FINAL REPORT

Document Sample
FINAL REPORT Powered By Docstoc
					                                                   1
Contracts number:
ESTEC - 13694/99/NL/US   Final Report
ESA S-UMTS study         Date           Revision
Final Report             2000-11-29     A




                    FINAL REPORT
                       S-UMTS
   Preparation of Next Generation Universal
  Mobile Satellite Telecommunications Systems
                                                                                                                                                                          2
Contracts number:
ESTEC - 13694/99/NL/US                                                          Final Report
ESA S-UMTS study                                                                Date                                   Revision
Final Report                                                                    2000-11-29                             A


1     INTRODUCTION .............................................................................................................................................. 4
    1.1      SCOPE AND GOAL .......................................................................................................................................... 4
    1.2      BACKGROUND ............................................................................................................................................... 4
    1.3      REFERENCES.................................................................................................................................................. 4
    1.4      LIST OF ABBREVIATIONS ................................................................................................................................ 4
2     CRITICAL REVIEW OF S-PCN ROLL-OUT .................................................................................................... 7
    2.1      INTRODUCTION .............................................................................................................................................. 7
    2.2      S_PCN ROLL-OUT: LESSONS AND CONCLUSIONS .......................................................................................... 7
    2.3      CONCLUSION ................................................................................................................................................. 9
3     THE S-UMTS/IMT-2000 ROLE ........................................................................................................................ 10
    3.1      S-UMTS VERSUS T-UMTS ......................................................................................................................... 10
    3.2      USER AND OPERATOR REQUIREMENTS ......................................................................................................... 12
    3.3      BUSINESS, MARKET AND FINANCIAL CONSTRAINTS ..................................................................................... 12
    3.4      TECHNICAL ISSUES ...................................................................................................................................... 14
    3.5      RELATIVE CAPACITY EVALUATIONS ........................................................................................................... 16
    3.6      REGULATORY ISSUES................................................................................................................................... 18
4     PRELIMINARY SYSTEM SCENARIOS ......................................................................................................... 20
    4.1      PROPOSED SYSTEM AND ARCHITECTURE ..................................................................................................... 20
    4.2      PAYLOAD DESIGNS TO MEET THE REQUIREMENTS OF THE S-UMTS SYSTEM ............................................. 25
    4.3      CAPACITY.................................................................................................................................................... 31
5     BUSINESS ASSESSMENT ............................................................................................................................... 36
    5.1      ANALYSIS OF THE COMPETITIVE ENVIRONMENT .......................................................................................... 36
    5.2      ASSESSMENT OF KEY SUCCESS AND RISK FACTORS ...................................................................................... 39
    5.2      BARRIERS TO ENTRY.................................................................................................................................... 40
    5.3      ANALYSIS OF THE MARKET MIX ................................................................................................................... 41
    5.4      INTERFACE BETWEEN MARKET AND TRAFFIC ANALYSIS AND SATELLITE SYSTEM SIZING ............................ 54
    5.5      EXPECTED EXPENSES AND REVENUES FOR THE SCENARIOS ......................................................................... 59
    5.3      FUNDING STRATEGIES ................................................................................................................................. 74
    5.4      IDENTIFICATION OF STRATEGIC PARTNERS .................................................................................................. 76
6     IDENTIFICATION OF KEY TECHNOLOGIES FOR S-UMTS...................................................................... 80
    6.1      GATEWAY SUBSYSTEMS .............................................................................................................................. 80
    6.2      SATELLITE PLATFORM, PAYLOAD AND ANTENNA ........................................................................................ 81
    6.3      USER TERMINAL BASE-BAND, RADIO FREQUENCY FRONT-END AND ANTENNA ............................................ 83
7     STANDARDISATION AND REGULATORY ISSUES ................................................................................... 93
    7.1      STANDARDISATION ...................................................................................................................................... 93
    7.2      LICENSING ISSUES ....................................................................................................................................... 94
    7.3      SPECTRUM ISSUES AND FREQUENCY SHARING ............................................................................................. 94
    7.4      CEPT / ITU ACTION PLAN ........................................................................................................................... 96
8     EUROPEAN S-UMTS DEVELOPMENT STRATEGY ................................................................................... 97
    8.1      SELECTION OF STRATEGIES.......................................................................................................................... 97
    8.2      THE WAY FORWARD ................................................................................................................................... 98
    8.3      DISCUSSION OF AN S-UMTS VENTURE...................................................................................................... 103
    8.4      ROLES IN A S-UMTS VENTURE ................................................................................................................. 110
    8.5      BENEFITS OF S-UMTS FOR THE EUROPEAN INDUSTRY .............................................................. 112
                                                                                                                                                               3
Contracts number:
ESTEC - 13694/99/NL/US                                                    Final Report
ESA S-UMTS study                                                          Date                                Revision
Final Report                                                              2000-11-29                          A


  8.6     WEAKNESSES AND RISKS OF S-UMTS ............................................................................................. 114
  8.7     CONCLUSION ........................................................................................................................................ 115
                                                                                                        4
Contracts number:
ESTEC - 13694/99/NL/US                            Final Report
ESA S-UMTS study                                  Date                   Revision
Final Report                                      2000-11-29             A




1 Introduction
This report presents the results from the ESA study contract ―Preparation of the Next Generation
Universal Mobile Satellite Telecommunications Systems (S-UMTS)‖. The work was undertaken by an
industrial consortium consisting of Ericsson AS, ASTRIUM , Telenor, Sintef and NERA ASA with
Ericsson as the consortium leader and the main contractor towards ESA

1.1     Scope and Goal
The study scope included seven defined tasks covering the following themes:

1.    Critical review of S-PCN roll-out
2.    The S-UMTS / IMT-2000 role
3.    Preliminary System Scenarios
4.    Business Assessment for the selected Scenarios
5.    Identification of Key Technologies
6.    Standardisation and Regulatory Issues
7.    European Development Strategy

Through performing these tasks the goal of the study has been to identify:

     Baseline system scenarios
     Need for key technologies and further research
     Market analysis and business profitability estimates
     Standardisation and regulatory issues that need to be addressed
     Possible European contributions


1.2     Background
The background of the S-UMTS project was to provide a study report on the subject ―Preparation of the
Next Generation Universal Mobile Satellite Telecommunications Systems (S-UMTS)‖. The main objective
of the study was to propose a strategy for maximising European chances to play a major role in this area.
The work was initiated by the European Space Agency (ESA).

1.3     References
      Submitted and approved work package reports:
       WP 1000 reports
       WP 2000 reports
       WP 3000 reports
       WP 4000 reports
       WP 5000 reports
       WP 6000 reports
       WP 7000 reports

1.4     List of abbreviations
3GPP         Third Generation Project Partners
A/D          Analogue to Digital
ACeS         Asian Cellular Satellite
ANSI         American National Standards Institute
                                                                        5
Contracts number:
ESTEC - 13694/99/NL/US                    Final Report
ESA S-UMTS study                          Date               Revision
Final Report                              2000-11-29         A


ARIB    Association of Radio Industries and Business
ATM     Asynchronous Transfer Mode
BCCH    Broadcast Control Channel
BER     Bit Error Rate
BOD     Bandwidth On Demand
BSS     Base Station Subsystem
CCIR    International Radio Consultative Committee (now ITU-R)
CDMA    Code Division Multiple Access
D/A     Digital to Analogue
EDGE    Enhanced Data Rates for GSM Evolution
ETACS   European Total Access Communication System
ETSI    European Telecommunications standard Institute
FCC     Federal Communications Commission
FDD     Frequency Division Duplex
FER     Frame Erasure Ratio
FPLMTS  Future Public Land Mobile Telecommunications System
FSS     Fixed satellite Services
G/T     Gain over noise Temperature
GB      Base Antenna Gain
GEO     Geostationary Earth Orbit
GGSN    Gateway GPRS Support Node
GHz     Giga Hertz
GM      Generic Service Module
GMPCS   Global Mobile Personal Communications Services
GMSC    Gateway MSC
GMSS    Global Mobile Satellite Standard
GPRS    General Packet Radio Service
GSM     Global System for Mobile Communications
GSO     Geostationary Orbit
GUI     Graphical User Interface
HEO     High Earth Orbit
HLR     Home Location Register
HSCSD   High Speed Circuit Switched Data
IFRB    International Frequency Registration Board (Now Radio
        Regulations Board)
IGW     Inter Gateway Link
IMT2000 International Mobile Telecommunications System 2000
INETS   Input Networks
IS-95   CDMA standard
ITU     International Telecommunications Union
ITU-R   International Telecommunications Union - Radio
ITU-R   ITU Radio communication sector
L-Band Generally 1-2 GHz
LCR     Least Cost Routing
LEO     Low Earth Orbit
Mbps    Mega bits per second
                                                                            6
Contracts number:
ESTEC - 13694/99/NL/US                       Final Report
ESA S-UMTS study                             Date                Revision
Final Report                                 2000-11-29          A


MEO          Medium Earth Orbit
MMS          Matra Marconi Space
MoU          Memorandum of Understanding
MPA          Multi Port Amplifier
MSC          Mobile switching Centre
MSS          Mobile satellite System
NCC          Network Control Centre
NGSO         Non Geostationary Orbit
NMT450       Nordic Mobile Telephone system 450
NMT900       Nordic Mobile Telephone system 900
PEST         Personal Earth Station
R&D          Research & Development
RF           Radio Frequency
RNC          Radio Network Centre
S-Band       Generally 2-4 GHz
SBS          Satellite Base Station
SCC          Satellite Control Centre
SGSN         Serving GPRS Support Node
SNMP         Simplified Network Management Protocol
S-PCN        Satellite Personal Communication Network
SSPA         Solid State Power Amplifier
S-UMTS       Satellite UMTS
SW-          Wide band Hybrid Code- and Time Division Multiple Access
C/TDMA       Scheme based Satellite standard
SW-          WCDMA based Satellite standard
CDMA
TACS         Total Access Communication System
TDD          Time Division Duplex
TG           Technical Group
TTA          Korean standards organisation
T-UMTS       Terrestrial UMTS
UMTS         Universal Mobile Telecommunication System
URAN         UMTS Radio Access Network
USRA         UMTS Satellite Radio Air Interface
UTRA         UMTS Terrestrial Radio Air Interface
VLR          Visitor Location Register
VPN          Virtual Private Network
VSAT         Very Small Aperture Terminal
WCDMA        Wide band CDMA
                                                                                                               7
Contracts number:
ESTEC - 13694/99/NL/US                                Final Report
ESA S-UMTS study                                      Date                    Revision
Final Report                                          2000-11-29              A




2 Critical review of S-PCN roll-out
2.1     Introduction
This part of the study represent a summary of the technical and economic issues faced by the first
generation of GSM like satellite systems targeting the mobile market in a broad sense. The main
intention of this review is to define the framework and find guidelines for a way forward for the next
UMTS based satellite systems

2.2     S-PCN Roll-out: lessons and conclusions

There are a number of key lessons to be derived from the current S-PCN roll out for European
participation in S-UMTS.

The terrestrial telecom network forces, driven by fierce competition among operators as well as mobile
phones and cellular network developments, were much underestimated at the time the satellite industry
made their initial optimistic business plans. Some key areas which were heavily misjudged, include:

     The rapid expansion of cellular area coverage.
     Bridging of cellular multi standards or multi frequencies in the mobile phone itself, which facilitates
      roaming between different cellular standards.
     Heavily reduced mobile phone prices often through subsidies.
     Big tariff reductions per minute for national cellular and long-distance traffic and also big reduction of
      international call charges.
     The data rates 2,4 – 9,6 kbps that S-PCN offers are not competitive with the rapid improvements in
      cellular data rates.
     The big cellular brand names are focused on mass markets and reluctant to develop dual band
      satellite/cellular phones for niche segments. Scarce R&D resources are used in more profitable
      areas.
      Even significant marketing efforts and expenses has failed to create the necessary customer interest
      for S-PCN services.

2.2.1 S-PCN Development Status
S-PCNs long time to market has allowed for considerable changes in market conditions to take place.
Marketing failures and technical difficulties, combined with high customer expectations derived from
cellular improvements, led to slow market acceptance for Iridium. The Globalstar roll out will prove if this
signals that the window of opportunity for S-PCNs is about to be shut. Investors‘ and creditors‘ growing
impatience and unwillingness to accept the considerable cost overruns and additional need for capital is
to be expected with the higher risks and uncertainty for returns.


2.2.2 Key Technical Issues

All the reviewed S-PCNs are based on proprietary solutions although the mobility solution and network
structure is GSM like. The air interface for ACeS and EAST are based on the GSM standards and
further work on a GEO GSM based air interface is carried out in ETSI.
                                                                                                                8
Contracts number:
ESTEC - 13694/99/NL/US                                Final Report
ESA S-UMTS study                                      Date                     Revision
Final Report                                          2000-11-29               A


The contemporary evolution towards open standards and an competitive environment, suggests that
evolving a proprietary system into an S-UMTS system may experience difficulties especially in trying to
address the horizontal market where the terrestrial systems dominate.

The S-PCN systems are using very different technology: GEO vs. LEO/MEO satellites, bent pipe vs. on-
board processing satellites, TDMA vs. CDMA air interface, call routing in private vs. public network, a
large vs. small number of gateways, etc. Although the technical characteristics vary a lot, all the S-PCNs
offers essentially the same services: voice, fax, low speed data and GSM based supplementary services.
A new complex system architecture supported by new technologies, such as with Iridium‘s solution,
represents considerable technical risk and the potential for cost and schedule over-runs. The decision to
route calls entirely within the space segment for Iridium introduced the need for inter-satellite links, on-
board processing, and a complex network control system. Such a high level of complexity in the first
generation represents a very ambitious technical programme. For future projects a lesson would be to
evolve the system from a precursor if possible – i.e. do not introduce new architectures and technologies
until confidence is very high that the system will provide the required services. Nothing like Iridium has
been flown before – hence major technical difficulties should have been anticipated.
The result for Iridium has been delays in introducing a service with the promised QoS (dropped calls are
reported to happen frequently and call completion rates are low), on-going technical problems which
prejudice credibility, and cost over-runs. Globalstar and ICO both represent less complex solutions. For
Globalstar, the call processing and switching are carried out by gateways on the ground and the signal is
relayed to its final destination via existing telecommunication networks. Hence, a first generation system
like Globalstar is more likely to be successful.
The choice of technical solution must be market driven – not technology driven. It is crucial in the rapidly
changing and converging marketplace to avoid getting ―side-tracked‖ by the cleverness of the technology
or the satellite industry interests alone. For the technical solution to be successful it is of essence that it is
compatible with the time to market demanded by the business plan.

2.2.3 Key Marketing Issues
The long time to market has changed the market situation, allowing alternative solutions to gain
considerable stronghold. The GSM penetration has risen service quality standards and set the agenda
for customer expectations (pricing levels and terminal sizes).
At service launch timing is critical and marketing efforts must be co-ordinated with distribution-efforts to
ensure sufficient availability of handset. A regional step by step rollout may be the key as seen in
Globalstar‘s example. Handsets are easier to sell on a regional basis since a smaller production volume
and distribution system is required compared to a global system. Moreover, telephones can be tailored
for each geographical region. For global sales of handsets a system of distributors, who knows the
product and can convince customers locally to buy into the solution, is needed.

2.2.4 Key Financial Issues
The high risk and large sums involved, together with the long time to market, means that financing for
satellite systems is difficult in general, and more so in particular when dealing with the amounts needed
for LEO/MEO constellations. A typical investment from a single bank would be $20M – hence, would
need 50 parties to raise $1B. Getting all the parties to sign up is an immense problem.
At the present time it is almost impossible for a new satellite project to obtain debt financing without full
recourse i.e. must be able to guarantee that the lender will always be able to recover the loan. Parent
companies can guarantee loans based on their own balance sheets. For example, Iridium has full
recourse to Motorola, which is a $39B company. In all mobile systems to date, only full recourse funding
has been obtained.

Strategic alliances and wealthy partners are needed to overcome financing hurdles. Investors looking
into satellite systems may choose from a wide range of projects. Financing is therefore very difficult
                                                                                                                 9
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A


without wealthy partners. Repeated postponements and efforts to obtain additional funding on top of an
already long time to market has left reluctant investors sceptical of further participation in the projects.

Cost overruns will happen and a significant error margin needs to be taken into account when break-
even levels are calculated and financing strategy is developed. Cost containment is crucial as costs often
are doubled from original estimates when projects are implemented.

2.2.5 Key Regulatory And Licensing Issues

It has been claimed that the main issue for the success of S-PCNs is regulation. The availability of
sufficient frequency spectrum and a smooth and efficient licensing system are prerequisites for the
successful introduction of such systems.

Fortunately, this is being taken seriously by the actors in the field, both among the authorities at different
levels and among the system suppliers, national service providers and operators. It is of paramount
importance that the interested parties are active in the various organisations, be it the ITU with its sub
bodies, the Regional regulators organisations, the standards organisations and the other fora where
issues of common interest are discussed and actions co-ordinated.

2.3    Conclusion
S-PCN systems have turned out to be high-risk projects of extreme proportions. Longer than expected
time to develop and severe budget miscalculations have resulted in business cases less attractive than
first expected. Within the telecom and datacom industry it is difficult to have a clear view about the
product, system and market developments for more than 2-3 years into the future, which is much less
time than S-PCN rollout from concept to commercial service takes. This is a major dilemma and problem
area where risk mitigation is extremely important.

A key problem is that time to market (general system availability) seldom meets time to customer
requirements. GEO solutions will in general provide a better, less risky business case than more complex
non-GEO systems. This is due to a shorter time to market period. GEO enables starting regional service
and earn revenues and learn about the market before going into global expansion, possibility for reuse of
existing ground infrastructure equipment, and the potential for alternative usage of satellites like;
broadcasting and fixed terminals in different frequency bands than S-PCN.
                                                                                                           10
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A




3 The S-UMTS/IMT-2000 Role

3.1    S-UMTS versus T-UMTS
      rd
The 3 generation mobile telephony system is called universal mobile telecommunications systems
(UMTS) in Europe. International Mobile telecommunications-2000 (IMT-2000) aims to provide wireless
telecommunications round the world, regardless of location, network or terminal used. It includes
proposals for a variety of mobile terminal types, linked to terrestrial and/or satellite networks and serving
mobile or fixed users.

Satellite mobile telecommunications systems inherently are the more costly (calculated in the unit
          2
kbps/km ), but once in place, a satellite system could provide uniform coverage over an enormous area.
True global coverage can be provided only by satellite systems.

To find the optimal mobile telecommunication system, there are choices, priorities and tradeoffs to be
made. Two types of S-UMTS usage is defined:
    a) Complement to T-UMTS
    b) Competition with T-UMTS

Starting with case (a), Complement to T-UMTS, the list below is suggestions of areas where S-UMTS
supplements T-UMTS by providing:
 Coverage extension; The UMTS coverage can be increased considerably using S-UMTS for areas
    less densely populated
 Coverage completion; S-UMTS can be used for filling gaps in the T-UMTS network. Long timeframe
    or ―window of opportunity‖.
 Global roaming; S-UMTS can extend T-UMTS operators coverage by offering roaming possibilities.
    This can also be used in areas with incompatible terrestrial systems. Long timeframe or ―window of
    opportunity‖.
 Rapid deployment; S-UMTS can provide coverage in areas still waiting for T-UMTS build-out. S-
    UMTS will probably be too late to make full use of this opportunity.
 Disaster-proof availability; In crisis areas telecommunications systems are often disabled, or non-
    existent. S-UMTS can easily be reconfigured to cover these areas. Not possible to base a business
    case on.
 Dynamic traffic management; S-UMTS is used to relieve temporary or permanent congestion in T-
    UMTS. Long timeframe or ―window of opportunity‖.

Continuing now with case (b), Competition to T-UMTS, are cases where S-UMTS has an advantage over
T-UMTS. Services and applications involving broadcast and point-to-multipoint messages can e.g. be
more efficiently (in terms of capacity) transmitted via S-UMTS instead of via several T-UMTS cells.
Satellite services should focus on non-real-time services like downloading, browsing, etc. to avoid the
problem with transmission delays.

3.1.1 Potentially viable S-UMTS scenarios
Two business concepts seem viable:
   A. Mass market with relatively low tariffs: basic T-UMTS complement
   B. Niche market with higher tariffs: extended services, for executives and other high-end users
       willing to pay a premium
                                                                                                        11
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                   Revision
Final Report                                         2000-11-29             A


Both of these concepts involve a basic S-UMTS package that offers:
 Same services and applications as T-UMTS
 a car and an indoor system
 specific car and broadcast/multicast applications
 General Packet Radio Service (GPRS) outside GPRS coverage
 communication at tariffs equal to or slightly higher than those for T-UMTS

In addition to the basic package, the niche market package also includes:
 executive applications
 UMTS outside of T-UMTS coverage areas, at higher tariffs


3.1.2 System architecture
There are several aspects to discuss regarding the system architecture of an S-UMTS system. A generic
architecture of T- and S-UMTS is shown in the figure below.
                                                                 Access networks Iu

                                                          Uu
                                                                        ISDN

                                                                        GSM

                                          Other
                                          cellular                      UTRAN
                                          networks
               USRA                                                                      Core Network
                            USRA
                                                               RNC      USRAN
                                                                                  1

                                                Network
                                                Control
                                                                        USRAN
                                                Centre                             n
               S-UMTS
               compatible
               terminal            S-UMTS compatible earth
                                   station



Figure 3-1 A simplistic architecture of a T- & S-UMTS system

The amendments required to incorporate a satellite system in UMTS may impact smaller or larger parts
of the 3GPP specifications. In the best case the amendments are limited to the air interface, then no
amendments to the CN are required. Minimal impact on the air interface is probably the best strategy for
an optimal synergy with T-UMTS.

In future satellite systems, on-board beam switching could provide an optimised way of linking to the best
running earth station but also to the users when and only when they need to communicate. Beam
switching could be used to direct the satellite beams only to those who request a connection, e.g.
coverage of the oceans and air or in the disaster case above.
                                                                                                                        12
Contracts number:
ESTEC - 13694/99/NL/US                                Final Report
ESA S-UMTS study                                      Date                     Revision
Final Report                                          2000-11-29               A


3.2    User and operator requirements
The UMTS vision is to integrate the complementary elements of satellite and terrestrial mobile systems to
provide universal services, where universal means world-wide including cellular + satellite, mobile
including handheld, and the services comprise bearer services and Tele-services in range and quality
similar to those planned for fixed telephony.

For S-UMTS to fulfil this vision, it has to provide many of the same services as T-UMTS, and within the
same range of QoS. The higher success of integration the more reasonable it is that the user will expect
more or less the same quality of service, irrespective of which network he momentarily is using.

S-UMTS services
Below are some of the expected requirements for S-UMTS regarding available teleservices and
associated services.

Table 3.2 Teleservices required
                                 Traffic type                  Services
INFORMATION TYPE
Telephony (voice)                Constant bit rate             Conversation, telephone-conferencing.
                                 Real-time                     Personal real-time services.
Messaging                        Dynamical bit rate            SMS, paging, fax, two-way messaging, e-mail, etc.
(text/data)                      Non real-time
Web browsing & Wireless          Dynamical bit rate            Access to Internet
Application Protocol (WAP)       Non real-time
browsing
Audio streaming                  Constant bit rate             Voice messaging (asymmetric), voice over IP (symmetric).
                                 Non real-time                 Music-on-demand.
Video streaming                  Constant bit rate             TV-on-demand, video-on-demand, information retrieval (videos
Data streaming                   Non real-time                 or e-mail with videos, possibly compressed).
Multicasting
Real-time video                  Constant bit rate             Video-telephony, video-conferencing, multi-party games,
(video/voice/data)               Real-time                     scanning, telnet, etc.
Data transfer                    Dynamical bit rate            Information acquisition (bulk) from databases
                                 Non real-time
Broadcasting                     Constant bit rate             Distribution applications without users control ―push‖
                                 Non real-time
E-commerce                       Mostly dynamical bit rate     Transactions (2-way interactive), mobile shopping/banking,
                                 Mostly non real-time          online payment.
Location-based services          Mostly dynamical bit rate     Location-precise traffic information, navigation, logistics,
                                 Mostly non real-time          telemetry (monitoring), etc.
Relay between T-UMTS cells BSs   Mostly constant bit rate      Relay of high-capacity services from BS to BS. Symmetric
                                 Mostly non real-time          services except for broadcasting.




