributions to the noise temperature of an earth-station receiving antenna........................... 202

Report 875-1 A survey of interference cancellers for application in the fixed-satellite service............................ 206

Report 212-3

Use of pre-emphasis in frequency-modulation systems for frequency division multiplex telephony and television in the fixed-satellite service............... 216

Report 384-6 Energy dispersal in the fixed-satellite s e rvice......... 220

Report 553-3 Operation and maintenance of earth stations in the fixed-satellite service................................... 248

Report 554-4

The use of small earth stations for relief operation in the event of natural disasters and similar emergene i e s ................................................. 255

Report 869-2 Low capacity earth stations and associated satellite systems in the fixed-satellite service................... 264

Ill

Page

Section 4D - Frequency sharing between networks of the fixedsatellite service - Efficient use of the spectrum and geostationary-satellite orbit

4D1 - Permissible levels of interference

Report 455-5 Frequency sharing between networksof the fixedsatellite service......................................... 279

Report 710-3 Interference allocations in systems operating at frequencies greater than 10 GHz in the fixedsatellite service......................................... 322

Report 867-2 Maximum permissible interference in single-channelcarrier and intermediate rate digital transmissions in networks of the fixed-satellite s e r v i c e............. 333

Report 1001-1 Off-axis e.i.r.p. density limits for fixed-satellite

service earth-stations.............

356

Report 555-4 Discriminations by means of orthogonal circular and linear polarizations...................................... 367

Report 1141

Polarization discrimination in interference calculation................................................ 398

Section 4D2 - Coordination methods

Report 453-5

Technical factors influencing the effciency of use of the geostationary-satellite orbit by radiocommunication satellites sharing the same frequency bands. General summary.................................................... 419

Report 454-5

Method of calculation for determining if coordination is required between geostationary-satellite networks sharing the same frequency b a n d s ......................... 455

Report 870-2 Technical coordination methods for communicationsatellite systems......................................... 494

Report 1000-1 Spectrum utilization methodologies....................... 507

Report 1135

Optimization methods to identify satellite orbital positions.................................................. 515

Report 1003

Methods for multilateral coordination among satellite networks................................................... 532

Report 1137

Stochastic approach in the evaluation of interference between satellite n e t works............................... 537

Report 557-2

The use of frequency bands allocated to the fixed-

satellite service for both the up link and down link

of geostationary-satellite systems

............... . 567

Report 999

Determination of the bidirectional coordination a r e a . .. 575

IV

Page

Report 1140 Satellite networks for more than one service in one or more frequency b a n d s ..................................... 589

Report 1138 Intra-service implications of using slightly inclined, geostationary orbits for fixed satellite service networks. Operational, sharing and coordination cons iderat ions............................................ 596

Report 1004-1 Physical interference in the geostationary-satellite o rb i t ...................................................... 617

Section 4D3 - Spacecraft station keeping - Satellite antenna radiation pattern - Pointing accuracy

Report 556-4

Factors affecting station-keeping of geostationary satellites of the fixed-satellite s e r v i c e .............. 623

Report 1002-1 Flexibility in the positioning of s a t e llites............ 627

Report 558-4 Satellite antenna patterns in the fixed-satellite service.................................................... 640

Report 1136 Geostationary satellite antenna pointing accuracy

697

Section 4E - Frequency sharing between networks of the fixed-satellite service and those of other space radiocommunications systems

Report 560-2 Sharing criteria for the protection of space stations in £he fixed-satellite service receiving in the band 14 to 14.4 G H z ....................................... 711

Report 872

Sharing criteria between inter-satellite links connecting geostationary satellites in the fixedsatellite service and the radionavigation service at 33 G H z ......................................... 713

Report 561-4 Feeder links to space stations in the broadcastingsatellite service......................................... 717

Report 712-1 Factors concerning the protection of fixed-satellite earth stations operating in adjacent frequency band allocations against unwanted emissions from broadcasting satellites operating in frequency bands around 12 G H z .............................................. 729

Report 873-2 An analysis of the interference from the broadcastingsatellite service of one region into the fixedsatellite service of another region around 12 GHz...... 740

V

Report 713-1

Spurious emissions from earth stations and space stations of the fixed-satellite s e r vice.................

Page 746

Report 874

Frequency sharing between the inter-satellite service when used by the fixed-satellite service and other . . space services............................................. 754

Decisions

Decision 2-7

Frequency sharing between radiocommunication satellites.

Technical considerations affecting the efficient use

of the geostationary-satellite o r b i t ....................

757

Decision 64-1 Updating of the handbook on satellite communications

(Fixed-Satellite Service)................................

761

Decision 70-1 Implementation of digital satellite s y s t e m s ............. 762

Decision 76-1 Satellite news gathering (SNG)............................ 763

Decision 87

Determination of the coordination area. Appendix 28 of the Radio Regulations. (See Annex to Vols. IV/ix-2).. 766

PAGE INTENTIONALLY LEFT BLANK PAGE LAISSEE EN BLANC INTENTIONNELLEMENT

VII NUMERICAL INDEX OF TEXTS

ANNEX TO VOLUME IV

Page

SECTION 4A: Definitions (There are no Reports in this S e c t i o n ) ...

1

SECTION 4B: 4B1:

Systems aspects - Performance and availability susceptibility to interference......................... Systems aspects..........................................

... 3

SECTION 4B2: Performance and availability............................ 91

SECTION 4C: Earth station and baseband characteristics - Earth

station antennas - Maintenance of earth stations

143

SECTION 4D: Frequency sharing between networks of the fixedsatellite service - Efficient use of the spectrum and the geostationary satellite orbit
4D1: Permissible levels of interference..................... 279

SECTION 4 D 2 : Coordination m eth o d s .................................... 419

SECTION 4D3: Spacecraft station keeping - Satellite antenna radiation pattern - Pointing accuracy................. 623

SECTION 4E: Frequency sharing between networks of the fixedsatellite service and those of other space radiocommunications systems............................ 711

REPORTS

Section

Page

REPORTS

Section

Page

208-7 212-3 214-4 384-6 390-6 391-6 451-3 453-5 454-5 455-5 552-4 553-3 554-4 555-4 556-4 557-2 558-4 560-2 561-4 706-2 710-3 712-1 713-1 867-2

4B2

91

4C

216

4B2

135

4C

220

4C

159

4C

143

4B1

76

4D2

419

4D2

455

4D1

279

4B1

3

4C

248

4C

255

4D1

367

4D3

623

4D2

567

4D3

640

4E

711

4E

717

4B2

130

4D1

322

4E

729

4E

746

4D1

333

868-1 869-2 870-2 872 873-2 874 875-1 997-1 998-1 999 1000-1 1001-1 1002-1 1003 1004-1 1134 1135 1136 1137 1138 1139 1140 1141 1237

4C 4C 4D2 4E 4E 4E 4C 4B2 4C 4D2 4D2 4D1 4D3 4D2 4D2 4B1 4D2 4D3 4D2 4D2 4B1 4D2 4D1 4B1

202 264 494 713 740 754 206
97 182 575 507 356 627 532 617
56 515 697 537 596
37 589 398
89

N o t e . - Decisions which already appear in numerical order in the table of contents, are not reproduced in this index.

PAGE INTENTIONALLY LEFT BLANK PAGE LAISSEE EN BLANC INTENTIONNELLEMENT

IX

INDEX OF TEXTS WHICH HAVE BEEN DELETED AT THE END OF THE STUDY PERIOD 1986-1990
(In order to facilitate the retrieval of a given text, the page number of Volume IV of the XVIth Plenary Assembly, Dubrovnik 1986, is indicated.)
ANNEX TO VOLUME IV

Text

Title

Page No. Vol. IV Dubrovnik, 1986

Report 204-6 Terms and definitions relating to space

1

radiocommunications

Report 205-4 Factors affecting the selection of

5

frequencies for telecommunications with

space stations in the fixed-satellite

service

Report 383-4 The effects of transmission delay in

15

the fixed-satellite service

Report 385-1 Feasibility of frequency sharing

208

between systems in the fixed-satellite

service and terrestrial radio services.

Criteria for the selection of sites

for earth stations in the fixed-

satellite service

Report 559

The effect of modulation

440

characteristics on the efficiency of

use of the geostationary-satellite

orbit in the fixed-satellite service

Report 707-1 Digital interface characteristics

122

between satellite and terrestrial

networks

Report 711-1 Criteria of efficiency of use of the

448

geostationary-satellite orbit

Report 871-1 Calculation of the equivalent satellite

335

link noise temperature and the

transmission gain

SECTION 4A: DEFINITIONS There are no Reports in this Section.

PAGE INTENTIONALLY LEFT BLANK PAGE LAISSEE EN BLANC INTENTIONNELLEMENT

Rep. 552-4

3

SECTION 4B: SYSTEMS ASPECTS - PERFORMANCE AND AVAILABILITY - SUSCEPTIBILITY TO INTERFERENCE
4B1 : SYSTEMS ASPECTS

REPORT 552-4

USE OF FREQUENCY BANDS ABOVE 10 GHz IN THE FIXED-SATELLITE SERVICE

(Study Programme 27C /4)

(1974-1978-1982-1986-1990)

1.

Introduction

This Report makes a preliminary examination o f some o f the technical factors which should be considered in the design o f systems o f the fixed-satellite service which are intended for use in frequency bands above about 10 G H z. Since the allocated bandwidth is generally wider at frequencies above about 10 G H z, the use o f these frequencies would facilitate the design o f high-capacity systems. The use o f the 3 0 /2 0 GHz bands, would facilitate the design o f very high capacity systems em ploying spot beam antennas.
The factors considered in this Report are:

- analogue system performance,
— system configuration strategies, - frequency sharing with terrestrial systems, — design considerations for system s in the fixed-satellite service.

2.

Analogue system performance

CCITTRecommendation G.222 (see sections 1.2.1,1.2.2 and1.2.3) states the required design objective for an analogue telephony HRC of 2,500 kmas:

- 10,000 pWOp for 20% of any month - 50,000 pWOp for 0.1% of any month - 1 x 10^ pWO for 0.01% of any month.

Reference to satellite systems is made by citation of Recommendation 353 of the CCIR which is:

” that the noise power, at a point o f zero relative level in any telephone channel in the hypothetical reference circuit as defined in Recommendation 352 should not exceed the provisional values given below: 1.1 10 000 pWOp psophometrically-weighted one-minute mean power for more than 20% o f any month: 1.2 50 000 pWOp psophometrically-weighted one-minute mean power for more than 0.3% o f any month; 1.3 1 000 000 pWO unweighted (with an integrating time o f 5 ms), for more than 0.01% o f any year; 11
CCIR Recommendation 353 has been developed to be in compliance with the requirements of the C C IT T , although there are some small differences. However, the concept of availability is not contained in the current version of the Recommendation and the following analysis shows the impact of its inclusion. The analysis is limited to 14/11 GHz systems since the performance of 6/4 GHz systems is not generally affected by propagation fades.

4

Rep. 552-4

Performance of 14/11 GHz systems compliant with Recommendation G.222
The 10,000 pWOp requirement for 20% of any month is interpreted as applying to the worst month*, i.e., for the poorest propagation month. The same interpretation is applied to the 50,000 pWOp clause.
A standard link concept is used for the analysis to correspond to the current practice of other terrestrial systems of allowing 1 pWOp/km for design, or a link of 10,000 km. The operational locations for such links are typically at 40 degrees latitude and 25 degrees elevation angle. The climates for these latitudes exhibit rain rates, for 0.01% of the time, between 30 and 60 mm/hour. A value of 50 mm is chosen for the analysis.Calculations of the rain attenuation are then made in accordance with the methods of Study Group 5.
Propagation availability factor (as defined in Report 997) is taken as 10% of the duration of fade which results in reaching the system threshold. Two cases are shown in Figure 1, one at 50,000 pWOp and one at 100,000 pWOp. The margin in the first case is 7 dB and is 10 dB for the second.
The performance for the path expressed in terms of the available time will meet all of the G.222 performance objectives for the climate and latitudes assumed in this study. For low antenna elevation angles and higher rain rates, it may be more difficult to meet G.222. Further studies are required for such cases.

* The definition of the worst month is provided in Recommendation 581.

Degradation in S/N relative to nominal (dB)

Rep. 552-

14/11 GHz noise performance as a function of available time

6

Rep. 552-4

3 # System configuration strategies
Scatter and absorption by cloud and precipitation increase rapidly at frequencies above about 10 GHz, and this adds considerably to the problems of designing such systems. Without the use of special techniques it may be quite impracticable to provide the large rain margins necessary to meet the required standards of performance.
Four possible ways in which the severe effects of precipitation at the higher frequencies can be overcome are: (a) the use of site diversity; (b) the use of a lower alternative frequency band to that normally used, and which is much less affected by
precipitation; (c) the use of adaptive systems which alter the transmission parameters during changing propagation conditions;
(d) the use of multiple narrow beam on-board antennas with possible extension to the single station per beam (SSPB) concept. In the first approach referred to in (a) above advantage can be taken from the fact that for earth stations
spaced a suitable distance apart (i.e. 10 to 30 km) the correlation of precipitation between them is almost negligible and the probability that both stations will be affected simultaneously by heavy rain is likely to be very small. The technique is to connect the two earth stations providing the diversity, by a transmission line free from the effects of precipitation, and select for operational use the earth station which is least affected. Diversity operation is discussed in detail in Annex I.
In the second approach referred to in (b) above, the assumption is that a number of earth stations within a system normally operate at frequencies which can be severely affected by precipitation, i.e. above about 10 GHz. However, since the probability of more than one station at a time being affected is likely to be small, the technique of switching into use a lower frequency band at the earth station badly affected by precipitation, can be employed (Mori et al., 1978]. To make a better utilization of the normally unused lower alternative frequency band, it may be possible to normally carry the traffic in the lower frequency band and interchange the operating frequency bands between stations operating in the lower frequency and those operating in the higher frequencies under adverse weather conditions [Kosaka, 1978]. Based on this concept, an experimental system using 30/20 G H z and 6/4 G H z bands was constructed [Kosaka et al., 1982].
In the third approach, referred to in (c) above, system performance of digital systems may be improved by reducing the information rate transmitted or increasing the transmitted power (up-link power control) during poor propagation conditions. Examples of this approach are given in Annexes II, III and IV.
Adaptive fade countermeasure (FCM) techniques give selective enhance­ ment to carriers undergoing fading. Some FCM methods require that the user is prepared to accept a lower data rate during fading, as in Annex II, but other methods allocate part of a shared resource overhead (eg. power, frequency, time) to any fading carriers within the network, and thus maintain the user rate (see Annex III and IV) . Adaptive methods use the shared resource efficiently by apportioning resource to carriers according to the depth of fading.
In the 30/20 GHz frequency range, even in temperate zones, fade depths for significant portions of time are too great for simple fixed fade margins to be a practical solution, so some FCM is essential if the bands are to be exploited. For applications requiring high availability in the wetter climatic zones, stations will suffer even more frequent and severe fading, and there is a practical limit to the fade depth which can be countered by an adaptive system, the deeper fades requiring unacceptable high levels of shared resource. Although further propagation studies are required, indications are that it is practical to operate a shared resource adaptive scheme for an availability corresponding to Recommendation 522 in climatic zone E, but for greater availability in the wetter regions, the diversity methods, which are not adaptive and may be expensive in the earth sector, seem the only suitable option for trunk satellite services.
In the fourth approach referred to in (d) above, the objective is to avoid complications of design and operation of earth stations, even at the expense of making the satellite more complicated due to the use of complex multiple beam on-board antennas with several narrow beams which however provide for both high satellite e.i.r.p. and high satellite G / T to compensate for the propagation effects.
Examples of various existing and planned system implementations in the 30/20 GHz frequency bands are given in Annex V .

Rep. 552-4

7

4.. Frequency sharing with terrestrial systems
At frequencies above about 10 G H z variations in the level o f the w anted and unwanted signals due to precipitation, and the effects o f scatter, have a greater influence on the minim um separation distance obtainable between earth stations o f the fixed-satellite service, and terrestrial stations o f the fixed service.
The effect o f scatter can be overcome by careful site selection to avoid beam intersection o f the two systems, and by using cross-polarization in the case o f linearly polarized waves and, since the basic transmission loss over a given path increases with frequency, the separation distance between stations o f the two systems can be less at the higher frequencies. By arranging that the separation angle between an earth station and terrestrial stations is more than about 20 to 30 degrees, the minimum separation distance can be reduced to a few kilometres and the effect o f the fluctuation o f the wanted apd unwanted signals caused by differential rain attenuation o f the two systems can be avoided to some extent.
5 . Design aspects of systems in the fixed-satellite service at frequencies above 10 GH z
For systems in the fixed-satellite service which use frequency bands above 10 G H z, the effects o f hydrom eteors, especially rainfall, are particularly important and must be taken into account when the systems are designed. The most reliable calculation o f the effects o f hydrometeors may be made on the basis o f experimental distributions o f attenuation due to hydrometeors against time. This distribution varies with the frequency and the time o f the year and depends on clim atic conditions at the site o f the earth station and the angle at which the satellite is visible.
It should also be borne in mind that the correlation between attenuations on the paths o f the satellite link declines with the distance between earth stations and increased intensiveness o f precipitations. A further de-correlating factor is the frequency difference between the up-link and the dow n-link.
The relevant data on propagation can be found in Reports 564 and 565. In addition to that, since 1969, continuous rain attenuation experim ents on earth-satellite paths have been carried out at various locations in the U nited States o f America. The measurement frequencies include 11.7, 13.6, 15.5, 17.8, 19 and 28.5 GH z. Interim results o f the 10 year (1969-1978) experiments have been published in various technical journals and conference proceedings. [Lin, et al., 1980] summarizes new results and the previously published results and discusses radio com m unication systems. The summary includes the geographic dependence, the frequency dependence, the diurnal, m onthly, and yearly variations o f rain attenuation statistics, the diversity im provem ent factors, the fade duration distributions, the dynam ic rain attenuation behaviour, the long-term (20 years) rain rate distribution for United State o f America locations and a sim ple empirical model for rain attenuation.
The data indicate that the 28.5 GHz earth-satellite radio link, assum ing 20 dB fade margin, will require site-diversity protection for most United States o f America locations to meet the conventional long-haul reliability objective. Operation in this or higher frequency bands w ould, therefore, probably require new network operation procedures.
On the other hand, the site-diversity protection may be avoidable for frequencies at or below 14 GH z where the antenna elevation angle is relatively high.
The earth-station satellite link at 19 GH z may or may not require site-diversity protection, depending on earth-station location and satellite orbital position. Other major findings are:
— Rain-induced outages on earth-satellite radio links have higher service im pact than multipath-fading-induced outages on terrestrial (6 /4 G H z) radio relays even if the two systems are engineered for equal total outage time. This is because multipath fading occurs m ostly during the early m orning hours o f low telephone activity. Furthermore, multipath fading is frequency selective and interrupts only a fraction o f the frequency band at a time. By contrast, about 35% o f rain outages will occur during telephone busy hours, and the outage will interrupt all traffic on an earth-satellite radio link at the sam e time.
— Site-diversity protection can reduce the rain outage time by at least one o r d er .o f magnitude if site separation exceeds 20 km. Orbital diversity protection, although effective against sun-transit outages, reduces rain outage time by less than 20%.

8

Rep. 552-4

However, at present, there exists little measured propagation data in most parts of the world. In many places data from radiometer measurements is also available, but since satellite beacons at 20 and 30 GHz have not been generally available there are few results which are reliable enough for system design purposes.
Examples of the magnitude of this problem are shown in Figure 2. The attenuation axis is indicative pf the rain margins which would be necessary to limit service unavailability to percentages of time corresponding to the probability axis.
In deriving link power budgets account shall be made of up-link and down-link noise, external interference and internal impairments, such as intermodulation, co-channel and adjacent channel interference, in the same way as for systems working below 10 GHz. However the relative importance of these contributions may be different and will usually vary significantly with the propagation conditions. The parameters of systems intended for international transmission should be defined according to CCIR Recommendations 353 and 522. These Recommendations specify transmission performance for three percentages of time. Link budgets should be calculated for these three percentages of time. For each earth station in the network the resulting parameters should be such as to satisfy the most stringent condition. Whereas at frequencies below 10 G H z the longer term condition is usually the governing one, at frequency above 10 G H z any of the three may be the most stringent depending on system requirements and on local climatic conditions. For this reason some of the earth station parameters usually have more than one specification to take into account the propagation phenomena. In this respect Annex VI gives a method of deriving an earth station antenna diameter from a G / T specification.

It should be noted however, that the choice of antenna diameter utilized in practice can depend on many factors, including up-link e.i.r.p. requirements, desired fade margins, cost, etc. It is possible in the design of a satellite system to minimize the cost of the earth station by optimization of the combined costs of the antenna, HPA, LNA, etc.

6.

30/20 GHz Band Technologies

The 30/20 GHz bands have yet to be exploited widely in the FSS even though they represent 3.5 GHz of available bandwidth for satellite communication applications. The principal technical problems are associated with severe rain attenuation of RF transmissions in this band. Table I summarizes some advantages and disadvantages of 30/20 GHz systems based on current technology.

"Rep. 552-4

9

Probability that ordinate is exceeded (%) (worst month)

FIGURE 2

Rain attenuation versus probability that ordinate is exceeded*

N rain climatic zone N M rain climatic zone M K rain climatic zone K F rain climatic zone F E rain climatic zone E

Frequency : 30 GHz

-----

elevation angle of 30 deg. (latitude 52.57 deg.)
elevation angle of 60 deg. (latitude 25.66 deg.)

The longitude of the earth station is assumed the same as that of a geostationary satellite.
Calculation based on Table I of Report 563, equations (8), (9) and (21) of Report 564, and Table I and equations (1), (2) and (3) of Report 721 in CCIR Volume V, 1986.

10

Rep. 552-4

TABLE I

Merits and Demerits of the 30/20 GHz

TtarH

rrmrnmication Systems

conpared with lower frecuercy bards

MERITS

DEMERITS

1. Wide bandwidth, high transmission capacities
2. Narrow beanwidth and large antenna gain.
3. Narrow required satellite orbital spacing between satellite antenna beazwidth (Large ranter of satellites cn the GSO), easy coordinaticn between satellites.
4. Parts of the spectrum are exclusive to satel­ lite services.
5. Easy introduction of multiple beam systems which increase system availability performance.

1. Higher rain attenuations carpared to lcwer frequency bands, i.e., large rain margins to guarantee the available time. Need for fade counter measures.
2. Relatively ccuplex hardware characteristics (e.g. high output pcwer and low noise figure).
3. Relatively high cost of equipment at this early state of equipment develop­ ment.
4. Relatively limited experience of frequency reuse.

Hardware technologies such as antennas, high power anplifiers (HPA), lew noise amplifiers (INA), and other cn-board sub-systems have substantial effects cn system design. Same Joey elements are summarized below.

Antennas at 30/20 GHz can achieve high gains in small physical sizes, compared with lower frequencies. However, surface roughness is more sericus, perhaps resulting in seme degradation in gain cr sidelobe characteristics.

High power amplifiers at 30/20 GHz will continue to improve over the next few years. She output pcwer from solid-state (FET) amplifiers, and the efficiency and size of travelling-vave-tube (TWT) amplifiers, will continue to lrprove.
lew noise anplifiers are also continuing to improve. Ihe noise temperatures of both FET (field effect transistor) and HEMT (hi^i electron mobility transistor) amplifiers should acntinue to improve significantly over the next five to seven years.
Other spaoe-segment sub-systems requiring improvement include satellite switching, and very small and light weight transponders. An cn-board switch, if inplemented, is necessary to connect signals frrn different beams in a TTm’i-HHo*™ system. Ch-board switching can be performed both at baseband and microwave level. In case of baseband processing, cn-board regeneration also needs to be implemented, which in turn improves the link budget performance. MICs and monolithic MICS will be used in transmitters and receivers to reduce size and
weight. On-board switch control equipment will be realized by adopting very large scale integration (VLSI). These technologies are now under development, and some experimental systems will use these technologies in the 30/20 GHz bands.

Rep. 552-4

11

By utilizing an optimum satellite arrangement to match the geographical distributicn of service areas, introducing multiple beam satellite antennas, using improved sideldbe mall earth station antennas and up-link pcwer control techniques, a large number of satellites with high cccnnunicatian capacities can be realized at 30/20 Giz. If it is practicable for new services, a moderate availability value should be selected far system design, cwing to the strong Hppprripnnp of system capacity cn its value at these frequencies.

7.

Frequency reuse in the 30/20 GHz frequency bands

In the context of the exploitation of frequency bands above 10 GHz, particular attention must be paid when considering the practical use of the 30/20 GHz frequency bands especially if compared to the case of 14/12/11 GHz frequency bands. In fact, because the propagation losses in heavy rainfall conditions are considerably greater (even of the order of 5-6 times in decibels) at 30/20 GHz than at 14/12/11 GHz, they may become the ultimate limiting factor for the efficient exploitation, in a cost effective way, of the 30/20 GHz frequency bands, in many geographical regions.

Primarily important in this respect is the extent to which extensive frequency reuse of the 3.5 GHz available bandwidth can be feasible at such frequency bands, for the provision of high capacity regional and domestic satellite systems.

Preliminary indications, deriving from initial system analysis [CCIR, 1986-90] carried out using typical up-link attenuation values for the "K" climatic zone are as follows:

a)

satellite antennas with a single coverage beam operating with

frequency reuse by orthogonal polarizations and earth stations

transmitting both cross-polarized signals, can provide

satisfactory performance with state of the art values of

polarization discrimination of on-board and earth station

antennas, with and without up path power control (UPC) (see Report 710);

b) frequency reuse by orthogonal polarization in the same single coverage beam from the satellite and different earth stations transmitting cross-polarized signals, assuming no rain . correlation between sites, is marginally feasible depending on the system configuration. In particular, it seems that frequency reuse is possible when UPC is used at all earth stations. If UPC is not applied, the possibility of frequency reuse is strictly dependent on specific climatic conditions and earth station operational elevation^angles;

c) frequency reuse by orthogonally polarized signals from the same earth stations operating with a multi-beam satellite antenna is possible in adjacent beams if:

spatial satellite antenna discrimination equal to or greater than 30 dB is used which in turn is achievable with an inter­ beam distance of about 3 times the 3 dB beamwidth from beam centre to beam centre;

by using UPC such inter-beam distance may be reduced to about 2 beamwidths;

d) frequency reuse by orthogonally polarized signals from different earth stations operating with a multi-beam satellite antenna is possible in adjacent beams only if:

UPC is used. If full UPC is used, less than 20 dB of spatial satellite antenna discrimination may be provided, but in any case an inter-beam distance of about 2 beamwidths will be necessary.

12

Rep. 552-4

However, other possible configurations can be foreseen in the practical exploitation of the 30/20 GHz frequency bands which could be of interest in the context of frequency reuse under severe rain attenuation conditions and further studies would then be required.
8 . Conclusions
At frequencies above about 10 GHz, scatter and absorption caused by cloud and precipitation have much greater significance. Moreover, there may be an appreciable difference between the up-link and down-link frequencies and careful design is needed to ensure that the necessary performance objectives are met in a balanced way.
Certain techniques, such as site diversity, the use of an alternative frequency band or the use of adaptive systems, could overcome the problems met due to large attenuation from precipitation for small percentages of time. At the 30/20 GHz frequency range a 99.95% availability may be achievable in the drier climatic regions by using one of the adaptive techniques; a composite adaptive system could also be considered. However, for greater availability in the wetter climatic zones, the only practical option for trunk satellite services at 30/20 GHz would seem to be one of the diversity schemes. Plans.exist to test fade mitigation techniques using both the Olympus and ACTS satellites.
The use of frequency bands around 30 and 20 GHz, where 3.5 G H z of bandwidth is available, would make possible the provision of very high capacity regional and domestic systems using spot beam antennas, and should make it possible for the earth stations of such systems to be located very close to traffic centres.
In the 14/11 G H z frequency range several satellite communication systems have already been in service for some time. They have demonstrated that reliable operation can readily be achieved at these frequencies.
Concerning the 30/20 G H z frequency range substantial results have been obtained through the communi­ cation experiments with Japan’s communications satellite (the medium capacity communications satellite for experimental purposes). As a result of these experiments, the design techniques of the 30/20 G H z band satellite and earth station hardware have been established. The prospects for practical use of a 30/20 G H z satellite communication system may be extremely good, provided that the minimum elevation angle is adequate according to the local climatic conditions [Hatsuda, et al., 1980]. In February and August, 1983, Japan’s operational communication satellites of CS-2a and CS-2b using the 6/4 G H z and 30/20 G H z bands were launched success­ fully. These satellites put the new 30/20 G H z frequency band into practical use for the first time in the world [Hayashizaki, 1983].
In the design and planning of systems in the fixed-satellite service using frequencies above about 10 GHz, there are a number of areas which require further study. These are, for example, the determination of earth station G/T, and the allowance to be made for propagation, including the effect on cross-polarization discrimina­ tion.
REFERENCES
HATSUDA, T., MOR1H1RO, Y. and NAKAJ1MA, S. [July-August, 1980] Link calculation method for 30/20 GHz band satellite xommunication system. Rev. Elec. Com m . Lab.. N T T , Vol. 28, 7 /8 , 604-619.
HAYASHIZAKI, H. [1983] Domestic satellite communications system — satellite communication using 30/20 GHz band. Japan T eleco m m . Rev., R-129, 57-65.
KOSAKA, K. [1978] Frequency switching TDMA system for countermeasure against rainfall attenuation. J. R adio Res. Labs.} (Japan), Vol. 25, 117/118.
KOSAKA, K., HAMAMOTO, N„ SUZUKI, R., UMEDA, Y. and ITO, H. [1982] Experimental frequency switching TDMA equipment for countermeasure- against rainfall attenuation. IEEE Global Telecommunications Conference, (GLOBECOM ’82), Miami, Fla., USA.
LIN, S. H., BERGMANN, H. J. and PURSLEY, M. V. [February, 1980] Rain attenuation on Earth-satellite paths - summary of 10-year experiments and studies. B ST J, Vol. 59, 2, 183-228.
MORI, Y., HIGUTI, 1. and MORITA, K. [1978] Estimation for frequency band diversity effect on Earth-satellite links in microwave and millimeter wave bands. Trans. Inst. Electron. C om m . Engrs. Japan, Vol. J61-B, 10.

Rep. 552-4

13

B IB L IO G R A P H Y
D A L G L E IS H , D. J. an d R E E D , A. C. [25-27 N o v em b er, 1969] Som e co m p ariso n s o f the traffic-carry in g cap ac ity o f co m m u n ication-satellites using digital techniques with the capacity o f satellites using frequency m odulation. Proc. IN T E L S A T /IE E C onference on Digital Satellite C om m unications, L ondon, 226-240.
H A T SU D A , T., FU K E T A , H. and IZ U M I, K. [February, 1974] System design o f m illim etre-w ave satellite com m unication systems. Tech. G roup on com m unication system, Inst. Electronics and C om m unications Engineers o f Japan, Paper CS 73-147.
IP P O L IT O , L. J. [F ebruary, 1971] E ffects o f p recip itatio n on 15.3 an d 31.65 G H z e a rth -sp a c e tran sm issio n s w ith ATS-V Satellite. Proc. IEE E International C onference on C om m unications (IC C '71), M ontreal, C a n ad a, Conf. Rec. Vol. 59, 2, 189-205.
IP P O L IT O , L. J. [19-21 Ju n e , 1972] T he 20 a n d 30 G H z co m m u n icatio n s system for th e A T S-F m illim etre w ave experim en t. IE E E In te rn a tio n a l C o n feren ce on C o m m u n icatio n s (IC C ’72), P h ilad elp h ia, Pa., U SA , C onf. Rec., Session 26 A ero ­ space and electronic systems, 26-2 - 26-7.
K .A PL U N O V , B. B. [1971] R a sp ro stran en ie m illim etrovykh voln v atm osfere i sp u tn ik o v a y a svyaz (A tm o sp h eric p ro p a g a tio n o f H E H F waves in satellite com m unications). Obzornaya informatsiya o zarubezhnoi tekhnike svyazi (Review o f foreign telecom m unication technology). M inistry of T elecom m unications o f the U.S.S.R., 5, 24-55.
C C IR Documents
[1986-90]: 4/36 (Italy).

ANNEX I SITE DIVERSITY OPERATION

1.

General design considerations

The performance required for the diversity earth stations is decided not only by the rain climate but also by the diversity configuration. The first kind of configuration is balanced diversity (diversity by two earth stations with equal performance). The other configuration is unbalanced diversity. In this configuration, the perform ance o f one earth station (main station) is m ade sufficiently high, so that the perform ance requirements o f the other station (sub-station) may be considerably alleviated. Such an unbalanced diversity configuration is expected when the main station an ten n a is equipped with a multiple-frequency ban d feed such as 6 / 4 G H z an d 14/11 G H z, a nd/or when the sub-station has to be simplified for technical and operational reasons.
TableU summarizes the results of sample calculations of the antenna diameter and the maximum transmit power required for the balanced diversity links with low elevation angles. Estimates are given for two assumed diversity links. (A) Yamaguchi-Hofu (diversity distance = 20 km) and (B) Y am aguchi-Ham ada (100 km); both in Japan.
It is seen from this Table that the antenna diameters required for the 14/11 G H z FM link (14 G H z for up-link and 11 G H z for down-link) are abo ut 28 m an d 19 m for cases (A) a n d (B), respectively. When the diameter of the main station can be made larger than those values, the required diameter for the sub-station becomes smaller. Values shown in this Table are derived using many of the link parameters established for Intelsat-V satellites, so that they are subject to change when the link parameters are different from those used here.
The calculation methods of the required performances (antenna diam eter and e.i.r.p.) for the diversity earth stations are different depending on the diversity configurations. In the design of the balanced diversity link, calculations have to be based upon the joint probability distribution of the rain attenuation at both locations, while in the case of the unbalanced diversity configuration, the cumulative time distribution of the rain attenuation and the conditional probability of the attenuation are required.
The conditional probability P ( L " / L ') is the probability with which the rain attenuation at the site o f the sub-station exceeds L " under the condition that the rain attenuation at the m ain site exceeds L ' [CCIR, 1978-82].
In order to perform reliable estimates of the earth station requirements, reliable statistics on the basis of the long-term propagation measurements are needed.

2.

Site diversity switch-over operation

To implement diversity earth stations, care should be exercised on the switch-over operation, because, in the event of switch-over, short duration of signal loss or overlap may occur due to the difference in path length of diversity routes or carrier phase discontinuity.

14

Rep. 552-4

TABLE JX - Sample calculations o f the required performances fo r balanced diversity links with low elevation angles (14/11 GHz)

Location
Elevation angle (degrees)
Diversity distance (km)
FM Required antenna diameter (m) Required transmit power(i) (W) (maximum value)
TD M A (2) Required antenna diameter (m) Required transm it power (W) (maximum value)

(A) Yamaguchi - Hofu

9.1

9.1

20

28 / 32 730

17 / 19 530

(B) Yamaguchi - H am ada

9.1

8.4

100

1 9 /2 2 510

11 / 12 400

(1) Values for 792 channel FDM -FM carrier (25 MHz). (2) Values for 4-phase CPSK at 120 M bit/s with forw ard error-correction.
A ssum ptions: Frequency: 14.5 (u p lin k )/11.7 (downlink) GHz O rbital position of satellite: 63°E, 0°N Satellite e.i.r.p .: 41.1 dBW Antenna diameters are estimated for two cases, namely: Ts = 50 K and Ts = 150 K Ts : System noise tem perature of earth-station antenna Efficiency o f the earth-station antenna: 65% Estimates are based on the rain-rate statistics obtained for those locations.

In analogue transmissions such as FM -FD M A , switch-over in transmitting will necessarily cause disconti­ nuity o f carrier phase which will result in signal hit at the d em o d u la to r o u tp u t in the receiving earth stations. Signal hit due to switch-over at the receive earth station may be avoided by carefully adjusting the electrical path length o f each diversity link measured from the switch-over equipment to the satellite.
In digital transmissions it is possible to avoid signal hit even in the event o f switch-over at the transmit earth station by providing dummy intervals in the transmitting signal sequence and making the switch-over during the dumm y interval. In the receiving earth stations the dummy intervals should be discarded no matter whether or not switch-over took place.
The hitless switch-over both in the transmitting and receiving of the diversity system may most conveniently be achieved in T D M A transmission [W atanabe, et al., 1978]. The d um m y intervals are built-in because TDMA transmission occupies only a part of the TDM A frame. Furthermore, TDMA demodulators are capable of receiving burst mode carriers of incoherent phase. Therefore, phase incoherency of TDMA carriers does not cause any difficulty. The only possible problem of site diversity operation of TDM A transmission would be the necessity of very precise transmit timing control even for the initial transmission from the stand-by station. This may be solved either by continuously transmitting a dummy burst from the stand-by station or by obtaining sufficiently accurate ranging data of the satellite which is possible when the T D M A system employs open loop synchronization. In TDM A transmission, the path lengths of diversity routes can be equalized using the reception timing of frame synchronization signals. The reception timing of signals from both diversity routes can be automatically equalized by controlling the variable delay line inserted in one of the diversity routes. An experimental system and the results of experiments using the dummy burst technique are described in [Fugono et al., 1979; Suzuki et al., 1983].
For route selection in diversity operation, it is necessary to m easure the transmission quality of diversity routes. Because the diversity effect may degrade depending on the choice of the measuring method of link quality, care should be taken on selecting the measuring time d u ratio n and achie vable accuracy [Suzuki et al., 1983].

Rep. 552-4

15

3.

Diversity interconnect link

A factor which must be considered is that the C C I R hypothetical reference circuit contained in R eco m m e nd atio n 352 an d the C C I R hypothetical reference digital path co n ta in e d in R ec o m m e n d a tio n 521 include the Diversity Interconnection Links (DIL) to the diversity switching point and any additional modulation/demodulation equipment required. This would mean that system noise budgets must include all the effects of the DIL.

3.1 Basic configuration

3.1.1 Physical aspects
There are a number of different specific configurations which can be considered and there could be reasons for preferring one of these. Two o f these are identified and described in this Annex as (see Fig. 3):
— a main site which contains the diversity switch an d the terrestrial interface. The diversity site is connected by a tw o-hop microwave D IL using either an active, o r passive, repeater. (A repeater site is assumed, since the likelihood of mutual visibility of the diversity sites is small.)
— dual diversity sites and a separate control site with the interface and diversity switch; single microwave hops for each site to the control site.
It may also be possible to employ cable or waveguide links for the DIL. When both FDM -FM and TDM (FDM A or TDMA) are used at an earth station, two parallel links would usually be required.

3.1.2 M odulation requirements
When F D M -FM is used, because the satellite link m odulation and baseband configurations are usually different from those conventionally used for terrestrial systems, rem odulation will be required. The main difference is associated with the channel packaging. The terrestrial system will usually combine the channels in one or more basebands in each direction a n d will use a relatively low m odulation index. The earth-station will break these basebands dow n into multiple, m ulti-destination, transm it basebands; different from those on the terrestrial system and using a different modulation index. The receive basebands are even more numerous and may only consist of a few channels and these must be re-combined into the terrestrial basebands. This process requires modulation/dem odulaton equipment at the main earth station site and at the diversity site where conventional design o f the D IL is used. All configurations can be implemented using the remodulation technique at the expense of providing duplicate equipm ent at the diversity site.
An alternative technique is to use the same m odulation arrangem ents on the terrestrial system as used on the satellite system. Such a technique would appear to be technically feasible although not conventional. The incentive is to save the cost o f remodulation equipm ent at perhaps some added cost to the terrestrial system, although savings may also result for this element as well. The use of such a technique is only applicable to the second configuration o f Fig. 3 . W hen T D M is used (F D M A or T D M A ), either technique could be employed. In the case of T D M A , diversity switching is perform ed between bursts (see A nnex I, § 2). The sam e m odulation could be used on the D IL as used on the satellite system although the data rates would not normally be those of a conventional terrestrial digital radio system.

3.2 Technical factors

3.2.1 Frequency selection
Frequency selection for a microwave DIL requires careful study to ensure that the required overall performance is obtained. Information on terrestrial microwave propagation is shown in C C I R Reports 338 and 720.

3.2.2 Bandwidth requirements
The bandwidth required to implement the DIL can be related to the earth station bandwidth by a factor which may be unity or less, depending upon whether re-m odulation is used o r not. If only frequency translation is used then b and w id th requirem ents must be M H z for MHz. By re-m odulating, a greater channel density can be achieved by using smaller FM modulation indices at the expense of a substantial multiplex interface.

16

Rep- 552-4

Terrestrial interconnect
FIGURE 3 - Diversity configurations
A: Main site B : Diversity sites C : Control site
3.2.3 Rain attenuation
A further factor is related to rain attenuation and site diversity characteristics as related to rainfall p h e n o m e n o n . A dry climate is preferred. Diversity action is a function o f the site spacing. -It is expected th at ab ou t 16 km spacing is the n om inal required. The best orientatio n o f a line connecting the sites may be assumed to be perpendicular to the direction of predominant weather patterns since the most severe attenuation condition would not be expected to affect both sites simultaneously and maximum diversity action w ould be obtained, [Gray, 1973; Hall an d Allnut, 1975; Davis a n d C ro om , 1974]. The weather effects on the microwave DIL must be accounted for if the higher frequencies are used for these links, ■although this should be a secondary consideration.
3.2.4 Variations in transmission delay due to diversity switching
A nother element o f im portance is associated with the differential transmission delay between the diversity signals as they arrive at the switching point. Variations in transmission delay due to diversity .switching are considered in Report 383.
3.3 *G eneral considerations
Two particular aspects of the DIL are important: — the contribution to the overall system noise budgets, and — the contribution to system outage.
These subjects are studied here to develop the effects of the important parameters and the interrelationship with the satellite link parts of the system.
The diversity link design can be made on two bases. If a re-m oduiation system is selected, then co n v en tio n al radio-relay designs can be used. If a translation system is selected, then the design can follow a different pattern an d will be very similar to the satellite system transm ission design. Fading margins and noise contributions must be accounted for in overall performance. In the special case where the same frequencies are used.for the DIL as for the satellite system, then interference noise allowances must also be made.