3.3    Business, market and financial constraints
In the 1990s, the telecommunications market changed radically and rapidly, and is constantly generating
new forecasts. Especially forecasts for wireless users have increased during the last few 6 month-
periods. The Ericsson September 1999 forecast estimated 1.1 billion wireless end-users by 2004, an
increase of 9% since last forecast. Using this as a base for 3G estimates, a bright future is expected for
3G mobile systems. This chapter discusses the possibility that satellite UMTS is among them.

The general view of satellite 3G systems capability, service, and quality is in some cases not completely
true. Below is an effort to correct the misconceptions:
                                                                                                                                 13
Contracts number:
ESTEC - 13694/99/NL/US                                       Final Report
ESA S-UMTS study                                             Date                        Revision
Final Report                                                 2000-11-29                  A


Table 3.6 Myths and Realities concerning S-UMTS
MYTH                                                             REALITY
The introduction of wireless data applications will create a     Most of the applications carried will be entertainment
demand for application accesses anywhere.                        applications or urban related services.
S-UMTS will be a carrier with QoS properties equal to T-UMTS     S-UMTS will have different QoS properties than T-UMTS and
and may thereby act as a transparent substitute for T-UMTS       will therefore need to focus on applications where these
carrying the same applications                                   properties are sufficient or even better suited (e.g. broadcast
                                                                 and multicast)

                                                                 A good solution for in-door coverage is a prerequisite for S-
                                                                 UMTS reaching the horizontal market.
S-UMTS can take on a role as a competitor to T-UMTS.             S-UMTS can only take a role as a complement to T-UMTS.

GPRS is not a competitor to S-UMTS.                              GPRS can offer the same services as S-UMTS and has good
                                                                 coverage properties and will as such compete with S-UMTS.
There is a direct relation between carried amount of data and    The pricing strategy will have major impact on the relation
revenue.                                                         between carried amount of data and revenue.
An S-UMTS operator can steal revenue from a T-UMTS/GPRS          Flat Fee pricing will guarantee the revenue stream to a T-
operator where the latter lack coverage.                         UMTS/GPRS operator.
The bearer service provider will get the entire user spending.   A major part of the user spending might go directly to the
                                                                 content provider.
There is a lot of money in carrying broadcast services.          Broadcast services are seldom critical to get access to
                                                                 anywhere.



Evolution or revolution of S-PCN towards S-UMTS?
Today, the most successful mobile satellite systems principally serve niche markets, whilst terrestrial
systems, such as T-UMTS and GPRS, aim at the mass markets. S-UMTS may address both niche and
mass market segments. Addressing the mass market will probably present a major challenge for satellite
systems. To the extent possible, S-UMTS should be an evolution from T-UMTS, principally to permit
reuse of terminals and inter-working with terrestrial mobile networks and services. Several S-PCN
systems, both existing and new, intend to introduce high data rate services before S-UMTS can be
available. This constitutes a threat to the business case for S-UMTS. Aiming at slightly different markets
can reduce the threat, e.g. S-UMTS can aim at the mass market.

It is assumed that S-UMTS systems will be generally available from year 2005 on. Terminal availability is
critical for the time to customer. The terminal market for T-UMTS and dual mode GPRS/T-UMTS will be
well developed by 2005. If the difference between air interface in T-UMTS and S-UMTS can be
minimised, terminal availability will not be a problem.

Competition scenarios
S-UMTS may face keen competition from other systems in trying to capture customers.
 T-UMTS. In areas with terrestrial UMTS coverage, T-UMTS may in some cases be a competitor. This
   is especially the case whenever T-UMTS offers the same sort of applications and services as S-
   UMTS.
 GSM/GPRS. In areas with GSM/GPRS coverage, it may compete with S-UMTS if it offers similar
   services.
 S-PCS systems may compete with UMTS but are not expected to be principal competitors. This is
   due to the constraints of their technologies, differences in offered services and applications, as well
   as roaming capabilities.
 Inmarsat may be the major competitor to S-UMTS. Inmarsat is scheduled to provide global UMTS
   compatible services in 2004, and will probably provide some of the services planned for S-UMTS.
   Inmarsat is therefore likely to capture some of S-UMTS‘s target market.
 Broadband satellites. The use of broadband satellites that connect remote T-UMTS islands, can
   compete with S-UMTS. Most likely, S-UMTS has insufficient capacity to connect T-UMTS islands.
                                                                                                             14
Contracts number:
ESTEC - 13694/99/NL/US                                Final Report
ESA S-UMTS study                                      Date                    Revision
Final Report                                          2000-11-29              A



Co-operation scenarios
In some cases, S-UMTS can be an attractive system to co-operate with, such as for:
 T-UMTS. Terrestrial operators may expand their coverage or offer special applications and services
    using S-UMTS.
 GSM/GPRS. S-UMTS can be used as an early implementation of UMTS-services.
 Inmarsat. Since Inmarsat could be a major competitor to S-UMTS, it will be an advantage for S-
    UMTS service providers to co-operate.


3.4     Technical issues

3.4.1 Global or regional coverage?
Requirements for coverage highly depend on the market scenario and, of course coverage of terrestrial
networks. There are also cost limitations and risks to consider, and a balance must be found of all these
parameters. Different requirements are set to coverage in different scenarios:

     Coverage completion:
      Satellite coverage will fill gaps in the terrestrial coverage, either created intentionally or due to bad
      planning. Global or regional coverage.
     Coverage extension:
      The satellite will cover vast remote areas where it is not possible or feasible to build a terrestrial
      network. Requires global coverage.
     Global roaming:
      S-UMTS services are offered everywhere, anytime. Roaming with terrestrial networks is a condition,
      and this service requires global coverage.
     Rapid deployment:
      S-UMTS provides the first step for full UMTS coverage in an area. A regional coverage with spot-
      beams providing high capacity is required.
     Disaster proof availability:
      S-UMTS provides backup service in case of disaster. Requires global coverage, but spot-beams
      providing extra capacity will be a valuable asset.
     Dynamic Traffic Management:
      Satellite is used to off-load some of the traffic from the terrestrial network. Regional coverage with
      spot-beams is required.

The best possible business case will include all the identified scenarios. This requires a satellite network
having:
 Global coverage
 Regional coverage with steerable spot-beams
 Low cost
 Early launch relative to implementation of T-UMTS


3.4.2 Comparison between GEO and Non-GEO Constellations
The success of an S-UMTS system, are decided by three key factors: a broad range of services
provided, a high Quality of Service, and a reasonable price. A comparison between satellite
constellations based on these factors is made.
                                                                                                               15
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A


General discussion
The LEO / MEO constellation sizing depends strongly on orbit altitude and minimum required elevation.
The fact that Non-Geo systems can provide high elevation at high latitudes is one of the principal
interests of these constellations. Geo satellites cannot provide high elevation coverage at high latitudes.
In some cases, though, this may not be needed.

An interesting aspect of LEO / MEO constellations is the use of satellite diversity. However, in order to
have satellite diversity at reasonable elevation angles it implies increasing the constellation size and
hence its cost. On the plus side, the QoS is improved, and system capacity can be increased. Satellite
diversity could also be employed on a Geo-system.

A large number of satellites is needed to provide a world wide coverage with a LEO constellation. Only
four GEO satellites are required to guarantee almost global coverage, except in the high latitudes areas.
A MEO constellation offers an interesting compromise between a LEO and a GEO.

LEO and MEO constellations provide the same capacity to all areas irrespective of the population
density. Large parts (oceans, deserts, etc.) of the total system capability is thus unused. A GEO
constellation has a more efficient resource utilisation, since the fixed orbital position permits coverage of
the most suitable market regions.

The propagation delay alone for a satellite hop via a GEO satellite is typically 240 ms compared to 5 to
20 ms for a LEO constellation or 70 to 100 ms for a MEO constellation. A GEO satellite system is thus
not well suited for applications requiring low delay times, but this does not matter for other applications,
such as e.g. web browsing.

Link budget and Capacity
A LEO or MEO satellite has certain link budget advantages with respect to a GEO satellite. A LEO
satellite communicating with a terminal at 20° elevation is some 14 times closer to the terminal than a
terminal at the same elevation for a GEO satellite. To achieve the same link quality with all other
parameters remaining the same, there is a gain of 23 dB on the link budget which can be transferred to
reduced satellite antenna size, or increased link margin, or both.


                    Earth's Surface              Slant Range= 2.5 for Leo
                                                   Altitude = 1.1 for Geo

                           Edge of
                           Service Zone

                                                           Altitude
                                                    LEO                     Geo Satellite

Figure 3-2 Service zone comparison.

Increasing the antenna gain is easier on a LEO satellite. E.g. moving from a 1m to a 1.5m antenna is
straightforward on a LEO satellite, whereas the equivalent for a GEO satellite is to move from a 14 m to a
21m reflector. The limitation in large reflector technology has serious impacts on the system architecture
for a GEO- system.

The consequence of these facts is that a LEO system can almost certainly outperform a GEO- system.

Technological Considerations
                                                                                                            16
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


Change in propagation time is important for synchronisation, particularly in the return link. Non-GEO
systems will require frequent updating of the system timing. For CDMA/LEO systems in particular, to
maintain the synchronisation required to assure that spreading codes remain orthogonal on the return
link, the rate of change is so rapid that precompensation schemes may be necessary.

A LEO constellation without beam scanning will require frequent hand-over and require a large signalling
overhead. In addition, a given user can never spend more than about 1 minute within a given beam
without requiring hand-over. For GEO systems hand-over occur due to user movement or possibly
between satellite coverage.

For a GEO satellite, a limiting factor is the complexity of the antenna and payload design.
Communication with handheld terminals requires a large number of spot beams, typically 250. The
design solutions consist of large reflectors (12-15 metre) and multiple-element (around 100) feed
assemblies. Particular attention has to be paid to the optimisation of antenna G/T, beam isolation and
frequency reuse, and the control of passive inter-modulation products.
The design solutions, which could be envisaged for a LEO system, are somewhat relaxed with respect to
the GEO solution.

Space Segment Cost
Despite the fact that GEO satellite systems will inevitably be more complex, the greater number of
satellites and launches required for LEO and MEO constellations imply a higher cost for these systems.
Geo systems can be expected to cost around USD 1 billion, whereas a Non-Geo system costs around
USD 3-5 billion.

Regulatory
Today licensing issues are leading to different spectrum allocations in different geographical regions. A
global system will have to adjust its frequency. Consequently, a LEO or MEO satellite will have to be
capable of operating over all possible allocation frequencies. A GEO system can more easily support
different frequency allocations in different areas.

Summary
Looking at the criteria which have been considered for the three satellite constellations, it would appear
that :
 A MEO system would be preferred when examining constellation-related issues
 A GEO system would be preferred when examining technological-related issues
 A GEO system would be preferred when examining regulatory issues.
 A LEO system would be preferred when Quality of Service is the driver.
 A GEO system would be preferred for a market and business driven strategy.

3.5    Relative Capacity Evaluations
An evaluation of the capacities of a LEO, MEO and GEO system is made. The results have to be treated
with caution, since some of the hypotheses are simplified, but will give an indication of the capacity
capabilities by the systems.
For the LEO and MEO solutions we assume that the LEO constellation is Globalstar-like and the MEO
constellation is ICO-like. The constellations are thus:

Table 3.7 Constellations used in simulations
                                     LEO                      MEO                       GEO
        Orbit Altitude             1410 km                  10390 km                  35786 km
        Active satellites             48                       10                     1 or more
                                                                                                                                                                                                         17
Contracts number:
ESTEC - 13694/99/NL/US                                                                                                Final Report
ESA S-UMTS study                                                                                                      Date                                    Revision
Final Report                                                                                                          2000-11-29                              A



The resulting capacities for Leo, MEO and Geo constellations are shown in the figures below.



                                                                                                            Capacity of LEO System
                                                                                                            Forward Link Capacities

                                                                 14000


                                                                                                                                                                         Capacity per
                                                                 12000
               Number of Users Served (8 kbps, handheld users)




                                                                                                                                                                         satellite


                                                                 10000
                                                                                                                        Suggested
                                                                                                                        capacity limit at
                                                                 8000                                                   c230 beams
                                                                                                                                                                                2 band, dual diversity
                                                                                                                                                                                1 band, no diversity
                                                                 6000
                                                                                                                                                    Capacity per
                                                                                                                                                    satellite
                                                                 4000



                                                                 2000



                                                                        0
                                                                            0        50        100              150               200             250              300
                                                                                                          Number of Beams




Figure 3-3: Satellite Capacity vs. Beam Numbers for LEO Option (Variable RF Power & Antenna
                                            Gain)


                                                                                                            Capacity of MEO System
                                                                                                            Forward Link Capacities

                                                                 9000


                                                                 8000
               Number of Users Served (8 kbps, handheld users)




                                                                 7000                                                                                                    Capacity per
                                                                                                                                                                         satellite

                                                                 6000

                                                                                                                              Suggested
                                                                 5000                                                         capacity limit at                                 2 band, dual diversity
                                                                                                                              c120 beams
                                                                                                                                                                                1 band, no diversity
                                                                 4000


                                                                 3000                                                                               Capacity per
                                                                                                                                                    satellite
                                                                 2000


                                                                 1000


                                                                    0
                                                                        0       20        40         60         80          100             120         140        160
                                                                                                          Number of Beams
                                                                                                                                                                                                 18
Contracts number:
ESTEC - 13694/99/NL/US                                                                                               Final Report
ESA S-UMTS study                                                                                                     Date                                Revision
Final Report                                                                                                         2000-11-29                          A


Figure 3-4: Satellite Capacity vs. Beam Numbers for MEO Option (Variable RF Power & Antenna
                                             Gain)



                                                                                                          Capacity of GEO System
                                                                                                          Forward Link Capacities

                                                                 25000

                                                                                                                                                         Capacity per
                                                                                                                                                         satellite
               Number of Users Served (8 kbps, handheld users)




                                                                 20000



                                                                                     Suggested
                                                                 15000               capacity limit at
                                                                                     c300 beams
                                                                                                                                                                        2 band, dual diversity
                                                                                                                                                                        1 band, no diversity
                                                                                                                                          Capacity per
                                                                 10000
                                                                                                                                          satellite




                                                                 5000




                                                                     0
                                                                         0   100   200           300           400          500     600         700          800
                                                                                                         Number of Beams



  Figure 3-5: Satellite Capacity vs. Beam Numbers for GEO Option (Fixed RF Power & Variable
                                          Antenna Gain).


The results tend to indicate that the capacity (in terms of user density served) of the Leo solution as
defined here is around 7 times that of a global geo solution. In order for a Geo system to be able to
complete with a Leo system in terms of capacity, very large reflectors would have to be employed (>20
metres) and the coverage would have to be severely limited to very specific geographical regions.


3.6    Regulatory issues

3.6.1 Frequency allocations
In Europe, the CEPT ERC has split the IMT-2000 band between terrestrial and satellite UMTS (T-UMTS
and S-UMTS) but the lowest 5 MHz of the band is allocated to DECT. The allocations for MSS, including
S-UMTS are: 1980-2010 MHz and 2170-2200 MHz.
However, frequency usage is not harmonised between national administrations, but most European
countries have allocated the MSS bands for this use now or from year 2005.

In Japan, the MSS bands are the same as in Europe, but the lower part of the IMT-2000 band (1895-
1918 MHz) will be used for second generation systems Personal Handiphone Service (PHS).

In the USA, the use of the IMT-2000 band differs from that in Europe and second generation Personal
Communications Service (PCS) is allocated to the lower part of the band.
The MSS bands are also shifted: 1990-2025 MHz (up-link) and 2165-2200 MHz (down-link).
                                                                                                        19
Contracts number:
ESTEC - 13694/99/NL/US                            Final Report
ESA S-UMTS study                                  Date                    Revision
Final Report                                      2000-11-29              A



The current frequency allocations for MSS are symmetric (same bandwidth for up-link and down-link). In
the future a need for asymmetric bandwidth will probably rise, as most interactive services needs more
bandwidth on down-link (downloading of data, video-streaming etc.).

3.6.2 Feeder link planning

According to the ITU Radio Regulations allocation table, feeder links may use various bands. For
geostationary systems (GSO) the feeder links are normally in one of the Fixed Satellite Service (FSS)
bands (S/C-band or Ku-band). The feeder link planning is thus regulated according to the rules
established for FSS and the planning is done via the ITU RS. EAST will use Ku-band, while ACeS and
Thuraya will use S/C-band for the feeder links.

Potential frequency bands for feeder links for non-geostationary systems (NGSO) are C-band, Ku- and
Ka-band. Globalstar and ICO use C-band for the feeder links, while Iridium uses Ka-band.

3.6.3 Frequency band and interference issues
In some countries the interference between terrestrial communications (fixed radio services, mobile
services and radar) and mobile satellite services (MSS) operating in the same frequency band may be
severe. Various satellite systems may also interfere with each other. Consequently, frequency sharing
issues are important in the implementation of a MSS. In addition Power flux density (PFD) limitations
may be imposed as given in ITU Resolution no.46. The PFD limits given for GSO and NGSO systems
are flexible limits but will trigger co-ordination.

After conducting studies of potential interference, the ITU has concluded that the sharing of frequency
bands between satellite and terrestrial systems is not feasible. For sharing between satellite systems the
ITU has concluded that:
 Sharing on a co-frequency, co-coverage basis between MSS networks employing TDMA or FDMA
    techniques is not feasible.
 Sharing on a co-frequency, co-coverage basis between MSS networks using FDMA or TDMA and
    networks using CDMA is not feasible.
 MSS networks using CDMA can share on a co-frequency, co-coverage basis.

3.6.4 Licensing aspects

The purpose of spectrum licensing is to divide the limited frequency spectrum, with a minimal amount of
interference, among as many systems as possible.
There are different types and combinations of licenses for MSS, e.g. license may be needed for service,
network, earth stations, frequency assignment or the radio-equipment.
From a MSS operator‘s point of view, it would be highly advantageous to receive all necessary licenses
from one organisation, according to the one-stop-shopping (OSS) principle. This may however be difficult
to accept for the national licensing authorities, which would lose some of their influence and national
independence.

Licensing authorities can chose several approaches when dividing spectrum. The total market driven
approach is one possibility: the spectrum is auctioned to the company willing to pay the highest price.
This is not a suitable approach for satellite systems, though. The best solution is probably to impose
some minimum entrance requirements of technical, financial and legal nature. Then there can be either a
competitive or a comparative bidding to determine which systems should be granted license. This is the
normal situation in most countries today for mobile licensing, although auctions are also used.
                                                                                                           20
   Contracts number:
   ESTEC - 13694/99/NL/US                             Final Report
   ESA S-UMTS study                                   Date                   Revision
   Final Report                                       2000-11-29             A




   4 Preliminary system scenarios
   4.1     Proposed system and architecture
   The figure below illustrates the overall architecture of an S-UMTS system. It consists of a geo-stationary
   transparent satellite payload, a number of gateways, network and satellite control centres, and the user
   terminals.



                                    GEO satellite

Gateways

                                                                                 User segment


                                                                                                        Vehicular
                                                                                                        Terminal




                                         SGF




                                                                         Palmtop/Laptop
                                                    Handheld                Terminal
  Core                                              Terminal
 Network

                       NCC


                                  Figure 4-1 S-UMTS system architecture


   4.1.1 System issues
   Selected air interface
   The recommended air interface for S-UMTS is an air interface based on W-CDMA. Only a single air
   interface is assumed to be provided for S-UMTS.

   A prerequisite for this selection is:
   ETSI S-UMTS WG takes T-UMTS specification and ‗converts‘ into satellite specifications. ETSI will thus
   make a satellite standard based on CDMA.

   Connectivity
                                                                                                                21
Contracts number:
ESTEC - 13694/99/NL/US                                Final Report
ESA S-UMTS study                                      Date                      Revision
Final Report                                          2000-11-29                A


Use of a transparent satellite is the baseline for an S-UMTS system. The system is assumed to have a
large (e.g. 250) number of mobile beams at S-band, and a smaller (e.g. 16) number of feeder beams.
Groups of mobile beams are connected to a specific feeder beam. To access a specific mobile beam, the
feeder beam connected to this mobile beam has to be used.
In each feeder beam, there can be one or more gateways. Each gateway serves a specific set of mobile
beams. These beams have to be among the beams connected to the feeder beam where the gateway is
located.

Single hop MS – MS connections
Single hop MS – MS connections are not considered vital, but if the implications are small, it should be
supported. If single hop connections would require modifications of the terminal with regard to T-UMTS, it
should not be provided.

Diversity
Use of diversity for improved satellite visibility is not part of the baseline as the benefits of diversity for a
geostationary satellite system is limited.

Synchronisation
The same approach as in T-UMTS should be used for synchronisation. The beams and gateways are not
synchronised to each other. (Assuming that a frequency block is not shared between the gateways).
Scrambling codes are allocated to each gateway, and channelisation codes are selected by the gateway.

Coverage
The mobile beam coverage and feeder beam coverage for Scenario 1 are shown below. The use of a
multi-beam feeder coverage allows reduction in the occupied spectrum requirement for the feeder, which
can be achieved via frequency reuse between beams. It also allows a reduction in Gateway traffic and
EIRP requirements (i.e. system load sharing between Gateways).




               Figure 4-2 : Mobile Beam Coverage (132 beam, extended satellite design)
                                                                                                     22
Contracts number:
ESTEC - 13694/99/NL/US                           Final Report
ESA S-UMTS study                                 Date                   Revision
Final Report                                     2000-11-29             A




                                 Figure 4-3 : Feeder Beam Coverage


4.1.2 Space Segment
The space segment consists of the satellite (or satellites) and the SGF (Satellite Ground Facilities).
The S-UMTS satellite payload has been designed around known satellite platform capabilities and is
characterised by :
 Two coverage scenarios investigated (Global and Regional)
 A high-performance S-Band Antenna with a large reflector (7-15 metres)
 a large number of feed elements (120) forming 132 or more high-performance beams
 Possibility to modify beamset in-orbit so as to optimise antenna performance to evolving traffic mix
    and distribution.
 Modular payload architecture to allow customised accommodation of mission-specific requirements
     antenna architecture (reflector diameter, number of feed elements (80-120))
     processor architecture
         transparent analogue or transparent digital processors can be envisaged
 High spectral capacity :
     320 x 5 MHz spectrum blocks (analogue processor)
     480 x 5 MHz spectrum blocks (digital processor)
 Resource allocation flexibility to address hot spot traffic (reconfigurable in-orbit)
     Flexibility of RF power allocation to beams
     Flexibility of spectrum block allocation to beams
         up to three blocks in one mobile beam
 Feeder antenna design optimised to cover gateway locations
                                                                                               23
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


   Sixteen Ka-band beams in receive and transmit from separate antennas


The primary characteristics of the payload are given in the following tables.