Rep. 552-4

17

3.3.1 Noise budgets fo r F D M -F M
The contributions of the DIL to the overall noise of the hypothetical reference circuit have to be made reasonably small in order to maintain the system performance in accordance with Recomm enda­ tion 353 o f the CCIR.
It seems reasonable to assume that the DIL noise contribution w ould be considered as part of the earth station budget (usually 1500 pWOp), as the D IL actually provides part o f the norm al earth station function. It only needs to be determined that such a contribution can be kept sufficiently small so that the total o f 1500 pWOp is not exceeded. The fading o f the DIL will contribute to the overall short-term noise budget of the link.
The noise contribution from the DIL would have a number of components depending upon the implementation configuration and the frequency bands used. The components are:
(a) Thermal noise
Conventional C C IR designs for radio-relay are 1 to 3 pWOp per km or less, and can be held to 10 pWOp or less, for a single hop. Special designs also achieve small contributions. T he time varying components due to multipath fading and rain attenuation are relatively large, but for short hops can be controlled to reasonable values. The thermal noise is dB for dB related to the fading from either mechanism.
(b) Basic intrinsic noise
This is baseband noise and is applicable only to re-modulating configurations. Noise levels o f 50 to 100 pWOp are usual for back-to-back basebands. The normal earth station noise budget provides for one such contribution while a re-modulation configuration would add a second contribution.
(c) Interference
A very small interference contribution would be present from other microwave systems operating in the same frequency bands in some cases. This contribution can be considered to be negligible. For the special case of using the same frequency re-use design, up-link and down-link contributions of interference at the earth station can be expected. Values o f the order o f 10 to 100 pWOp are estim ated for norm al operation. In addition, certain fading situations may be accompanied by increases in this noise for very short time periods along with the thermal noise. This configuration does not require re-modulation, so all extra noise associated with item (b) is eliminated.
(d) Intermodulation
A re-modulating design will have an extra mod-demod pair plus IF amplifiers, while the translation design is all conventional earth station equipment and therefore contributes very little IM noise.
The following table illustrates a possible noise budget:
TABLE H I

Thermal Baseband intermodulation Interference Intermodulation (RF)
Total (pWOp)

Sample budgets - Free space conditions

Re-modulation (2 hops)

Low (pWOp)
2 50
-
100

High (pWOp)
20 100
-
200

152

320

Frequency translation (1 hop)

Low (pWOp)
1
-
10 20

High (pWOp)
10
-
100 50

31

160

18

Rep. 552-4

3.3.2 Error budget fo r TDM A
The contributions of the DIL to the overall error rate of the hypothetical digital reference path have to be made reasonably small in order to maintain the system performance in accordance with Recom m endation 522.
It should be noted that in the case o f the re-modulating D IL the errors will be additive whereas in the case o f frequency translation the noise effects will be additive.
3.3.3 Frequency considerations
The fading characteristics as a function of frequency, climate and path length for rainfall can be derived from conventional microwave designs. Rain attenuation and multipath fading are independent events — in fact they are almost mutually exclusive.
Since the expected spacing o f a diversity pair o f earth stations is o f the o rde r o f 16 to 24 km and since it is also expected that either a repeater or a com m on site will be needed, the individual path lengths o f the D IL will probably not exceed 16 km. The margins for such a path length can norm ally be m ad e sufficiently high to ac c o m m o d a te short-term outages as low as 0 .0 0 1 % o f the time.
REFERENCES
D A V IS, P. G. an d C R O O M , D. L. [N ovem ber, 1974] D iversity m easurem ents o f a tte n u a tio n at 37 G H z w ith solar track in g rad io m eters. Electron, a n d Lett., Vol. 10, 482-3.
F U G O N O , N ., Y O S H IM U R A , K. a n d H A Y A S H I, I. [O ctober, 1979] J a p a n ’s m illim eter w ave satellite co m m u n icatio n pro g ram . IE E E Trans. Com m ., Vol. C O M -27, 10, 1381-1391.
G R A Y , D. A. [22-24 A ugust, 1973] E arth to space path d iv ersity ; d ep en d en c e o f b aselin e o rie n ta tio n . IE E E G A P In te rn a tio n a l S ym posium .
H A L L , J. E. an d A L L N U T , J. E. [7-10 A pril, 1975] R esults o f site diversity tests a p p lic a b le to 12 G H z satellite co m m u n icatio n s. I EE C o n feren ce P ublication 126.
S U Z U K I, Y., S H IM A D A , M., M IN E N O , H „ S H IN O Z U K A , T., K U R O IW A , H. a n d IN O M A T A , H. [Septem ber, 1983] T D M A site diversity sw itching ex p erim en ts with Ja p an ese CS. IE E E J. Selec. A reas C om m ., Vol. SA C -1, 4, 674-680.
W A T A N A B E , T., S A IT O H , H., O G A W A , A., an d D E A L , J. H. [O ctober, 1978] Space d iversity system fo r T D M A satellite link. 4th International Conference on Digital Satellite C om m unication, Session IX, M ontreal, C anada.
CC IR Documents [1978-82]: 4 /2 6 (Japan).

Rep. 552-4

19

ANNEX II VARIABLE INFORMATION TRANSMISSION RATE SYSTEMS

1.

Introduction

System performance of digital satellite communication systems can be

improved by reducing adaptively the information transmission rate during poor

propagation conditions. Variable parameters (clock rate and number of phase

states) of PSK modulation and the variable coding rate of FEC (forward error

correction) can be used for variable information transmission rate. A

synchronization sum method for a demodulated PSK signal has also been

appliedto

a variable transmission rate TDMA system.

It should be noted to reduction in information necessary in these cases.

that public services may not be able to be subjected rate, and that other fade countermeasuresmay be

2.

Variable parameter system of PSK modulation

Various kinds of general purposes PSK modems have been developed using LSI type digital signal processors. These modems have two modes of operation, BPSK and QPSK, and their transmission clock rates are continuously variable. They are suitable for a variable transmission rate system although high transmission rates are not achievable because of the limitation on operational speed of the digital signal processor. For example, a modem with a maximum transmission rate of 400 kbauds has been developed [Suzuki et a l ., 1987]. This modem can be applied to a burst mode signal. In another example, a modem with a maximum transmission rate of 6 Mbauds has been developed [Iwasaki et a l .. 1988]. This transmission rate is variable, but it takes a significant amount of time to reach a stable state after resetting the parameters.

When the reduction ratio of the transmission rate is 7 , improvement of C/N is given by:

A(C/N) = -10 log 7

dB.

3.

Variable coding rate system

Convolutional encoders and Viterbi decoders are suitable for a variable coding rate system because these codecs are expected to be constructed economically in the form of LSI and their coding gains are high. For example, LSI type codecs of constraint length = 4 or 7, coding rate = 1/2 or 2/3 or 3/4 or 7/8, and maximum transmission rates = 20 to 25 MHz have beendeveloped [Suzuki et a l ., 1988], [Kubota et a l ., 1987]. One of these is a general purpose codec with a selectable coding rate.

20

Rep. 552-4

C/N improvement, when a codec is applied to the variable information transmission rate system, is given as:

A(C/N) = 10 log R0 /Ra + Ga - G0

dB,

where R0 : coding rate;

Gc : coding gain at operation in clear sky conditions;

Ra , Ga : same as above for operation in rain conditions.

4.

Variable transmission rate system using spread spectrum and

synchronization sum techniques

There is another method in the variable information transmission rate system [Yamamoto et a l ., 1986]. A baseband data bit stream (information bit or error corrected bit) is scrambled by a constant clock rate PN code and then PSK modulated. The transmission rate can be varied by changing the ratio of the data bit rate and the clock of the PN code. The selected ratio must be 1/n where n is a positive integer. In a receiver, the scrambled signal is PSK demodulated at the clock of the PN code and descrambled by the PN code. The data bit is detected after synchronizing the sum over the length of the data bit at the rate of the PN code clock.

This technique has been applied to a variable transmission rate TDMA

system in which the transmission rate is adaptively variable at each TDMA burst.

It has been confirmed by experiments that the degradation of biterror rate

performance, compared with theoretical performance in a gaussian

channel,is

less than 2 dB when the TDMA system operates at a rate of 8 /n Mbit/s (n = 1, 2,

4, 8 , 16, 32).

This technique may be considered as modulation and demodulation capable of varying transmission rates using a constant clock or a variable coding rate with a coding gain of 0 dB.

5..

• Conclusion

Three kinds of variable information transmission rate techniques are discussed as methods of maintaining the signal quality of a digital satellite communication system during poor propagation conditions.

The variable coding rate technique using a convolutional encoder and a Viterbi decoder is suitable for a simple communication system with relatively economic and/or simple earth station equipment.

The other techniques are to be used in combination with a forward error correction technique. In the variable parameter system, however, an adaptively variable transmission rate modem with a maximum rate of more than 400 kbauds has not been developed. The variable transmission rate technique using the synchronizing sum fits a TDMA system rather than an SCPC system.

Rep. 552-4

21

References
SUZUKI H. et al.(1987) . MODEM and FEC LSIs for COMPACT EARTH STATION. GLOBECOM 1987, 8.3.
IWASAKI M. et al (1988) Development of Variable Rate Digital Modem for Satellite Communications, in Japanese. 1988 AUTUMN NATINAL CONVENTION RECORD, EIC, SB-2-1, JAPAN.
SUZUKI H. et al. (1987) On LSI-Implementation of Error Correcting Techniques, in Japanese. 1987 JOINT CONVENTION RECORD OF INSTITUTES OF ELECTRICAL AND INFORMATION ENGINEERS, 26-6, JAPAN.
KUBOTA S. et al.(1987). A proposal of universal-coding-rate Viterbi decoder. ICC 1987, 24.3.1-24.3.6.
YAMAMOTO M. et al.(1986). Variable Transmission Rate TDMA Communication System for Small Earth Stations Network, in Japanese. NATIONAL CONVENTION RECORD, 1986., IEICE, JAPAN, S25-1.

A N N E X III UP-LINK TRA N SM ITTIN G POWER CON TRO L

1.

Introduction

Up-link power control (UPC) can be used as a means of reducing the effect of up-link attenuation in the higher frequency bands (for example 14/11 and 30/20 G H z bands) satellite communication systems. This technique could be used to achieve efficient operation of a satellite communication system and to decrease interference to other satellite and terrestrial links by reducing clear-sky e.i.r.p.

2.

Implementation of UPC

There are various methods of achieving UPC. The most commonly used methods are as follows.

2.1 Open-loop UPC m ethod
O pen-loop U PC is a method whereby a beacon signal from the satellite is used to measure the down-link rain attenuation. Owing to the correlation between the up-link and down-link rain attenuation, this measurement is used to estimate the up-link rain attenuation level and hence the UPC control values. Most predicted attenuation values coincide with actual values; however, some values differ because of such environmental co nditions as wind velocity or rain drop-size distribution. Table IV shows an ^example o f potential errors in estim ating up-link (14 G H z) attenuation from a dow n-link (11 G H z) m easurem ent.

22

Rep. 552-4

TABLE IV - Example o f potential errors in estimating up-link (14 GHz) attenuation fr o m a dow n-link (11 GHz) measurement is tabulated below

a) Up-link attenuation o f less than 1.0 dB
Elevation angle:
Equipment error(i) Ice attenuation W ater vapour/diffusive Clear-sky level
Maximum up-link error (dB)

5°
0.725 0.05 0.20 0.10
± 1.075

15°
0.725 0.05 0.10 0.10
±0.975

25°
0.725 0.05 0.05 0.10
±0.925

b) Up-link attenuation o f between I and 6 dB
Elevation angle:
Equipment error(i) Ice attenuation Raindrop-size distribution Water vapour/diffusive Clear-sky level Polarization error P ath length error Melting layer
Maximum up-link error (dB)

5°
0.725 0.05 0.10 0.20 0.10 0.10 0.20 0.05
±1.525

15°
0.725 0.05 0.075 0.10 0.10 0.075 0.10 0.05
±1.275

25°
0.725 0.05 0.05 0.05 0.10 0.05 0.05 0.05
± 1.125

c) Up-link attenuation in excess o f 6 dB
Elevation angle:
Equipment error(i) Ice attenuation Raindrop-size distribution Water vapour/diffusive Clear-sky level Polarization error Path length error Melting layer
Maximum up-link error (dB)

5°
0.725 0.05 0.20 0.10 0.10 0.20 0.40 0.05
±1.825

15°
0.725 0.05 0.15 0.075 0.10 0.15 0.25 0.05
±1.550

25°
0.725 0.05 0.10 0.05 0.10 0.10 0.15 0.05
±1.325

(i) The equipm ent error o f ±0.725 dB assumed above is estim ated on the basis o f a ± 0 .5 dB error being encountered at 11.7 G H z (on the down-link) and assuming a. 1.45 scaling factor between 11.7 GH z and 14 GH z. The ± 0 .5 dB error was obtained using available data and needs further verification by additional measurements.

Rep. 552-4

23

Some potential error sources have been excluded as being too small to estimate (e.g., antenna tracking error, satellite antenna pointing error, pre-emphasis error, antenna gain degradation, refractive effect at low elevation angles, rapid rainfall rate fluctuation). Also excluded are error sources of a very rare type (e.g., large accumulations of wet snow on the antenna, failure in the control or measurem ent circuits). Various combinations of these additional error sources could, potentially, make the cumulative up-link power level error larger.

2.2

Closed-loop UPC method

Closed-loop UPC is a m ethod whereby the beacon signal from the satellite is c o m p ared with the loop-back C /N or S / N of a pilot signal or special channel signal. In this way the up-link rain attenuation and the UPC control value can be determined with high accuracy. A disadvantage o f the ap p ro a c h , however, is that separate control channels in addition to the communication channel are necessary.

3.

Up-link power control (U P C ) experiment

An o p en-loo p U PC experim ent was con du cted using the 3 0 /2 0 G H z b a n d [Isobe et al., 1982], with the result shown in Fig. 4 . In this experiment, the U PC values were determ ined from the values of down-link attenuation. Figure 4 a shows the beacon level, Fig. 4 ^ the H PA transmitting pow er level, and Fig. 4 c the satellite receiving level. As shown, the variation in total C /N values can be kept within 1 dB (peak-to-peak) except in the period in which the required transmitting power exceeds the maximum transmitting power.
A closed-loop UPC experiment was also conducted using the 30/20 G H z band with the result shown in Fig. 5. The control error was kept within 0.3 dB (peak-to-peak).

a) Beacon level

b) Earth-station tran sm ittin g power

c) Satellite receiving signal level

20:00

22:00

00:00

02:00

Local time (Lhr)

04:00

06:00

FIGURE 4 - Experimental results o f open-loop UPC

24 I ~
e E C§Oc0Qd>STw3

Rep. 552-4

5 E Su 3w
c/5

Local time (Lhr)

FIG URE 5- E xperim ental result o f closed-loop U PC

4.

Open-loop UPC using a radiometer

Uplink power control can be achieved by using a radiometer to measure the energy emitted by rain along the propagation path to the satellite. No beacon or pilot signal is required. Errors introduced by beacon receivers, such as the variation of gain with LNA temperature, etc., are eliminated.

The relationship between,slant-path precipitation attenuation and antenna temperature has been examined by several investigators ['Strickland, 1974]. Path attenuations calculated from measurements of the antenna temperature are generally accurate to better than 0.5 dB for attenuations less than 6 dB (at-12 GHz, in Canada). In a practical system, the uplink power will not likely be increased appreciably more than 6 dB. Thus, the radiometer can be used to compute path attenuations over the entire range of practical interest.

Sun transits will occur for a few days near the equinoxes when the declination of the sun is approximately that of the satellite. To distinguish between these increases in antenna temperature and those due to rain attenuation at other times, the satellite and solar look angles are calculated frequently. When the angular separation between the radiometer antenna axis and the sun is less than a chosen angle, the increase in antenna temperature is assumed to be due to the sun and uplink power control is inhibited.

An uplink power control system has been developed for the 14/12 GHz band in which a radiometer measures the antenna temperature in a band of frequencies below the uplink band, calculates the path attenuation at the desired uplink frequency and controls the signal strength applied at IF to the

Rep. 552-4

25

up— converter. The antenna temperature is measured by a new type of radiometer. The principle of operation differs fundamentally from that of the conventional Dicke radiometer and results in a very stable measurement of antenna temperature. The entire radiometer is contained within a cylinder mounted at the prime focus of a parabolic reflector. The radiometer frequency must differ from the uplink frequency so that transmitted energy which is backscattered from rain along the path is not detected by the radiometer, a radiometer frequency of 13.3 GHz is used.
In an experiment using the system described above, the loopback signal strength was compared with the received signal strength of the satellite beacon. The signal strengths were well correlated, indicating that the uplink signal strength as received at the satellite was almost constant, independent of rain attenuation. Additional operational experience will be gained with the two uplink power control systems currently being installed in Canada.
5 . Conclusion
UPC is one of the most important techniques for establishing higher frequency band satellite comm unica­ tions systems. By using UPC for the higher frequency ban ds, interference betw een neig hbouring satellite systems and terrestrial networks can be reduced. As a result, efficient utilization o f the geostationary-satellite orbit and efficient system operations can be achieved.
Detailed studies will be necessary for more accurate UPC methods.
REFERENCES
ISO B E,. S., S A SA O K A , H., K O SA K A , K. and OTSU , Y. [1982] Sm all traffic dom estic satellite co m m u nicatio ns svstem with a K -band transponder. IEE E G lobal T elecom m unications Conference (G L O B E C O M , '82), M iam i, Fla., USA.
STRICKLAND, J.I. [1974] - The measurement of slant path attenuation using radar radiometers and a satellite beacon. Journal de Recherches Atmospheriques, 8 , Nos. 1-2.

ANNEX IV FADE COUNTERMEASURES USING TIME-DIVISION-MULTIPLE-ACCESS TECHNIQUES (FCM-TDMA)

1.

Introduction

FCM-TDMA is a method of countering the severe effects of precipitation at the higher frequencies; it is an adaptive system which allocates an additional time resource to fading carriers in a TDMA network to thus provide an acceptable error rate in the .degraded C/N environment which occurs during fading.

A FCM-TDMA system has a portion of the frame designated as a shared resource, which is made available to fading carriers. This means that the frame efficiency, and therefore capacity, of a FCM-TDMA system is less than that of an

26

Rep. 552-4

equivalent conventional TDMA system under clear-sky conditions. The frame period is not normally a variable but any bursts suffering fading are expanded in time within the frame. This means that a burst retains the same number of information (user) bits when expanded, therefore the information rate is not changed. The technique is therefore particularly suited to public switched services/networks, where variable information rate techniques (see Annex II) may not be appropriate.

Each burst need only be expanded to the degree necessary to counter the fading experienced on a particular routing, be it onthe up-link ordown-link, or both to maximize the efficiency of the system.

2.

Methods of using an expanded burst to provide a level of noise

immunity

There are several ways of using the additional time provided by an extended time slot to give a greater degree of noise immunity, the following are examples:

a)

FEC

Differing rates of FEC overhead can be introduced in stages with increase in fade depth, the time slot being extended as necessary.

b) Reduction in transmitted data rate

The transmitted data rate can be reducedandthe sameinformation conveyed by increasing the burst length. If the transmitted data rate is reduced, the noise bandwidth at the receiver can also be reduced, giving increased noise immunity.

rate '

c) Replication of the user data within the burst

A burst suffering fading can be repeated (replicated) a number of times, and a sophisticated demodulator employed to interpret the received signal by taking a mean value for each symbol.

The above techniques all have implications on the modem design and care must be taken to ensure clock and carrier synchronization is retained during a fade. Furthermore, if the symbol rate is changed (resulting in a variation of spectra and power flux-density between bursts), there may also be interference implications.

The depth of fade which can be countered varies according to the method employed and the degree of sophistication which can be built into the modem. In practice, a composite FCM-TDMA system may be preferable; for example, adaptive FEC could be used together with any of the other methods outlined, and there is a case for using permanent FEC with method (c).

3.

Svstem control

FCM-TDMA systems will need robust protocols and control mechanisms to identify the onset and level of fading on any routing, to determine which bursts need to be expanded and by how much, and to implement such expansions together with any time-plan revision.

:Rep. 552-4

27

4.

Conclusions

An FCM-TDMA system must be tailored to need. There are many system parameters which must be determined, for example, the maximum expansion which is to be given to any particular burst, the expansion step size, the implementation or reaction time to onset of fading, the percentage of the frame to be allocated as the shared resource, etc. The sizing of these parameters will depend on the nature of the network, the climatic region, the maximum depth of fade to be countered, and on the number and data rate mix of carriers.

It may also be possible to combine FCM-TDMA with other fade countermeasures systems; for example, the FCM-TDMA protocols could be developed to incorporate up~link power control, or FCM-TDMA could be combined with a frequency"diversity system whereby bursts subject to severe fading are transmitted in an alternative TDMA frame at a lower frequency.

ANNEX V SYSTEM EXAMPLES AT 30/20 GHz

1 -

Introduction

In the following, a brief description of existing and on-going development of 30/20 GHz systems- is presented.

2 .

Japanese communication satellite (CS)

The medium capacity communication satellite for experimental purposes (CS) was launched in 1977. Through the CS experiment, valuable data were obtained.

Two communication satellites (CS-2a and 2b) were launched in 1983 for commercial use. The CS-2 satellite communication system was introduced mainly for maintaining important communication links in case of disasters, providing communication links for remote islands and setting up temporary links.

Two communication satellites (CS-3a and 3b) were launched in February and September, 1988, respectively from Tanegashima using a Japanese H-I rocket, as successors to CS-2a and 2b. Public organizations and private companies, a total number of 14, use CS-3a and 3b. They are employed for satellite communication systems to provide telephone network services, dedicated/user oriented communication services, and private communication circuits. 113 earth stations were in operation at the end of 1988 for commercial services. For the telephone network, dynamic channel assigning and routing satellite aided, digital networks (DYANET) have been newly introduced that combine satellite and terrestrial circuits to reduce network cost and enhance network reliability.

28

Rep. 552-4

Four major types of systems were brought into commercial service:

trunk transmission system utilizing earth stations with 11.5 m diameter antennas and PA-TDMA equipment and those with 4.2 m diameter dual-beam antennas and DA-TDMA equipment;
back-up communication system for local telephone links utilizing small earth stations with 4.2 m diameter antenna and 20 Mbit/s TDMA equipment;
temporary transmission system for TV and telephone utilizing transportable earth stations with 2.7 m diameter antenna and FM modulation equipment;

dedicated/user oriented communication system utilizing small earth stations with 4.2 m diameter antenna and 20 Mbps TDMA equipment.

3.

Olympus

The Olympus-1 satellite, which was launched on 11 July, 1989 to its

orbital location at 19°W, is a three-axis stabilized satellite capable of

carrying a variety of payloads. It has a rectangular 2.1 x 1.75 x 5.3 metre body

with a flexible solar array which can provide up to 3.5 kW of electrical power

at the

endof life with an expansion capability up to

7.0 kW forfuture

satellites. The launch mass of Olympus-1 is 2,600 kg; however, the structure has

been designed to handle a lift-off mass of up to 3,500 kg.

The Olympus-1 payload consisis of the following four packages:

• a two-channcl high power (63 d B W ) T V broadcast payload operating in the 11.7-12.5 G H z band for dircct-to-home transmission in Europe;
• a four-channel 12/14 G H z SS/TDMA payload for specialized or business services in Europe;

• a 30/20 G H z communications payload; and

• a propagation beacon payload operating at 12.5, 20 and 30 GHz.

The 30/20 G H z payload consists of a transponder which provides two 40 M H z narrow band channels and one 700 M H z wide band channel via two independently steerable linearly polarized transmit/receive spot beam antennas • with a 0.6° nominal coverage. The propagation payload consists of three unmodulated C W beacons which transmit via individual h o m antennas at 12.5, 20 and 30 GHz.
The experimental program to be carried out through the 30/20 G H z payload consists mainly of data transmission tests for typical business, tele-education and tele-medicine applications performed using earth stations with antennas of 2 - 3 m diameter. The beacons will be used to collect comprehensive statistics on attenuation and depolarization from rain at 12.5, 20 and 30 G H z from numerous locations throughout Europe and eastern North America.

4.

ITALSAT

In 1990; the launch of the ITALSAT pre-operational geostationary satellite is planned.

Rep. 552-4

29

This system is a forerunner of the future telecommunication domestic satellite, foreseen to be operative within the end of the century in the context of the Italian terrestrial ISDN.

The satellite system incorporates both real-time demand assignment to achieve increased efficiency and traffic rearrangement (non real-time) for reallocating the satellite capacity between traffic stations to match diurnal, weekly, seasonal and unforeseeable variations in traffic demand.

The ITALSAT satellite will carry three different payloads:

a 30/20 GHz multibeam regenerative payload with on-board baseband switching for the provision of digital telephony and data circuits;

a 30/20 GHz domestic coverage transparent payload for digital business services;

a 40 and 50 GHz package for propagation experiments.

The multi-beam payload will adopt a 147.5 Mbit/s "SS-TDMA" access technique with DSI to increase the satellite capacity. The domestic coverage payload will provide business services, by using a 25 Mbit/s TDMA system.

The ground segment will consist of 3.5 m circular antennas and 3.5 x 7m elliptical antennas depending on the climatic zones. The selected antenna configuration is a multireflector off-set type with shaped subreflectors and with provision of frequency selective surface.

5.

Kopernikus (DFS)

In December 1983^ the Deutsche Bundespost awarded a contract for the establishment of a national communication satellite network to a consortium of German firms. The project is called "Deutscher Fernmeldesatellit Kopernikus (DFS)". The DFS Kopernikus was launched on 6 June 1989 to its orbital location at 23.5°E.
The principal mission in 30/20 GHz band will be experimental and subsequent commercial use of the 30/20 GHz range for TV broadcasting with the aid of large and small mobile earth stations.
In the process, 19.73 - 19.83 GHz signals,with horizontal polarization, will be received and 29.53 - 29.63 GHz signals, with vertical polarization, transmitted.
Two large earth stations with 11 m antennas at Usingen and in Berlin (West) are ready for operation.
Additionally, it is intended to put into operation transportable stations with antennas 2.5 m in diameter.
After a two-year pre-operational trial phase, the 30/20 GHz system will go into commercial service.

6. Advanced Cfacminicatian Technology Satellite (ACTS)
ACTS is an experimental 30/20 GHz satellite ccmmunicaticns system cinder develcposnt by the National Aeronautics and Space Administration (NASA) of the United States. This system, ■which is expected to be launched in 1992, will include the following features:

30

Rep. 552-4

(a) STEERAELE ANTENNA - A mechanically steerable antenna with a 1 degree half­ power beamwidth and 1 degree per minute slew rate.

(b) MJIHTEZAM ANTENNA (MBA) - Rapidly reocnfigurable pattern of helping beams

and fixed spot

with 0.3 degrees half-pcwer beamwidth.

(c) CK-BQARD S’lURED BASEBAND SWITCHED TEKA (QSBS/TEMA) - A high speed digital base—band pmnp^gnr (EBP) on board the satellite viiich stores, regenerates and routes individual, circuit-switchert,messages.

(d) SAIEIHTE SWITCHED TEMA (SS/TCMA) - A dynamic reccnfigurable intermediate frrar-jivrryy microwave switch matrix (MSM) which routes high volume point-to-point traffic and point-to-nnltipoint traffic.

(e) RF CCMKNENIS - Flight and ground segment hardware at 20 and 30 <3iz,

(f) NETWORK CCNTRDL - Advanced algorithms to provide flexible, efficient Demand Assignment Multiple Access cxjirnunicaticns.
(g) ADAPTIVE CCKFENSAITCN FOR SIGNAL LEVEL CHANGES DDE TO RAIN - Techniques such as Forward Error Correction and uplink power control (increased pcwer, reduced burst rate) to automatically adjust far fades.
The apace segment of the ACTS system makes use of tvro systems, a remodulating on-board processor that provides demand assigned time division multiple access (TEMA) cx'iiuiuTnicaticns, and a microwave switch ratrix (MSM) that enables high burst rate satellite switched TEMA oa mnmi cations. Both of these systems mate use of spot antenna beams, hopping spot beams far the demand assigned TEMA system and fixed spot beams far the satellite switched TEMA system. The ground segment of the ACTS system is made up of the master control staticryTCF terminal, located near Cleveland, Ohio and various experimental terminals, located throughout the United States.
In the baseband processing (BBP) mode, ground terminals can use 2.4-meter, 3 -meter or 5-meter disb antennas and must use Serial Minimum Shift Keying (SMSK) modulation. Data are uplinked at burst rates of 27.5 mega symbol per second (M synfool/s) or 110 M symbol/s depending on terminal capacity requirements, and are downlinked at 110 M symbol/s. The BBP mode network is designed to accommodate 15 dB or uplink fade and 6 dB of downlink fade and still achieve end-to-end bit error rate (BER) of 1 x 10*6 . Fade compensation is implemented automatically by reducing the symbol rate by a factor of 2 and using a rate 1 / 2 convolutional FEC code to produce a 10 dB gain in performance.
In the MSM network, the NASA Link Evaluation Terminal (IET) uses a 4.72 meter dish antenna and SM3K modulation though other modulation schemes may beused in the MSM mode of operation. Data are uplinked and downlinked at 110 or 220 m symboi/s. The mem network is designed to accommodate 12 dB of uplink fade to maintain a BER of 1 x 106. Fade compensation is implemented automatically by increasing the uplink pcwer by 8 dB. Pfcdes are sensed by monitoring beaocn signals transmitted fxan the spacecraft, one in the up-link band and two in the cbwn-lirik band.
Figure 6 shews a diagram of the principal cxxrpcnents and functions of the ACTS space station, earth stations and ground facilities.
The technical features of the ACTS system support unique network architectures with the following advantages:
a. Efficient use of satellite capacity - Adaptive control of beeping beam dwell time in any one geographical location, allows the capacity of the satellite to be efficiently matched to the instantaneous demand far traffic;

Rep. 552-4

31

b. individual voice circuit switching - Stare-end-farward in oaribinaticn with baseband processing allows both time and space switching. This capability allows switching at individual voice circuits cn-board the satellite, thus avoiding double heps and the attendant delay. Shis is particularly advanta­ geous when ground stations with low to medium traffic are employed, as in the case of networks employing custraiRr premises terminals;
c. Efficient use of spectrum, and smaller ground terminals - Satellites that produce multiple narrow spot beams with relatively low side lobes levels allows frequency band reuse among the beams over a relatively mull geographical area, and the use of smaller, lower-cost ground terminals.

6c
H igh B u rst R ate (H B R ) E arth T erm inals
F i g u r e 6 - ACTS f u n c t i o n a l o v e r v ie w b l o c k d i a g r a m

Rep. 552-4
BIBLIOGRAPHY
MO RIHIRO, Y. , OKASAKA, S., NAKASHIMA, H. and O H N U K I , M. [1987] - A dynamic channel assigning and routing satellite aided digital networks. G L O B E C O M ' 87.
KATO, S., INOUE, T. and H O R I , T. [1987] - 30/20 GHz band compact earth station for LSI implemented TDMA terminal, transponder hopping HPA and dual-beam antenna. GLOBECOM'87.
ISHIDA, N. [1988] - Effective application of satellite communication for public telecommunication network (DYANET), JTR, April.
SHERTLER, R.J. [1986] - ACTS experimental program, IEEE Global Telecommunication Conference (GLOBECOM'86) Houston, United States.

Rep. 552-4

33

ANNEX V I
M ETHOD OF D ETER M IN IN G EARTH STATION A N TEN N A CH A RA CTER ISTIC S A T F R E Q U E N C IE S A B O V E 10 G H z

1.

Introduction

In comm unication-satellite systems operating at frequencies above 10 G H z , the specifications o f the earth stations, in particular the figure of merit must take account of G /T losses due to atmospheric effects and precipitation. These losses are generally specified for a percentage of time determined by the desired quality of the system.
The specification of the G / T must take account of losses:
— in the first place directly, since they lead to an increase in the required G / T \ — in the second place indirectly, since they entail an increase in the noise te m p e ra tu re T.
The formulae given below are designed to standardize the methods used in determining the antenna characteristics from the standpoint o f losses. Inform ation on the general characteristics o f earth station antennas is given in Reports 390 and 8 6 8 .

2. Specification o f the figure of merit
T he general form ula used to specify the G / T o f earth station a n te n n a s at frequencies above 10 G H z is usually written as follows:

| - L, > ( k , + 20 log T j

d B (K -‘)

(U

in the receiving ban d o f the frequencies F for atleast (100 — P,-)%o f thetime. Li, expressed in dB, is the additional loss onthe down-linkcaused by the climatic conditions specific to
the site of the earth station concerned referred to nominal clear sky conditions. The following examples may be cited:
(a) The following dual specification for earth stations belonging to the European network ECS:

- L, > (39 + 20 log 2 L )

dB (K _1)

for at least 90% o f the worst month.

T2- L^ { 31+ 20,0^ )

dB(K-1)

for at least 99.99% of the year.

(b) T he following dual'specification for IN T E L S A T sta n d a rd C earth stations:

« + 20 log

dB(K" M

to be met by all earth stations for at least 90% of the year

j r ~ L 2 * (32.5 + 20 log T J

dB(K )

to be met by earth stations in Europe for at least 99.983% of the vear.

j r - L , > (29.5 + 20 log 2 L )

jrwir- H

to be met by earth stations in North America for at least 99.983% o f the vear.

Note. — The different values for each region for the small percentage of time recognizes the different propagation conditions which exist in the two areas, while at the same time allowing the use of reasonably similar antenna design. Values for other regions are still under study.

34

Rep. 552-4

3.

Calculation model

It is proposed to establish a relation D = f (L,,

TR) which may be used to determine the circular

r* aperture diameter D for the antenna of an earth station with a — specified according to formula (1) and taking

* i

account of the receiving equipment noise temperature TR.

Taking into account the expression for antenna gain G :

kD F \2 G = 10 log M — )

formula ( 1) may be expressed as follows:

20 log D > (Li + Ki) dB + 10 log 7; - 10 log r\ + 20 log —

(2)

where:

D :

antenna diameter (m),

c:

speed of light: 3 x 10s m /s,

F0 :

frequency (GHz),

r) :

an te n n a efficiency at receiving port at frequency F0,

L ,:

atmospheric attenuation factor (referred to clear sky conditions) expressed in dB,

K j:

value specified for clear sky figure o f merit at frequency F0, expressed in d B ( K -1 ),

T j:

noise temperature of the earth station, referred to the receiving port, expressed in kelvins.

T h e earth station noise tem peratu re 7) is fairly accurately represented by the formula:

Tc + Ts + ( a —1) Tphys + Tr

K

(3)

where: Tc : Ts : Talm: Tphys '• Tk : a > \: L\ ^ 1

antenna noise temperature due to clear sky, antenna noise temperature due to ground, physical temperature of atmosphere and precipitations, physical temperature of the non-radiating elements o f the antenna feed, receiving equipment noise temperature, resistive losses due to non-radiating elements of the antenna feed, losses due to atmospheric effects and precipitation ratio.

L 'j = 1 0 io where L, is expressed in dB.

Formula (3) may conveniently be expressed as follows:

Tj = TA + (A Ta ) + Tr

(4)

where: Ta :

antenna noise temperature in clear sky conditions (L, = 0 dB)

Tc + Ts , a - 1 t P1'.VS

(5)

( A Ta ) = ad ditio nal a n te n n a noise tem peratu re caused by atm o sph eric an d precipitation losses.

(A TA) = l^ - ( T alm- T c)

(6)

a L,

Rep. 552-4

35

Inserting relation (3) or relation (4) into relation (2), we can solve D = f (L„ K„ Tr )
using additional data relating to the typical characteristics of earth station antennas operating in the frequency band considered [ESA, 1976; H a n s s o n , 1 9 8 1 1 .

4.

Sample calculation

In the following example the diameter D of an INTELSAT standard C station antenna meeting the dual specification o f paragraph 2(b) is calculated.

4.1 A ssum ptions

— the calculations are m ad e at F0 = 11.2 G H z for an elevation angle o f a b o u t 30° ab o v e the horizon

— the a n te n n a p erfo rm an ces at receiving port at the frequency F0 are:

T] = 0.67

Te = 15 K Ts = 10 K

(typical values of contribution to antenna noise temperature at an elevation angle of 30°) at F0 = 11.2 G H z

Tam - 270 K Tphys= 290 K a = 1.122 (resistive losses = 0.5 dB) — the specifications are; K x = 39 dB , K2 = 32.5 dB

4.2 Calculation results
Figure 7 shows two series of curves D = f (Lj) parametered according to the dual specification for the clear sky figure of merit ( K t) and according to three receiving e q u ip m e n t noise tem peratu re values TR (130 K. — 160 K — 190 K). In the case o f the exam ple given above, if 77? = 160 K and if we wish to install a station at a site where the propagation data are such that: L) < 1 dB for 90% of the year, L 2 < 6 dB for 99.983% of the year, we obtain the following two values for the antenna diameter: D\ = 16.80 m, D2 = 17.20 m, so that we must select D > 17.20 m.

36

Rep. 552-4

1

2

3

4

5

6

7

8

9

10

A ttenuation L, (dB)

FIGURE 7 - Variation in antenna diameter D as a function o f attenuation L j
F or two Figures o f merit at 11.2 G H z: (a): G /7 i = 39 dB(K->) (b): G /T 2 = 32.5 d B (K -])
and for three receiving equipm ent noise tem perature values TR: A : Tr = 130 K B : Tr = 160 K C : Tr = 190 K

REFERENCES
E S A 'JJan u ary , 1976] A tm o sp h eric a tte n u a tio n an d noise in satellite system s at 11-14 G H z. D oc. T E C /9 0 3 9 /S E D /S A , Rev. 2.
HANSSON L .I8 -1 0 A pril, 1981] C orrelating OTS beacon m easurem ents with radiom eter an d rainfall data: IE E International
Conference on Results o f Tests and Experim ents with the E uropean OTS Satellite, L ondon, UK.

Rep. 1139

37

REPORT 1139

GENERAL SYSTEM AND PERFORMANCE ASPECTS OF DIGITAL TRANSMISSION IN THE FIXED-SATELLITE SERVICE (Question 29/4 and Study Programme 29A/4)

(1990)

1.

Introduction

As the world-wide evolution of digital telecommunications systems continues, an increasing amount of different types of digital services are being carried by satellites as part of the evolving digital network. An example of this trend is the integrated services digital network or ISDN. This report presents material on system and performance aspects of other digital transmission systems in the FSS, in general, and, in addition, for systems carrying ISDN traffic but operating at rates greater than 64 kbit/s. Performance factors affecting ISDN transmission, that is connections operating at 64 kbit/s and designed to meet CCITT Recommendation G.821 are treated in Report 997.

2.

Performance considerations

This section presents the effects of coding on the performance of a digital satellite link describing, as an example, the effects of burst errors on performance of a 64 kbit/s channel.

All of the aspects of a satellite digital transmission system can have an effect on the performance realized by the customer or end user. The performance of satellite digital transmission systems has traditionally been characterized by bit error ratio or bit error probability. Average bit error ratio represents a good measure of system performance but the system designer must be aware of the effects of other system aspects, such as "burst errors", on the performance as realized by the end user. Note that a "burst error" does not necessarily denote a group of consecutive errors, but more generally refers to an error event where the errors occur in a cluster.

38

Rep. 1139

2.1

Error performance at transmission rates greater than 64 kbit/s

The error performance objectives currently specified by CCITT and CCIR relate to the 64 kbit/s channel level and it is necessary to consider how these objectives can be applied to higher transmission rates. When con­ sidering this question it is important to distinguish between systems carrying separately identifiable 64 kbit/s channels and those carrying wideband services (e.g. at 2.048 or 1.544 Mbit/s). The sensitivity of particular coding schemes and frame structures must also be taken into account.

In this section, we report a generalization of performance objectives of other bit rates.

The methods described in section 2 of Report 997 for the 64 kbit/s rate can be easily generalized to accommodate other bit rates. We show below how the probability P (< E) of obtaining E errors or less during a time T, when transmitting at a rate R is given by:

P «E) = 1 - P (2RT x BEP, 2E + 2)

(1)

where.BEP is the Bit Error Probability of the system and the number of
errors is simply given by E = RT x BER. J? (X , v) is known as the
cumulative chi-squared (X^) distribution function with v degrees of freedom. In this formalism the calculation of BEP can easily be
achieved by looking at tables of percentage points of the X^ distribution which are readily available in the literature.

This approach shows that the relevant parameter is neither the transmission rate, R, nor the time interval considered, T, but rather the product of the two. RT refers to the total number of bits transmitted in the interval considered, and this block of data is the effective parameter when taking measurements. It is therefore natural to consider a constant block size when defining a recommendation, so that such a recommendation will be independent of the transmission rate and of the time interval considered. Thus, Recommendation 614 dealing with EFS at 64 kbit/s can be generalized to read:
i) P-^% of the transmitted blocks (RT bits) must be error free,

and the recommendation on BER is generalized to:

ii) P2 % of the transmitted blocks (RT'bits) shouldhave a BER<B.

with the parameters in both recommendations being related by

thevalue

of BE?.

The 64 kbit/s EFS recommendation is simply a particular case of

i) with R = 64000 and T = 1 second. The degraded minutes recommendation

is a particular case of ii) with T' = 60 seconds and B = 10“^, while the

SES recommendation refers to ii) with T' = 1 second and B = 10“^.

The error free interval defined on equal size blocks of data provides a universal measure of performance which can be applied to any transmission rate. In general it will be natural to select as the blocks of data those commensurate with the protocol used by the system of interest. This will undoubtly avoid the problems associated with locking error performance recom­ mendations to a particular rate or time interval. However, given the present predilection for, and the widespread use of, EFS as a determinant of the quality of a digital transmission system, the level of performance must be selected to conform to expectations for the particular block that corresponds to one second for the transmission rate of the system.

Rep. 1139

39

One immediate application of these results is the performance recommendation at the Primary Rates (1.544 Mbit/s and 2.048 Mbit/s). The error free block performance corresponding to an acceptable EFS level of the Primary Rates may be different from that selected for 64 kbit/s. The specific EFS level for the Primary Rates must be based on specific ISDN appli­ cations and customer needs, as well as on practical technical limitations and costs of satellite systems. Other parameters can then be derived using the appropriate statistical and propagation models. Again the specific additional parameters must be selected based on Primary Rate applications. For example, customers transmitting video teleconferring signals may desire 10 minute inter­ val specifications.
An immediate consequence of the approach stated above is that the knowledge of the BEP of the system allows the calculation of the probability of obtaining any given number of errors in any given time interval at any given transmission rate. BEP is the parameter that determines the performance of the system. This, in turn, provides for a simple way to relate recommendations between two different rates. All we need to do is to calculate the value of the BEP by using equation (1) with the known values of the parameter at one rate, and recalculate the performance objectives at the new rate by means of the same equation and the same BEP, with the low rate parameters.
The Poisson statistical distribution, as a limiting case, has provided illuminating results. However, it is necessary to reexamine this assumption by looking at more bursty distributions of errors and study how the results chanqe from the ones quoted above. See Annex IV of Report 997 for a treatment of the effects of bursty errors on 64 kbit/s ISDN connections.
In Annex I of Report 997 a Poisson distribution formula is presented that is used to generate the graphs that apply to the specific case of the 64 kbit/s rate. The formulae given below are generalized to accommodate any transmission rate.
If we denote, by R the transmission rate of the system and by T the time interval under consideration, the total number of transmitted bits is given by N=RT. The probability of having E errors or less, when transmitting these N bits is given by:

P«E)

(RT X BEP) - RT x BEP

(2)

n=o

;rr“ e

where BEP is the Bit Error Probability of the system and the number of errors E=N x BER.
Equation (2) can be transformed into a form more amenable to analytical inversion. Calculating the derivative of the sum with respect to BEP, and expressing it in terms of the original sum, one can find a differential equation that can be easily integrated to obtain:

P«E) = 1-P(2RT x BEP, 2E+2 )

(3)

* O

p

where P(X ,v) is the cumulative chi-squared (X ) distribution

function with v degrees of freedom, and it is defined in terms of the

integral:

1

r X 2/£ v. -i

p ^ ) = — ); j

**

—

at

(4)

40

Rep. 1139

Since what is needed is the variable BEP as a function of R for a fixed value of the cumulative function, one only needs to look at tabulations of the percentage points of the x distribution, which are widely available. Notice that the transmission rate also appears in the variable E; for a given value of R , one can easily obtain v (2E + 2) and read the x^ value from the statistical tables without the need for further extrapolation. That is, from the value of x^ the Bit Error Probability is readily obtained.