                                                                Regional
                                                                                      Global
                                                               Augmented

             Forward Direction
                Feeder Side
                     Uplink Frequency                               27500 - 30000 MHz
                     Receive G/T                                   6 dB/K       0.3 dB/K
                Mobile Side
                     Downlink Frequency                               2170 - 2200 MHz
                     EIRP (Average gain, peak RF)                 75.4 dBW        68 dBW
             Return Direction
                Mobile Side
                     Uplink Frequency                                 1980 - 2010 MHz
                     Receive G/T                                  16.7 dB/K      9.5 dB/K
                Feeder Side
                     Downlink Frequency                             17700 - 20200 MHz
                     EIRP (EoC)                                   58 dBW        52.5 dBW

                         Table 4.2 Payload Performances (132-beam version)
                                                                                                        24
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                 Revision
Final Report                                         2000-11-29           A



                                                             Mass         Power         Dissipation
                                                             (kg)          (W)              (W)

     Global Analogue Option
      - characterised by global coverage, 120 feed           1031          9300            6316
     elements, analogue processor
     Global Digital Option
      - characterised by global coverage, 120 feed           927           10068           7084
     elements, digital processor
     Regional Analogue Option
      - characterised by European coverage, 80 feed          899           7216            5120
     elements, analogue processor
     Regional Digital Option
      - characterised by European coverage, 80 feed          771           7984            5888
     elements, digital processor
     Augmented Regional Analogue Option
     Augmented Regional Digital Option
      - characterised by European coverage, 120              968           10068           7084
     feed elements, digital processor

                     Table 4.3 Payload Mass, Power, and Dissipation Summary


4.1.3 Ground infrastructure
The ground network should be based on the T-UMTS architecture to the extent possible. The core
network should be identical to T-UMTS core network.
The ground infrastructure will consist of gateways giving access to the terrestrial network through the
standard cellular core network nodes, and a network control centre providing some common functions
(e.g. resource management) for co-ordination between the gateways.
The satellite gateway – in general providing the same functionality as a RNC/Node B in T-UMTS - should
interface the core network through the standard Iu interface. As the gateways are normally considered to
be part of different PLMNs, it is not considered useful to have a direct interface between the gateways for
S-UMTS (Iur interface).
To keep the total system cost down, the system should have only a few gateways. Typically, there will be
a minimum of one gateway per feeder beam.
User data should be routed via terrestrial public networks to the gateway covering the S-UMTS user.
―Least cost routing‖ using special arrangements to minimise the terrestrial part of the connection by
selecting an ―optimum‖ gateway is not considered useful. Terrestrial routing costs are small, and not
deviating from the T-UMTS standard is more important.

4.1.4 Terminals
A large number of S-UMTS terminals can be envisaged for different applications, services and user
groups. It is impossible to pinpoint how S-UMTS terminals will look like 5 to 10 years from now, anyway
five different types of terminals have been defined:
 Handheld terminals
 T-UMTS terminals with extended S-UMTS capabilities
 Transportable terminals (palmtop-sized)
 Transportable terminals (laptop-sized)
                                                                                                           25
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                   Revision
Final Report                                       2000-11-29             A


     Vehicular terminals

The different terminal types can be categorised in different terminal classes in terms of functionality.
Single-mode terminals are pure S-UMTS terminals, i.e.; they may not be used for communications with
any terrestrial system. Multi-mode terminals are capable of communicating with both S-UMTS systems
and terrestrial systems. A wide range of multi-mode terminals can be envisaged. Some of these are:
 Single-mode S-UMTS terminals
 Generic multi-mode terminal, capable of connecting to a arbitrary number of systems
 Dual-mode S-UMTS/T-UMTS terminals
 Dual-mode S-UMTS/GPRS terminals

In general, an S-UMTS terminal should be as identical to a T-UMTS terminal as possible maximising
reuse of hardware and software units. Such an approach is believed to be the best way to provide low
cost and attractive terminals for S-UMTS.


4.1.5 Conclusion
The primary selection criteria for the S-UMTS system architecture should be:
1. Low total system cost
2. Attractive terminals
3. Short time to market.
These targets are obtained through reuse of T-UMTS components (terminals and ground network,
system specifications).

These requirements are set out in two scenarios the first dealing with a regional coverage that provides
service over Europe, North Africa and the Near East. This is achieved by defining a coverage area
consisting of an ellipse 10 E/W and 7 N/S centred on a point on the earth at 20 E and 31 N as seen
from a geostationary satellite. This is shown in Figure 4-2.
The second scenario is based on providing a global service using three satellites placed in
geostationary orbits at 10E, 140E and 100W.
Within these two scenarios, two approaches have been taken to the satellite channelisation. Both
approaches consider transparent or ‗bent pipe‘ designs in which the frequency of the call channels is
changed before re-transmission. In one case an analogue filtering and routing process (using SAW
filters) has been investigated, and in the second case a digital processing approach is taken where the
CDMA channels are digitally filtered and routed before being allocated to their destination down link.


4.2     Payload Designs to Meet the Requirements of the S-UMTS System

4.2.1 Design trade-off Evaluation
The initial satellite design activity was based on a generic design using a 10 m S-band reflector. As a
result from business case optimisation activities, the reflector design was extrapolated to a 15 m
diameter using 120 feed elements.

4.2.1.1 Sizing The Transponder
The traffic carrying capacity of the S-UMTS transponder is dependent upon two parameters. The first is
the total radiated power (EIRP) generated by the output amplifiers and the antenna and the second is the
number of through channels that the processing portion of the transponder is able to handle.
Because of the difficulty of generating RF in a space environment, the transponder traffic capacity must
be EIRP limited, in other words, the number of channels provided is much greater than the total EIRP
can support. Some attempt was made to assess the EIRP required per traffic channel and an
                                                                                                              26
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                     Revision
Final Report                                         2000-11-29               A


approximate figure was arrived at but this still under debate and is a function of the final application of the
system.
A figure of 240 Mb/s has been defined in WP 3100 as the required total data throughput of the
transponder. If the assumption is made that all of this traffic will be in the form of voice calls then the total
number of call channels required is 30,000. If it assumed that each 5 MHz satellite channel s is capable
of carrying 250 CDMA traffic channels then the number of 5 MHz channels required is 125.
The minimum number of beams required to cover the earth with an acceptable antenna gain is 160. It
was proposed that each beam should be allocated at least one 5 MHz segment of spectrum (channel)
with other beams allocated two segments and the beams facing the greatest traffic three. Since the
modular construction of existing analogue processors can contend with multiples of 16 channels the
value for the number of channels of 320 was decided upon. This gives sufficient for one spectrum
segment per beam and 160 segments that can be allocated to the busy areas.
Thus the maximum number of CDMA traffic channels that can be handled by the transponder is 80,000
or an equivalent total capacity of 640 Mb/s if each traffic channel is supporting 8 kb/s. If this is taken as a
maximum then the mean demand will be 266 Mb/s.
Both the global and regional coverage transponders were designed with this figure in mind.
When considering the use of digital processing techniques in the transponder it became apparent that a
fully regenerative processor was not a feasible proposition because of the weight and power
requirements. It was therefore decided to design a transparent system using the digital processing
techniques being applied in other mobile satellite systems.
Having made this decision the specification for the digital on board processor was made comparable to
that for the analogue, that is for 320 channels carrying 500 traffic channels each a total of 160,000.
For the analogue transparent transponder the capacity could be most conveniently reduced in size by
multiples of 80 channels and the power and mass required of the processor reduced pro rata.
In the case of the digital equivalent, the rules are not so simple as there is a significant overhead in
power and mass which is essential, no matter how many channels are dealt with.
The difference between the two systems is that a given number of beams to the mobile stations will be
associated with a gateway in the case of the analogue processing transponder whereas with the digital
processor there will be significantly greater flexibility in the allocation of channels.


4.2.1.2 Payload EIRP and G/T Performance at S-Band
The EIRP and G/T Performance of the various options are given below:

Performance of the Global Transparent Transponder:

                     Antenna diameter                                    7.33 m
                     Number of feed elements (helices)                   120
                     Number of beams                                     160
                                 Table 4.4 Rx / TX ANTENNA SIZES
                     Edge of coverage gain                               36.9 dBi
                     RF Output from each element                         22.2 W
                     Output network loss                                 3.0 dB
                     EIRP                                                68.1 dBW
                                 Table 4.5 TRANSMIT PERFORMANCE
                     Edge of coverage gain                 36.9 dBi
                     Receive Noise Temperature             27.5 dBK
                     G/T                                   9.4 dB/K
                                Table 4.6 RECEIVE PERFORMANCE
                                                                                                   27
Contracts number:
ESTEC - 13694/99/NL/US                          Final Report
ESA S-UMTS study                                Date                   Revision
Final Report                                    2000-11-29             A



Performance of Global digital Transparent Transponder:

                    Antenna diameter                              9.0 m
                    Number of feed elements (helices)             120
                    Number of beams                               271
                                Table 4.7 Rx / TX ANTENNA SIZES

                    Edge of coverage gain                         37.0 dBi
                    RF Output from each element                   22.2 W
                    Output network loss                           3.0 dB
                    EIRP                                          68.2 dBW
                              Table 4.8 TRANSMIT PERFORMANCE
                    Edge of coverage gain                  36.9 dBi
                    Receive Noise Temperature              27.5 dBK
                    G/T                                     9.5 dB/K
                                Table 4.9 RECEIVE PERFORMANCE

Performance of Regional Transponders (analogue and digital Transparent.):

                    Antenna diameter                              10.0 m
                    Number of feed elements (helices)              80
                    Number of beams                               157
                               Table 4.10 Rx / TX ANTENNA SIZES
                    Edge of coverage gain                         41.5 dBi
                    RF Output from each element                   22.2 W
                    Output network loss                            3.0 dB
                    EIRP                                          71.0 dBW
                              Table 4.11 TRANSMIT PERFORMANCE
                    Edge of coverage gain                 41.5 dBi
                    Receive Noise Temperature             27.5 dBK
                    G/T                                   14.0 dB/K
                               Table 4.12 RECEIVE PERFORMANCE


4.2.1.3                   Payload EIRP and G/T Performance at Ka -Band
There are two separate designs for the antenna for the global and regional coverage. The EIRP and G/T
performances are given below:

Global Coverage

                                                   Transmit                       Receive
      Reflector Diameter                           300.0 mm                       193 mm
      Feed Aperture                                51 0 mm                        33.8 mm
      Edge of Coverage Gain                         28.9 dBi                      28.8 dBi
                                                                                                         28
Contracts number:
ESTEC - 13694/99/NL/US                            Final Report
ESA S-UMTS study                                  Date                   Revision
Final Report                                      2000-11-29             A


      Number of Beams                                   16                             16
      RF Output per Beam                               20 W
      Output Network Losses                           1.5 dB
      EIRP                                           52.5 dBW
      Receive Noise Temperature                                                     28.5 dBK
      G/T                                                                           0.3 dB/K
                                    Table 4.13 global performance
Regional Coverage

                                                     Transmit                        Receive
      Reflector Diameter                             560.0 mm                       360.5 mm
      Feed Aperture                                   48 0 mm                        30.9 mm
      Edge of Coverage Gain                           34.5 dBi                       34.4 dBi
      Number of Beams                                    16                             16
      RF Output per Beam                               20 W
      Output Network Losses                            1.5 dB
      EIRP                                           58.0 dBW
      Receive Noise Temperature                                                     28.5 dBK
      G/T                                                                           5.9 dB/K
                                   Table 4.14 regional performance

4.2.2 The Analogue Transparent Global Coverage Transponder

4.2.2.1 The analogue transparent payload
The function of the S-UMTS payload is to provide communication services to mobile terminals. The main
aim is for the transmission of high digital data rates compatible with Multi-Media and Internet access to
larger mobile earth stations.
The technologies involved in the design have been chosen, as far as possible, for their heritage and are
developed from concepts used on such programmes as Artemis, Inmarsat 3, Horizons and EAST.
In the case of designs which have not previously been used an attempt has been made to identify the
state of the art in each area involved and to identify developments in associated technologies.
Key features of the payload design are:
 A reflector based S-band active antenna designed for stable sidelobe performance in the presence of
     phase errors over lifetime.
 Mechanically simple helix elements for the S-band mobile antenna feed overcoming the PIM
     problems associated with crossed dipoles and patches
 S-band solid state power amplifiers based on extensions of the design philosophy developed for the
     L-band amplifiers used on Inmarsat 3, to be used on Inmarsat 4 and proposed for use on EAST.
 Multiport amplifiers incorporating Butler matrices providing minimum impact of SSPA phase and
     amplitude errors on the antenna performance.
 The use of SSPA designs that are mass, power and linearity efficient.
   Combined S-band transmit and receive antenna providing high beam congruency on the
    ground.
In studying the requirements of the payload two major scenarios have been addressed.
Scenario 1 defines a system that will cover a particular geographical area or zone (mainly Europe, the
Near East and North Africa) and Scenario 2 which covers the visible area of the earth so that 3
spacecraft will provide world wide coverage.
                                                                                                         29
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


This payload description is constrained to global coverage that will provide communications to all mobile
users for which the satellite elevation is greater than 20º from satellites positioned at 20º E, 140º E and
100º W.
These coverage‘s are achieved with multi-beam antennas operating in the frequency bands defined by
the ITU for the S-UMTS system; S-band for the links between the mobile earth station (MES) and the
satellite and Ka-band between the satellite and the Gateway Station (GS).
The actual frequency bands allocated are:
         BAND                      PATH                                   FREQUENCY RANGE
         S BAND                    Spacecraft to MES                      2170 to 2200 MHz
                                   MES to Spacecraft                      1980 to 2020 MHz

        Ka BAND                  Spacecraft to BS                   17.7 to 20.2 GHz
                                 BS to Spacecraft                   27.5 to 30.0 GHz
                                Table 4.15 Allocation of frequency band

As well as these two major approaches, the study has also addressed two types of transponder within
the scenarios, the transparent or ―bent pipe‖ and the digital processing transponder.
In the transparent transponder the S-band allocation is divided into 5 MHz bands or spectrum segments
(channels) which separate the sets of traffic channels for individual amplification and direction to the
appropriate beam and its associated coverage area.
 In the final performance analysis it has been assumed that the traffic channels use spread spectrum
techniques and are transmitted in code division multiple access (CDMA) so that a number of users can
access the same 5 MHz segment of spectrum.
 In the transparent case the user signals are amplified and retransmitted without modification hence the
term ―bent pipe‖.
The use of CDMA and coding permits the development of another type of transponder, one in which the
signals received by the spacecraft are decoded (de-spread in this case) before being re-spread and
routed using digital techniques to the appropriate beam via a beam forming network (BFN). A further
approach is where digital techniques such as analogue to digital conversion, digital filtering and digital
multiplication are used to isolate individual segments of the spectrum (5MHz channels) and to route
these to specific beams after digital to analogue conversion.
This description of the transponder payload will be aimed mainly at the global coverage; analogue
transparent transponder with additional sections aimed specifically at the other three design approaches.
An outline block diagram of the transponder is shown in Figure 4-4.
                                                                                                                               30
Contracts number:
ESTEC - 13694/99/NL/US                                    Final Report
ESA S-UMTS study                                          Date                         Revision
Final Report                                              2000-11-29                   A


                                            Automatic Level Control
                                                                                 Pilot Tone
                                                                                  Receive               S-Band      S-Band
                                      Feeder                          Ka-Band       Unit                Rx/Tx       Rx/Tx
                                        to                            Downlink                           Feed      Reflector
        Ka-Band                       Mobile                                                             Array
         Rx/Tx
        Antennas
                                                                                 Beamformer
                           Ka-Band                                                                120
                                                       Forward IF                 & S-Band
                    16     Payload    16                                 160
                                                       Processors                  Payload
                           Receive
                                                                                  Transmit
                           Section
                                                                                   Section




                                                                                 Centralised
                                                                                 Frequency
                    16                  Mobile
                                          to                                     Generator
                                        Feeder
                                                                                         LOs


                                                                           160   Beamformer       120
                           Ka-Band         16
                                                                                  & S-Band
                           Payload                      Return IF
                                                                                   Payload
                           Transmit                    Processors
                                                                                   Receive
                           Section                                                 Section



                                                                                                                 120
                                                                                  Pilot Tone
                                                                                   Injection
                                                                                      Unit


Figure 4-4 OUTLINE BLOCK DIAGRAM

4.2.3 The Digital Transparent Global Coverage Transponder
The digital transparent global coverage transponder shares the RF equipment arrangement with the
analogue transparent equivalent. The main differences are the S-band antenna which uses a total of 271
beams generated from 120 feed elements and within the Ka-band transmit section where the TWTA‘s
are grouped as Multiport amplifiers with 4X4 Butler matrices combining the out puts of four TWTA‘s.

4.2.4 The Transparent Regional Coverage Transponder
The transparent regional coverage transponder uses the exactly the same technology and layout as the
global coverage transponder. The main differences being the antennas which must meet the requirement
of scenario 1 of the S-UMTS system.
The region covered is defined from a satellite positioned at 20 E and is the projection on the earth‘s
surface of and ellipse 10 E-W and 7 N-S centred on a point on the earth 20 E and 31 N.

4.2.5 The Digital Transparent Regional Coverage Transponder
The digital transparent regional coverage transponder uses the same technology as that described for
the global digital transparent system. The major difference is in the design and performance of the
antennas to provide satisfactory coverage of the defined region.
The region concerned is an ellipse 7 N-S by 10 E-W centred at 20 E and 31 N viewed from a
geostationary satellite stationed at 20.

This equipment is identical to that used on the global digital transponder. However, there are major
differences in those equipment‘s which provide regional coverage such as the Ka-band antenna and the
S-band antenna.
                                                                                                          31
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


4.3    Capacity
One of the important issues, when elaborating the business plan for an S-UMTS system, is the capacity
of the system, in order to get an upper (physical) limitation of the addressable subscribers base.
In fact, if maximum loading is reached, then revenues are directly proportional to the sellable capacity of
the system (i.e. from a service provider's point of view what can finally be sold to the customer).
In case maximum loading is not reached, capacity is an important parameter for marketing and pricing
issues. Actually, revenues can be maximised given the effective load of the network by studying the
margin between the actual value and the maximum achievable load (capacity). If load is low, prices can
be lowered in order to boost subscriber base. To the contrary, if load is high (close to system overload)
prices can be rose in order to increase revenues.

Nevertheless, capacity is not an intrinsic value of a telecommunication system but is dependent on a
number of parameters such as:
- traffic repartition
- user profiles (resource usage along time)
- terminal performances
- required quality of service (BER, FER, blocking probability, interruption probability, etc...)

As several methodologies can be adopted when assessing capacity for a multiple access
telecommunication network, we will consider in this document two of them:
1. statistical link budget taking into account the number of users in the system
2. dedicated software tool simulating the considered network access method (―Capacity Analysis Tool‖)

4.3.1 Traffic Scenarios
As 3G networks should not start commercial service before 2002 in Europe (2001 in Japan), no operator
could presume about expected traffic for those future networks.
Existing models (built on some key parameters as population density, GNP, fixed-lines penetration) from
traditional telecommunications world _such as fixed-data or mobile-voice networks_ can be adapted but
                                                           rd
no background experience exists for this new generation -3 G- of mobile networks carrying data.

Two traffic models have been envisaged for this study:
1. uniform traffic repartition on the service area (with an optional segmentation by country)
2. detailed traffic scenario based on classical parameters as described here above

Data for this second model proceed from Telenor traffic scenario studies performed in the scope of
WP4200.
Horizontal market potentials for S-UMTS are envisaged for every part of the world, basing on the
targeted population, itself derived from the total population of the considered country scaled with the
GDP per capita of the country.
Values of global traffic requirements per country are converted in geographical traffic (per km², for
instance) for each location of the considered service area. This is done thanks to a mapping of the
service area in 1° latitude / 1° longitude cells.
These values are strongly dependent on the market segments addressed by the marketing plan and
should thus be refined once the S-UMTS market is accurately modelled.

For this capacity assessment, we have considered the regional scenario in which one satellite covers
Western and Eastern Europe, Middle East and North Africa. The resulting capacity figures are thus given
for one satellite. For the global coverage also considered as a possible S-UMTS scenario, the resulting
capacity would be obtained region by region (i.e. on a per satellite basis).
                                                                                                          32
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


4.3.2 Capacity Analysis Tool (CAT)
In the course of the study, a software tool (CAT) was developed to perform a detailed analysis of
capacity in a CDMA system, taking into account real antenna patterns and random user positioning. The
results obtained with this tool were somewhat lower than had been expected from simplified Excel
spreadsheet methods. Numerous investigations were carried out on the tool, but no error was detected,
though due to the difficulty of defining a test case which could rigorously be applied to detect errors, this
does not imply that no error exists. The investigations carried out on the CAT indicate that if an error
exists, it is within the calculation of interference modules.
ESTEC offered to perform an analysis of the Astrium antenna design using their own tool, CELIA
(Capacity EvaLuation and Interference Analysis). The results given by CELIA are more in line with the
results predicted by Excel, and are also coherent with comparative analyses performed on other studies.
This tends to indicate that an error exists within the CAT tool. Further work is therefore required on the
CAT tool to correctly identify the origin of the problem and correct it, along with other improvements in the
tool which have been identified as being necessary for a full-blown capacity analysis tool.
For this reason, the results presented here are based on Excel spreadsheet results, following correlation
between the CELIA analysis, which was used to examine the case of a single terminal type, and Excel.
                                                                                                                                                                      33
Contracts number:
ESTEC - 13694/99/NL/US                                                                     Final Report
ESA S-UMTS study                                                                           Date                              Revision
Final Report                                                                               2000-11-29                        A




4.3.2.1 Uniform traffic
Considering the case of a uniform traffic over the service area permits to reach the upper bound of the
system capacity. Uniform users repartition permits to statistically minimise average C/I and therefore
maximise the number of users that can be provided with the service in the system. To the contrary, an
uneven traffic repartition results in poorer capacity results.

4.3.2.2 Non-uniform traffic (from prognosis)
We have performed simulations for a single terminal type (laptop) for both 8 kbps and 64 kbps services.
Results are strongly dependent on the number of simulations performed. As a matter of fact, the more
users positioning are simulated the more accuracy is obtained. For reasonable CPU-time usage, we
limited our simulations to 100 user positioning drawings.
Due to the error identified in the CAT software tool, results for non-uniform traffic distributions are not
reproduced here.


4.3.3 Overall capacity
The overall capacity has been derived from correlation between Excel analyses and the CELIA tool. A
comparison of power vs. capacity for 8 kbps services to a park of Palmtop terminals is shown in Figure
4-5.


                                                                    Power Use

                               1000.0


                                900.0

                                800.0


                                700.0
  Satellite RF Power (Watts)




                                                                    CELIA :
                                                                    8020 to
                                600.0
                                                                    8370 users                                                             ASTRIUM Excel Tool Results
                                                                    @ 440 W                                     Excel :                    ESTEC Excel Tool Results
                                500.0                                                                           9350 users
                                                                                                                                           2-block use
                                                                                                                @ 440 W
                                                                                                                                           Various CELIA runs
                                400.0


                                300.0

                                                                                                       Excel :
                                200.0
                                                                                                       8020 users @ 290 W
                                                                                                       8370 users @ 320 W
                                100.0


                                  0.0
                                        0.0       2000.0   4000.0     6000.0        8000.0        10000.0       12000.0          14000.0
                                                                         Number of Users



                                              Figure 4-5 : Comparative Analysis of CELIA vs Excel for Palmtop Terminal

The correlation results shown above do not include a number of other effects : the CELIA tool uses
directivity data rather than gain data; it does not include degradation of the overall link by the feeder link;
and it does not include the power robbing effect present in transparent repeaters due to the amplification
of satellite front end noise. Such considerations can be readily incorporated into Excel link budgets,
                                                                                                              34
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                    Revision
Final Report                                        2000-11-29              A


though in the example above these effects have been removed from the Excel results so as to provide a
like-with-like comparison with the CELIA tool.

However, the above comparison demonstrates that real capacity is lower than predicted by Excel, which
is due to a breakdown in the simplifying hypotheses used in the Excel link budgets. To achieve a
capacity in the region of 8000 users, the CELIA indicates that an RF power of 440 watts is required,
whereas Excel indicates that around 300 watts is sufficient (a factor of 1.4 difference). At higher capacity
levels, the difference can be even greater.

Following considerable discussion both internally and with ESTEC, it was decided to use a peak-to-
average RF power factor of 2 for all estimations of capacity with Excel budgets. Based on the above
comparison, this figure probably results in lower capacities than could be obtained in reality.

For a 3G system, it is also important to account for the gain, in terms of the number of users served, that
is to be had from the utilisation of packet switching with respect to that of a circuit switched system. The
number of users that can be simultaneously served by a packet-switched system is derived from the
performed simulations whose results are reported in WP3000 and from packet-switched to circuit-
switched capacity ratio calculated in WP3400, in accordance with principles outlined in ETSI S-UMTS
documents.
For 8 kbps transmission, the capacity gain is supposed to be around 3, while it is estimated about 12 for
64 kbps transmissions.

In accordance with these results, the average number of users that can be simultaneously served based
on a single service with a single terminal type for a uniform traffic repartition is presented in table 4.16.
The results in the forward direction are based on Excel values after correlation with the CELIA tool. They
incorporate feeder link degradation (0.5 dB), antenna loss factor of 1 dB, and also include power robbing
effect (which results in poor results for handheld terminal) :

   Terminal           Service          Forward              Return           Forward             Return
                                       average             average            average            average
                                       channels            channels        simultaneous simultaneous
                                                                               users              users
Handheld              8 kbps               582              5462                1746              16386
Palmtop               8 kbps             11058              6789               33174              20367
Palmtop              64 kbps              1358               206               16296               2472
Laptop                8 kbps             15908              6832               47724              20496
Laptop               64 kbps              1940               262               23280               3144
Vehicular             8 kbps             12222              8610               36666              25830
Vehicular            64 kbps              1940               912               23280              10944
  Table 4.16: single service / single terminal / uniform traffic for baseline satellite design (97-beam, 10
                                    metre reflector, 80 feed elements).

When introducing a detailed traffic scenario with global traffic requirements per country, we noticed
degradation of the order of 80 % with the CAT tool. The degradation would be the same for all the
services and all the terminals. These figures remain to be confirmed.
                                                                                                                    35
Contracts number:
ESTEC - 13694/99/NL/US                                 Final Report
ESA S-UMTS study                                       Date                       Revision
Final Report                                           2000-11-29                 A


4.3.3.1                     Capacity improvement results
In an initial capacity estimation, the number of users that could be served by the system appeared to be
quite low, and the associated business case was very pessimistic. We therefore modified a number of
parameters, including pushing the satellite design to the limit of realistic possibilities, given the baseline
constraints which had been imposed. The modifications were :
                            Use of 2 blocks of frequencies per beam
                            Increase of the satellite antenna diameter to 15metres to increase the
                               number of beams within the coverage area
                            Increase of number of radiating elements (SSPAs) to 120
The capacity analysis of such modifications were undertaken using Excel spreadsheet methods,
following the previously described correlation between CELIA and Excel. This approach, which assumed
a scaling up of the antenna performances (e.g. unchanged C/I performance), avoided the need to reenter
a complex and long antenna redesign phase.
The results obtained by this approach for the "Enhanced" satellite design are produced below.


 Terminal           Link Data Rates      Forward Average         Forward Average Users              Capacity
                                            Channels                    Served

                                      132 beam, 15m reflector,   132 beam, 15m reflector,    132 beam, 15m reflector,
                                           120 elements               120 elements                120 elements
                                             2-blocks                   2-blocks                    2-blocks
 Handheld               8 kbps                 N/A                        N/A                         N/A
 Palmtop                8 kbps                 22440                      67320                     180 Mbps
 Palmtop               64 kbps                 2904                       34848                     186 Mbps
 Laptop                 8 kbps                 34056                     102168                     272 Mbps
 Laptop                64 kbps                 4488                       53856                     287 Mbps
 Vehicle                8 kbps                 27720                      83160                     222 Mbps
 Vehicle               64 kbps                 4224                       50688                     270 Mbps




 Table 4-20 System capacity estimation with satellite improvement (132 beam, 15metre reflector).
                                                                                                         36
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A




5 Business assessment
5.1    Analysis of the competitive environment

5.1.1 The S-UMTS opportunity
The factors likely to support or drive the deployment of S-UMTS coverage include:

Demand for Universal Mobile Telecommunications System (UMTS) coverage outside urban areas.
Most operators plan to start initially with limited UMTS services, with a focus on urban centres, and
gradually move to total coverage. In effect, there will be UMTS islands supported by Global System for
Mobile (GSM) networks in surrounding areas.

High cost of Terrestrial UMTS (T-UMTS) infrastructure. The build-out of T-UMTS infrastructure will be
many times more expensive than any second-generation GSM evolution; it will require new base stations
and smaller cells to support the higher data rates. Implementing new mobile services in areas previously
having no coverage is cheaper and quicker with S-UMTS than with T-UMTS (or other terrestrial radio
systems), principally because the satellite signals are already there.

High cost of T-UMTS licences. UMTS license auctions have driven prices so high that some analysts
are unsure whether operators can afford to build the infrastructure needed to provide the service in all the
planned areas. For operators having spent these large sums on licences, S-UMTS may be an alternative
to provide fast coverage and generate revenue streams early, before the planned T-UMTS can be
implemented. Higher prices and lower usage in the beginning may make satellite feasible while terrestrial
infrastructure is put in place.

Synergies with T-UMTS. S-UMTS will be able to support and create synergies with the development of
terrestrial UMTS by means of allowing terrestrial network operators to offer extended, even world wide,
service coverage for their subscribers.

Complementary coverage for car terminals. The car industry is on the verge on a revolution when it
comes to the possibilities offered by Internet access in moving cars. The drive to get optimal systems in
place by all major car manufacturers represents a major opportunity for S-UMTS, especially as a
complementary solution to T-UMTS.

Political pressure to make UMTS available. Conscious of the global advantage that Europe has
enjoyed up to this time in mobile technologies, and therefore of the need to get European Third-
Generation (3G) services up as early as possible, the EU has requested all member states to introduce
UMTS services no later than January 1, 2002. Most EU members will call for bids during 2000, and
winners of UMTS concessions will be obliged to place orders for equipment by the end of 2000.
                                                                                                        37
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                   Revision
Final Report                                       2000-11-29             A




5.1.2 Factors influencing the competitive environment
The competitive environment for S-UMTS is influenced by the following:

Deregulation for increased UMTS competition. A key objective of most regulators is to try to increase
competition to ensure rapid low-cost deployment, avoid simply handing UMTS in its entirety to incumbent
GSM operators, and establish conditions that will result in innovative and creative service provision.
Some countries intend to award at least five licenses, or award both national and regional licenses, so as
to ensure that the market is not dominated by incumbent operators. Others have created auctions that
allow new entrants to bid and defeat incumbents. In the roaming area authorities must balance the need
to encourage network rollout against the need not to allow existing operators to continue dominating by
virtue of their existing infrastructure ownership.

The positioning of global players. The presence of ever larger operators is shifting the competitive
scene as UMTS moves up on the agenda. The right calibre S-UMTS would undoubtedly fit well with the
global player strategy, and may be even more attractive to the smaller competitors, which may need S-
UMTS to fill in the gaps.

The arrival of new players. The IT, telecom and media scene is undergoing rapid changes. In addition,
new players may include unexpected ―invaders‖, unanticipated at this time. Examples from allied fields,
such as the piggybacking of telecommunication channels on electric power lines, indicate that invasions
may come from any sector.

Full focus on the ground (on terrestrial UMTS). All established mobile operators are currently planning
GPRS implementation and at the same time positioning themselves for the T-UMTS licences. And when
a T-UMTS license is won, an operator must begin building the network as fast as possible to pay its high
licence costs. To the mobile vendors, the need to have a solution available as soon as possible means
full focus on T-UMTS over the next two years. Even if there is synergy between T-UMTS and S-UMTS,
any disturbance to their main business case is risky.

Alternative technology developments. Ultimately, the aim of UMTS is to provide end users with
wireless access to the widening range of e-commerce services, entertainment, and information that is
being made available on the Web. However, UMTS has no monopoly in an area. It's clear that GPRS,
WAP, Bluetooth, and the dedicated mobile portal, all of which will be deployed before UMTS, can provide
much of what UMTS can offer.

Mobile satellite systems. Currently there are a handful mobile satellite systems (MSS) with plans to be
operational in the coming years: Globalstar, ACeS ICO/Teledesic, Thuraya and Inmarsat, which were
reviewed in the WP 2000 part of this Study.

It remains to be seen how strong these players‘ position is in 2005, but with the current performance of
MSS, it may become difficult to convince investors that significant amounts of additional capital should be
spent on investing in system upgrades towards 3G capabilities.

The following areas were considered of importance when assessing the future of MSS:
                                                                                                               38
Contracts number:
ESTEC - 13694/99/NL/US                                  Final Report
ESA S-UMTS study                                        Date                   Revision
Final Report                                            2000-11-29             A


MSS survivals in 2005
The following matrix outlines the possible status in 2005 of the MSS players currently on the scene.

System              Current situation:                    Current situation:              Likely status in 2005
                    strengths                             weaknesses
Iridium             100% global coverage, first           Financially bankrupt, short     Not here
                    to market, few ground                 lifespan, low capacity,
                    stations                              system complexity
Globalstar          Greatest capacity                     Not true 100% global            Still niche provider.
                                                          coverage, delays / launch       Needs system upgrade
                                                          schedule, partial service       to data to be viable,
                                                          launch, slow uptake in          lacks the capital to do
                                                          sales                           so?
ICO / Teledesic     Longest satellite life, financial     No US license, late to          Market uptake slow,
                    backing, gradual build ICO            market, potential look-         may be struggling to
                    then Teledesic                        angle issues.                   justify spendings.
ACeS                Lowest cost for users                 Time delay, no back-up.         Needs to upgrade for
                    (handsets and minutes),               Does not support data,          data for long term
                    smallest handsets                                                     viability
Thuraya             Middle eastern backers and            Not yet in the market           Upgrades needed for
                    markets.                                                              data, but sufficient bus
                                                                                          case from voice
Inmarsat            Profitable already today.             Limited capacity.               Strong presence in all
                    Upgrades planned.                                                     profitable niche
                                                                                          markets.


5.1.2.1                    Alternative applications
Some industry analysts suggest that the best case for MSS is to shift focus to fixed telephone services,
thereby tapping into the huge potential for pay phone services in less developed but heavily populated
areas of the world.


5.1.2.2                    Impact of a postponed decision by the European Council of Ministers
In the discussions leading to EU‘s European Council of Ministers decision to postpone its mobile satellite
phone frequency decisions for two years, it was stated that EU officials doubted whether any EU country
is interested in pressing for a satellite phone frequency agreement.

For a continent (as Europe), there is precedent. In the 1980s, Inmarsat became operational with three
overlapping satellite footprints covering the globe to serve maritime traffic in the Atlantic, Pacific and
Indian Ocean Regions. Together, the three footprints also covered the land areas of the globe, save for
the Poles. At the time, Inmarsat supported no land mobile services. Accordingly, some larger countries
protested the Inmarsat frequency allocations, which they saw as unfairly wasted for their large land
masses. For them, the frequency band allocations for satellite services robbed them of a scarce resource
which otherwise might have been available for terrestrial systems. Indeed, the politics of scarcity have
become a major issue in the development of radio communications systems.


5.1.2.3                    Inmarsat 4
Inmarsat already has ―S-UMTS-like‖ services in MobMan and M4. The possibility of collaboration may
need to be discussed further. Inmarsat-B has supported 64 kbit/s/56 kbit/s since ca. 1996, so Inmarsat
                                                                                                         39
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


has experience with and customers for it. New broadband systems are being developed and schedules
for their implementation have been set.


5.1.2.4                    VSAT technologies
VSAT may be the principal competition, both technically and commercially. There are many VSAT
manufacturers world wide, and even dealers in used VSAT equipment, such as antennas. Throughputs
can be high. Systems are now being implemented to extend Satellite News Gathering (SNG) techniques
to simpler, cheaper terminals operating at lower speeds. As most VSAT systems operate within national
boundaries, they need no international agreements (aside from frequency allocations) to operate. So one
might view VSAT as a form of ―pre-S-UMTS‖. The drawback, of course, is that the higher operating
frequencies at C band and Ku (future Ka) bands are not as amenable to hand-held terminals as are
frequencies at L band.



5.2    Assessment of key success and risk factors

5.1.3 Key success factors
The following are factors key to the success of S-UMTS:

Timing. Most operators believe that true 3G services will not emerge until 2003. Commercial launch of
S-UMTS should take place no later than 2005 in order to take advantage of the growing market
awareness likely to have evolved in the second fully operational UMTS year. By 2005 it is reasonable to
expect increased availability and variety of terminals and applications, reducing the risk of slow uptake
often found in the very early stages of new systems launch.

Sufficient demand. Demand for S-UMTS services will depend heavily on the uptake and success of the
T-UMTS implementation, which again will depend on content and applications.
The two year period from the UMTS launch should be sufficient to generate an attractive offering of
UMTS content for users in areas outside terrestrial coverage.

Three key factors support the belief that the demand for UMTS will be strong enough to warrant the
added coverage of S-UMTS:
 Telecom forecasts have for the past years consistently underestimated actual demand;
 The biggest players spending this much money on licences indicates the huge potential;
 High-bandwidth multimedia mobile access to the Internet is imminent. UMTS is how.

Competitive satellite system solution. The solution must have attractive capacity, dynamic coverage,
and allow services at a competitive price close to or the same as for T-UMTS. To realise the potential of
S-UMTS, one or more terrestrial operators need to be convinced of the benefits of satellite solutions. This
is a success factor requiring a major effort in terms of human and monetary resources.

Reuse of T-UMTS service networks and applications. S-UMTS needs to be as compatible as possible
with T-UMTS to permit synergy through the reuse of T-UMTS service networks and applications. It is
necessary to start the plans for the S-UMTS system as an integral part of an upcoming T-UMTS
implementation. A separate stand-alone S-UMTS will not work.

Terminals. It is vital that the S-UMTS system gives the same ‖look and feel‖ for the user as the terrestrial
version of UMTS. The user should through a virtual home environment have access to the same services
                                                                                                            40
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                     Revision
Final Report                                       2000-11-29               A


as experienced in his normal environment in the terrestrial network. In a multi-systems world,
manufacturers will need to develop multi band (including S-UMTS), multimode handsets to enable global
roaming during the initial implementation of 3G. Such handsets will be more expensive than equivalent
second-generation (2G) handsets.

Applications. Content and applications are key areas for whether UMTS becomes a success, and an
important factor for satellite access is the availability of as many of the T-UMTS functions as possible.

Funding and strategic partners. For any new S-UMTS solution to get off the ground the business
model will be a key factor for success or failure. As seen in the case of several MSS insufficient funding
can mean delayed time to market and limitations on funds for marketing.


5.1.4 Risk factors
Too late to market. The time frame 2003 – 2007 outlines the main rollout years for UMTS. In this period
it is perceivable that GPRS will capture the users outside the area covered by T-UMTS. However, by
2005, S-UMTS still has the opportunity to capture market shares from the gradually built out GPRS. The
later commercial services get started, the greater the risk for demand already being met by alternative
solutions.

Market growth slowing. While all forecasts only show upwards sloping curves for the years ahead, it is
feasible that for some unanticipated reason the explosive growth in cellular will taper off and slow down
in the next 5 to10 years.

S-UMTS capabilities not competitive. It is important that the quality of service offered by S-UMTS is at
an acceptable level in relation to users expectation levels. Ever more user-friendly terminals and
solutions continue to be developed for terrestrial terminals, and the gap between these and the satellite-
able terminals may become excessive. Of critical importance is the question of how the users in 2005 will
accept a terminal with no in-house penetration. Another key issue is capacity. By 2005 user expectations
will have grown to require higher levels of quality of service, and to be less patient with capacity problems
if the system is congested.

Market and investors sceptical after Iridium failure. The current status for mobile satellite systems will
continue to have a negative effect on investors and potential users well into the future.

Spectrum lost to T-UMTS. In light of satellite being perceived as a communication technology best
suited for niche applications requiring very limited capacity, there are forces who argue for freeing the S-
UMTS frequencies so that these may be used for T-UMTS instead.

Development costs higher than budgeted. Going over budget before system launch means additional
financing is needed, often at a very high cost. Such overruns also dig into allocations for operations,
marketing and sales needed in the early stages.



5.2    Barriers to entry
                                                                                                          41
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


5.2.1 Frequency
Today it would appear that an S-UMTS system such as is suggested in this study would be limited to 10
or possibly 15 MHz spectrum in Europe. Such a low level of spectrum limits the capacity and this
consideration has to be incorporated into the business case. Additionally, the impact of frequency co-
ordination needs to be taken into account.
A sound business case should rest upon a detailed analysis of the various aspects of frequency
discussed above. In particular, frequency co-ordination, fixed service use, and spectrum availability
issues should be explored further. Such matters are beyond the frame of this Study.


5.2.2 Licensing
The allocated spectrum for S-UMTS (or the satellite component of IMT-2000) is limited to 2x30 MHz in
Europe and most parts of the world. Consequently, the use of the spectrum must be regulated to ensure
judicious licensing procedures.

Licensing practices vary from country to country, so a satellite operator may face a formidable task in
acquiring the licences necessary to provide services within the footprint of a satellite. For 2 GHz MSS
(including satellite IMT-2000), the US Federal Communications Commission (FCC) proposed four main
spectrum assignment options (Report no. IN 99-12, March 1999). The options were:
 A flexible band arrangement: 2.5 MHz of uplink and downlink spectrum given to each system, group
     systems in segments based on the particular technology used, and provide expansion spectrum
     between the assigned segments for additional system requirements.
 A negotiated entry approach: Licence all the applicants across the entire band and leave it to them to
     co-ordinate their operations with FCC being available to resolve disputes.
 A traditional band arrangement: Divide spectrum equally among the applicants.
 An auction of licences: Could be done if none of the preceding three options are viable.

Some of these options could be used outside the USA also. However, an auction for S-UMTS spectrum
may be inadvisable, as it will increase the initial cost of a system and limit the competition.




5.3    Analysis of the market mix

5.3.1 Prognosis for mobile coverage penetration
The study estimates for mobile coverage penetration are based on a combination of internal and external
sources, and are set conservatively to reflect the uncertainty of market estimates so far into the future. As
shown in Fig. 4.1, the penetration levels presented by the UMTS forum, as well as those presented in the
most recent consultant reports from OVUM and Strategy Analytics, indicate a steeper growth curve than
what is used in the study.
                                                                                                         42
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A




5.3.2 S-UMTS market scenarios
Conclusions from a workshop half way in the study recommended continuing the work on two business
concepts, defined below and referred to elsewhere in WP4000 as scenario 1 and scenario 2.

Scenario 1
The scenario 1 business idea entails offering a mobile broadband service in areas with no terrestrial
UMTS. Compared to T-UMTS, it
 offers the same services and applications as T-UMTS,
 offers GPRS service in areas outside GPRS coverage,
 offers a car/indoor system,
 offers UMTS specific car and broadcast applications, and
   has the same tariffs or slightly higher.

In countries with no T-UMTS service, S-UMTS will be the initial UMTS service. This can be either through
a local GPRS service provider or a global operator.

Scenario 2
Scenario 2 entails offering a mobile broadband service to end users who are willing to pay high tariffs for
services provided outside T-UMTS coverage areas or for other specific services.

Scenario 2 comprises an executive mobile broadband service for a niche market. It:
 Offers a car/indoor system
 Offers UMTS specific car/broadcast/++ executive applications
 Offers UMTS services outside T-UMTS coverage at high tariffs


5.3.3 Coverage and penetration of the two market scenarios.
Coverage will be 100% within the footprint of the satellite providing S-UMTS.
                                                                                                                                                        43
Contracts number:
ESTEC - 13694/99/NL/US                                                       Final Report
ESA S-UMTS study                                                             Date                                       Revision
Final Report                                                                 2000-11-29                                 A



In the complementary scenario 1, the penetration of S-UMTS depends much on coverage and
penetration of GPRS and T-UMTS. As an example, Fig. 4.2 shows how Norway in 2005 –2006 is
expected to be provided with high coverage of GPRS and smaller ‖islands‖ of T-UMTS.




The study has derived the following total penetration prognosis for scenario 1.
                                2000      2001   2002      2003      2004          2005          2006        2007        2008         2009      2010
Penetration of people in area of coverage
   GSM speech                   32 %      43 %   53 %      62 %      69 %       75 %           79 %           82 %        84 %         84 %      85 %
   GPRS                                    2%     6%       13 %      24 %       38 %           52 %           65 %        74 %         80 %      83 %
   T-UMTS                                                   1%        2%         4%             8%            13 %        17 %         22 %      25 %
   S-UMTS                                                                     0,60 %         0,85 %         1,17 %      1,43 %       1,61 %    1,64 %


These estimates include areas without GPRS / UMTS coverage, assuming 10% penetration in these
areas compared to the levels in the areas with terrestrial coverage, it also includes traffic generated by T-
UMTS subscribers travelling outside terrestrial coverage areas.

In scenario 2, the penetration of S-UMTS has been set to 1/100 of T-UMTS, delayed by two years. This
gives the prognosis for the penetration in areas with coverage.

                                  2000    2001      2002          2003      2004          2005      2006         2007        2008       2009     2010
Penetration of people in area of coverage
   GSM speech                     32 %    43 %      53 %          62 %      69 %        75 %         79 %        82 %         84 %      84 %     85 %
   GPRS                                    2%        6%           13 %      24 %        38 %         52 %        65 %         74 %      80 %     83 %
   T-UMTS                                                          1%        2%          4%           8%         13 %         17 %      22 %     25 %
   S-UMTS                                                                             0,01 %       0,02 %      0,04 %       0,08 %    0,13 %   0,17 %
                                                                                                         44
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                   Revision
Final Report                                       2000-11-29             A


5.3.4 Services, applications and terminals
The same applications and usage as described for T-UMTS are valid for S-UMTS in scenario 1. The
same applications are also valid in scenario 2, but usage will be different because of higher tariffs.

High volume data applications are not expected to suffer greatly from the reduced quality of service in S-
UMTS. The traffic volume of S-UMTS will probably not deviate very much from T-UMTS. Most handheld
terminals will be dual-mode GSM/UMTS and voice services are not expected to be essential for neither
T-UMTS nor S-UMTS. Multimedia services such as video conferencing are expected to be less user-
friendly in S-UMTS because of greater delays and lower data rates than in T-UMTS.

An Ericsson prognosis for type of services, average usage, average sessions and percentage users for
T-UMTS in 2005 is summarised in Table 4.10. This prognosis is a result of discussions with a number of
operators and service providers.

Table 4.10 Supplemented Ericsson prognosis.
Application                kBytes per session,            Sessions per day,           Percentage users
                           average                        average
E-mail without attachment  5                              5                           100
E-mail with attachment     250                            5                           100
www.laptop                 500                            5 x 5 min.                  80
www.minibrowser            40                             10                          30
Real time video            5000                           0.3 x 20 min.               25
Distribution/file transfer 1000                           1 x 3 min.                  25
Information push           10                             10                          15
E-commerce                 1                              0.3                         50

An extended version of the table to include prognoses for voice, audio streaming, video streaming,
broadcasting and location based services is needed for a complete picture, but similar data from
discussions with service providers are not available for these services.


5.3.4.1 Vehicular system services
The vehicular market can be divided into three segments with different usage of applications:

Professional vehicular market
1. Company intranet access (mail with attachments, multimedia)
2. Message services (broadcasting, mail)
3. Internet and web.browsing (www.laptop, www.minibrowser, e-commerce)
4. Entertainment (audio streaming, video streaming, information, web browsing)
5. Navigation and maps (location based services, information) 6. Customer service and remote diagnosis
(information, file transfer)

Business travellers
1. Company intranet access (mail with attachments, multimedia)
2. Internet and web.browsing (www.laptop, www.minibrowser, e-commerce)
3. Entertainment (audio streaming, video streaming, information, web browsing)
4. Navigation/maps (location based services, information)
5. Customer service and remote diagnosis (information, file transfer)
                                                                                                          45
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


Mass market
1. Entertainment (audio streaming, video streaming, web browsing)
2. Information (information push, broadcasting)
3. E-commerce
4. Navigation and maps (location based services, information)
5. Customer service and remote diagnosis (information, file transfer)

Navigation, customer service and remote diagnosis may be important applications, but they are expected
to be seldom used and will generate little traffic. The time the drivers will be in or around their vehicle
differs considerably for the three market segments, so the generated traffic volumes also will differ.