BEP = x2/*2 R T)

'

This method avoids the need for inverse extrapolation techniques and for numerically calculating the probability function. It can be used more effectively and permits simple generalization to any transmission rate.
For the purpose of evaluating error performance objectives normalized to 64 kbit/s on the basis of measurement results obtained at the bit rate of a primary digital system or higher order systems, the following method may be used:

an error substream corresponding to the 64 kbit/s channel is formed by selective demultiplexing from the error stream extracted from the signal transmitted over the system;

the 64 kbit/s channel error signal thus obtained is processed in accordance with the algorithm given in Annex B to CCITT Recommendation G.821.

The error stream selective demultiplexing method can also be used to evaluate the performance objectives of various services with bit rates exceeding 64 kbit/s (e.g. sound broadcasting or television) which are component parts of a high bit rate signal.

Error characteristic measurements are also considered in Report 613.

2.2

Burst error events

As mentioned previously, average bit error ratio or probability gives a good measure of the long term performance of a digital transmission system and forms the basis for system design. The performance as realized by the end user can be greatly affected by the distribution of the errors. Systems using multiplexed 64 kbit/s channels experiencing "burst errors" may demonstrate degraded performance on some channels and acceptable performance on others when using octet interleaved multiplexing. Systems using bit interleaved multiplexing, will experience equal degradation of all channels. Channels with bursty error statistics may also cause problems in end to end signalling.

The systems that are connected to the transmission system may also be affected by the statistics of the bit errors. Digital television coder/decoders (codecs) that use bandwidth compression techniques to improve the efficiency of digital television transmission can exhibit picture degradation and loss of codec synchronization due to "bursty" errors. Digital circuit multiplication equipment (DCME) can also experience the loss of system synchronization due to "bursty" errors. The degree of degradation is of course dependent upon the specific equipment architecture which is connected to the transmission system. Source coding used at the terminal equipment may improve or degrade the sensitivity to "bursty" errors.

Rep. 1139

41

2.3

Loss of frame-synchronization

Loss of synchronization (for example loss of multiplex frame alignment) can happen when a long error burst is encountered. The garbled output during the re-framing time may not be detected upstream, and performance monitoring at the higher bit rate may then not represent the performance at the enduser bit rate. System designers should ensure that the frame synchronization structure be sufficiently robust to minimize this problem. Any increase in margin in the performance recommendation to cover de-synchronization will require further study.

3.

Synchronous Digital Hierarchy

CCITT Recommendations G. 707,708, and 709 relate to the Synchronous Digital Hierarchy (SDH).

One of the major features of this hierarchy is the new bit rates which transmission media will have to accomodate. Specifically, the lowest transport rate is 155 Mbit/s and so the -implications of this on the design of satellite networks need to be considered.

If a satellite transmission network has to support the full 155 Mbit/s signal, new transmission equipment will have to be developed and it may also be necessary to increase the bandwidth above the typical 72 MHz currently in use.

Alternatives could be developed to interface between the SDH and existing transmission rates, as well as a sub 155 Mbit/s rate of appoximately 50 or 40 Mbit/s. It this case some of the features of the SDH, particularly those relating to network monitoring and control, would be lost/reduced.

4.

Broadband Integrated Service DigitalNetwork (B-ISDN)

The Broadband Integrated Service Digital Network (B-ISDN) is a network concept that is developing from the Narrowband (64 kbit/s) ISDN (N-ISDN) and will be a network that will provide broadband services such as large data base enquiries, broadcast TV distribution, High definition TV, and video telephone,as well as those services currently being provided by the N-ISDN standard. These are expected to be implemented in the 1990's and it can be expected that connections in the B-ISDN will involve satellite radio connections particularly for intercontinental connection.

Transmission bit rate is a primary factor in determining how efficiently service such as B-ISDN can be transmitted via satellite.

This section examines existing and planned satellite systems that are scheduled for installation in the 1990's, and based on this examination, system bit rates that would allow efficient satellitetransmission are identified.

42

Rep. 1139

4.1

B-ISDN Transmission Rates

CCITT Broadband Task Group (BBTG), at its January 1989 meeting, established the Asynchronous Transfer Mode (ATM) as its target mode for Broadband ISDN. The broadband chanel rates which must be supported, per CCITT Recommendation I. 121, are: H21 (32.768 Mbit/s), H22 (43 - 45 Mbit/s), H4 (132 138.240 Mbit/s). The user-network interface would be at 150 Mbit/s, and possibly 600 Mbit/s. ATM can be supported by the Synchronous Digital Hierarchy (SDH) described in section 3.

4.1.1

Digital Transmission of Television Signals

Broadband Integrated Services Digital Networks (B-ISDN) are currently being defined by the CCITT. One application of B-ISDN will certainly be the distribution of television signal. Due to the point-to-multipoint nature of communications satellite systems, and the need to distribute television signal over wide areas, satellites are uniquely suited to the distribution of digital television signal. By careful application of image coding and channel coding techniques, where necessary, digital television transmission rate requirements can be made compatible with communications satellite transmission capabilities.

4.1.2

Satellite Transmission Rates

As described above, a variety of channel rates have been specified in B-ISDN Recommendation 1.121. These rates will however be supported by both the existing and new digital hierarchies.

For the existing hierarchy, a common standard exits at 139.264 Mbit/s and is in use on a wide variety of transmission systems (eg TAT 8 ). Multiplexing schemes are specified so that the various rates of both the North American and European hierarchies can be accommodated.

For the SDH, the comparable rate is 155.52 Mbit/s.

It seems logical that from the discussion above the transmission rate used for B-ISDN would be between 140 to 160 Mbit/s. A rate within this range would be compatible with both existing and currently planned communication satellites. Satellite transmission at 140 Mbit/s through a 72 MHz transponder has already been demonstrated. In order to accommodate transmission rates of up to 160 Mbit/s, a coding rate of 8/9, as opposed to the presently tested coding rate of 7/9, could be used. This increase in coding rate implies a decrease in coding gain and hence the necessity to make up this loss somewhere in the satellite transmission path.

4.2

Satellite Capabilities

4.2.1 Satellite Transponders

The transmission capacity of satellite transponders is basically charaterized by the available output power and the usable bandwidth. The Table I shows the transponder bandwidths and powers to the INTELSAT satellites that are already in service or are planned for service in the 1990's. The earth coverage regions of these transponders are also shown.

Carriers bearing B-ISDN traffic would most likely require full transponder bandwidth and, for reasons of efficiency, these carriers would operate close to the saturated transponder EIRP. As an example of current transmission systems, the INTELSAT TDMA system operates at a transmission rate of 120 Mbit/s in a 72 MHz bandwidth transponder at an EIRP of 27 dBW. Tests have demonstrated satisfactory performance of aCoded Octal Phase Shift Keyed (COPSK) system operating at an information rate of 140 Mbit/s and channel rate of 180 Mbit/s in a 72 MHz bandwidth transponder at an EIRP of 28.5 dBU.

Rep. 1139

43

4.2.2

Use of Spot Beam Antennas

The use of spot beams owing to its high satellite antenna gain will enable satellite to handle higher bit rate digital signals in proportion tothe number of beams to cover a certain service area. In Japan, an experimental multibeam communication satellite, ETS-VI, will be launched in 1993 which covers the Japanese main islands by 13 spot beams. At the present, a single beam system carries 20 Mbit/s digital signal by using a 4.2m diameter earth station antenna, it is possible to transmit as high as 250 Mbit/s digital signal by using the same RF facility. Experimental applications of satellite communication to the B-ISDN are planned to be carried out using the ETS-VI.

As the rates that seem probable for B-ISDN have been indicated and it has been shown that both existing and currently planned telecommunications satellites are capable of accommodating these transmission rates, satellites will be able to form integral parts of B-ISDN connections.

4.3

Asynchronous Transfer Mode

A B-ISDN is a network that can support a wide range of audio, video and data applications, such as video telephony, video surveillance, television (standard, enhanced and high definition), file transfer, high speed data, etc, in addition to the traditional ISDN services.In order tobe able to handle such a wide variety of signals represented by various bit rates, the CCITT has established the Asynchronous Transfer Mode (ATM) as the target transfer mode solution for implementing B-ISDN.

ATM is a packet-oriented switched network in which multiplexed information is organized in fixed-size blocks called cells. A cell consists of a user information field and a header. The primary role of the header is to identify cells belonging to the same virtual channel - i.e. the same connection. At its June 1989 meeting, CCITT Study Group XVIII establisheda cell size of 53 8 -bit bytes (or octets), consisting of an information field of 48 octets and a header field of 5 octets.
The header contains only the information required to transfer the information field through the ATM. The following are some of the functions of the header:
Virtual channel identification (VCI), to identify the cells belonging to the same virtual channel Virtual path identification (VPI) Access Control Field (ACF) Priority and congestion control (PR) Header error control (HEC) Payload identification (PI), which indicates whether the information field contains network or customer data.
The HEC can correct single errors in the header field andcandetect multiple (2 or 3) errors. There is no check of the information bits. As a result,the primary types of errors are:
detected but corrected errors detected but uncorrected errors, which could result in erroneous addresses and thus loss, undetected errors, resulting in delivery to the wrong address.

44

Rep. 1139

Information is needed on the acceptability of the various kinds of errors, and also of errors in the information fields. Till now most of the activity has concentrated upon the structure of the ATM. It is necessary to address the question of ATM performance specification, i.e. signal impairment levelswhich correspond to acceptable qualityof services. These specifications may then be related to the contributions, ifany, expected from the satellite portion of any connection.
For example, in the cases of video services, user-perceived criteria could be:
events affecting 1 or 2 TV lines events affecting 2 fields, events affecting multiple fields
These could be affected by individual and burst errors, short interruptions, information losses and delay, resulting in cell losses, cell transfer delay and information errors.
Cell losses may be due to header error as discussed above and to buffer over-flows;
Cell delay may be due to queueing delays in ATM and to propagation;
Information errors may be due to a variety of causes.
The need for standards on these should be further studied in the CCIR as paced by decisions in the CCITT.
TABLE I - Equivalent isotropicallv radiated power and usable bandwidth of existing or planned international communications satellites

Satellite Designation
INTELSAT V (currently .in-service)

Usable Transponder
Bandwidth
36 MHz 41 MHz 72 MHz 77 MHz 241 MHz

Minimum Transponder
EIRP*

Antenna Beam Earth
Coverage

23.5 dBW 23.5 dBW 29.0,29.0,41.1 dBW 29.0,29.0,41.1 dBW 41.1 dBW

Global Global Hemi ,Zone,Spot Hemi,Zone,Spot
Spot

INTELSAT VI <(initial launch
1989)

36 MHz 41 MHz 72 MHz 77 MHz 150 MHz

26.5,28.0 dBW 26.5 dBW
31.0,31.0,41.7 dBW 41.7 dBW 44.7 dBW

Gobal,Zone Global
Hemi,Zone,Spot Spot Spot

INTELSAT VII (planned launched 1993)
* Saturated EIRP

36 MHz 41 MHz 72 MHz 77 MHz 112 MHz

26.5,33.0.33.0 dBW 29.0 dBW
33.0,33.0,42.0 dBW 33.0,33.0,42.0 dBW
42.0 dBW

Global,Hemi,Zone Global
Hemi,Zone,Spot Hemi,Zone,Spot
Spot

Rep. 1139

45

5.

Interference and sharing constraints

In designing a digital satellite circuit to meet the ISDN performance objectives in CCITT recommendation G.821, the system designer must include in his link budget some allowance for interference. One possible method of determining the permissible level of interference received from fixed-satellite services and terrestrial services would be to assume the interfering system will be accomodated on the basis of Recommendations 523 and 558. However, because of the improved performance (relative to recommendation 522) and the unique requirements of ISDN digital connections interference should be considered on a different basis, especially, systems operating in the 14/11-12 GHz bands.

Studies have shown that 14/11-12 GHz ISDN satellite connections, operating in moderate rain climatic regions (e.g. CCIR rain climatic zone 'K') will be generally limited by the short term performance objective, (BER < 1 x 10" 3 for 0.03% of any month). In such cases the clearsky operating point may be much better than 1 x 10" 7 BER long term requirement as stated in Recommendation 614. Therefore an interference budget similar to Recommendation 523, based on the 10" 7 performance objective would not be related to the clear sky or the degraded operating conditions, nor the availability objectives.

Another consideration in developing an interference criteria for ISDN connections is the effect of interference on systems utilizing FEC. Most digital systems are utilizing some forms of FEC codingas a means of lowering the required C/N to achieve agiven performance objective. However, by utilizing FEC a system becomes more sensitive to changes in the C/N. In one example it was found a 25% increase in system noise due to interference could degrade a systems BER with FEC about 6 times more than a system without FEC.
Figure 1 gives an example of the INTELSAT 2.048 Mbit/s service operating at 14/12 GHz in a fading environment. It is assumed that the up link is fading in this example and more information can be found in section 4 of CCIR Report 710 on the assumptions that are made.

The network is designed with a C/I of 16.7 dB, to meet the CCIR proposed allocation of interference noise for 10% of any month. With fading, this C/I will be reduced to 10.3 dB which is only 3.6 dB above the C/N required to meet CCIR Recommendation 614 for a BER of 10~ 3 at 0.2% of the worst month. In other words, the effect of this interference has become a significant factor, in terms of C/I, at the operating point of 10- 3 BER.

Figure 1 also shows that the system performance is controlled by the short term performance requirement at 0.2% of the month. The carrier level was chosen just to meet this short-term performance, i.e. the unfaded carrier level had to be increased to take account of fading due to rain in order to meet the 0 .2 % point.
The effect of increasing the carrier level could have an impact on orbit efficiency. However, if some form of fade compensation, such as up link power control, is used the carrier during clear weather can be reduced back to its nominal level and thereby not affect the utilization of the orbit. Up link power control can also be used toimprove the short performance in order to reduce the total noise as well as the interference noise at the 0 .2 % of the month point. Figure 1 shows how using 4 dB up link power control can improve the performance at 1 0 "^ BER.
Table II summarizes the amount of up link power control that may be needed for a moderate climate. It also shows the difference between a system with and without FEC. The system without FEC is not used by INTELSAT and is only

46

Rep. 1139

shown for a comparison. It also shows the sharp roll-off for the FEC performance between a BER of 1 0 and 10'^. A more complicated case, that of two mutually interfering systems, both using UPC, requires further study.
These considerations show that existing systems interference criteria, when applied to ISDN connections, do not reflect the unique requirements of ISDN connections and other approaches may be needed. One approach would be to base the interference on a percentage of the system noise, thereby eliminating the referencing of interference to any specific BER performance objective or to the use or non-use of FEC. Another approach would be to base the interference on the additional percentage of time the system can be degraded.
A similar consideration of the interference from terrestrial systems is also required.
It is considered that urgent and intensive study is required to develop an appropriate interference recommendation for an ISDN Satellite connection. This study might also have to take into account the need to maintain existing orbital efficiences while still making ISDN connections cost effective. Annex I is a possible form of such a Recommendation.

Rep. 1139

47

CCHfe40446—1.DWG

Carrier-to-noise ratio, (dB)

Percent of any month (%) FIGURE 1 - 2.048 Mbit/s C/N with fading, region K with FEC at 14/12 GHz
A CCIR INTERFERENCE ALLOWABLE, (C/I) Rec. 523
^ HDRP PERFORMANCE, (C/N) Rec. 614 * NOTE: EQUIVALENT C/I AND C/N IMPROVEMENT CAN BE ACHIEVED BY
INCREASING THE UNFADED UPLINK POWER BY 4 dB OR OTHER MEANS

48

Rep. 1139

T A B L E II I N T E L S A T 2 . 0 4 8 Mbit/s SERVICE AT 1 4 / 1 2 G H z . K- C L I M A T E

W i t h .FEC

Without FEC

Fading Depth at 0.2% of the month w i t h r e s p e c t to 10% of the month
Difference in C/N between BER 10"7 & 10'3 in Demodulator*
Advantage due to.operating in the non-linear region of power amplifier at sa turat ion

6.4 dB -3.0 dB -0.0 dB

r O

o

6.4 dB -4.6 dB
dB

1

I

Amount of Fade Compensation Required**

3.4 dB

1.8 dB

As measured over the entire system
** R e q u i r e d in o r d e r to n o t i m pinge on the s a t e l l i t e spacing set by the long term interference allowance, at 10% of the month

6.

Other Applications of Satellite Links in the ISDN

There are various ways to apply satellite links within the ISDN which can lead to different system performance objectives. This section gives some considerations on the applications of satellite links within the ISDN.

Such applications of satellite links as a part of the ISDN can be classified into two basic forms as follows:

(1 ) applied to one or more sections of the digital path in place of terrestrial systems;

(2 ) applied to one or more sections of the digital path in parallel with terrestrial systems.

Figure 2 shows an example of the first form while Fig. 3 shows an ^example of the second form.

6 .1

Performance Objectives

6.1.1 Error Performance

When a satellite link is applied to one or more sections of the .'digital path in place of terrestrial systems, overall end-to-end error ■performance requirements can be met if the performance objective of the ^satellite link equals the sum of the error performance objectives allocated to ‘the corresponding section of HRX as given in Report 997. Figure 4 shows several .examples of error performance objective allotments for a satellite link used in ;.this way.

Rep. 1139

49

6.1.2

Availability

The unavailability objectives of a satellite HRDP due to equipment and propagation are given in Recommendation 579 and would be applicable to satellite links used in any of the above situations.

In satellite communication systems using frequency bands above 10 GHz, site diversity technology is an effective countermeasure against severe rain attenuation. In satellite links which are applied in parallel with terrestrial systems, earth stations are not required to process the site diversity function because of its dynamic routing capability. If the terminating earth station nearest to the destination subscriber encounters heavy rain attenuation, a satellite channel is assigned to the unaffected earth station second nearest to the destination subscriber.

Local

Medium

High grade

Medium

Local

grade

grade

>. ✓

grade

grade

------ O ------------- V ------------- O --------- O -------- C

T

LE

' '

Satellite link

LET

( a ) A s a t e l l i t e l i n k a ppl i e d t o medium grade and hi gh grade p o r t i o n

Local grade

Medium
grade \ ,

High grade

Medium grade

Local grade

o ; j------ V — --------------- o -------- o -------o

T

/ x

LET

Satellite link ( b ) A s a t e l l i t e l i n k a p p lie d to l o c a l g rad e , medium grade and high grade p o rtio n

using user premise earth stations

FIGURE 2 - Configuration examples of s a t e l l i t e lin k s a p p lie d to one o r s e v e ra l s e c tio n s of the di git al path as substitution for t er r es t r i a l systems

Local

Medium

High grade

Medium

Local

grade

grade

grade

grade

O --------- O -------------O -------------------- C ----------- O ------- — O

T______ L E ________________ |_____________

|

L E

T

Satellite lin k

( a ) A s a te llite lin k anplied to high grade p o rtio n in p a ra lle l with

t e r r e s t r i a l systems

Local

Medium

High grade

Medium

Local

grade

grade

grade

grade

O -------O --------- O ---------------O -------- O ------- O

T

L E

,

LET

Satellite lin k ( b ) A s a t e l l i t e lin k a p p lie d to medium grade and high grade p o rtio n
in p a ra lle l w ith t e r r e s t r i a l systems
FIGURE 3 - C o n fig u ratio n examples of s a t e l l i t e lin k s a p p lie d to one or s e v e ra l sectio n s of the d ig ita l path in p a ra lle l w ith t e r r e s t r i a l systems

50

Rep. 1139

Local grade
T
^ 15%

Medi um

Hi gh g r a d e

Me di um

Local

grade LE

\ /

grade

grade

- x - -------- o -------- o ------- o

LE

' T

Satellite link

;_____________ 553

^ 15%

153 |

( 1 ) A satellite link applied to medium grade and high grade portion

Local

Medium

High grade

Medium

Local

grade

grade

\ ✓

grade

grade

o ------- j---------- | - V ------------c -------- o -------o

/ \ Satellite link

LET

^___________________ 70S

^

153

15S j

( 2 ) A s a t e l l i t e lin k a p p lie d to lo c a l g rad e , medium grade and high grade port ion

FIGURE A - Exampl es of e r r o r p e r f o r ma n c e o b j e c t i v e a l l o t m e n t s f o r t he HRDP u s i n g satellite links

6 .2

An example of satellite communication systems applied to parts

of the ISDN with terrestrial systems

In Japan, a satellite transit network with dynamic channel assignment . and routing capability has been introduced to the public switched telephone network since October 1988. This system combines terrestrial-based circuits and satellite circuits.

Figure 5 shows the system configuration and Table m s h o w s major parameters of this system. It uses the Japanese domestic communication satellite CS-3. The frequency band used in this system is 30/20 GHz band. This system consists of a number of traffic earth stations and one control earth station. The control earth station transmits and receives call control signals to/from terrestrial switches, and selects the destination traffic earth station and switch.

In this system, the satellite circuit group can be used commonly in units of 64 kbit/s circuit when the terrestrial circuit group is fully occupied. This can be realized by demand assigned TDMA techniques.

The earth station has a 4.2 m diameter dual-beam offset type antenna with torus-type reflector, and can access two satellites; CS-3a and 3b, simultaneously. As multiple transponders are used for many small traffic earth stations in this system, it is necessary for each station to access multiple transponders to connect with each other. Therefore, transponder hopping technology, one of the schemes available to them with a single set of TDMA equipment, transmitter and receiver, is effectively applied to the system in order to economize on earth station cost.