5.3.4.2 Broadcast services
Standardisation of broadcast-services for T- UMTS are planned for UMTS Phase II in 2001 with a
possible rollout in a 2005 timeframe. The point-to-multipoint service types described below may exist in
different combinations depending on application served.

Location based services       Location based services directed towards specific business locations with
                              broadcast message filtering based on actual terminal position (within some
                              practical granularity). As an example, decision on which frames are
                              relevant within a broadcast multiframe would depend on your location
                              within a beam.
Multicast Services            Service aimed at closed user-groups or membership based groups for
                              distribution of specific information (company Newsletters, music files, ISP
                              logon pages, WEB server broadcasting, etc.)
―Thin client‖ Information     Broadcasting of news, financial data, sports, traffic, weather etc. using mini-
push                          browser approach, WAP etc. to lower needed bandwidth.

Traffic generating            The broadcast of different ISP‘s / businesses homepages with some form
broadcast services            of attractive presentation to the user. Broadcast of "tickers" or news-
                              headlines especially designed to trigger the interest of users. Both network
                              operators and content providers might be interested in supporting
                              financially such a service for S-UMTS.

Success factors for broadcast services over S-UMTS
Staying as close as possible to the emerging terrestrial standards is a major guideline.
The ―feel and look‖ equality (characteristics) of applications and services between the satellite and
terrestrial based access.
Terminal costs and development effort, as synergies with terrestrial terminals (mass market) are crucial
in the relative small satellite market.
The use of low traffic periods for bulk broadcasts combined with some minimum continuous service.
Special attention should be given to the ―thin client‖ type of services that mainly would address the hand-
held and palmtop roaming users.
An attractive broadcast channel will enhance the system value to the user. The effect could be an
increase in subscriber database, increase in general usage of the S-UMTS system and income from
closed user groups paying for dedicated services.




5.3.5 Terminals
                                                                                                       46
Contracts number:
ESTEC - 13694/99/NL/US                            Final Report
ESA S-UMTS study                                  Date                   Revision
Final Report                                      2000-11-29             A


To serve the same market as T-UMTS, S-UMTS must offer handheld terminals, transportable terminals
and vehicular terminals.

Handheld terminals
To be successful in the market, handheld terminals must be multimode and differ little terrestrial
terminals. The ideal solution would be for a user to have a single GSM/UMTS with the adaptation to
satellite done in a ‖modem/antenna‖ module.
To be competitive in scenario 1, the requirements for a S-UMTS-terminal from a market point of view are:
 size, price and performance similar to terrestrial terminals.
 several makes and models of terminals on the market.
   user interface and applications offered (service network) should be the same as for
    terrestrial terminals.

Moreover, scenario 2 might advantageously fulfil these requirements, but might also offer specific
satellite solutions for a niche market.

Transportable terminals, palmtop sized or laptop sized
In T-UMTS connectivity products will be developed to connect data terminals without own W-CDMA
interface such as PDAs, handheld personal computers and messaging devices. The data terminals will
connect with a mobile device using Bluetooth to create complete integration without physical
connections.




                                                      Bluetooth
                                                                      General data terminals



Fig. 4.3 S-UMTS transportable terminals.

The same model is valid for S-UMTS as shown in Fig. 4.3. The transportable terminal in S-UMTS will be
a combination of an antenna and a connectivity product. Consequently, the user interface from a data
terminal will be the same in S-UMTS and T-UMTS.

The main disadvantage of this solution is the need for an antenna to be fitted outdoors. For nomadic
usage in a niche market like scenario 2, this should be fully acceptable. To be a mass market alternative
for scenario 1, a fixed antenna mount must be attractive and the antenna/modem module must be
inexpensive.

Vehicular terminals
                                                                                                         47
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                   Revision
Final Report                                        2000-11-29             A


Several car manufacturers now offer mobile solutions via GSM. In the future, a vehicular S-UMTS
terminal might be integrated in the vehicle as shown in Fig. 4.4.

                          S-UMTS satellite




                             S-UMTS air i/f
                                                      Cellular network

                                       Satellite antenna
                                     (integrated  in the roof)

                                               Bluetooth radio i/f



                                   Satellite               UMTS
                                   terminal                terminal
Fig. 4.4 Future S-UMTS terminal integrated in a car.

The modem may also be integrated in the car to provide access to S-UMTS inside or near the car via a
Bluetooth interface. By 2005, most data and multimedia devices are expected to have Bluetooth
interfaces, so users could use their personal devices in the car. Moreover, this solution will support car
specific applications, such as audio streaming to the car radio. The antenna/modem module should be
capable of switching between T-UMTS and S-UMTS. This is especially important for scenario 1. As for
the other terminals, it would be favourable from a market point of view if handheld GSM voice terminals
could communicate via antenna/modem module via S-UMTS.

Several car manufacturers plan to implement communication facilities in future models. So efforts should
now be initiated to promote S-UMTS/T-UMTS system as the optimal solution.

Terminal mix
The terminal mix depends on the market segments addressed by S-UMTS service providers. However,
two conclusions concerning usage may be drawn from the discussion above.
 First, voice service is essential in S-UMTS, but use of it arguably will be minimal. This implies a
    relatively small market for handheld terminals.
   Second, the features of a vehicular system, and particularly those in scenario 2, suggest
    that the traffic from vehicular terminals may be dominant.

These conclusions lead to a likely terminal mix (usage) as listed in Table 4.12.

Table 4.12 Most likely terminal mix (usage).
Terminal    Scenario 1     Scenario 2
Handheld:   10 %           10 %
Palmtop:    20 %           10 %
Laptop:     20 %           10 %
Vehicular:  50 %           70 %
                                                                                                          48
Contracts number:
ESTEC - 13694/99/NL/US                                Final Report
ESA S-UMTS study                                      Date                     Revision
Final Report                                          2000-11-29               A


5.3.6 Coverage and distribution
Most promising markets

UMTS is prompted by the high growth of Internet usage combined with a mature mobile market. This
indicates that Western Europe should be the first priority for coverage. This is also the ―home-market‖ for
UMTS as a system. A satellite covering Western Europe may also cover Eastern Europe, the Middle
East, North Africa and Africa from the same orbital position.

To maximise the business potential and minimise the expenses and financial risks, the S-UMTS
coverage should consist of the following three phases.

Phase           Coverage
Phase 1         Western Europe, Eastern Europe, the Middle East and North Africa. This will cover 33% of
                the global target population. These are also regions of high car density, and cars have
                been identified as an important application for S-UMTS.
Phase 2         North America
Phase 3         Asia


5.3.7 Traffic requirements on S-UMTS
S-UMTS traffic model

                           GEO Satellite traffic variation
The current traffic variation during a day for a global satellite system supporting both voice and data
traffic is shown in Fig. 4.6.
     12

     10

      8                                                                    L2M (Voice)
                                                                           M2L (Voice)
      6
                                                                           L2M (Data)
      4                                                                    M2L (Data)
      2

      0
          1    3    5   7    9   11   13   15   17   19   21   23

Fig. 4.6 Variation of GEO satellite communications traffic during day, hours UIT.

Most traffic in UMTS is expected to be data and multimedia, and this traffic occurs more evenly
throughout the day than voice traffic. However, both voice and data traffic apparently account for some
10% of the day‘s traffic during busy hour.

Busy hours will differ by time zone, which will help spread the total traffic via the satellite.

Expected distribution of data rates

A prognosis of T-UMTS applications in 2005, including expected requirements on data rates, is given in
Table 4.14.
                                                                                                          49
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A



Table 4.14 Prognosis of T-UMTS applications in 2005.
Application                kBytes per         Sessions per                Percentage      Data rate
                           session, average day, average                  users           (kb/s)
E-mail without attachment  5                  5                           100             9.6
E-mail with attachment     250                5                           100             64
www.laptop                 500                5 (5 min. each)             80              64 - 144
www.minibrowser            40                 10                          30              9.6 - 64
Real time video            5000               0.3 (20 min. each)          25              64 - 384
Distribution/file transfer 1000               1 (3 min. each)             25              64 - 144
Information push           10                 10                          15              9.6 - 64
E-commerce                 1                  0.3                         50              9.6

Several of the applications, such as E-commerce and E-mail require only 9.6 kb/s. All the applications
listed in the table above have an average data rate of less than 64 kb/s. The only application that may
require 384 kb/s or higher data rates is real time video

In scenario 1, S-UMTS should offer higher data rates than GPRS to be a complementary alternative to T-
UMTS. As a complement to GPRS and for the car scenario, high data rate requirements are fewer.

Accordingly a proposed average data rate model for S-UMTS is given in Table 4.15.

                         Table 4.15 Proposed data rate model for S-UMTS.
                            Average data rate   % of total traffic
                            9.6 kb/s            20
                            64 kb/s             50
                            144 kb/s            20
                            384 kb/s            10

However, because applications based on packet switching will dominate S-UMTS, such a model is of
limited interest. Each application will be transmitted at the highest available bitrate unless means such as
higher tariffs for higher bitrates are introduced to reduce the data rate.
                                                                                                                                                 50
Contracts number:
ESTEC - 13694/99/NL/US                                                   Final Report
ESA S-UMTS study                                                         Date                           Revision
Final Report                                                             2000-11-29                     A


Total traffic requirements per scenario

The traffic prognosis starts with a base case for GSM speech, in which the average GSM user is
expected to place 5.5 minutes of outgoing calls per day in 2000, growing 2% per year. The volume for
GPRS traffic per user is then calculated by assuming that the GPRS traffic relative to the speech traffic
(measured in Mbit per user per day) will grow linearly to 150% in 2010, per GSM/GPRS user. T-UMTS
traffic volumes are assumed to be 50% more than GPRS traffic volumes per user.

In scenario 2, S-UMTS traffic volumes are assumed to be 10% of T-UMTS traffic volumes per user. This
gives the traffic prognosis per user shown in Table 4.16.

Table 4.16 Scenario 2 S-UMTS traffic volumes.
                                2000   2001   2002     2003     2004        2005    2006      2007          2008   2009        2010
Traffic assumption per user per day
     GSM speech minutes         5,51   5,62   5,73     5,85     5,96        6,08    6,20      6,33          6,45   6,58        6,71
     GPRS Mbit                         0,42   0,85     1,30     1,77        2,26    2,76      3,29          3,83   4,40        4,99
     T-UMTS Mbit                                       1,95     2,66        3,39    4,15      4,93          5,75   6,60        7,48
     S-UMTS Mbit                                                            0,34    0,41      0,49          0,58   0,66        0,75


In scenario 1, S-UMTS volumes are assumed to be equal to T-UMTS volumes per user (as a
complementary service to T-UMTS) and equal to GPRS volumes (as a complementary service to
GPRS). This gives the traffic prognosis per user shown in Fig. 4.17.

Table 4.17 Scenario 1 S-UMTS traffic volumes.
                                2000   2001   2002     2003     2004        2005    2006      2007          2008   2009        2010
Traffic assumption per user per day
     GSM speech minutes         5,51   5,62   5,73     5,85     5,96        6,08    6,20      6,33          6,45   6,58        6,71
     GPRS Mbit                         0,42   0,85     1,30     1,77        2,26    2,76      3,29          3,83   4,40        4,99
     T-UMTS Mbit                                       1,95     2,66        3,39    4,15      4,93          5,75   6,60        7,48
     S-UMTS Mbit                                                            2,40    3,43      4,24          4,94   5,84        6,84


Traffic variation diagrams for GSM as shown in figure 4.5 and for satellite as shown in     figure 4.6
indicates less than 10% of the traffic in busy hour. By assuming 10% busy hour traffic for S-UMTS, the
traffic per user in a busy hour will be as listed in Table 4.18.

Table 4.18 Traffic per user in a busy hour.
Traffic per user / busy hour                         2005              2006        2007              2008          2009               2010
Scenario 1                                           240 kbit          343 kbit    424 kbit          494 kbit      584 kbit           684kbit
Scenario 2                                            34 kbit           41 kbit     49 kbit           58 kbit       66 kbit            75 kbit

The target population of Western Europe is 382 Million (Table 4.13). Even if the satellite may cover more
than Western Europe the market potential is calculated based on the Western Europe population. Using
the potential penetration figures of §2.3 with uniform distribution, the busy-hour satellite traffic
requirements is as given in Table 4.19.

Table 4.19 Busy-hour satellite traffic requirements in Western Europe.
                                                     2005              2006             2007               2008               2009            2010
Scenario 1 S-UMTS subscribers                    2.28 million      3.24 million     4.47 million       5.46 million       6.15 million    6.26 million
           Traffic                               548 Gbit          1113 Gbit        1894 Gbit          2700 Gbit          3592 Gbit       4282 Gbit
Scenario 2 S-UMTS subscribers                    0.16 million      0.30 million     0.49 million       0.65 million       0,84 million    0.96 million
           Traffic                               5.28 Gbit         12.36 Gbit       24.18 Gbit          37.63 Gbit        55.49Gbit       71.97Gbit
The traffic requirements of Table 4.19 assume the same busy hour throughout Western Europe.
However, busy hours vary by country and thereby by satellite beam, which will reduce overall busy hour
traffic in the satellite.
                                                                                                                               51
Contracts number:
ESTEC - 13694/99/NL/US                                    Final Report
ESA S-UMTS study                                          Date                                Revision
Final Report                                              2000-11-29                          A




5.3.8 S-UMTS tariff estimates
Future UMTS tariffs are difficult to predict, for two reasons. First, future UMTS operators have yet to set
tariff schedules. Second, the high licence fees recently paid in Western Europe may be reflected in
tariffs. Analyses in the UK indicate that the T-UMTS licence holders may need to charge users USD 150
to 300 a month just to cover costs. If so, service providers must find new ways of charging for attractive
applications.

This Study uses T-UMTS tariffs that are based on an evolution of the present mobile tariffs. New
methods of charging for applications may enhance the business case considerably.

Tariff prognosis for mobile multimedia communications is calculated using an extrapolation of tariffs for
GSM speech in Norway, with an annual 20% price reduction for traffic and no reduction in subscription
fees. Average tariffs, not including value-added tax (VAT) in 2001 are assumed to be USD 0.13 per
speech minute and USD 0.20 per Mbit for GPRS data. Average subscription fees for GSM speech is
fixed at USD 15.6 per month, whereas it is assumed that having GPRS in addition to GSM speech will
cost 50% more than GSM speech alone. T-UMTS is assumed to cost twice as much as GSM with GPRS
for traffic and 20% more for subscription. Whereas S-UMTS in the vertical segments is assumed to cost
five times as much as GSM with GPRS. This gives the following tariff prognosis for scenario 2 of Table
4.20.

Table 4.20 Tariff prognosis for scenario 2.
                         2000    2001    2002    2003    2004      2005       2006     2007       2008     2009     2010

USD per speech minute    0,13    0,10    0,08    0,06    0,05      0,04       0,03     0,03       0,02     0,02     0,01
USD per Mbit, GPRS               0,20    0,16    0,13    0,10      0,08       0,07     0,05       0,04     0,03     0,03
USD per Mbit, T-UMTS                             0,26    0,21      0,17       0,13     0,11       0,09     0,07     0,06
USD per Mbit, S-UMTS                                               0,42       0,33     0,27       0,21     0,17     0,14

USD per month, speech   15,63   15,63   15,63   15,63   15,63     15,63      15,63    15,63     15,63     15,63    15,63
USD per month, GPRS             23,45   23,45   23,45   23,45     23,45      23,45    23,45     23,45     23,45    23,45
USD per month, T-UMTS                           28,14   28,14     28,14      28,14    28,14     28,14     28,14    28,14
USD per month, S-UMTS                                            117,26     117,26   117,26    117,26    117,26   117,26


For scenario 1 the subscribers can be divided into four categories, depending on prime subscription type.
Pricing for the four main subscriber categories is listed in Table 4.21.

Table 4.21 Pricing for the four main subscriber categories.
Prime Sub-       GSM/GPRS               T-UMTS                            S-UMTS (roaming)                  S-UMTS
scription
Monthly cost     $18 – $25/month        $25-$30/month +                   N.A (normal roaming)              $20 -$25/month for S-
                 +$12/month for S-      $7/month for S-UMTS                                                 UMTS
                 UMTS                   (Assumes partnership              (Assumes no partnership
                                        between T-UMTS and                between local T-UMTS
                                        S-UMTS operator)                  and S-UMTS operator)
Traffic cost     Additional 10 – 20%    Additional 10-20% of T-           Additional 50-100% of T-          Additional 10 – 20% of
                 of T-UMTS traffic      UMTS traffic charges              UMTS traffic charges              T-UMTS traffic charges
                 charges

Scenario 1 is a competitive scenario, and the price levels reflect this. The following table lists the
assumptions made for the purpose of this study in assessing monthly subscriber rates and the portion of
users subscribing to the corresponding rates.
                           GPRS w/S-UMTS               T-UMTS w/S-UMTS               S-UMTS
                                                                                                                                                       52
Contracts number:
ESTEC - 13694/99/NL/US                                              Final Report
ESA S-UMTS study                                                    Date                                   Revision
Final Report                                                        2000-11-29                             A


S-UMTS user                         30%                                     20%                                       50%
breakdown
Monthly rate USD                    $ 12                                    $7                                        $ 25

The table should be read as follows. Of all the S-UMTS subscribers, 50% use S-UMTS as their main 3G
provider, 30% buy S-UMTS as an add-on or package deal to their GPRS subscription, and 20% buy S-
UMTS as an addition to their T-UMTS subscription. This corresponds quite well with the penetration
figures for S-UMTS from Table 4.5, Table 4.6 and Table 4.7.

The traffic revenues is determined based on the following assumption:
 80% of the S-UMTS traffic generated is charged at a rate 20% higher than the expected rates for T-
   UMTS.
 20% of the S-UMTS traffic generated is assumed of be roaming traffic without agreement and
   charged at a rate 100 % higher than T-UMTS.

This gives the tariff prognosis for scenario 1 as listed in Table 4.22.


Table 4.22 Tariff prognosis for scenario 1.
                             2000      2001    2002        2003      2004        2005       2006       2007       2008      2009         2010

  USD per speech minute      0,13      0,10    0,08        0,06      0,05        0,04       0,03       0,03       0,02      0,02         0,01
  USD per Mbit, GPRS                   0,20    0,16        0,13      0,10        0,08       0,07       0,05       0,04      0,03         0,03
  USD per Mbit, T-UMTS                                     0,26      0,21        0,17       0,13       0,11       0,09      0,07         0,06
  USD per Mbit, S-UMTS                                                           0,23       0,18       0,15       0,12      0,10         0,08

  USD per month, speech     15,63     15,63   15,63    15,63        15,63    15,63        15,63       15,63      15,63     15,63     15,63
  USD per month, GPRS                 23,45   23,45    23,45        23,45    23,45        23,45       23,45      23,45     23,45     23,45
  USD per month, T-UMTS                                28,14        28,14    28,14        28,14       28,14      28,14     28,14     28,14
  USD per month, S-UMTS                                                      17.5         17.5       17.5        17.5      17.5      17.5


With these assumptions, the average revenue per user per year (USD) for scenario 1 is as listed in Table
4.23.

Table 4.23 Average USD revenue per user for scenario 1.
S-UMTS                              2005       2006               2007           2008                2009                2010
Traffic                              201             225           232                  216                214              171
Subscription                         210             210           210                  210                210              210
Sum                                  411             435           442                  426                424              381

Based on the tariff prognosis for scenario 2 in Table 4.20 the average revenue per user per year (USD)
for this scenario is as listed in Table 4.24.

Table. 4.24 Average USD revenue per user for scenario 2.

      ARPU (USD per year)             2000    2001    2002         2003     2004         2005       2006       2007      2008      2009         2010
      GSM speech
         Traffic                      252     206      168         137       112          92         75         61        50        41           33
         Subscription                 188     188      188         188       188         188        188        188       188       188          188
         SUM                          439     393      356         325       300         279        263        249       238       229          221
      GPRS
         Traffic                               31       50          61        65          66         63         58        52        45           39
         Subscription                          94       94          94        94          94         94         94        94        94           94
         SUM                                  125      144         155       159         159        157        152       146       139          133
      T-UMTS
         Traffic                                                   185       202         206        202        193       180       166          150
         Subscription                                              338       338         338        338        338       338       338          338
         SUM                                                       523       540         544        540        531       518       504          498
      S-UMTS
         Traffic                                                                           52         51         48      45        41          38
         Subscription                                                                   1 407      1 407      1 407   1 407     1 407       1 407
         SUM                                                                            1 459      1 458      1 455   1 452     1 449       1 445
                                                                                                                 53
Contracts number:
ESTEC - 13694/99/NL/US                                          Final Report
ESA S-UMTS study                                                Date                    Revision
Final Report                                                    2000-11-29              A



The above ARPU and penetration assumptions apply to Western Europe only, but as a first
approximation, these figures may be used to calculate the revenue prospects in other parts of the world
as well. For this purpose, a ―target population‖ is derived from the population and the average purchasing
power (GDP) per capita in the respective countries.


Average revenue per user in the car scenario

The yearly charge for cars system might well depend on its use. As an initial estimate, with no
specification of use, the traffic ARPU figures from T-UMTS have been used, as in Table 4.25.

Table 4.25 Yearly charge for S-UMTS car system (ÙSD).
Yearly charge for S-UMTS car system (USD)
                             2000   2001   2002   2003   2004     2005    2006   2007       2008   2009   2010
   West Europe                -      -      -      -      -        206     202    193        180    166   150
   West + East Europe         -      -      -      -      -        206     202    193        180    166   150
   Eur + Mid. East + N Afr    -      -      -      -      -        206     202    193        180    166   150
   World                      -      -      -      -      -        206     202    193        180    166   150



Pricing of services in S-UMTS

The general trend in telecommunications is toward lesser revenue generation from minute pricing. To
compensate for this loss of revenue flat rate in combination with pricing of services is expected.

In Table 4.26 a projected combination of different price mechanisms are given for T-UMTS applications
in 2005.
                                                                                                               54
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                    Revision
Final Report                                        2000-11-29              A



Table 4.26 Different pricing mechanisms for T-UMTS in 2005.
Application                      kBytes per       Sessions per              Percentage            Type of fee
                                 session, average day, average              users
E-mail without attachment        5                5                         100                   Fee per session
E-mail with attachment           250              5                         100                   Fee per month
www.laptop                       500              5 (5 min.)                80                    Fee per minute
www.minibrowser                  40               10                        30                    Fee per month
Real time video                  5000             0.3 (20 min.)             25                    Fee per minute
Distribution/file transfer       1000             1 (3 min.)                25                    Fee per minute
Information push                 10               10                        15                    Fee per session
E-commerce                       1                0.3                       50                    Fee per session

For some applications the end user will only see the price of the content, and service and content
provider have to share the income. It is difficult to say whether this will make the business case better or
worse for the service provider.



5.4    Interface between market and traffic analysis and satellite system sizing


5.4.1 Market and traffic analysis
The conclusions from the market analysis that are important for this elaboration are summarised in
Tables 5.1 and 5.2

Table 5.1 Scenario 1 market analysis.
                                 2005             2006       2007        2008          2009          2010
 Potential S-UMTS subscribers,      2.28            3.24       4.47        5.46           6.15         6.26
 million
 Traffic per user / busy hour         240 kbit   343 kbit   424 kbit    494 kbit       584 kbit     684kbit

Table 5.2 Scenario 2 market analysis.
                                 2005             2006       2007        2008          2009          2010
 Potential S-UMTS subscribers,      0.16            0.30       0.49        0.65           0.84         0.96
 million
 Traffic per user / busy hour          34 kbit    41 kbit    49 kbit     58 kbit        66 kbit      75 kbit

The number of available channels per GEO satellite was delineated in Task 3 (WP3400) and is an
important factor for the calculation of number of subscribers that can be supported. Results from
simulations and estimated capacity improvements are used for the calculations below.


5.4.2 Satellite system sizing
Table 5.3 System capacity estimation with satellite improvement.
                                           CAT Results                 Excel Spreadsheet Results


 Terminal           Link Data Rates     Maximum average      Intermediate Capacity         High Capacity
                                        satisfied demands         (I/N=0.5 limit)         (pk/av factor=2)
                                           CAT Results         2-spectrum blocks         2-spectrum blocks
                                                                                                                    55
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                        Revision
Final Report                                         2000-11-29                  A


                                            (channels)


                                      97 beam, 10m reflector, 218 beam, 15m reflector, 120   132 beam, 15m reflector,
                                           80 elements                elements                    120 elements
                                             2-blocks                  2-blocks                     2-blocks
    Handheld             8 kbps                3096                                                   N/A
    Palmtop              8 kbps               6620                       15260                        22440
    Palmtop             64 kbps                888                       1962                         2904
    Laptop               8 kbps               6620                       15260                        34056
    Laptop              64 kbps                888                       1962                         4488
    Vehicle              8 kbps               8308                       18748                        27720
    Vehicle             64 kbps               1246                       2834                         4224


Input from the ―High Capacity‖ results in this table is used in case 1 and 3 below.

System capacity with various terminals and services has been studied in simulations 3 in Task 3
(WP3400) and can be assessed from the nominal results given in the Table 5.3. Assuming that the 2.4
kbps service will be accessed exclusively through the 10% of handheld terminals (moreover, this service
will be the only one considered for handheld terminals), the capacities are as listed in Table 5.4.


Table 5.4 Overall capacity scenarios 1 and 2.
                   Absolute capacity Maximum average                 Absolute                Maximum
                   FWD                 satisfied demands             capacity                average satisfied
                                       FWD                           RT                      demands - RT
Scenario 1         688                 896                           566                     794
Scenario 2         693                 910                           783                     1014

The figures in Table 5.4 refer to the low capacity system and has to be modified to reflect the increased
system capacity performance described in Table 5.3.
With this modification for scenario1, maximum average satisfied demands, a total of 6333 channels can
be served, whereof:
 307 handheld terminals using the 2.4 kbps service
      1185 palmtop terminals, from which 407 use the 8 kbps service and 778 use the 64 kbps
       service
      1820 laptop terminals, from which 617 use the 8 kbps service and 1203 use the 64 kbps service
      3021 vehicular terminals, from which 994 use the 8 kbps service and 2027 use the 64 kbps service

These figures are used in case 2 below.
                                                                                                           56
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                  Revision
Final Report                                         2000-11-29            A



5.4.3 S-UMTS subscribers per GEO satellite

Fig. 5.1 Throughput as a function of traffic load.


                    %

             100
              90
              80
              70
              60
              50
              40
              30
              20
              10
               0                                             WWW load
                        1     2        3        4      5
The curve in Fig. 5.1 is a result of GPRS lab simulations with traffic generation from GPRS 4-slot
mobiles. A similar curve for satellite system throughput with different data rates will deviate somewhat
from that of Fig. 5.1, but nonetheless the curve shown provides a good initial estimate.

Calculation of users per GEO satellite for Case 1

Based on input from Task 3 (WP3400) and Chapter 4 (WP4200-1) of this Task, the subscriber capacity
for the case 1 simulation are calculated below. Case 2 and 3 can be found in the study report.

1/2.    Determine the average number of PDCHs per beam (pdch) and predict the mix of multislot
classes.

Assume that all channels are offered as 64 kb/s channels to vehicular terminals.

According to Task 3 (WP3400), the average number of forward channels is 4224, and the number of
beams is 132. It is assumed that 3564 (27x132) channels are used for packet switching and the rest for
voice. One 64 kb/s channel is 8 pdch.

3.      Predict the user application bit rate (avgbitrate).
The average bit rate per user in a busy hour is given in Chapter 4 (WP4200-1) and listed in Table 5.2.
Packet traffic accounts for 90% of average bit rate in busy hour.
 Avgbitrate = average bit rate busy hourx90%/3600.

4/5     Assume the average acceptable throughput per multislot class and find the traffic load from the
graph of Fig. 5.1.

The traffic load depends on required medium throughput and in this case we will look at two alternatives:
Medium throughput 60 kb/s ---> traffic load = 2
Medium throughput 45 kb/s ---> traffic load = 4

6.       Calculate the number of supported packet users.
                                                                                                    57
Contracts number:
ESTEC - 13694/99/NL/US                                Final Report
ESA S-UMTS study                                      Date                      Revision
Final Report                                          2000-11-29                A


The number of users per beam is then calculated to:
Users = 27x8xtraffic loadx132 / [avgbitrate(table)/3600x90%]

The result is shown in the graph below.
 Subscribers (1000)

     14000
     13000
     12000
     11000                                                                       Scenario      1/
     10000                                                                       60kb/s
      9000
      8000                                                                       Scenario      1/
      7000                                                                       45kb/s
      6000                                                                       Scenario      2/
      5000
      4000                                                                       60kb/s
      3000                                                                       Scenario      2/
      2000                                                                       45kb/s
      1000
         0
                    2005 2006 2007 2008 2009 2010                           Year

                                      2005    2006    2007      2008     2009       2010
          Sce na rio   1/   60kb/s     950     663     538       464      391        333
          Sce na rio   1/   45kb/s    1900    1326    1076       928      782        666
          Sce na rio   2/   60kb/s    6709    5591    4674      3960     3456       3049
          Sce na rio   2/   45kb/s   13418   11182    9348      7920     6912       6098

Fig. 5.2 Total number of packet subscribers per satellite as a function of traffic per user.

The remaining traffic channels are available for voice. Maximum number of channels per beam:
       (4224 – 3564)*8/132 – 2 (control) = 38

In Chapter 4 (WP4200-1), busy hour traffic in a GSM network is estimated to 16 mErlang, and the
acceptable Grade of Service (GoS) is assumed to be 2 %. Erlang calculations show then that 2375
subscribers can be supported per beam and the total number of voice subscribers one satellite can
support is:

         2375*132 = 313 500 subscribers

Because only 10% of traffic is voice, the voice capacity should be sufficient for a network with
approximate 3 000 000 subscribers. Note that voice capacity can be increased or decreased by changing
the ratio between packet channels and circuit channels.


5.4.4 Elaboration of the results
                                                                                                          58
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


The calculation above shows that by 2005 roughly 1 900 000 subscribers can be supported by one GEO
satellite in scenario 1 and 13 500 000 subscribers in scenario 2. These figures will decrease over time as
traffic per user is expected to increase.

The traffic per user is the critical parameter; using the expected traffic figures from T-UMTS (scenario 1)
results in a number of subscribers limited to 1 000 000 by year 2007/2008. The mix of data rates and
terminal types has moderate influence on the total subscriber figures. Evaluating this together with the
conclusions from the market analysis leads to two principal conclusions:
   One GEO satellite cannot fulfil the estimated market potential for the mass market (scenario
    1) for Western Europe but can fulfil a reasonable share of the market potential.
   One GEO satellite can fulfil the market potential for the executive market (scenario 2) for Western
    Europe.



5.4.4.1 Ways of meeting the market potential or enhancing the business case for
   scenario 1.
   Increased tariffs in the first years of operation may cut the traffic per user and thereby
    enable the satellite to support more subscribers. Differentiated tariffs may be even better for
    the business case.
   More than one GEO satellite can be a good solution to meet the increasing demand.
   Rather than cover all of Western Europe, a single GEO satellite may cover a selected few
    countries.
   A new generation satellite with a twofold or fivefold increase in capacity could meet the
    challenge of S-UMTS, meet the market potential and effect a very good business case.
   Broadcast and multicast services are a promising segment for S-UMTS but since neither
    the technical standard is finalised nor market estimates from T-UMTS are available we have
    not been able to estimate the business potential.


5.4.4.2 Risks in calculation of number of users per satellite
Table 5.3 lists capacity figures for three alternative satellite system designs. Figures from the highest
capacity system has been chosen for the calculations above. The other alternatives and especially the
low capacity system design will reduce the business case significant.
Packet switching throughput calculations are based on experience from terrestrial GPRS. A good
estimate for satellite GPRS traffic will not be available until satellite GPRS is standardised and traffic
simulations performed.
For simplicity we have used a simple application (www traffic) in our traffic model. In a real network there
will be a mix of service classes which will reduce the throughput per user or the number of users.
                                                                                                            59
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                    Revision
Final Report                                         2000-11-29              A



5.5    Expected expenses and revenues for the scenarios
S-UMTS can be realised only if its business potential is sufficiently attractive to be of interest to operators
and other investors, and system capital costs are kept low enough to be recouped reasonably rapidly by
the operating revenues.



5.5.1 Revenue potential


5.5.1.1 Key assumptions
Table 6.1 outlines the key parameters and assumptions underlying the business case calculations in this
chapter.

Table 6.1 Key parameters and assumptions
Parameter          Assumption                                           Comment
Total population   382 million people in Western Europe, 99%            Coverage beyond Western Europe
                   of the total population, has an average GDP          not included here could improve
                   of $21000.                                           bottom line further.
                   1603 million people globally meet the                Global coverage is only considered
                   minimum requirements for literacy and                for scenario 2, the niche scenario.
                   annual earnings.
Market potential   Ch 5 elaborates the following total                  4 types of users:
                   penetration levels in %:                              Prime S-UMTS subscribers
                   Year 2005 2006 2007 2008 2009 2010                    GPRS subscribers with satellite
                   Sc1: 0,60 0,85 1,17 1,43 1,61 1,64                       add-on
                   percent of the total of 382 million                   T-UMTS subscribers with
                   Sc2: 0,01 0,02 0,04 0,08 0,13 0,17                       satellite add-on
                   % of the total of 382 million & 1,6 billion           Roaming usage without
                                                                            agreement
                       This gives the following total market
                       potential (in million people):
                       Year 2005 2006 2007 2008 2009 2010
                       Sc1: 2,29 3,25 4,47 5,46 6,16 6,27
                       Sc2: 0,03 0,07 0,15 0,31 0,49 0,65

                       Sc2 global (in million people):
                              0,16 0,32 0,64 1,28 2,08 2,73
                                                                                                          60
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                   Revision
Final Report                                        2000-11-29             A


Parameter               Assumption                                     Comment
Capacity limits         The elaboration in Ch 5 for case 1 lists the   The cases in Ch 5 show little
                        max number of users for two levels of          difference in capacity limits
                        degree of service, 45 and 60 kb/s              regardless of mixture of use and
                        throughput per 64 kb/s channel. In the         terminal types. Only case 1 is
                        chapter the terms 45 kb/s or 60 kb/s service   therefore used.
                        level or only 45 kb/s and 60 kb/s is used.
                        These set the constraints for which the
                        revenue potentials are calculated from.
Pick up & retention     The pick-up rate is set at 50% in year 1, 90   By 2005 a significant level of UMTS
rates                   % in year 2 and 100% in year 3. The            applications, terminals and coverage
                        retention rates are 97% and 95% in 2009        will ensure a rapid pickup of S-
                        and 2010.                                      UMTS. Reuse of T-UMTS
                                                                       infrastructure and close co-operation
                                                                       between T-UMTS and S-UMTS will
                                                                       also support rapid pick-up levels.
Revenues per user       Revenues are derived from both traffic and     Tariff levels for scenario 1 reflect
                        subscription levels described in Ch 5.         competitive pricing, for scenario 2
                        Average annual traffic revenues (in USD        niche prices have been applied.
                        per year) per user:                            Traffic revenues varies from year to
                        Year 2005 2006 2007 2008 2009 2010             year because usage increases as
                        Sc1: 201 225 232 216 214 171                   price per Mbit gets cheaper. 80% of
                        Sc2:     52     51    48     45    41    38    the traffic is charged at a rate 20%
                                                                       higher than the expected rates for T-
                        Annual subscription (in USD) per user:         UMTS. The rest is assumed to be
                        Year 2005 2006 2007 2008 2009 2010             roaming traffic without agreement,
                        Sc1: 210 210 210 210 210 210                   charged at a rate 75% higher than
                        Sc2: 1407 1407 1407 1407 1407 1407             the T-UMTS rate.

                                                                       Subscription rates are kept constant
                                                                       at monthly rates of USD 17,50 for
                                                                       scenario 1 and 117,25 for scenario 2



5.5.1.2 Scenario 1
Scenario 1 involves a mass market scheme for Western Europe.

Table 6.2 Scenario 1 revenue potential
                                2005        2006        2007         2008         2009         2010
Population w. Europe         382 516 762 382 516 762 382 516 762 382 516 762 382 516 762 382 516 762
Penetration S-UMTS                0.60%       0.85%        1.17%        1.43%        1.61%        1.64%
Market potential               2 295 101   3 251 392    4 475 446    5 469 990    6 158 520    6 273 275
Pick up & retention rates*         100%        100%         100%         100%         100%         100%
Number of users                2 295 101   3 251 392    4 475 446    5 469 990    6 158 520    6 273 275
Traffic revenues per user            201         225          232          216          214          171
Total traffic revenues       461 315 215 731 563 307 1038 303 499 1181 517 774 1317 923 252 1072 730 007
Subscr. revenues per user            210         210          210          210          210          210
Total subscr. Revenues       481 971 120 682 792 420 939 843 684 1148 697 836 1293 289 172 1317 387 728
                                                                                                                             61
Contracts number:
ESTEC - 13694/99/NL/US                                Final Report
ESA S-UMTS study                                      Date                          Revision
Final Report                                          2000-11-29                    A


Revenues in USD millions             943      1 414             1 978              2 330             2 611             2 390
*Note: the 100% pick-up and retention rates used for maximum revenue potential in the mass market
scenario

5.5.1.3 Scenario 2
This niche market scenario also focuses on Western Europe, but as a complementary service to T-
UMTS.

Table 6.3 Revenue potential for scenario 2 with regional coverage
                              2005         2006             2007          2008              2009             2010
Population w. Europe        382 516 762 382 516 762 382 516 762 382 516 762 382 516 762 382 516 762
Penetration S-UMTS               0.01%       0.02%       0.04%       0.08%       0.13%       0.17%
Market potential                 38 252      76 503     153 007     306 013     497 272     650 278
Pick up & retention rates          50%         90%        100%        100%         97%         95%
Number of users                  19 126      68 853     153 007     306 013     482 354     617 765
Traffic revenues per user            52          51          48          45          41          38
Total traffic revenues          994 544   3 511 504   7 344 322 13 770 603 19 776 499 23 475 054
Subscr. revenues per user         1407        1407         1407        1407       1407        1407
Total subscr. Revenues       26 910 054 96 876 195 215 280 434 430 560 867 678 671 567 869 194 751
Revenues in USD millions             28         100         223         444         698         893



5.5.1.4 Vehicle scenario
As a subset of scenario 2 there is also revenue potential derived for a vehicle scenario.

Table 6.4 Revenue potential for a vehicle scenario
                                                    2005       2006        2007            2008        2009           2010
Number of existing cars (mill cars)                  197           199      202            204          206            208
Number of new cars (mill cars)                        18           18        18             18           19             19
Penetration existing cars                          0.03%      0.09%       0.18%         0.32%         0.46%         0.59%
Penetration new cars                               0.66%      1.73%       3.63%         6.31%         9.27%         11.82%
Accumulated S-UMTS systems in existing cars        59 388    274 466     760 852    1 671 529      3 113 403   5 084 850
New S-UMTS systems in new cars sold               117 530    310 569     658 503    1 157 546      1 720 219   2 217 388
Annual charge for S-UMTS in cars                     100           80        64             51           41             33

Total, USD million                                    18           47        91            145          198            239




5.5.2 Effects of capacity limitations on the revenue potential
Chapter 5 elaborates various capacity cases where number of subscribers is a function of the data traffic
speeds and how many channels are available.
The following analysis is based on case 1, as results will differ only marginally if either of the other two
cases are used.
                                                                                                     62
Contracts number:
ESTEC - 13694/99/NL/US                            Final Report
ESA S-UMTS study                                  Date                   Revision
Final Report                                      2000-11-29             A


Case 1 assumes that all of the beams offer twenty-seven 64 kb/s packet communication, with the
remainder for voice.


5.5.2.1 Capacity limitation effects on the scenario 1 revenue potential

The capacity limits applied to the revenue potential are elaborated in Chapter 5 §1 for case 1.

Table 6.6 Scenario 1 revenue potential with capacity limitations
                                   2005        2006        2007        2008        2009        2010
Market potential (W. Europe)      2 295 101   3 251 392   4 475 446   5 469 990   6 158 520   6 273 275
Max no of subscribers             1 900 000   1 326 000   1 076 000     928 000     782 000     666 000
Pick up & retention rates              50%         90%        100%        100%         97%         95%
Subscriber base                     950 000   1 193 400   1 076 000     928 000     758 540     632 700
Annual traffic rev per user             201         225         232         216         214         171
Traffic revenues                190 950 000 268 515 000 249 632 000 200 448 000 162 327 560 108 191 700
Annual subscription per user            210         210         210         210         210         210
Sum subscription rev            199 500 000 250 614 000 225 960 000 194 880 000 159 293 400 132 867 000
Total revenues - USD millions           390         519         476         395         322         241



Possible effect of increased tariffs

In Table 6.7 usage tariffs are increased by 35% for the years 2005 - 2007. Subscription prices
are unchanged. It is assumed that the number of subscribers stays the same. The effect is an
increase of over 10 % for the total revenue stream in the first three years:

Table 6.7 Scenario 1 revenues with a 35% increase in traffic tariffs for 2005 - 2007
                                   2005        2006        2007        2008        2009        2010
Market potential (W. Europe)      2 295 101   3 251 392   4 475 446   5 469 990   6 158 520   6 273 275
Max no of subscribers             1 900 000   1 326 000   1 076 000     928 000     782 000     666 000
Pick up & retention rates              50%         90%        100%        100%         97%         95%
Subscriber base                     950 000   1 193 400   1 076 000     928 000     758 540     632 700
Annual traffic rev per user             271         304         313         216         214         171
Traffic revenues                257 782 500 362 495 250 337 003 200 200 448 000 162 327 560 108 191 700
Annual subscription per user            210         210         210         210         210         210
Sum subscription rev            199 500 000 250 614 000 225 960 000 194 880 000 159 293 400 132 867 000
Total revenues - USD millions          457         613           563         395       322         241



Possible effect of adding 2 more satellites
By deploying 2 more satellites in 2006 and 2007 & keeping the 35% increase in tariffs for the
first three years the revenue prognosis gives is a 138% increase in the total revenue stream.
Each satellite is assumed to handle 50% of its capacity in the launch year, 90% in year 1 and
100% in year 2.

Table 6.8 Scenario 1 revenues with increased tariffs and three GEO satellites.
                                                                                                                           63
Contracts number:
ESTEC - 13694/99/NL/US                                              Final Report
ESA S-UMTS study                                                    Date                   Revision
Final Report                                                        2000-11-29             A


                                                    2005        2006        2007        2008        2009        2010
Market potential (W. Europe)                       2 295 101   3 251 392   4 475 446   5 469 990   6 158 520   6 273 275
Max no of subscribers                              1 900 000   1 989 000   3 012 800   2 784 000   2 346 000   1 998 000
Pick up & retention rates                               50%         90%        100%        100%         97%         95%
Subscriber base                                      950 000   1 790 100   3 012 800   2 784 000   2 275 620   1 898 100
Annual traffic rev per user                              271         304         313         216         214         171
Traffic revenues                                 257 782 500 543 742 875 943 608 960 601 344 000 486 982 680 324 575 100
Annual subscription per user                             210         210         210         210         210         210
Sum subscription rev                             199 500 000 375 921 000 632 688 000 584 640 000 477 880 200 398 601 000
Total revenues - USD millions                            457         920       1 576       1 186         965         723

Overall, the capacity limitations bring about a significant gap between the market potential and
revenue estimates, as illustrated in Fig. 6.1.
Fig. 6.1 Comparison of total potential vs. revenue prognosis limited by capacity.


                                             Scenario 1 revenue potential vs capacity limits


                                  2500
                                                                                                       Revenue potential

                                  2000



                                  1500                                     3 GEOs
           USD Millions




                                  1000



                                    500

                                                                                                            1 GEO

                                       0
                                                 2005        2006          2007     2008        2009        2010
                   Revenue potential             1105        1670          2342     2330        2611        2390
                   1 GEO base                     390        519           476      395         322          241
                   1 GEO increased tariffs        457        613           563      395         322          241
                   3 GEOs                         457        920           1576     1186        965          723




5.5.2.2 Capacity limitation effects on the scenario 2 revenue potential
The capacity limits have no effect on the revenue potential for scenario 2, where the estimated capacity
is between 6,2 million and13,5 million subscribers for a single GEO satellite system, and the market
potential is estimated at max 1,3 million subscribers.
                                                                                 64
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date             Revision
Final Report                                         2000-11-29       A



5.5.3 System cost estimates

5.5.3.1 Key assumptions
Cost assumptions for a single vs. a 3 GEO system are compared in Table 6.10.

Table 6.10 Key assumptions for a GEO S-UMTS system.
GEO S-UMTS                       Global Regional
Number of satellites                       3           1
Cost per satellite                       210         250
Insurance per launch                      64          64
Number of satellites per launch            1           1
Launch costs per satellite               120         120
Total number of launches                   3           1
Life expectancy of satellites yrs         15          15
Life expectancy of ground segment         15          15
Minimum time to market                     3           3
Commercial operations period               6           6




5.5.3.2 System costs and expense projections for a GEO S-UMTS system
Table 6.11 System costs and expense projections for GEO S-UMTS.
                                        Regional Global
(USD million)
INVESTMENTS
System development costs                        50     100
Space segment costs                            370     990
Launch insurance                                64     192
Long lead items                                 60      60
Ground segment costs                            75     150
Net system costs                               619    1492
Cost of initial order of terminals             250     500
Total system costs                             869    1992

OPERATIONS
Sum pre-operational expenses                 100       150
Total operating expenses                     240       360
Total depreciation                           495      1240
Total financing and interest expenses        209       478
Marketing and G&A                            200       400
Sum operating expenses 2002-2010            1244      2628

Average annual operating expenses              207     438
                                                                                                       65
Contracts number:
ESTEC - 13694/99/NL/US                            Final Report
ESA S-UMTS study                                  Date                      Revision
Final Report                                      2000-11-29                A



5.5.3.3 Capital required
Table 6.12 Capital required for a GEO S-UMTS system.
                                                  Regional      Global
(USD million)
Total system costs                                        869     1992
Operating expenses for the first 3 years                  614     1269
Total capital required                                   1483     3261

Financing
Debt                                                      890     1957
Equity, preferred and common stock                        593     1304
Total financing required                                 1483     3261

The total required capital must be supplied in 2002-2004. The debt to equity ratio is assumed to be 60 /
40 %. For simplicity interest rate on debt is set at 10%.


5.5.3.4 Break even levels
The initial break even level given the total system costs and operating expenses can be compared as
follows.

Table 6.13 Required annual break even levels.
                                                           Regional       3 GEOs
(USD million)
Total system costs / 6 years period                                 145      332
Sum operating expenses / 6 years period                             207      438
Average annual revenue required to break even                       352      770

As a first approximation these average break even levels may be used to compare the average revenue
schemes listed in Table 6.14.