Rep. 1139

51

CS : C o m m u n i c a t i o n Satellite ES : Earth S t a t i o n E q u i p m e n t TES: Traffic Earth Station CES: C e n tra liz e d Control Earth S ta tio n SCU: Satellite ch an n el Control Unit SXU: Satellite co m m u n ic a tio n tr a n sit E x ch an g e Unit

IS : Local Switch TS : Toll S w i t c h
: Traffic channel — : Control channel

FIGURE 5 - Sy s t e m co nf i g u r a t i o n of trunk t r a n s m i s s i o n s y st em c o m b i n e d s a t e l l i t e sy stems and terrestrial svstems

TABLE III - Major parameters of satellite trunk transmission svstem

Item Frequency Multiple access Earth stations
Modem FEC
Clock rate Transmission capacity TDMA-Terrestrial network interface Channel assignment

Content 30/20 GHz bands TDMA Reference stations; 2 Traffic terminals ;30/transponder QPSK-Coherent demodulation Rate 1/2 convolutional encoding Viterbi decoding (constraint length: 4) 25.024 MHz 160ch/transponder (in 64 kbit/s channels) 8.192 Mbit/s
Demand assignment

U.I.T.

52

Rep. 1139

7.

Conclusions

This report has presented the general system and performance aspects of satellite digital transmission, and specifically, for systems carrying ISDN traffic operating at rates greater than 64 kbit/s.

Performance considerations have been discussed, noting that bit error ratio is still the preferred method of specifying digital system performance in the fixed-satellite service. System designers are cautioned to be aware of the effects on the performance of an end-to-end connection and end user equipment that can be caused by statistically non-random error distributions.

The application of satellites in the developing framework of digital transmission by the synchronous digital hierarchy (SDH) and in the broadband ISDN (B-ISDN) have also been considered.

Specific attention has been given to the effects of interference on ISDN performance, primarily for narrow-band ISDN (N-ISDN). It is anticipated that this information will be developed into a Recommendation providing guidelines on interference into satellite portions ISDN connections.

Finally, the applications of satellite links, other than those specifically covered by the hypothetical reference digital path (HRDP) given in CCITT Recommendation G.821, in the ISDN have been considered. It was shown that, when satellites are applied to digital sections normally carried by terrestrial means, more of the error performance allotment can be apportioned to the satellite part of the connection.

It is hoped that this report will serve as the basis for the development of new or improved Recommendations on digital transmission in the fixed-satellite service.

Some areas that require further study are:

effects of bursty errors on equipment and systems that use bandwidth compression techniques;

effects of interference on systems that use UPC;

performance studies on packet-oriented switched networks.

BIBLIOGRAPHY
1.. C C I T T , COM X V I I I N° D . 1 6 4 / W P 6 , J u n e 1 9 8 9 .
2_ MINKIN,V.M. and ITKIS, G.E. [1989] - Eksperimentalnoe issledovanie oshibok v tsifrovykh lineinykh traktakh (Experimental study of errors in digital systems), Trudy NIIR. No. 1.

Rep. 1139

53

ANNEX I
Elements for the preparation of a possible New Recommendation for consideration during the next Study Period
MAXIMUM PERMISSIBLE LEVELS OF INTERFERENCE IN A GEOSTATIONARY-SATELLITE NETWORK FOR A HYPOTHETICAL REFERENCE
DIGITAL PATH WHEN FORMING PART OF THE INTEGRATED SERVICES DIGITAL NETWORK IN THE FIXED-SATELLITE SERVICE CAUSED BY
OTHER NETWORKS OF THIS SERVICE BELOW 15 GHz
(Study Programme 28C-1/4)

The CCIR,

C O N SID ERIN G

(a) that geostationary-satellite networks in the fixed-satellite service operate in the same frequency bands:

(b) that interference between networks in the fixed-satellite service degrades the e r r o r p e r f o r m a n c e relative to its value in the absence o f frequency sharing;

(c)

that it is desirable that the i n t e r f e r e n c e in networks in the fixed-satellite service caused by transmitters of

different networks in that service should be such, as to give a reasonable orbit utilization efficiency;

(d) that the overall performance of a network should essentially be under the control of the system designer;

(e) that it is necessary to protect a network in the fixed-satellite service from interference by other such networks;

( f ) that it is necessary to determ ine the m axim um permissible interfering radio frequency pow er in a satellite system to establish space station and earth station characteristics such as required protection ratios and minimum orbital spacing;

(g) that netw orks in the fixed-satellite service may receive interference both into the space station receiver and into the earth station receiver;

(h) th a t it is desirable that the increase in ---------------------------------- in terfe rence from o th e r satellite netw orks should be a controlled fraction of the total n o i s e t h a t w o u ld g i v e r i s e t o a b i t e r r o r r a t i o , as s e t o u t in Recom m endation 614;

j ) that the levels of interference between geostationary-satellite networks in the fixed-satellite service below 10 G H z are n o t expected to exhibit a large variation with time, an d u n d e r these c o n d itio n s it is preferable to define the permissible interference limit as a fraction of the pre-demodulator noise power, as this allows multiple interference entries to be superimposed on each other on the basis of RF power addition;

k)

that in frequency b a n d s betw een 10 a n d 15 G H z where very high p ro p a g a t io n a tte n u a tio n m ay occur for

short pe riods o f time, it w ould generally be desirable for systems to m ake use o f ------------------------------------------------

s o m e f o r m o f f a d e c o m p e n s a t i o n ---------- to cou n teract signal fading a n d th a t u n d e r these circumstances

the levels of interference from other satellite systems would also not undergo a large variation with time.

54

Rep. 1139

RECOM M ENDS

1.

that netw orks in the fixed-satellite service operating in the same frequency b and s below 15 G H z , an d using

geostationary satellites be designed and operated in such a m anner that the total interference to a 64 k b i t / s

c o n n e c t i o n in the fixed-satellite service caused by the earth station and space station transmitters of all other

networks, should conform provisionally to the following limits:

ISDN

1.1 in frequency b a n d s in which the netw ork does not practise frequency re-use, the interference pow er level, s h o u l d n o t e x c e e d , f o r m o r e t h a n 10% o f a n y m o n t h a s r e f e r r e d t o RECOMMENDS 1 . 1 o f R e c o m m e n d a ti o n 6 1 4 , 25% o f t h e t o t a l n o i s e p o w e r l e v e l a t t h e i n p u t t o t h e d e m o d u la to r w hich w ould g iv e r i s e to a b i t e r r o r r a t i o of 1 in 1 0 ^ ;

1.2 in frequency bands in which the network practises frequency re-use, the interference power level,
s h o u l d n o t e x c e e d , f o r m o r e t h a n 10% o f a n y m o n t h a s r e f e r r e d t o RECOMMENDS 1 . 1 o f R e c o m m e n d a ti o n 6 1 4 , 20% o f t h e t o t a l n o i s e p o w e r l e v e l a t t h e i n p u t t o t h e d e m o d u la to r w hich w ould g iv e r i s e to a b i t e r r o r r a t i o of 1 in 1 0 ^ ;

2. that the m ax im u m level o f interference power in any 64 k b i t / s ISDN c o n n e c t i o n caused by the transmitters of another fixed-satellite network, s h o u l d n o t e x c e e d , f o r m o re t h a n 1 0 % o f a n y m o n th a s r e f e r r e d t o RECOMMENDS 1 . 1 o f R e c o m m e n d a t i o n 6 1 4 , 6 % o n a p r o v i s i o n a l b a s i s o f th e t o t a l n o is e pow er le v e l a t th e in p u t to th e d em o d u la to r w hich w ould g iv e r is e to a b it e rro r r a tio of 1 in 10^ ;

3.

that the maximum level of interference noise power caused to that network should be calculated on the

basis o f the following values for the receiving earth station a n te n n a gain, in a direction at an angle ip (in degrees)

referred to the main beam direction:

G = 32 — 25 log cp dBi for 1° < <p < 48°

G = -10

dBi for 48° < <p < 180°

except when the actual gain is know n and is less th an the above value, in which case the actual value should be used ;

4.

that the following notes should be regarded as part of this Recommendation:

N ote 1. — For the calculation o f the limits quoted in §§ 1.1, 1.2 an d 2 it should be assumed tha t the total noise p ow er at the input to the d e m o d u l a to r is o f thermal nature.

N ote 2. — It is assum ed in this R ecom m endation that the interference from other satellite netw orks is o f a c o n t i n u o u s n a t u r e a t f r e q u e n c i e s b e lo w 10 GHz: f u r t h e r s t u d y i s r e q u i r e d w ith r e s p e c t t o c a s e s w h e re i n t e r f e r e n c e i s n o t o f a c o n ti n u o u s n a t u r e a b o v e 10 GHz.

N o t e 3 : F o r e x i s t i n g n e t w o r k s u s i n g 8 - b i t PCM e n c o d e d t e l e p h o n y ; s e e N o t e 3 o f R e c . 5 2 3 .

N ote 4. — In some cases it may be necessary to limit the single entry interference value to less than the value quoted in § 2 above in order that the total value recommended in § 1 may not be exceeded. In other cases, particularly in congested arcs of the geostationary-satellite orbit, administrations may agree bilaterally to use higher single entry interference values than those quoted in § 2 above, but anyinterferencenoise power in excess of the value recommended in § 2 should be disregardedincalculatingwhether the total valuerecommended in § 1 is exceeded.
N ote 5. — T he provisional single-entry value o f 6 % in § 2 has b e e n p r o v i s i o n a l l y t a k e n p e n d i n g t h e results o f studies to determine the most appropriate value, taking into account the increase in the effective number of interferences contributing to the aggregate interferers because of the increasing use of spot beam antennas at space stations. Study of the relationship between the single-entry interference value quoted in § 2 above and the aggregate interference values quoted in § 1 is required as a matter o f urgency.

Rep. 1139

55

N ote 6. — There is an urgent need for study o f the acceptability o f an increase in the m a x im u m total interference noise values recommended in § 1 and more particularly those given in § 1.2 for satellite networks in which frequency re-use is practised.

N ote 7. — In segments o f the geostationary-satellite orbit not likely to be c row d ed, interference allowances less than those recommended in § 1 above, may be utilized, allowing a corresponding increase in other noise c o n trib u tio n s within total acceptable noise limits. However, § 1.1 and 1.2 above should n orm ally be evaluated with the assum ption that the total power noise level present is that which produces the specified bit error ratio under unfaded conditions of the received signal.

N ote 8. — A lthough this R e c o m m e n d a tio n has ------------------------------

u p p e rfrequency limit o f 15 G H z, in the

frequency range from 10 to 15 G H z short term p ro pagation data are not available uniformly th ro u g h o u t the w orld

and there is a continuing need to examine such data to confirm the appropriateness of the interference noise

allowances.

N ote 9. — T here is a need for urgent study to be given to the interference noise allow ances ap p ro p ria te to systems o p era tin g at frequencies abo ve 15 GHz.

N ote 10. — The interference power levels indicated in § 1 and 2 above apply only to the transmission of digital services (see CCIR Recommendation 614 and CCITT Recommendation G.821). Further study by CCIR Study Group 4 is required regarding the performance objectives appropriate to the transmission of digital services other than 64 kb/s digital transmission forming part of a ISDN connection, as information on the performance requirements of such services becomes available to the CCIR.

Mote 11 - The principles of this Recommendation may also be applied to satellite networks providing long-term performance objectives different from those in Recommendation 61t. This is a subject of further'study.

56

Rep. 1134

REPORT 1 1 34

DIGITAL SATELLITE DEDICATED NETWORKS (Study Programme 29C/4)

(1990)

1.

Introduction

Satellite communication systems are characterized by their flexibility to provide various network configurations such as, distribution and multi-access, as well as direct network access using relatively small earth stations. Based on these features, Digital Satellite Dedicated Networks (DSDN) are established to offer a class of digital services to dedicated user points, by , allocating a specific portion of the satellite capacity.

Often the term "business" is associated with the definition of these networks because most applications are for business purpose. Nevertheless Dedicated Networks may be established either in a standardized manner which may or may not be compatible with the public switched network and the ISDN, or in ‘a non standardized manner which is usually not interconnected with the public switched network and is designed to meet specific customer requirements.

2.

User requirements

For networks not interconnected with the public switched network, user requirements may vary widely in speed, message length, traffic variation, delay tolerance, etc. These requirements are restricted by cost and network hardware/software constraints. *

3.

Network topology

Basic network topology is categorized into the following:

point-to-point

point-to-multipoint

multipoint-to-point

multipoint-to-multipoint

The implementation of these topologies can be achieved by various network architectures (e.g. direct link, star network, mesh network).

Rep. 1134

57

4.

Link and network management schemes

Link control schemes closely relate to multiple access schemes. Multiple access schemes are primarily categorized into pre-assignment, demand assignment, random access and so on.

Network management schemes must be determined based on the throughput and delay characteristics of the multiple access schemes and the constraints imposed by the services offered.

5.

User/network interface

For the non standardized networks, user/network interfaces for digital satellite communication systems for establishing dedicated networks, and direct-to-user's circuits, must be determined from the economical and technical standpoint based on the following factors:

adaptability to satellite link characteristics such as delay, quality and availability

compatibility with general use terminals

-

efficient utilization of satellite resources, such as spectrum and

orbit, allocated to the fixed-satellite service.

6.

System performance objectives

There may be different quality performance requirements in dedicated satellite networks depending on the application.

6 .1

Error performance objectives

The types of Digital Satellite Dedicated Networks cover a wide variety of applications which are characterized by different error performance requirements and design objectives often related to the type of services carried in these Networks.

Annex I gives examples of existing systems and illustrates how broad the range of applications can be. This can be observed by comparing the following three categories of networks which can all be defined as DSDN:

Standardized SCPC connections via INTELSAT IBS or EUTELSAT SMS.

Independent TDMA networks, as the Japanese SDCS system or the French TELECOM 1, which offer well structured forms of digital access for business users.
Non-standardized connections on INTELSAT IBS or EUTELSAT SMS, such as VSAT based networks, designed Co meet specific service requirements or users' requests.
Reference error performance objectives are specified in some cases, such as the "open network" applications in the INTELSAT IBS or the EUTELSAT SMS systems (see sections 3.2 and 4.0 of Annex I) which offer two well defined "grades of service", but often the error

58

Rep. 1134

performance objectives are fixed by the end user on a case by case basis. It can be observed from the examples in Annex I that "closed network" applications for unidirectional transmission of press or financial data (e.g. distribution of news photographs, texts, stock exchange data) usually require a much higher error performance quality than the standard quality of the "open network", and that closed network applications for bidirectional services in VSAT based networks (using TDM outbound from hub ter VSATs, and TDMA inbound from VSATs to hub) are often designed to objectives slightly more stringent than those of open network applications.

It is therefore rather difficult to provide forms of unified description or guidelines for the error performance requirements of digital satellite networks in dedicated applications.

There are however two cases of possible utilization of these networks which may be better characterized from the error performance viewpoint if further studies are performed.

These are given in the following subsections.

6.1.1

Interconnection of satellite dedicated networks with the ISDN

Digital Satellite Dedicated Networks can provide interconnection between points external to the ISDN and the point of access to the ISDN. In this case,the satellite path forms part of the subscriber terminations which ard external to the ISDN HRX -according to the definition given in CCITT Recommendation G.821. The quality of service experienced by the end users may be affected by the error performance and availability objectives of these links, thus end user requirements should be given due consideration in link design.

6.1.2

Digital satellite dedicated networks providing ISDN equivalent performance

There could be a requirement for the establishement of Dedicated Networks offering end-to-end performance equivalent to the ISDN, namely the overall G.821 error performance objectives.

Such networks, fully external and independent from the ISDN, could' interconnect terminals designed for the ISDN and provide services of ISDN type to a closed group of users before the switched network is ready.

The overall link between the terminal equipments would be represented by a Hypothetical Connection (See Fig. 1) comprising one satellite portion and15terrestrial tails at both ends.,

In this case the apportionment of the overall objectives to the satellite path would depend on the length and the performance of the terrestrial tails. As an example, Figure 2 presents three ways of making the apportionment.

Figure 2c represents the worst case condition for the satellite portion in an ISDN-based business system. Nevertheless, this apportionment is a relaxation of the objectives set for a satellite link in the public ISDN.

Figure 3 illustrates this relaxation by comparing BEP model b from Report 997 with two BEP models (1 and 2) which are examples based on the apportionment of Figure 2c [CCIR, 1986-90]. Using the method described in Report 997, these BEP models can be translated to the corresponding G.821 parameters see Table I. It is worth noting that, although these calculations are based on a Poisson distribution of errors, the SES (Severely Errored Seconds) calculation results in a significant margin when compared with the G.821 based objective and this margin offers protection against the effects of error bursts.

Rep. 1134

59

i

Table I : Calculated G.821 parameters for models 1 and 2

Model 1 Model 2
possible allowance (G.821)

EFS (%)
97,70 96,78
> 96

DM (%)
4,53 4,88
< 5

SES (%)
0,03 0,005
< 0.06

G.821 error performance

« ------- M — terrestrial
_ ., tail

1------------------------------- > - < ----- ►

, ...... , . . satellite link

terrestrial_ ., tail

TE : Terminal Equipment

FIGURE 1 - Hypothetical connection includine a satellite link which offers ISDN equivalent quality (e.g. end-to-end error performance objectives of Rec. G.821)

60

Rep. 1134

99.2%*

93.6% EFS*

99.2%*

te p - i ” — ,--------------- a --------------- t * * — □ TE

10%** I

80% **

I 10Z**

satellite link

a)

earth station at the subscribers premises

98.8%* j
15%**
b)

99.2%*

94.8% EFS*

99.2%*

EFS ,--------------- » --------------- — d t e

10%**

65%**

10%**

one "remote" subscriber

2°/.*1 99 2%

€FS EFS

1

i z P ----------------- *■

_ isy.** -tov.

1

^

3C,0?. E F S
.
v~
So%

93,1°/.
FFS EFS

4sy*

LE

t

c) two "remote" subsribers

T : reference point TE: terminal equipment LE: local exchange
* Percentage of time ** Percentage of End-to-End (Recommendation G.821) HRX performance
FIGURE 2 - Examples of apportionment of the Rec. G.821 error performance objectives in a point-to-point connection including a satellite business system"

Rep. 1134

61

4tamp117 0,2

Percentage of the Worst Month Total Time FIGURE 3
BEP performance models for DSDN which meet CCITT Recommendation G.821

62

Rep. 1134

6 .2

Availability

A business satellite network is a communications system where one customer's premises is directly connected to another customer's premises with dedicated facilities. These networks are alternatives to using the switched network and are generally constructed, or entered into, for economic reasons i.e. communication via a business satellite network is less expensive than communications via the switched network. Since these systems are being utilized by businesses, the period of heaviest use is during the business day and 24 hour a day availability may not be absolutely necessary.

~From a practical viewpoint, most business satellite networks are being implemented at 14/11 GHz Band rather than 6/4 GHz Band with relatively small and therefore inexpensive antennas. Propagation conditions at these higher frequencies are generally more difficult than they are at 6/4 GHz Band. In the light of these facts, it is logical to expect that the availability of a business satellite network may be less than that of a switched network connection and thus a relaxed availability objective may be appropriate for business satellite networks.

The service is considered available if a given BER is not exceeded. The customer fixes a percentage of time for availability and does not care what happens during the remainder of the time.
The way business customers put requests to service providers is often different from what CCIR definitions and methods would suggest. For instance in certain cases only interruptions exceeding given durations (e.g. 2 hours, or 10 min) are considered unacceptable. These aspects are strictly related to the nature of each service and require further investigation.

It should be noted that some customers lease capacity on business satellite networks in order to overcome service availabilityproblems associated with the public switched network (especially during "busy hour" periods). These customers are not generally prepared to relax their requirements and indeed may require a higher level of service availability than that which applies to links in the public network.

7.

Earth stations

Implementation of economical earth stations is essential to expand the satellite communication applications. Optimization for earth station parameters, frequency bands and applied technologies such as modulation, coding and multiple access protocols must be achieved under the interference constraints to other satellite and terrestrial systems. The likely use of very small aperture terminals in the earth segment with poor side-lobe gain, may require careful analysis of the interference environment and appropriate provisions for coordination purposes.

References CCIR Documents [1986-90]: 4/362 (Federal Republic of Germany)

Rep. 1134

63

AMEX I

Examples of existing systems used in dedicated network applications

1.

The Japanese SDCS system

The Satellite Digital Communication Service (SDCS), one of the new application services designed expressly for high-speed digital transmission, has been brought into commercial service using CS-2 and CS-3, Japan's communications satellites, in early 1985. The SDCS system is referred to as an example of digital satellite communication systems.

The SDCS system configuration is shown in Figure 4 and the major system parameters in Table II. a SDCS satellite channel is connected with digital terrestrial circuits which form access circuits. Data bit rates offered to subscribers range from 64 to 6144 kbit/s. Signals from each subscriber are conveyed to the SLT (Subscriber Line Terminal) by radio-subscriber lines or by metallic or optical fibre cables at speeds of 1544 or 6312 kbit/s. These signals are then multiplexed into 2048 or 8192 kbit/s signals and sent to the TDMA equipment.

Figure 5 shows the SDCS channel assignment scheme called the Multi-Access Closed Network (MAC-Net). The MAC-Net has three principal features:

1)

The satellite channel is allocated to each user group using

pre-assignment mode.

2) The "S" bit in service information channel controls signal transmission, differing from the conventional pre-assignment channel allocation scheme. No satellite channel signals (bursts) are transmitted if the "S" bit is OFF. That is the TDMA equipment transmits subscriber signals only when the "S" bit is ON.