Table 6.14 Summary of potential revenue schemes by annual averages.
(USD Millions)                                      1 GEO    3 GEOs
Scen1 base case                                                    391
Scen1 + tariffs                                                    432
Scen1- 3 GEOs                                                                971
Scen 2 reg                                                         398
Car scenario                                                       123
Scen 2 reg + car scenario                                          521

The scenarios all fall comfortably above the average break even levels.
                                                                                                                       66
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                       Revision
Final Report                                        2000-11-29                 A



5.5.4 Financial performance of the selected scenarios

5.5.4.1 Cash flow analysis for the scenarios
Combining the revenue projections with capacity limits and system cost information gives the following
net cash flow elaborations for each of the scenarios.

Table 6.15 Cash flow statement for scenario 1
Scenario 1 Western Europe
(In USD millions)                     2002   2003     2004     2005     2006       2007     2008    2009 2010
System development costs               15     15        20
Space segment costs                   100    100       170
Launch insurance                                        64
Long lead items                                         60
Ground segment costs                    10     30       35
Cost of initial order of terminals                     175     75
Pre-operational expenses                30     30       40
Operating expenses                                              40      40      40         40       40         40
Marketing and G&A                                       40      40      24      24         24       24         24
Interest expenses                                       20      37      12       5          4        4          4
Sum cash outflows                     155    175       624     192      76      69         68       68         68
Sum cash inflows                                               390     519     476        395      322        241
Net cash flow                         -155   -175     -624     198     443     407        327      254        173
Accumulated cash flow                 -155   -330     -954    -756    -312      94        421      675        848

Table 6.16 Cash flow statement for scenario 1 with higher tariffs
Scenario 1 - higher tariffs
(In USD millions)                     2002   2003     2004    2005      2006     2007       2008       2009     2010
Sum cash outflows                     155    175       624     192        76        69       68         68       68
Sum cash inflows                                               457       613       563      395        322      241
Net cash flow                         -155   -175     -624     265       537       494      327        254      173
Accumulated cash flow                 -155   -330     -954    -689      -152       342      670        923     1096

Table 6.17 Cash flow statement for scenario 1 higher tariffs and 3 GEOs
Scenario 1 - 3 GEOs
(In USD millions)                     2002   2003      2004      2005     2006       2007       2008     2009       2010
Sum cash outflows                     150    480       1480      379   131           116         115      115        115
Sum cash inflows                                                 457   920          1576        1186      965        723
Net cash flow                         -150   -480     -1480       78   789          1460        1071      850        608
Accumulated cash flow                 -150   -630     -2110    -2032 -1243           217        1288     2138       2746


Table 6.19 Cash flow statement for scenario 2
Scenario 2 Western Europe
(In USD millions)                     2002   2003     2004     2005     2006       2007     2008    2009 2010
Sum cash outflows                     155    175       624      192       76        69       68         68   68
Sum cash inflows                                                 46      147       314      589        896 1132
Net cash flow                         -155   -175     -624     -146       71       245      521        828 1064
                                                                                                            67
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                   Revision
Final Report                                         2000-11-29             A


Accumulated cash flow                  -155   -330     -954   -1100 -1029       -784   -263   566 1629


Fig 6.6 Comparison of the scenarios based on accumulated cash flows projections




In figure 6.6. the comparison of the different accumulated cash flows show that all scenario 1
projections break even by 2007, while the scenario 2 breaks even by the end of 2008.

5.5.4.2 Sensitivity analysis for scenario 1

A good rule of thumb in budget development for new ventures is to overestimate costs and be
pessimistic about revenues. This should be particularly the case when estimating future
revenues and costs of a system of so high risk as S-UMTS.

In the following sensitivity analyses the possible outcomes of a change in total inflow vs.
change in outflow is shown. For regional or 1 GEO systems the change in revenues ranges
from a reduction of 30% to an increase of 10%, and in expenses from an increase of 30% to a
decrease of 10%. For global coverage or 3GEOs scenarios the additional uncertainty is
accounted for by reducing income by 40% and increasing expenses by 40% (worst case).

The resulting cash flow scenarios, as well as the internal rate of return (IRR) and the net present value
(NPV) when discounted at 10%, are included in the tables for each projection.

Scenario 1 - Western Europe 1 GEO
                                                                                                    68
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                       Revision
Final Report                                         2000-11-29                 A



Table 6.21 Cash flow sensitivity analysis for scenario 1
                     2002 2003 2004 2005 2006 2007 2008 2009 2010
Cash inflows
Low: 30% less               0      0      0    273   363    333     277   225       169
Original prognosis          0      0      0    390   519    476     395   322       241
High: 10% more              0      0      0    429   571    523     435   354       265

Cash outflows
Low: 10% less             140    158    562    173    68     62      61    61        61
Original prognosis        155    175    624    192    76     69      68    68        68
High: 30% more            202    228    811    250    99     90      88    88        88

Net cash flow                                                                          IRR NPV
Low in - low out          -140   -158   -562   101   295    271     216   164       108 8% -42
Low in - no change out    -155   -175   -624    81   287    264     209   157       101 4% -148
Low in - high out         -202   -228   -811    24   265    243     188   137        80 -7% -468
No change in- low out     -140   -158   -562   218   451    413     334   260       180 24% 352
Original prognosis        -155   -175   -624   198   443    407     327   254       173 19% 246
No change in- high out    -202   -228   -811   141   420    386     307   233       153 8% -75
High in - low out         -140   -158   -562   257   503    461     374   293       204 28% 483
High in - no change out   -155   -175   -624   237   495    454     367   286       197 23% 377
High in – high out        -202   -228   -811   180   472    433     346   265       177 12%   57

The resulting best case provides a net present value (NPV) of USD 483 million and an IRR of 28%. In the
worst case the NPV is USD – 468 million with an IRR of -7%. This reaffirms the original scenario 1 as
only a moderately interesting investment object, considering the high risk factor and degree of
uncertainty inherent in the underlying assumptions. In the following two figures the above results are
shown graphically.

Fig 6.7 Cash flow sensitivity for regional coverage scenario 1
                                                                                                      69
Contracts number:
ESTEC - 13694/99/NL/US                           Final Report
ESA S-UMTS study                                 Date                   Revision
Final Report                                     2000-11-29             A




Fig 6.8 Range of accumulated cash flows for regional coverage scenario 1




As seen in Fig. 6.8 the break-even year may range from the best case in 2006 or in the worst case not at
all within the time period.
                                                                                                                   70
Contracts number:
ESTEC - 13694/99/NL/US                                   Final Report
ESA S-UMTS study                                         Date                         Revision
Final Report                                             2000-11-29                   A


Scenario 1 with higher tariffs

Table 6.22 Cash flow sensitivity for scenario 1 with higher tariffs - case 1 (45 kb/s)
                     2002 2003 2004 2005 2006 2007 2008 2009 2010
Worst case              -202   -228     -811      70     330      304    188    137        80 -3% -361
Original prognosis      -155   -175     -624     265     537      494    327    254       173 25% 399
Best case               -140   -158     -562     330     606      557    374    293       204 34% 652

The resulting best case provides a net present value (NPV) of USD 652 million and an IRR of 34%. In the
worst case the NPV is USD -361 million and the IRR is -3%. In the following figures 6.9 and 7.0 the
above results are shown graphically.



Table 6.23 Sensitivity analysis for scen 1 case 1 with higher tariffs and 3 GEO (45 kb/s)
                              2002 2003 2004 2005 2006 2007 2008 2009 2010
Worst case                       -210     -672 -2072       -256     368 783 551             418   273 -6% -1,214
Original prognosis               -150     -480 -1480         78     789 1460 1071           850   608 23%    926
Best case                        -135     -432 -1332        162     894 1630 1201           958   692 32% 1,462

As seen in table 6.27, the additional 2 satellites improves total earnings but the additional cash outflows
digs into the return of the investment. However, it should be noted that the results here accounted for
does not include future earnings beyond 2010.




5.5.4.3 Scenario 2 – Western Europe 1 GEO

Table 6.25 Cash flow sensitivity for regional coverage scenario 2 with car scenario
                                 2002 2003 2004 2005 2006 2007 2008 2009 2010
Worst case                       -202     -228    -811    -217       4    130    324       539 704 3% -337
Original prognosis               -155     -175    -624    -146      71    245    521       828 1064 19% 433
Best case                        -140     -158    -562    -122      94    283    587       925 1184 25% 690
                                                                                                 71
Contracts number:
ESTEC - 13694/99/NL/US                         Final Report
ESA S-UMTS study                               Date                 Revision
Final Report                                   2000-11-29           A



5.5.4.4 Summary of findings

When comparing different investment opportunities, the net present value expresses the
present value of an investment's future net cash flows discounted to the present at certain rate,
in this case 10%.

Figure 7.9 illustrates in sum the various outcomes for the different scenarios in terms of net
present value.

Scenario 1 with increased tariffs and added GEOs yields the best results at the medium and
high end, but carries the biggest downside at the low end, which shows NPVs based on
increased system costs and decreased revenues.

The basic regional scenario 1 and 2 appear to be only moderately interesting as investments,
leading to a negative return if the underlying business case assumptions are too optimistic.
However, the risks and possible returns are well within what is suggested for the terrestrial
UMTS build-out in the first 5-10 years of operation.




Fig 7.9 Sensitivity comparison of net present values
                                                                                                      72
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                      Revision
Final Report                                         2000-11-29                A




                                   Range of outcomes - internal rate of return

                           40
                           35
                                                                                          Scen1-3GEOs
                           30                                            Scen1-tariffs     Scen2-global
                                                                                           Scen1-base
                                                   IRR = 25%
                           25                                                              Scen2-base
                           20
    IRR (%)




                           15
                           10
                              5
                              0
                              -5
                         -10
                                           Low                    Medium                  High
              Scen1 base                    -7                      19                    28
              Scen1 tariffs                 -3                      25                    34
              Scen1 3 GEOs                  -6                      23                    32
              Scen2 base                    3                       19                    25
              Required return               25                      25                    25


Fig 7.10 Sensitivity comparison of internal rate of returns

Figure 7.10 illustrates a comparison of the different IRR levels for each of the scenarios. At the
Low end we see the outcome if cash outflows increase by 40% and cash inflows decrease by
40%, an extremely unfortunate situation if so should happen. Here the single GEO regional
scenario 2 still manages to maintain a positive return.

The High end shows the situation if revenues are increased by 10% and expenses decreased
by 10%. This would certainly improve the business case for all scenarios.

The Medium is the most likely outcome based on the assumptions used in this report.
However, if the yield of minimum 25% was the cut-off, none of the scenarios would be
considered acceptable. Scenario 1 with higher tariffs comes closest.

Overall it is apparent that the scenarios fare within a reasonably close range of each other, and
only somewhat would be of interest from an investors viewpoint.

However, other motives than investment return levels may support deployments of any of the
scenarios here since the rates projected do not argue against it.
                                                                                                           73
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                    Revision
Final Report                                        2000-11-29              A



5.5.5 Conclusions

5.5.5.1 Scenario 1

The market potential of scenario 1, when not limited by capacity issues, is significant and
capturing sizeable portion of it provides an attractive business opportunity.

None of the other scenario 1 projections yield returns that may be considered sufficiently
attractive to a savvy investor, as the risks involved in the satellite industry ideally should return
much higher yield levels. However, from a telecom strategic point of view, none of the
scenarios yield returns that are too low to rule S-UMTS out of an integration plan for overall
UMTS coverage.



5.5.5.2 Scenario 2

The basis of scenario 2 is niche pricing, and it may be quite risky to assume that a high pricing
scheme can be carried through all through 2010. Thus, scenario 2 carries an acceptable yield
only if the investor is an operator with objectives other than pure financial return.



5.5.5.3 A word of caution
Overall, the word of caution is not to view the business case as a lone argument for (or against) the
viability of S-UMTS.

In retrospect, the flaw in the business logic underpinning the Iridium and Globalstar projects was a failure
to correctly predict how fast and how thoroughly the burgeoning analogue and then digital cellular
systems, such as GSM, would capture the lucrative mobile market in high population density areas. The
speed with which Iridium and Globalstar were financially outpaced is also in part due to the long
development and implementation time for a satellite system compared to a terrestrial system. To
oversimplify, a transceiver can be put on a rooftop far faster than a satellite can be put in orbit. Moreover,
its technology can be more rapidly modified, as historically confirmed by the transition from analogue to
digital cellular systems that came about as Iridium and Globalstar were being developed.

The analyses of this Chapter have endeavoured to avoid pitfalls such as these. The figures employed are
realistic and are based in part on the known financial performances of systems now in profitable
operation, such as the terrestrial GSM. The focus has been on plausible scenarios for S-UMTS, based
on clearly identified user preference trends.

In summary, the analyses have shown that return on investment is tenuous, in view of the long time
frame and high risks involved. The assessment of revenues and costs together comprise factors
indicating a sufficiently viable business case for S-UMTS only for scenario 1 with higher tariffs and
additional satellite capacity.
                                                                                                           74
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                    Revision
Final Report                                        2000-11-29              A


5.3    Funding strategies
Satellite communication, like other sectors in telecommunications, is increasingly accommodated in the
private sector. S-UMTS may draw investors from many sectors, in two broad groups, finance and
industry.


5.3.1 Potential investors in the industrial sector
Traditional operators
Traditional operators may be divided into three categories: fixed, mobile and satellite operators.

Fixed operators
The fixed operators can by investing in S-UMTS seemingly operate as a mobile operator, allowing new
types of applications and services as well as global communication concepts. By making roaming
agreements, the fixed operators will earn additional revenue whenever traffic from S-UMTS is routed
from a gateway into their terrestrial network.

Mobile operators
Mobile operators may invest in S-UMTS to expand the area of coverage, to take advantage of satellite
broadcast and multicast capabilities, and to provide additional revenues accrued from increased traffic
whenever S-UMTS traffic is routed into their network. Satellites may also be used to transfer traffic from
the terrestrial network to the satellite segment and thereby releasing resources and capacity in the
terrestrial network. This is particularly attractive for applications and services where the natural broadcast
and multicast capabilities provided by the satellite can be exploited.

Satellite operators
Satellite operators may through S-UMTS offer a new generation of T-UMTS-compatible services and
applications to a larger customer base. Traditionally, satellite systems has often been based on
proprietary technology, but by offering T-UMTS compliance, new and yet unexplored markets are
opening up.

Content providers
Global content providers may use S-UMTS to build a global niche network. Applications and services that
could be provided include music (radio channels and downloading of music on request), Internet access
and navigational services such as global positioning and maps.

Car manufacturers
S-UMTS vehicle applications and services will include navigation, positioning and maps, car service and
maintenance, entertainment, Internet access and first aid (such as an alarm with a call-back routine if an
air bag is released). Some systems exist today that offer some of the applications mentioned above, but
new, more sophisticates systems will soon be possible, several via satellite.

Equipment manufacturers
Manufacturers of telecommunications equipment may invest in S-UMTS to gain exclusive rights to deliver
the needed equipment. This includes the manufacturers of terminals (handheld, palm-top, laptop and
car), gateways and network control centres. In the case of S-UMTS terminals, the manufacturers can
gain an advantage over producers of other UMTS terminals do to satellite interoperability.

Satellite manufacturers
                                                                                                         75
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


Satellite manufacturers have a self-interest in investing in the S-UMTS system. As an owner, the satellite
manufacturer will be able to influence the choice of which satellites will be used. Satellite manufacturers
need to be active in such ventures to position themselves in the market and procure contracts.

Launch companies and consortia
Launch companies and consortia usually possess considerable financial strength. They can use S-UMTS
as a way to access the market, promote their products and build a reputation.


5.3.2 Potential investors in the public sector
Government agencies
Considering potential governmental influence in S-UMTS, several issues can trigger the interest of
governmental financiers.

Defence agencies
In many countries, defence budgets are now being cut back. The new budget constraints may oblige
military forces to use more standard communications products, which may offer opportunity for private
communication systems. Instead of investing in planning and developing new systems and being
dependent on expensive equipment, an attractive solution to use commercially available communications
systems such as S-UMTS to provide reliable, global access to a communications network. In this case,
the military will impose strict requirements regarding security, availability, and capacity.

Health care
S-UMTS may improve services provided by the health care sector. In addition to the possibility of one
common communication platform, which provides secure global interconnection of hospitals and other
medical institutions, medical personnel can have vehicles equipped with an S-UMTS terminals to ensure
access when travelling in areas outside terrestrial coverage. Each vehicle can also be used as a base
station, with a Bluetooth radio interface to communicate with the terminals. Possible services and
applications include speech, transmission of images, video and data.

Public services
It is possible that similar systems will be implemented by the time S-UMTS is operational. As an
example, several fire, health and police departments plan to use the terrestrial trunked radio (TETRA)
standard to establish a common radio communications network. S-UMTS can be used as a supplement
to these systems, which enhances the system performance by providing new, improved services, in
addition to extended coverage.

International aid organisations
International aid organisations need reliable global coverage. This group will require similar services and
applications to those provided to the Health-care sector.


5.3.3 Inmarsat as a potential partner

Inmarsat is an example of a satellite system initiated by governmental funds. When it was
founded in the late 1970s, Inmarsat was a joint venture of governments, with their signatories,
in most cases the various national telecommunications administrations, contributing capital and
bearing the risk involved. However, Inmarsat has been privatised and become a limited
company.
                                                                                                            76
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A


Inmarsat has been successful, in part because many businesses are involved to provide Inmarsat-related
products and services. The relevant relationship is known as the Inmarsat partnership, which includes
international investor organisations, equipment manufacturers and their agents, distributors and system
integrators, software developers and service providers. Today, the Inmarsat partnership programme
includes about 1200 companies.

Inmarsat‘s strategy is to be placed in the convergence between computers, communications and
mobility. Inmarsat will still serve its traditional markets in the maritime, aeronautical and land mobile
communities, along with new markets for personal and multimedia mobile satellite communications.


5.4     Identification of strategic partners

5.4.1 The need for a broad approach
S-UMTS can not be developed by itself. It must be developed in partnership with other actors.
Consequently, several issues are involved. First, there must be a customer base and an access to the
targeted market. Second, all regulatory issues must be resolved, and roaming agreements must be made
to provide seamless global access. Finally, terminals of appealing design, supporting services and
applications that the customers are willing to pay for must be available.

5.4.2 Search for partners by sector
Financial institutions
The interest of major financial institutions is contingent in part upon the visible technological
trustworthiness of S-UMTS. Consequently, large companies, such as global network operators and large
satellite manufacturers, should be brought in to lend credibility to the project. Large network operators
have the size, capacity and competence to provide its services to the international market.

Satellite manufacturers
Partnerships with satellite manufacturers most likely will ensure access to the latest equipment. A
dialogue is needed between those specifying the system requirements and the manufacturers of the
satellite segment so that requirements reflect what actually can be delivered.

Terminal manufacturers
Terminal manufacturers should be involved to exploit the advantages of a common platform for the
terminals, and thereby reap the benefits of scale in production, including making terminals more
affordable on the market.

Network operators
Operators of both fixed and mobile telecommunications systems may be involved, with key advantages
in:
 Customer base. Established operators have their customer bases, which they can exploit in
    introducing an S-UMTS system.
 Market access. Established operators can provide an easy access to the targeted market. Operators
    also have existing sales and marketing structures, and they can also provide operation and
    maintenance functions. This can be used to further speed up the S-UMTS implementation.
     Network infrastructure. Established operators usually have proven network infrastructures,
      with roaming and routing facilities. These facilities can be used in implementing S-UMTS.
      For instance, mobile operators may gain through allowing roaming into their networks as
      well as being thereby able to deliver value added services to their customers. Likewise,
                                                                                                           77
Contracts number:
ESTEC - 13694/99/NL/US                            Final Report
ESA S-UMTS study                                  Date                    Revision
Final Report                                      2000-11-29              A


    fixed operators can open their networks in ways that permit routing traffic from gateways
    into terrestrial networks.
   Licensing. Established operators usually have one or more licences, and consequently can assume
    responsibility for further licensing, as of S-UMTS, within their spheres of operation.

Satellite operators
Satellite operators are important partners because of their experience and knowledge in planning,
developing and operating systems. Moreover, many have terrestrial infrastructures that can be used for
S-UMTS purposes.

Content providers
In S-UMTS it is vital to develop good applications and services that the users are willing to buy. The
limitations of S-UMTS terminals imply that special applications are needed to attract content providers.

Car manufacturers
Involving car manufacturers may help generate a sizeable S-UMTS customer base. Co-operation with
content providers would ensure the development of useful applications and services.

5.4.3 Co-operation scenarios
The challenge
S-UMTS arguably cannot survive as a stand-alone system. But through co-operation, it has a better
chance of success. Several co-operation scenarios have been identified.

Inmarsat-4
Inmarsat-4 is scheduled to be fully operational in 2004 and to be IMT 2000 compliant. Consequently,
Inmarsat could be a major competitor that would capture some of S-UMTS targeted customer base. So it
would be advantageous for S-UMTS to co-operate with Inmarsat and thereby offer complementary
services and applications. Inmarsat plans to provide UMTS compatibility but at present not S-UMTS
compatibility. However, it is possible that Inmarsat also may migrate to S-UMTS as it is standardised.

Fixed operators
A fixed operator co-operating with S-UMTS can offer new types of services and applications (value-
added services) and increase the coverage. Consequently, a fixed operator can seemingly operate like a
mobile operator without investing in a network, thus expanding market share. In addition, the operator will
earn additional revenue due to the traffic that is routed from the gateways and into the fixed network.


T-UMTS operators
T-UMTS operators can co-operate with S-UMTS in order to expand their coverage and to offer mobile
broadband services to a larger customer base. Roaming agreements will also provide additional revenue
when S-UMTS subscribers use the terrestrial network. T-UMTS operators may also exploit the broadcast
and multicast capabilities of satellites to offer new sets of applications and services. Moreover,
transferring traffic to the satellites may help relieve capacity in the T-UMTS network, whereby S-UMTS
can be used to support early implementation of T-UMTS in some areas.

GSM/GPRS operators
Co-operating with S-UMTS will enable GSM/GPRS operators to offer UMTS services and applications on
a global scale. GSM/GPRS coverage will be increased, and the operators will be able to offer new and
improved applications and services. S-UMTS can be used as an early rollout for T-UMTS.
                                                                                                             78
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                    Revision
Final Report                                         2000-11-29              A


Car manufacturers
Car manufacturers may co-operate with S-UMTS to provide their customers with access to tailor-made
broadband applications. The cars will be equipped with dual mode terminals that handle both terrestrial
and satellite communications as well as support global coverage.
However, many car manufacturers already offer or soon will offer services and applications similar to
those of the S-UMTS car scenario. The question then is whether the long time to market for S-UMTS will
lessen the attractiveness of it for car scenario investors/users, or whether alternative, financially less
risky technologies may be available by the time an S-UMTS system is operational.


5.4.4 Where and how to open for competition
Mobile users of satellite communications need not depend on local infrastructure. So the opening for
competition is determined upon where the potential customer base is located. If S-UMTS is used to effect
an early implementation of T-UMTS, the satellites must provide dynamic resource allocation. As soon as
T-UMTS coverage is established in the area of interest, there is no need for a large S-UMTS capacity,
and this capacity can be transferred to other areas. The S-UMTS system is designed to provide an early
entry of the UMTS market and to allow its operator a competitive edge in this attractive market before the
opportunity is lost.

There are several issues to consider in assessing the routes that open for competition.
Local operators in every country. It could be an advantage to have local operators in every country of
interest. These operators often have access to the market, and can handle marketing and sales, along
with the operation of the local networks. Each of these national operators is responsible for obtaining the
necessary national licences in each of the nations of interest.
Mobile operators. Will terrestrial mobile service providers be permitted to offer S-UMTS services, or will
S-UMTS be reserved for competing service providers? By allowing terrestrial mobile service providers to
offer S-UMTS services, it can be easier to obtain the roaming agreements required.
One satellite operator. An alternative to have local operators in every country is to have one global
satellite operator that provides all services (such as operation, marketing and sales).
Virtual operators. Should virtual operators be allowed to access the network? Virtual operators produce
and sell services without owning a network, and they sell access to the network through their subscriber
identification module (SIM) cards. It is not certain that the market is large enough or sufficiently profitable
to attract virtual operators. There is considerable risk connected with entering this market, so if the
anticipated profit is marginal, virtual operators may not be interested.