3) No burst collision control for bursts in the pre-assigned channel is provided by the network.

The frame at the user/network interface consists of four components: information channel, signalling channel D, frame alignment signal F and service information channel with four indicators of DNR, UNR, S and SEND. The indicators of DNR, UNR, S and SEND indicate: circuit failures, unassigned satellite circuit, transmit demand and failures at DTE, or DSU to DTE, respectively.
It is easy to communicate in the following three ways using the MAC-Net:
a) point-to-point,
b) multipoint-to-point, and
c) half-duplex, in which transmitting points change alternately.
Regarding the application, there are circuit switched services and multi-point TV conference as well as pre-assignment (PA) services in which users directly control the "S" bit.

64

Rep. 1134

BIBLIOGRAPHY MORIHIRO, Y. [1984] - "Satellite Digital Communication System for New Business Use", JTR, Vol. 26, No. 4, pp. 270-277.
NAKASHIMA, H. et al. [1986] - "Satellite Digital Communication'Service (SDCS) Using CS-2 in Japan" AIAA Proc., 86-0626, pp. 138-143.

TABLE II
Major system parameters

1

Frequency band

30/20 GHz

Multiple access

TDMA

Modem

QPSK coherent demodulation

FEC

R— 1/2 Convolutional encoding

Viterbi decoder

Clock

24.556 MHz

Transmission capacity

320 ch/transponder

No. of stations per transponder

Reference stations - 2 Traffic stations - 50

Bearer rate

64, 192, 384, 768, 1536, 6144 kbit/s

Rep. 1134

65

cs 3

User/Network interface point

t — rS

T

Subscriber radio system

s L

—

n M

TA

.TA DSU— (J

L-J

Optical fiber or Metallic cable

Telephone office
Relay station

DIE

Itelephon

office Subscriber radio system

Relay station

^ r— -i DIF i^-jcsu|-[ [.R4 \

Cptical fiber or Metallic cable

Access circuit (High-speed Satellite circuit

Access circuit (High-speed

digital transmission systen (multi-access, uni-directional) digital transnission system,

bi-directionnal)

bi-directionnal)

FIGURE 4

SDCS circuit configuration

DSU: Digital Subscriber Unit TA: Terminal Adaptor DIE: Digital terminal Equipment SLT: Subscriber Line Terminal

FIGURE 5 Multi-Access Closed Network (MAC-Net) configuration

66

Rep. 1134

2.

The French business system TELECOM 1

A fully digital system using the French TELECOM 1 satellites and specially designed for multi-service corporate communications is in operation in France and Europe.

This system enables business, industrial, governmental, etc. organizations or companies to establish direct multi-purpose data links between their various premises. All types of data link, either bidirectional (symmetrical or unsymmetrical duplex) or unidirectional (with multi-user broadcast capability) are possible, ranging from low or medium bit rate (2.4 kbit/s to 64 kbit/s) to high bit rate (up to 2 Mbit/s) data transmission.

The system is entirely based on TDMA/DA transmission at 25 Mbit/s between a number of traffic stations forming the nodes of a mesh network. Synchronization and dynamic demand assignment (DA) management are carried out by a central station.
The Earth segment comprises of the central station located at Mulhouse (eastern France) and the traffic stations. In 1987, about 60 traffic stations were in operation, mainly in France but also in Danemark, Germany (FRG) , the Netherlands, the United Kingdom and Ireland.

The unattended traffic stations are composed of:

an outdoor radio, sub-system comprising a 3.5 m non-tracking antenna (including the LNA) and shelter-contained telecommunication equipment.

an indoor TDMA terminal comprising the modem, the common logical units and the interface modules.

The main system characteristics are summarized in Table III.

The TDMA terminal interfaces are provided at 2 Mbit/s, through the TDMA interface modules (TIM).

They are three types of TIM:

2 Mbit/s unstructured TIM used to transmit a data stream transparently at 2 Mbit/s.
2 Mbit/s framed or superframed TIM used to transmit voice channels, or data channels, at 64 kbit/s or 32 kbit/s (or combinations thereof) on a reserved basis (with subscriber line signalling transparency) or, on a call-to-call basis.

TIM-X 50: this type of TIM is used to transmit data from 2400, 4800, 9600 bit/s terminals in the TDMA frame.
Furthermore, a TIM/T1 at 1.5 Mbit/s is also available.

For more detailed system information, see CCIR Handbook on Satellite Communications - FSS (Second Edition, Geneva, 1988) § 5.6 .3.3, "A typical example of a medium bit rate terminal: Telecom 1 .

Rep. 1134

67

TABLE III Main characteristics of the French business svstem (TELECOM 11

Frequency bands
Multiple access Frame duration Modem
Bit rate Number of transponders
User data bit rate

down-link: 12.5 - 12.75 GHz up-link: 14 - 14.25 GHz
TDMA
2 0 ms
Phase modulation (2/4 PSK) with differential demodulation and differential encoding
24.576 Mbit/s
6 transponders accessible through frequency hopping at receive side
2.4, 4.8, 9.6 kbit/s 32, 64 kbit/s n x 64 kbit/s 2 Mbit/s 91.5 Mbit/s optional)

68

Rep. 1134

3.

INTELSAT Business Services (IBS) Network

INTELSAT Business Services (IBS) digital carriers utilize a Quadrature Phase-Shift-Keying (QPSK) modulation with Frequency Division Multiple Access (FDMA) technique. The service is designed for communication between INTELSAT Standard A, B, C, E and F earth stations and facilitates the use of national gateway, urban gateway, and customer-premise types of earth stations, but is not intended to be used for public switched telephony.

3.1

Grades of services (Basic and Super)

Two grades of service are offered: Basic and Super. Basic IBS is designed to maximize channel capacity in both 6 /A GHz and 14/11 GHz Band transponders. As an option, Super IBS is offered to provide an availability at 14/11 GHz Band equivalent to 6/4 GHz Band through an increase in uplink e.i.r.p. Super IBS has been designed to meet the requirements of CCIR Recommendation 614 and therefore offer ISDN equivalent quality. AQ comparison of Basic and Super IBS is shown below.

Comparison of Basic and Super IBS

Performance Objective
Service
Unavailability
Minimum Clear Sky BER Threshold BER
System Margin

Ot
00

6/4 GHz Band Uplinks (6/4 & 6/11-12 GHz)
Basic 0.04
1 0 -3 3.0

14/11 GHz Band Uplinks (14/11-12 & 14/4 GHz)

Basic
1.0

Super* % 0.04*

10’8

CO
1
O
(O
1
O

1 0 -3

2.5

7.0

Units % per year
dB

3.2

Categories of svstem usage (closed and open networks)

There are two general categories of system usage for IBS: a closed network and an open network. Regardless of the category of IBS usage, the earth station antenna and RF characteristics are the same.

a) The closed network is intended to provide freedom to the user in selecting the digital system required for his particular needs. The performance characteristics for this type of service do not require specifications related to interconnection with other users and can be defined in terms of ‘RF transmission characteristics. In general, the only mandatory requirements are those needed to ensure that one user's emissions will not interfere with others.

b) An open network requires a certain degree of common terminal features to be defined in order for one user's network to interface with another. Carrier parameters, e.i.r.p.'s and other requirements which are necessary to ensure the compatibility of equipment are thus mandatory.

* Basic IBS provides a BER of 10” 6 for 99% of the time; at 14/11 GHz Band Super IBS provides a BER of 10" 6 for significantly greater than 99% of the ;tlme.

Rep. 1134

69

3.2.1

Closed network characteristics

Example transmission parameters for Rate 3/4 FEC are shown for closed network operation in Table IV. Although these types of FEC are common in the closed network, their use is not mandatory, no coding may be used in some circumstances. A wide range of bit rates is possible in addition to the examples shown.

3.2.2

Open network characteristics

Example Transmission Parameters

Example transmission parameters for open network operation are shown in Table V . A wide range of other bit rates are also possible.

Bit Error Rate Performance

In IF back-to-back mode, with FEC and data scrambling, the channel unit is required to meet the performance requirements given below. The effects of any carrier slips must be included.

BER better than
1 0 "3 io-A 1 0 "6 io-®

Eb/N0 (dB)
4.2 4.7
6.1
7.2

The Eb/N0 is referred to the modulated carrier power. The data rate equals the information rate plus overhead.
Forward Error Correction
Rate 1/2 convolutional encoding with Viterbi decoding is used by all carriers in the IBS Open Network. The constraint length of the coding is seven. The soft decision Viterbi (maximum likelihood) decoder must use adequate quantization to achieve the required BER performance.
Because of the fairly common use of Rate 3/4 FEC in the Closed Network, a particular coding process is recommended by INTELSAT which permits easy switching between Rates 1/2 and 3/4, if desired.
Scrambler
A 15-stage synchronous data scrambler of generator polynomial 1 + X'1A + X ~ 15 is employed to ensure adequate energy dispersal in accordance with Recommendation 358.
Satellite Link Encryption
The use of satellite link encryption is optional, and the encryption method and algorithm are subject to bilateral agreement.

70

Rep. 1134

TABLE

- Example IBS reference transmission parameters for rate 3/4 FEC* (Closed network, 10% overhead)

Information Rate
(bit/s)

Transmission Rate
(bit/s)

Occupied Bandwidth
Unit (Hertz)

Allocated Bandwidth
Unit (Hertz)

C/T

C/N0

(dB(W/K)) (dB(W/Hz))

1 0 "8

10"8

64 k

94 k

56 k

67.5 k -171.0

57.6

1.544 M

2.3 M

1.38 M

1.643 M -157.1

71.5

2.048 M

3.0 M

1.80 M

2.138 M -155.9

72.7

8.448 M

12.4 M

7.44 M

8.708 M -149.8

78.8

* Depending upon the actual transponder and link conditions, INTELSAT may establish the clear sky setting of the link at a C/N better than or equal to 10.1 dB in order to ensure adequate margins. The C/T and C/N0 values for 10" 3 and 10" 6 are 3.5 dB and 1.1 dB less than those shown for 1 0 " 8 respectively.
Reference Unit = An integer multiple of the smallest carrier size (94 kbit/s x n) where n = 1 to 132. Example reference units are shown in this Table. In the case of 1.544 Mbit/s and 6.312 Mbit/s, which are not integer multiples, allocated bandwidth based on n = 25 and 99 respectively will be assigned.

TABLE V - IBS open network tr an smission parameters* (Open network, 1/15 (about 6.7%) overhead)

Information Rate
(kbit/s)

Data Rate IncIud ing Overhead (kb it/s )

Transmission Rate
(kbit/s)

Occupi ed Bandwi dth
Unit (Hertz)

A IIocated Bandwidth
Unit (Hertz)

No. of 22.5 kHz Slots for Allocated Bandwidth

C/T (dB(W/K))
j 10'8

C/N0 (dB(W/Hz))
10'8

64
1544

68.3 1 1638.4

137
32 77

CO CO

1 .97 M

112.5k 2 .3 18 M

5

j -172.5

103

j -158.7

56.1 69.9

2048
______

2184.5

4369

1

I

1

„1

2.62 M

3.082 M

I

I1 • 137

J

. 1

1 _

j -157.5

.1

__

J

71 .2

D e p e n d i n g upon the ac tu al t r a n s p o n d e r and link c o n d i t i o n s , I N T E L S A T ma ye s t a b l i s h the cl ea r skys e t t i n g of the link at C/N b e t t e r t h a n or e q u a l to 6 . 8 dB in o r d e r to e n s u r e a d e q u a t e m a r g i n s . The C/T and C/Novalues for 1 0 ~ 3 and 1 0 ” 6 are 3.0 dB and 1.1 dB less than those shown for 10'®.

NOTES :

1.

T h e a s s u m e d d a t a r a t e ( i n c l u d i n g o v e r h e a d ) E ^ / N q is 7 . 6 dB f o r a BE R of 10 - ft .

2.

T r a n s m i s s i o n R a te = ( I n f o r m a t i o n R a te pl us 1/15 o v e r h e a d ) x 2.

3. T h e b a n d w i d t h a l l o c a t e d t o th e c a r r i e r in th e s a t e l l i t e t r a n s p o n d e r is a mu It ip i e of 2 2 .5 kHz.

72

Rep. 1134

4.

EUTELSAT Satellite Multiservice Svstem (SMS)

The EUTELSAT Single Channel Per Carrier (SCPC) Satellite Multiservice System (SMS) offer fully digital service for business and other applications using capacity on the EUTELSAT I satellites. The "open network" configuration offers standardized forms of access and error performance levels.

The customer bit rates can range from 2.4 kbit/s to 2048 kbit/s, while carrier information rates are of 64 kbit/s and multiples thereof up to 2048 kbit/s.

Two grades of service are offered: standard grade with a BER lower than 10" 6 for 99% of the time and high grade giving Recommendation 614 quality.

Three types of earth stations are standardized with antenna diameters of 5.0 to 5.4 m, 3.7 and 2.4 m respectively. The main parameters and characteristics of the EUTELSAT SMS earth stations are provided in Table 5.XXX of Appendix 5-1, Chapter 5 of the CCIR Handbook on Satellite Communications FSS (Geneva 1988, Second Edition).

The SMS transponder can also be used for other network architectures tailored to customer requirements referred to as "closed networks". Table VI gives a list of existing SMS closed network applications, with the relevant BER and percentages of time used as design criteria. Figure 6 gives the representative points of all the various performance objectives on logarithmic coordinates of BER versus percentage of time. All percentages of the time have been expressed in terms of the worst month, by using the conversion formula from annual to worst month statistic given in Report 564.
For illustrative purpose the representative points are grouped into four classes which correspond to different service requirements. In particular class A comprises closed network applications for data distribution (news photographs, texts, stock exchange data etc.) and for computer to computer interconnection, while class B comprises closed network applications for b i ­ directional services in VSAT based networks, e.g. using TDM outbound •(from hub to VSATs) and TDMA inbound (from VSATs to hub), as well as data collection and some data distribution applications.

Rep. 1134

73

TABLE VI

- Design objectives for non standard applications (closed networks') in the EUTELSAT SMS system. BER not to be exceeded for more than a given percentage of time

Type of network

Bit rate

(kbit/s)

BER

Distribution, star point to multipoint, uni­ directional (financial data)

128 BPSK

io-9

Unidirectional, star (distribution of news text + photo)

19.2 BPSK

io-7

Unidirection, point to multipoint (stock exchange data distribution)

64 BPSK

io-7

Unidirectional, point to multipoint (stock ex­ change data distribution)

19.2 BPSK

io-7

Point-to-point b i ­ directional (computer inter­ connection)

64 QPSK

. io:6

Bidirectional, star
(Registration and delivery of documents)

9.6 BPSK
64 QPSK

io-6 io-6

Interactive TDM (outbound) TDMA (inbound)

512 BPSK, TDM
64 BPSK, TDMA

io-6

Bidirectional star (computer to terminals)

64 QPSK

1 0 "6

Unidirectional star point to multipoint (distri­ bution of documents)

64 QPSK

io-6

Interactive partly meshed (civil air traffic control)

9.6 BPSK

1 0 "6

% of time y = year wm = worst month
90 y 99.9 y
99.9 y
99.9 y 99.0 wm
99 y 99.3 y 99 y 99 y 99 y 99.5 y

74

Rep. 1134

TABLE VI (cont'd)

Type of network

Bit rate

(kbit/s)

BER

Interactive TDM/TDMA star - collection of
environmental data - emergency
communications

2048 QPSK-SCPC 512 TDM-64 TDMA
BPSK

Interactive star TDM/ TDMA (Terminal to central computer connection)

256 TDM-BPSK
56 TDMA-BPSK

Fully meshed TDMA (voice, data, video­ conference)

1544 QPSK

Interactive star TDM/ TDMA (on trial)

512 BPSK-TDM
128 BPSK-TDMA

Unidirectional point to

19.2

multipoint spread spec­ (2.4576 Mchip/s)

trum (Data distribution) BPSK SP. SP.

Unidirectional Point to multipoint -(distribution of news) -(stock exch. data,
reduced quality)

19.2 (2.4576 Mchip/s)
BPSK SP. SP.

Unidirectional point to multipoint (experimental)

19.2 (2.4576 Mchip/s)
BPSK SP. SP.

Interactive star TDM/ TDMA

256 BPSK-TDM
64 BPSK-TDMA

Unidirectional point to multipoint broadcast star (radio sound distribution)

1920 QPSK

io-6
1 0 "6 1 0 "6
1 0 "7
io-7 1 0 "7 io-7
1 0 "7 io-7 io-7 io-6

% of time y = year wm = worst month
99.5 y
99 y 99.5 y
99 y
99 y 99.9 y 99 y
99 y 99 y 99.9 y 99 y

Rep. 1134

75

Bit error probability

Percentage of the worst month
FIGURE 6 - Error performance objectives for digital satellite dedicated networks in the EUTELSAT SMS svstem. The points given in the figure are used as design objectives for different applications
A: Very high quality closed network B: High quality closed network C: High grade open network D: Standard grade open network

76

Rep. 451-3

R E P O R T 451-3

FACTORS AFFECTING THE SYSTEM DESIGN AND THE SELECTION OF FREQUENCIES FOR INTER-SATELLITE LINKS OF THE FIXED-SATELLITE SERVICE

(Question 3 1 / 4 )

(1970-1974-1978-1982)

1.

General

T he use o f radiocom m unication links between space stations in the fixed-satellite service is a means of interconnecting space networks. It is an alternative to employing multiple earth station antenna systems or multiple-hop circuits. Several experiments using this technique have been successfully conducted.
In recognition o f the potential usefulness o f inter-satellite links, the 1971 W A R C -S T defined the inter-satel­ lite service. The WARC-79 allocated to this service frequencies above 22 GHz.
This Report considers, in broad terms, some of the concepts involved in the development of inter-satellite links and the technical requirements of inter-satellite link operation.
Although the use of laser beams between spacecraft may be possible, their extremely narrow beamwidths and other limitations tend thus far to favour the use of millimetre wave or other radio links, and only the frequency range o f 3 G H z to 300 G H z is considered in this Report.

2.

Advantages and disadvantages o f inter-satellite links

There are at least four broad benefits provided by inter-satellite connections, viz:
— a reduction in the number of earth stations a n d /o r associated antennas needed; — better circuit utilization of available capacity on paths between Earth and space may be obtained; — the provision of extensive (global) connectivity for earth stations accessing satellites through spot beams; — increased flexibility of network arrangements.
In order to illustrate these points consider the extreme hypothetical example of a system containing, for every earth station, a satellite fully dedicated to connecting with only one earth station (a “tethered” satellite). Circuit connections between earth stations in the system would be made through inter-satellite communication links in the orbital arc. In effect each satellite would be a very long extension o f its associated earth station.
In this extreme example of tethered satellites, any inefficiencies in actual capacity utilization of available power and bandwidth due to:
— circuit grouping, — connection constraints, — switching, — multiplex, etc.,

Rep. 451-3

77

are removed from the earth-space links to the geostationary arc. Within the geostationary arc and removed from the terrestrial links, the greatly increased available spectrum at submillimetre or optical wavelengths would be more tolerant of inefficiencies in utilization generated by real system circuit connection constraints.
Furthermore, since there would be only one earth station for each satellite, the satellite antenna coverage requirement would be minimized. Conceptually, this allows using a very narrow beam earth oriented satellite antenna providing good off-axis rejection of received up-link interference signals from other earth stations, and minimizes generation o f down-link interference to other earth stations. With a reduced level o f interference flux density generated by low antenna sidelobes, closer inter-satellite spacings could be allowed. Also, with only one bothway connection to an earth station, additional capacity increases beyond that provided by narrow antenna beams would be obtained from the single carrier per transponder mode o f operation (i.e., reduced intermodulation and little or no transponder back-off).
Such a system conceptually would allow maximization of the space to earth spectrum efficiency. With the greater spectrum availability at the higher frequencies for the inter-satellite links, it is reasonable to expect that orbital capacity would still be dominated by the capacity achieved on the earth space links even in their most advantageous configuration. Consequently, the idealized model described, would appear to offer an upper limit to achievable geostationary orbital capacity. Less advanced systems using inter-satellite links to connect satellites, each of which serves several earth stations, would give the benefits listed above to a lesser degree.
However, C C IT T Recommendation G.114 specifies a limit o f 400 ms as an acceptable telephone channel signal propagation time. The signal delay in a link using a single geostationary satellite and including an allowance for the delay in the terrestrial end connections is generally abo u t 290 ms. Hence the permissible signal delay in a telephone channel on an intersatellite link should not exceed a figure o f the o rd er o f 110 ms. The separation angle between the two geostationary satellites is defined by the formula:

0 ^ 2 arc s i n -----t-'-C------2 (R e+ H)

where t:

permissible signal delay in a telephone channel on the satellite-to-satellite section,

C : speed of radiowave propagation,

R e : E a rth ’s radius,

H : height o f satellite above E a rth ’s surface.

To meet C CITT Recommendation G.114, the separation angle between two geostationary satellites in an inter-satellite link for telephony should not exceed a figure of the order of 50°.

In the near term, with the object of economically augmenting existing capacity with minimum impact on existing systems, consideration must be given to more limited concepts o f inter-satellite link facilities. Typically, such links might provide only limited inter-satellite circuit capacity. Then near-term systems would probably have the following characteristics:

— relatively short inter-satellite spacing, — limited capacity, — few systems in service, — minimum impact on existing spacecraft technology: structure, pointing/orientation, and RF components.

3.

Inter-satellite link design considerations

The elements which must be taken into account in the design of an inter-satellite link are the following:
— geocentric satellite spacing tp (degrees): — frequency f (MHz); — antenna diameter D (metres) of equivalent circular aperture of about 55% efficiency; — receiving system noise temperature T (K); — available RF power p (watts); — required bandwidth b (Hz).

78

Rep. 451-3

Based on these parameters, the performance o f an inter-satellite link may be expressed in terms o f its predetection carrier/noise ratio which can be approximated by the equation:

C /N = 3.72 pD7 * f 2 - IQ" 18

(1)

q? kT b

where k = 1.38 x 10-23 J / K is B o ltzm an n ’s constant. This equation holds for small orbital separations. For <p > 1 0 °, the term <p2 should be replaced by the term 1.31 x 10 4 x sin2 (tp/2 ).
In equation (1), the receiving system noise tem perature is a function o f frequency. Noise temperatures of about 1000 K are typical for frequencies around 6 G H z, an d the following frequency dependence is postulated:

T { f) = 31.6 ( / / 6 ) ,/2

K

(2 )

In a real system design the param eters C / N a n d tp are usually system co nstraints; the former determined by the required system performance, the latter by inter-satellite interference considerations. It is of interest to establish the relationship between power requirements and antenna size in inter-satellite links.
Figure 1 shows relative available power per unit bandwidth ( p / b ) as a function o f frequency with antenna d iam eter as a param eter, in dB against an arbitrary reference and for fixed values o f <p an d C /N . Also shown in Fig. 1 are the beamwidths o f inter-satellite link antennas for the various com binations o f D and f Increasing the antenna diameter on inter-satellite links reduces power requirements in an inverse 4th power relationship. As antenna size increases, beamwidth decreases and mutual pointing requirements between the two satellites become increasingly more stringent.

Bands allocated to the inter-satelllte service

00 = 1°

Frequency (GHz)
FIG URE 1 — Relative power per unit bandwidth required for an inter-satellite link as a function o f frequency fo r various antenna diam eters: no pointing or tracking losses
D: A ntenna Diameter 8q: Half-Power Beamwidth

Rep. 451-3

79

For example, with 20 watts o f available power, at a satellite spacing o f 3 degrees, a 100 M H z wide inter-satellite link at 54 G H z with a postulated carrier/noise ratio of 30 dB would require antennas of 0.68 metre diam eter. The be am w id th o f such an te n n a s is about 0.56°, an d pointing accuracies o f a b o u t 1 /10 o f that, or ab o u t 0.06°, would be required to utilize the full antenna gain. The carrier/noise ratio o f 30 dB has been assumed on the basis that the inter-satellite link noise should not significantly affect total performance.
Hence, the use o f reasonably sized antennas to minimize power at 54 G H z or beyond is likely to require mutual tracking between satellites with the present state of the art in spacecraft attitude stabilization and orbital elem ent matching. T racking and beam steering technology is well developed u p to optical frequencies.
Nevertheless, to minimize spacecraft complexity and improve reliability it is attractive to consider, e.g., for INTELSAT applications, inter-satellite links which can be maintained without mutual tracking.

4.

Non-tracking inter-satellite links

When considering inter-satellite links the antennas of which do not track each other, two additional system constraints need to be taken into consideration:
— allowable signal level variation on the inter-satellite link, A C(dB); — antenna pointing error due to relative attitude tolerances of the spacecraft, 5 (degrees).
Typical in-plane and plane-normal pointing error budgets for nominally geostationary spacecraft are derived in Annex I. W ithin the state o f the art net error angles between 0.5° a n d 1° can be realized.
The main lobe gain degradation due to pointing offset, g(Q )/go, for a simple feed antenna may be approximated by:
s (0)/^o 3510_ 12 (0/0o)

where 0 is the angle o ff boresight (in degrees) and 0O the half-power beam w idth (in degrees) which, in turn, is related to the aperture diameter D by:

0O=

• 104

degrees

(4 )

where,
/ = MHz D = metres.
Considering that the pointing uncertainty involves a boresight error of 0 = 8 and that two antennas are involved, each o f which may go through gain variations between g 0 and g(0), one may combine equations (3) abd (4) to obtain:

20 log ( g /g o ) = A C * 5.24 ( 5 / D )2 10" 8

dB

(5)

Combining equations (1) and (2), solving the new expression for m inim um required power spectral density p /b , and inserting equation (5) as an additional attenuation due to pointing errors, one obtains, in dB notation:

10 log ( p / b ) = P0 = C / N + 20 log <p — 15 l o g / — 40 log D +

+ 5.24 (8f D ) 2 10- 8 - 43.2

dB (W /H z)

(6 )

Plots o f P0 versus / for 5 = 0.5° an d 8 = 1° a n d an ten n a diam eters o f 0.25, 0.5,. 1 a n d 2 metres (Fig. 2) show that minimum power requirements may be realized at frequencies below 54 GHz, the lowest frequency presently allocated to the inter-satellite service. Figure 2 uses an arbitrary reference for P0 a n d considers C / N an d (p to be system constants. How close to the optim um (i.e., minim um P0 ) frequency an inter-satellite link can be

80

Rep. 451-3

o perated depends on the signal level tolerance A C one is willing to accept. Limits for A C = 1 dB and 2 dB are sh ow n in Fig. 2. T h e effective o p tim u m frequency is that at which the P0 an d A C curves intersect. Absolute m inim um power requirements are realized for A C of about 3.4 dB; hence, there is no advantage to be gained by allowing a greater level variation than about 3 dB.

6 = 1°

Frequency (GHz)
FIGURE 2 — Relativepower requirements in a non-tracking inter-satellite link as afunction offrequency with antenna diameter (D), net pointing error (8) and allowable level variation (AC) as parameters
6 = 1°
8 = 0.5°

These considerations indicate that for antenna sizes greater than 0.25 metres the optimum frequencies for non-tracking inter-satellite links between closely spaced satellites may lie below 54 G Hz, the larger antenna sizes favouring the lower frequencies.

5.

Frequency re-use among inter-satellite links with tracking antennas

Annex II to this Report presents the results and conclusions of an analysis, based on certain assumptions, o f the degree to which inter-satellite links between geostationary satellites can re-use the same frequencies.

6.

Choice of frequencies for inter-satellite links

Two distinct categories of use can be envisaged for inter-satellite links operating between geostationary communication satellites:
(a) Links between satellites separated quite widelyin the geostationary satellite orbit (e.g. 60°) to extend the geographical coverage of a system without relaying at an earth station.
(b) Links between satellites relatively close together (e.g. 3° to 5° orbit spacing) a n d having virtually the same coverage area, but serving different communities of earth stations.
The optimum technical parameters for these two types of link, which would normally be quite different, are considered below.

Rep. 451-3

81

6.1

Long inter-satellite links

. In category (a) links, a high path loss will exist between the satellites, co m p a ra b le with the path loss from a geostationary satellite to the Earth, and to keep the link transmitter power to a practicable value it would be necessary to use high gain transm itting a n d receiving antennas. Fora m ax im u m a n te n n a d iam ete r of, say, 1.2 m governed by satellite launch vehicle shroud dimensions, this high gain would call for operation at as high a frequency as possible, typically in excess of 20 GHz, and preferably in the inter-satellite band already allocated at 54.25 to 58.2 GHz. If station-keeping errors and orbital inclination are kept to small, bu t currently achievable values, variations in the direction from one satellite to the other will be quite small at the wide separation distances involved in these systems, allowing the use of narrow beam antennas even in the absence of antenna tracking facilities.

6.2 Short inter-satellite links
For the short-hop type o f link described in category (b) the relative proximity of the satellites results in large angular variations in the link path for quite small values of orbital inclination. For example, the maximum angular variation in a link between satellites spaced 4° apart and each having orbit inclination tolerances o f ± 0 . 1° will be:
0.2 2 arc tan — - 5.7
4

This assumes a “ worst-case” situation where the latitudinal excursions of the two satellites are in complete antiphase. This condition would normally be avoided by suitable choice of orbital parameters and, as shown in Annex I, it should be possible to reduce the an gu lar variation to ± 1°. If a n te n n a track ing were available, high gain antennas could be used for short-hop inter-satellite links. However, antenna tracking would involve rather sophisticated satellite devices, probably not currently achievable in a reliable, low-mass form. It is therefore necessary to use an ten n a s with beamw idths equal, in the example quoted above, to 2 .0 ° plus an allowance o f perhaps 0.15° for satellite attitude errors.
These results a n d Fig. 2, indicate that non-tracking antennas one metre in diam eter w ould be best served by frequencies o f the o rd er o f 10 G H z. If the an gu lar variation could be reduced to a fraction o f o ne degree by an antenna tracking system, the optimum would be raised to somewhere in the region of 20 to 30 GHz. The tracking problem could be relieved to some degree by the use of smaller antennas, but the transmitter power requirement would be substantially increased.
Because o f interference considerations it would not be possible to use fixed-satellite space-to-Earth or Earth-to-space bands for space-to-space links. However since link transmission paths will be directed well away from the Earth, there may be no problem in sharing with terrestrial services, and this point is examined in Report 387.
6.3 Device technology
Today the 11/14 and 2 0 /3 0 G H z technology is well-developed for space stations. It would be very desirable for this space-proved hardware or similar equipment to be made usable for inter-satellite links. Frequencies above 50 G H z may be too high for early implementation.
6.4 Interference and co-ordination problems
The following substantial problems have been identified:
— The inter-satellite service will need full duplex capability. Pairs o f bands o f equal width separated by a gap in frequency, will therefore be necessary.
— The inter-satellite ban d s should be separated in frequency from the up and the d ow n link bands. It is advisable to have gaps between any two of these bands to minimize the interference problems. An inter-satellite frequency is used on one satellite for transmission a n d on a n o th e r for reception. With up or do w n path frequencies very close to an inter-satellite frequency it would be possible to get very large level differences between a transmitter and a receiver adjacent in frequency. This leads to very severe isolation requirements on b oard the satellite, so severe, in fact, that it would be necessary to leave part o f the bands unused to create a gap between the transmitting and the receiving frequencies.
— In general the inter-satellite frequency bands should not be shared with other bands used by satellites. However, they can be shared with bands used for some terrestrial services. See Report 791.

62

Rep. 451-3

— The beams of all inter-satellite links would be aimed in the plane of the geostationarysatellite orbit and the angular separation between the beams of inter-satellite links belonging to differentsystems may be nogreater th a n 'th e angular spacing between adjacent satellites. In order to achieve frequency re-use of the spectrum allocated for inter-satellite links it will be necessary to design for very na rro w beamwidths, low side-Iobe levels and highly accurate a n ten n a pointing capability inrespect o f inter-satellite link applications. It is also clear that there is a need to devise techniques and criteria for the co-ordinationo f frequencies assigned to inter-satellite links.

7.

Conclusion

From the preceding discussion, it is reasonable to expect increasing interest in inter-satellite links. Some of the major technical issues identified as requiring further study are:
— design and specification of inter-satellite link transmission parameters, technology and design; — inter-orbital trunking arrangements and switching concepts/technology; — impact on spacecraft mechanical design requirements.
A possible demand can be foreseen for “short-hop” inter-satellite links to provide interconnection facilities between communication-satellite networks, and the optim um frequency o f operation is found to be substantially below the lowest frequency allocated for the inter-satellite service, and preferably betw een 15 and 35 GHz.

B IB L IO G R A P H Y
K U R A K O V , P. S. [1980] V ybor p aram etro v m ezhsputnik ovoi linii svyazi (C hoice o f in ter-satellite link p aram eters) Elektrosviaz, 2, 11-14.

ANNEX I
GEOM ETRY AND POINTING ERRORS FOR N O N-TRACKING A N T E N N A S IN AN IN TER -SA TELLITE LIN K
The inter-satellite link antennas on each space station are assumed to be fixed to a de-spun platform and are initially aligned so that each space station is in the centre of the other space station antenna's beam pattern under nominal synchronous orbit conditions. The pointing errors caused by deviations from nominal conditions are separated into their in-plane and out-of-plane com ponents in Tables I and II. The total error is the root o f the sum o f squares (RSS) of the two component classes and affects each satellite independently.

TABLE I — In-plane pointing errors
Error source
I. Long-term variation Antenna thermal distortion In-track position error Cross-track position e rro r: Nominal longitude separation: 2° 3° 4° 5°
2. Bias 3. Short-term variation
Despun platform pointing

Error magnitude
-0.0 2 ° ± 0 .0 5 °
± 0 .4 7 ° ± 0.31 ± 0 .2 3 ° * ± 0.19° ± 0.20°
± 0 .2 5 °

Rep. 451-3

83

TA BLE II — O u t-o f-p la n e p o in tin g errors

Error source

Error m agnitude

1. L o ng-term variation Antenna thermal distortion A ttitude uncertainty A ttitude precession with respect to orbit norm al Out-of-plane position e rro r: Nominal longitude separation : 2° 3° 4° 5°
2. Bias M echanical alignment
3. Short-term variation Spin wobble and nutation

± 0 .0 4 ° ± 0.01° < 0 .1 0 °
± 0 .3 1 ° ± 0 .2 1 ° ± 0 .1 5 ° ± 0 .1 2 °
± 0 .2 0 °
± 0 .0 6 °

The angular pointing errors contributed by orbital motion are a function of the nominal longitude separation between satellites. The in-plane position error comprises an in-track error, caused by departures from nominal longitude separation, and a cross-track error caused by one or both satellites not being at their nominal altitude. The out-of-plane error arises from motion not taking place in the same plane. These errors are assumed to exist due to tolerances allowed in the orbital elements and not from statistical uncertainty in the individual satellite positions.
The in-track position error makes a relatively small contribution to the an g u la r pointing error since it is equal to half the allowable error in satellite separation. A tolerance of ± 0 .1 ° in the satellite separation can be maintained at a small additional expenditure of propellant beyond that required for normal east-west stationkeeping, if great care is taken. The tolerance is also large enough to allow flexible scheduling o f individual manoeuvres.
The cross-track error arises because neither satellite can be maintained in a perfectly circular orbit without daily velocity corrections. Thus the altitude o f each satellite will undergo a 24-hour variation ab o u t a mean value. The m ean altitude o f each satellite will be nearly the same as a consequence o f the east-west m otion being nearly in phase, but the individual variations will, in general, have different amplitudes and phase.
The capability exists to control the amplitudes of the individual altitude variations, or alternatively, to control their phase, such that the cross-track position error never exceeds several kilometres. To do so, however, would require additional velocity corrections in the presence of in-plane velocity coupling from north-south manoeuvres, and would require close co-ordination of the normal east-west manoeuvres performed on the two satellites. W ithout introducing these additional complexities, it is estimated that the position error can be maintained within 12 km.
The out-of-plane pointing error arises if the two satellites do not move in exactly the same plane. It is estimated that the position error can be maintained within 8 km, which corresponds to an angle of approximately 0.01° between the individual orbit planes. This error reflects the expected uncertainty in the velocity corrections obtained during the individual north-south station-keeping manoeuvres performed on each satellite.
Each north-south station-keeping manoeuvre will alter the inertial orientation o f each orbit plane by 0.2°. In order to keep the relative orientation nearly the same, the individual manoeuvres must be closely co-ordinated. For a longitude sep aratio n o f 3° between satellites, the manoeuvres must be p e rform ed within 12 minutes o f each other (equal to 4 minutes for each degree of longitude separation).

84

Rep. 451-3

Pointing error
The various elements in the pointing error budget have been reduced to a single effective value by the following rule: — All in-plane, long-term contributions are summed algebraically to give a single value. Likewise all cross-plane,
long-term contributions. — The in-plane and cross-plane long-term errors are added on an RSS basis to give a single long-term error
value. Since the two contributors are equal an d assumed to be uncorrelated, the long-term RSS error is circular. — The in-plane, short-term contributions are summed algebraically. Likewise the cross-plane, short-term contri­ butions. — The in-plane and cross-plane contributions and the RSS long-term error are summed on an RSS basis to give the total pointing error. — Since each term of the total sum is individually a low probability, nominally 3a, value, the total pointing error is likewise a low probability, nominally 3a, value.
W hen the above rule is applied to the error budgets given above, the total error at various satellite-to-satellite spacings is found to be as given in Table III. These are the pointing errors that have been used to estimate transmission performance of the link.
T A B L E III — P ointing error versus satellite spacing

Spacing

Effective pointing error

2°

1.01°

3°

0.833°

4°

0.739°

5°

0.693°

The total effective error is assumed to be circular, although this is not precisely true. The computational convenience of making the assumption outweighs the small error involved.

A N N E X II
FREQUEN CY RE-USE AM ONG GEOSTATIONARY IN TER-SATELLITE LINKS USING TRA CK IN G ANTENNAS

1.

Introduction

The major eventual development of the inter-satellite service will require the use of mutually tracking communications antennas on board different linked space stations.
Stipulating widespread use of inter-satellite links between geostationary satellites, the questions arise as to how m a n y such links could share the same frequencies and what factors affect the frequency re-use potential in the inter-satellite service.

2.

Link definition

A geostationary inter-satellite link is characterized by two geostationary space stations at geocentric angular spacing 0 , with each space station having a high gain transmit and receive antenna (which may be the same antenna) point their main beams at the other space station. Pointing is m aintained through tracking with negligible tracking error. East-west and west-east transmit frequency bands are assumed to be different but sufficiently close to each other so that antenna gains and path losses may be assumed to be the same for both, ( |/ i — ./; | < / ) ■ T he perform ance requirem ent { C / N ) in both directions is assumed to be the same, as is the required protection ratio (wanted-to-unwanted carrier ratio (C //)) against interference from emissions of other inter-satellite links.

Rep. 451-3

85

The antenna patterns for both transmit and receive antennas are assumed to be identical and are characterized by the pattern equation:

g i f ) = 6750 (p0 1 + 1 —

dPo

whichever is

the greater

(7)

g&) = o.i

where 2 <p0 is the half-pow er b eam w idth o f the transmit an d receive an te n n a a n d q is the angle o ff the main beam

axis

(discrimination angle).

3.

Interference between identical short inter-satellite links

We consider first the simple case of identical (homogeneous) short inter-satellite links.Two suchlinks can be arranged in four different ways as shown in Fig. 3 which alsoidentifies the variouspossibleinterference paths.

3.1

Counter-directional frequency assignm ents

When considering two inter-satellite links, their frequencies may be assigned to be pair-wise counter-direc­ tional (Figs. 3 a) and b)). For such configurations it can be shown by detailed analysis that their spacing in terms o f the geocentric angle 0 ' between the two “left” satellites of either link may be quite small; for appropriate assum ptio ns regarding orbit eccentricity, a n d isolation o f the order o f 30 dB, it is generally less th an 1° o f arc. It can also be shown that, under the assumptions, a space station may originate an eastward and a westward link on the same frequencies.

Ai
a) Frequencies counter-directional; links separated in the orbit.

c) Frequencies co-directional; links separated.

b) Frequencies counter-directional; links interleaved.

B 2
d) Frequencies co-directional; links interleaved.

FIGURE 3 — Interference geometries between inter-satellite links
(Angles marked are antenna discrimination angles. Interference paths are shown in broken lines.)

When arranging more thatf two inter-satellite links with counter-directional frequency assignments along the orbit, every other link will, necessarily, have co-directional frequency assignments.

86

3.2

Co-directional frequency assignments

Rep. 451-3

These correspo nd to the arran g em e n t illustrated in Figs. 3 c) an d d). W hen more than two links are so arran g e d it can be show n that the link separation angle 0 ' is a function o f the link “ length” 0 , o f the inter-satellite link a n te n n a beam w idth 2tp0, o f the necessary isolation C /7 , and o f the m ax im u m radial deviation A h o f the space stations from the normal orbit altitude (a function of orbit eccentricity and the maximum east-west drift rate betw een station-keeping manoeuvres). Representative currently achievable values for A h vary between 15 an d 100 km.
Detailed analysis shows that there is minim um link “length” 0 which allows such links to be “interleaved” : see Fig. 3d). Further, the greater link length 0, the m ore identical inter-satellite links may originate within any given link’s occupied arc. Figure 4 shows the interleaving “c u t- o f f ’ link length, an d the num ber of interleaved links n which may originate within a link arc.

Inter-satellite link “ length” , 6 (degrees)
FIG U RE 4 — N um ber o f interleaved links n p er link “length” 6 as a fu n c tio n o f link ‘‘length’ f o r various com binations o f (p0, A h and C /I
A : tp0 = 0.25°; A h = 40 km B: <p0 = 0.25°; A h = 15 km C : <p0 = 0.25°; A h = 100 km D: <p0 = 0.125° \ A h = 40 km E: <p0 = 0.5°; A h = 40 km A ': as A, but with C / I = 25 dB
Non-interleaved co-directional links can be established end-to-end but require some non-zero spacing between adjacent satellites except for very small link lengths for which, with decreasing link length, this spacing must be increased.

Rep. 451-3

87

4.

Orbit utilization by short identical links

Figure 5 shows the relationship between the number of links which can be accommodated along the entire geostationary orbit (right-hand ordinate) and link length, for various parameter assumptions. The left-hand ordinate shows relative frequency re-use density (orbit utilization) against an arbitrary reference. The curve sections to the left of the major vertical “steps” reflect non-interleaved links, those to the right increasingly greater link interleaving. The actual curves would be in the form o f steps indicating the link-by-link addition, as shown for one example; for all other cases only the step envelopes are shown.

Inter-satellite link “ length” 9 (degrees)

FIGURE 5 — M axim um relative frequency re-use density (orbit utilization) f o r identical inter-satellite links as a fu n ctio n o f link "length ” 0, f o r various com binations o f (p0 and A/i

A: <p0 = 0.25°; B: <p0 = 0.25°; C : <p0 = 0.25°; D: <p0 = 0.125° E: <p0 = 0.5°;

A/i = 40 km A/i = 15 km A/i = 100 km A/i = 40 km A/i = 40 km