5.4.5 Recommended approach
S-UMTS should base the business case on capturing the horizontal market. The services offered should
be similar to the services offered by T-UMTS, and additional services should include specific car and
broadcast applications. The economies of scale can also be achieved in this case by extended reuse of
T-UMTS technologies.

For S-UMTS to survive, several conditions must be present and must interplay, as illustrated in the
diagram of Fig. 8.1.




           Content           Portal           Network          Access            Terminals
                                                                                                   79
Contracts number:
ESTEC - 13694/99/NL/US                          Final Report
ESA S-UMTS study                                Date                    Revision
Final Report                                    2000-11-29              A



Fig. 8.1 Conditions and their interplay essential to S-UMTS survival.


Content and portal, including the applications and services that will be offered to the customer, and
could include companies, such as Time Warner, AOL, large software houses, and European car industry
firms, such as Mercedes and BMW.

Network and access. To handle the network and access aspects, a co-operation between one or more
large operators must be present. This could also be a combination of fixed, mobile, and satellite
operators.

Terminals. Companies such as Ericsson and Microsoft could address the terminal aspects, which
include availability, functionality and design.


5.4.6 Conclusions
Private funding is the most viable approach to make S-UMTS happen. The main financiers should be one
or more large operators. Due to the high prices of T-UMTS licences, operators may gravitate toward S-
UMTS as an alternative way of supporting UMTS services, thus S-UMTS may become attractive to the
losers at T-UMTS licence auctions. The remainder of the funds might be raised from car manufacturers,
banks, and content providers.
Leading terminal manufacturers may be reluctant to become involved with S-UMTS, due to the small
production levels and the risk implied by the recent failures of other satellite systems, most notably
Iridium and Globalstar.
In order to have a chance of success, S-UMTS must be approved by several governmental agencies,
including the European Commission (EC) and the European Space Agency (ESA). Part of the approval
entails ensuring the allocation of the requisite part of the UMTS frequency bands necessary for S-UMTS
and not for T-UMTS operations. In addition, EC and ESA are important players in standardisation
activities.
                                                                                                           80
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                    Revision
Final Report                                        2000-11-29              A




6 Identification of key technologies for S-UMTS
This section of the document deals with the technologies that will need to be addressed and developed
before a successful S-UMTS system can be implemented.

The final report from WP5200 concentrated on the hardware that is critical for this but there is an
important analytical study that will need to be carried out if the new hardware is to be effectively applied
to the system. This analysis involves the design of the S-band antenna which defines the beams
covering the zones which are complementary to the cells in the terrestrial system This topic will be dealt
with in the following section.

6.1    Gateway subsystems

6.1.1 Key technologies
We can see the following key technologies for S-UMTS ground network/ gateways:
 Resource management for efficient use of the resources in a system with a large number of beams,
   variable capacity needs between beams, terminals with varying QoS demands and services with
   constantly changing resource needs.
 IP traffic via satellite. Interface with T-UMTS core network. How and where to locate any TCP/IP over
   satellite adaptations/ enhancements.
 QoS control for services provided in the satellite network. The satellite access network can in many
   cases not support the same QoS parameters as the terrestrial network (delay, bitrate). Support of
   services between networks with different capabilities.

6.1.2 Challenges
Some key challenges are:
 Satellite gateways will be produced in relatively small numbers. They shall be connected to core
   network infrastructure developed for T-UMTS. The T-UMTS market is very much larger. The testing
   of satellite infrastructure is as complex as for terrestrial. A very effective test approach must be done.
   Is it cost effective to ensure that a gateway can be connected to any vendors core network as done
   for terrestrial?
 T-UMTS standards converted to S-UMTS needs verification and comprehensive testing. A satellite
   standard may have a reduced foundation compared to terrestrial.
 Iu interface compatibility. Will the Iu interface specifications be adapted to take into account any
   needs emerging from S-UMTS specification work, and will it be done in time?
 Terrestrial UMTS has significantly larger teams working on the development of UMTS than what is
   possible for S-UMTS, as the S-UMTS market is so much smaller than the terrestrial market.
   Therefore there is an additional challenge for S-UMTS developers to do similar work as for terrestrial
   for a reasonably low cost with much fewer people.

6.1.3 Standards
There are currently no standards for S-UMTS, but work is in progress. However, the work is behind that
of 3GPP and the terrestrial segment. The ‗drive‘ is very high in T-UMTS and it is likely that S-UMTS
services can only be introduced years after T-UMTS.
ETSI has a small WG under TC SES working with S-UMTS. The number of people and the resources
available for S-UMTS standardisation is much less than what is the case for the terrestrial segment.
Therefore one has to take the approach to follow and not to lead. However, adopting terrestrial
specifications to satellite directly may be sub-optimal, and one could also risk poor performance in some
                                                                                                            81
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A


aspects unless things are properly simulated and tested with field trials. Further, full interconnectivity and
inter-working with the terrestrial segment is difficult to test before the terrestrial segment is there.

Standardisation work at ETSI is industry driven, and unless there is a drive from the ETSI members to do
the work it will not be done. There seems however to be a reluctance from many potential players to get
into satellite UMTS at this point in time. This situation may delay S-UMTS operations


6.2    Satellite platform, payload and antenna

6.2.1 S-band Equipment

6.2.1.1 The S-band Antenna Analysis
If the capacity of the S-UMTS system is to be fully utilised the problem of inter-beam interference must
be solved.

The proposal for the air interface for S-UMTS is a wide-band synchronous CDMA system. In such a
system all the user channels use the same frequency slot and the only isolation between them is
achieved by the orthogonality of the codes. In the case of any one beam all the codes used in that beam
are drawn from a set which because they are synchronous do not interfere with each other. In the case of
the other beams they will also use a common set of codes but they will not be synchronous and will
cause interference in the classical manner of CDMA system.

Thus, the design of the antenna must be such that the total interference into one beam of all the other
beams must be at a level which will not reduce the capacity of the first beam in terms of the number of
user channels it is capable of carrying. This factor not only effects the acceptable sidelobe levels in any
beam but also the gain at the overlap region of the beams.

It is normally assumed that the classic beam shape from a circular aperture will be sufficient but this may
not give a fast enough role-off of gain from the peak to provide the necessary isolation. The possibility of
using a shaped beam giving something like a quasi-Chebychev shape factor may have to be examined.
There is no doubt that this would reduce the absolute gain of the antenna and the EIRP of the payload
but it could allow the channel capacity to be more effectively utilised.

6.2.1.2 The S-band Antenna Reflector.
The S-band antenna will involve the use of a very large reflector with a diameter of 10 metres or more.
This will need to be stowed during launch and mechanically deployed in orbit. This is a highly specialised
field which has been abandoned in Europe. However there are two manufacturers in the US who have
been working in this field for the United States Department of Defence and who have been released to
bid on the Inmarsat 4 programme. These companies are Harris Corporation of Florida and TRW
Aerospace of California. Both companies have a record of design and fabrication of this type of antenna
spanning decades.

6.2.1.3 The S-band Antenna Feed Elements.
The S-band reflector must be illuminated with a matrix of feed elements in order to generate the multiple
beams required of the S-UMTS system.

This is a problem familiar to the designers of mobile satellite antennas but at L-band. The size of the
elements or at least the area occupied at the base of the radiator is important because of the limited
                                                                                                       82
Contracts number:
ESTEC - 13694/99/NL/US                            Final Report
ESA S-UMTS study                                  Date                    Revision
Final Report                                      2000-11-29              A


amount of space available on any launch platform. In the case of the Euro 3000 bus this limits the
number of elements to 120 using cup backed helices perhaps the most efficient element in this respect.
The gain of the element is also important as it has to be related to the requirements of the reflector.

Another important factor governing the choice of element is the generation of passive inter-modulation
products. This again points to the helical element due to its simple structure with very few welded or
soldered joints. It is fortunate in the case of the S-UMTS system that the order of inter-modulation
products from the transmit band which will cause problems on receive is the 13th which is considered
comparatively safe. Also since the Frequency power distribution of the W-CDMA signal is spread and
noise like in its structure the inter-modulation products will also be noise like but of much greater
bandwidth reducing still further the risk to the system.

There are a number of companies working on this topic but at L-band. These are mainly ESA funded
programmes such as ―The Advanced Array Assembly‖ (AAA) by Saab Ericsson, Goteborg and
―Advanced Mobile Payload Equipment‖(AMPE) by DSS ,Ottobrun. Work has also been carried out by
EMS Technologies, Montreal who have been working on the elements for Inmarsat 4 and the Telesat,
Canada spacecraft Anik F3.

6.2.1.4 S-band Solid State Power Amplifiers
The design of solid state power amplifiers at L-band is very mature with the equipment of Inmarsat 3 with
devices designed and manufactured by Astrium (the old Matra Marconi). The plan is for the company to
manufacture the over 400 amplifiers required for the Inmarsat 4 programme.

These amplifiers will be to a new design using GaAs FET devices and great caution is being exercised to
avoid any problems analogous to the Mc Donald effect which effects bi-polar devices. This is particularly
important in the case of S-UMTS with the noise like characteristics of the W-CDMA signals giving a very
high peak to mean probability power distribution. This fact and the need to abandon the conventional ‖
Noise Power Ratio‖ test in favour of the ‖Walsh Power Ratio‖ will mean that the back-off of the operating
point of the amplifier will be even more critical.

Because of the closeness of the two frequency bands L- and S- the transition of the design to the slightly
higher frequency should not pose any real problems.

6.2.1.5 Surface Acoustic Wave Filters
The design of Surface Acoustic Wave (SAW) filters at frequencies around 200 MHz is now very mature
and it is unlikely that any further work will be needed in this area.

However, the need to come down to an intermediate frequency to carry out this filtering adds greatly to
the complexity of a transponder using conventional analogue channelization. It is unfortunately
unavoidable as the first step in channelization of a digital processing system.

The concept of transferring the beam forming network from the satellite transponder to the ground has
been mooted as a possible way to reduce the complexity of the transponder. In this case a complete
transponder path is provided through the payload for each antenna element. In producing such a design
the possibility of carrying out Channelization at S-band is very attractive. The drawback is that the
technology for producing SAW filters at S-band is possible but the detailed performance of the filters is
not good enough at the moment with the transition region being excessively wide.
                                                                                                          83
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                    Revision
Final Report                                        2000-11-29              A


6.2.2 Ka-band Equipment
There would appear not to be a technology problem with the Ka-band equipment since this discipline is
now very mature.

If there is to be problem it will be in fabricating travelling wave tube amplifiers with the output power
levels required for the return link of the S-UMTS transponder. The possibility of connecting tubes in
parallel is an interesting prospect but this is a power supply problem and not an RF problem.

6.2.3 The Digital Processor
The digital processor which was proposed for the S-UMTS transponder was a transparent design
comparable with that to be implemented on the Inmarsat 4 system. The S-UMTS system is unique in the
size of the task the processor is to be asked to perform in terms of the number of potential channels to be
served.

The use of even partial despreading of the W-CDMA system for routing purposes was rejected on the
basis of complexity.

There is ongoing work in this field in many companies both in Europe and the US where the military have
been applying these techniques for several years. A great deal of internal R&D is being carried out in the
laboratories of Astrium in Stevenage who are in the process of developing ASICs for A/D converters and
other elements in the digital processing scheme. ESTEC are aware of the problem with their recent issue
of an RFP for work on this topic so little further can be said.


6.2.4 Frequency Generators
The generation of a coherent set of frequencies from a single highly stable source is one of the mature
features of a satellite communications payload. In contemporary systems the local oscillator signals are
all generated in a single unit and distributed by coaxial cable to the user units.

This approach is not entirely appropriate in the case of the S-UMTS transponder where many of the local
oscillator signals are themselves in Ka-band (20GHz or higher) and at frequencies which cannot be
carried by conventional co-axial cable and SMA connectors. The result is the process of generating the
local signal will need to be a distributed process with the signal taken to the unit at a lower frequency and
the final multiplication taking place in the user unit.

6.3    User terminal base-band, radio frequency front-end and antenna


6.3.1 Introduction
The following terminal types have been studied:
 Handheld terminals
 Transportable terminals (palmtop-sized)
 Transportable terminals (laptop-sized)
 Vehicular terminals

The terminals have also been grouped into classes as follows:
 Single-mode S-UMTS terminals
                                                                                                                   84
Contracts number:
ESTEC - 13694/99/NL/US                                Final Report
ESA S-UMTS study                                      Date                      Revision
Final Report                                          2000-11-29                A


   Generic multi-mode terminal, capable of connecting to a arbitrary number of systems
   Dual-mode S-UMTS/T-UMTS terminals
   Dual-mode S-UMTS/GPRS terminals
   Tri-mode S-UMTS/T-UMTS/GPRS terminals

Four different subsystems of the terminal architecture are considered:
 RF front-end ( including antenna)
 Up and down conversion (including A/D and D/A conversion)
 Base band processing (BBP) unit
 Connection between units


6.3.2 Terminal types
In Table 1, the principal technical parameters of the terminals are summarised.
Table 1 Key parameters of antenna
                              Handheld             Palmtop-sized          Laptop-sized            Vehicular
                              terminal                terminal              terminal              terminal
Average transmit
                                0.6 W                    1W                     1W                   2W
power
Max. burst power                 2W                     4W                     4W                    4W
                            El: cardioidal        2 elements: 3 dB           3 dB: 40           El: Covering
                                                     40 degrees              degrees            5-85 degrees
Antenna diagram
                             Az: Omni-            1 element: 3 dB            3 dB: 40             Az: Omni-
                             directional             80 degrees              degrees             directional
Antenna gain                Peak: 0.5 dBi
(including feeder          Average: -1 dBi              7 dBi                 10 dBi                 4 dBi
loss)
Receiver G/T                  -25.5 dB/K             -18.5 dB/K             -15.5 dB/K             -19 dB/K


6.3.2.1 Handheld terminals
A display of a decent size is incorporated to show images and video. Examples of such screens are
those on camcorders and PDAs. In addition to containing a user interface, the handheld terminal is likely
to contain a short-rang wireless interface (SRWI) enabling it to connect to other devices like PDA‘s or
laptop PCs. However, the main usage of the terminal will be via the user interface, not via the SRWI.

The terminal should be multi-mode permitting the terminal to connect to terrestrial networks. In urban
areas, users would probably connect to terrestrial network due to lack of line-of-sight (LOS) to the
satellite, higher data rates offered by the terrestrial network, or lower price of terrestrial services. In-call
hand-over between satellite and terrestrial modes must be provided.


6.3.2.2 T-UMTS terminal with extended S-UMTS capabilities
The idea behind this terminal type is to increase the potential S-UMTS market by providing access to
certain S-UMTS services on T-UMTS terminals. If manufacturers of T-UMTS terminals were persuaded
to incorporate S-UMTS functionality in all T-UMTS terminals, the potential market for S-UMTS would
increase tremendously.
                                                                                                           85
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                    Revision
Final Report                                        2000-11-29              A


Due to the low G/T and EIRP of T-UMTS terminals, only low data rate services like messaging may be
offered in S-UMTS mode. This will be in contrast to the capabilities in T-UMTS mode.

Transportable (or nomadic) terminals are assumed to be fixed during communications. Two sizes are
defined: palmtop-size and laptop-size. The palmtop-sized terminal is about 10 x 17 cm, the laptop-sized
terminal is about twice as large (20 x 20 cm).

Short-range wireless interface (SRWI)
The short-range wireless interface may be Bluetooth, DECT, or similar. Some base band processing will
be done in the S-UMTS terminal and some in the connected device. One alternative would be to do
everything, including modulation/demodulation in the terminal, and merely use the connecting device as
a user interface. Another alternative is to use the terminal as a repeater, leaving as much as possible of
the base band processing to the connecting device. The first alternative might be optimal if the device is
a PDA with very little processing power. The second alternative might be best it the device is a T-UMTS
terminal, especially if the S- and T-UMTS standards are very similar so that the T-UMTS functionality‘s
could be used also for S-UMTS.

Another option would be to use an S-UMTS interface. This seems however not to be a good idea for
several reasons. Firstly, such an interface will probably not be implemented in many PDAs, laptop PC
and similar devices, limiting the usability of the terminal. Secondly, handheld S-UMTS has a low data rate
capability compared to the transportable terminals. Hence, the capabilities of the terminal would not be
fully exploited. Finally, the receiver chain of the terminal is very sensitive, and positive feedback from the
short-range radio transmitter would occur and damage the terminal.

Services
The palmtop-sized terminal will support data rates up to 144 kbps. Its capabilities should include web
browsing, audio streaming, video/data streaming, real-time video, data transfer (bulk), broadcasting, and
e-commerce.
The laptop-sized terminal should support all the services provided by the S-UMTS system.

6.3.2.3 Vehicular terminals
The antenna is fixed to or integrated into the car body. The best position is probably on the roof, although
it might be a problem to put a ski box or something else on top of the antenna. Another possibility is to
place the antenna on the hood, or to scatter antenna elements over several parts of the car body. At
locations where the elevation angle to the satellite is low, the latter alternative has advantages, as one
antenna element always would be directed more or less towards the satellite. Several antenna elements
would however increase the complexity and cost of the terminal.

The rest of the terminal is located inside the car. The size of the terminal should probably be
standardised in the same way as car stereos are standardised today.

The terminal contains a SRWI similar to that of transportable terminals. Hence, users inside and in the
vicinity of the car may use it to connect their portable PC, T-UMTS terminal of some other device to the
S-UMTS system.

The vehicular terminal is a multi-mode satellite/terrestrial UMTS terminal. In urban areas, the terminal
connects to terrestrial networks (T-UMTS, GPRS), while it connects to the S-UMTS system when it is
outside the coverage area of terrestrial networks. It is also possible that the S-UMTS mode is preferred
when both S- and T-UMTS networks are available. In any case hand-over between terrestrial networks
and S-UMTS will occur.
                                                                                                         86
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


There are already a number of systems offering emergency and convenience services to cars. The
services may include:
 Emergency services, including air bag deployment notification, remote door unlock theft protection
    etc.
 Convenience services, including listing of hotels, restaurants etc. with online reservation possibilities.
 Route support providing customers with directions to find shortest route, closest gases station or
    ATM, avoiding jams etc.
 Radio and possibly TV broadcasting and multicast
 Other Internet or Intranet services, including e-mail and web browsing

The required data rate for TV broadcasting may be too large for the available spectrum in S-band,
especially if a number of programs are to be provided. Small screens may however reduce the needed
bandwidth.

6.3.3 S-UMTS signal parameters
A set of signal parameters for the S-UMTS terminal is presented to compose a framework for the
terminal work. The transfer frame composition is assumed to be an adaptation of the T-UMTS (IMT-DS).
T-UMTS parameters are used as reference values.

6.3.3.1 T-UMTS parameters
The air-interface parameters of interest we refer to T-UMTS FDD specifications. They represent the
December 1999 version and are subject to changes and enhancements.

6.3.3.2 S-UMTS parameter adaptation
In this section we briefly look at T-UMTS parameters in light of the S-UMTS environment. It is mainly RF
parameters that are affected, since the basic transfer frames and corresponding parameters are the
assumed to be almost the same for S-UMTS and T-UMTS.

Frequency range values for T-UMTS FDD and S-UMTS is given in Table 6.2.


                    Table 6.2 Frequency parameters of T-UMTS FDD and S-UMTS
                                               T-UMTS                           S-UMTS
Up-link freq. range [MHz]                     1920-1980                        1980-2010
Down-link freq. range [MHz]                   2110-2170                        2170-2200
 freq. max [MHz]                                250                              220
 freq. min. [MHz]                               130                              160


The T-UMTS has a system bandwidth of 60 MHz, while the S-UMTS has a bandwidth of 30 MHz. For a
dual mode UMTS terminal this may affect the RF filtering requirements depending upon the
implementation architecture. For a single mode UMTS terminal the S-UMTS requirements in terms of Q-
factor is more relaxed.

The frequency span of the T-UMTS is both larger and smaller than for the S-UMTS. This implies a
possible reuse of the T-UMTS channel frequency synthesiser. The shifted centre frequency may impact a
redesign for S-UMTS depending upon the up and down converter architecture.

The S-UMTS and T-UMTS systems will offer different maximum user bit rate ranges with similar
minimum bit rates. Here we look at the signal level differences at the antenna connector.
                                                                                                         87
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A



With a 0 dBi antenna gain, the EIRP classes for S-UMTS and T-UMTS differ. We have the ranges of the
maximum EIRP in Table 6..

                        Table 6.Range of maximum EIRP with 0 dBi antenna.

                                     S-UMTS                                T-UMTS
Highest class [dBm]                  36                                    33
Lowest class [dBm]                   36                                    21

We see that the satellite part is expected to have a more powerful output amplifier.

A T-UMTS terminal has a dynamic range of 80 dB. This is for severe fading environments. An S-UMTS
terminal will not experience such a large dynamic signal range, since the relative distance variations is
much lower. The propagation conditions include either a large line-of-sight components or the signal is
blocked with loss of communications as a result. Required dynamic range is therefore expected to be
less than 10 dB that is considerably smaller than the terrestrial range.

6.3.4 Generic terminal
The term generic terminal reflects a signal flow description and function review from antenna to base
band. For convenience, three major sub-systems have been defined: the antenna/RF subsystem, the IF
subsystem, the analogue-to-digital (A/D) and the base band unit. The terminal architecture described in
this section in generic, and may apply to T-UMTS terminals as well as S-UMTS terminals. This is
particularly the case for the analogue part, as the RF front-end does not differentiate between different
standards.

The analogue part of the receiver consists of the antenna unit, RF front-end and up/down conversion
chains. In Figure 6-1, the receiver chain is illustrated. Between the antenna unit and the LNA there is a
pre-selection filter H1 with bandwidth BRF equal to the system bandwidth of the selected mode. The LNA
must match the required bandwidth of the selected system.
The down converter may contain one or two stages. The bandwidth of the output filter H3 depends both
on the A/D converter (ADC) and the signal of interest. The minimum value is given by the signal
bandwidth BS, and the maximum value is fs/2, where fs are the sampling frequency of the ADC. A
possible advantage of increasing the bandwidth significantly above BS is that part of the channel
selection then can be done digitally in the base band unit. Moreover, two different modes could share the
same down-converter chain if the bandwidth(s) of H3 (and H2) are dimensioned for the mode with the
largest signal bandwidth. A disadvantage of using a larger bandwidth than necessary is that the
requirements to the dynamic range of the ADC will increase because of the extra out-of-band signals and
noise that will be present.
There exist several trade-offs between the various parts that must be exploited in further detail in a final
design. It is also necessary to take into account the expected performance (sampling rate and resolution)
of the AD/DA converters.
                                                                                                                           88
Contracts number:
ESTEC - 13694/99/NL/US                                    Final Report
ESA S-UMTS study                                          Date                               Revision
Final Report                                              2000-11-29                         A


                     RF FRONT END    DOWNCONVERTER

                                       fc1                                        fc2
           From
           Antenna                                                                                            To
                               LNA                                                                            AD/DA Unit
           Unit
                H1                                 H2                                         H3
                                             f2                                         f3
      f1
             Bandwidth = BRF                      Bandwidth  BS                             Bandwidth  BS

                                                        Optional coverter stage



 Figure 6-1 RF/IF Unit. Frequency f2 represents the optional first IF of a double conversion chain.
                               fc2 is the corresponding second LO

To parameterise the elements of Figure 6-1 and to refine the structure, a set of aspects must be
considered. Among these are:

   Direct down-conversion. In the case of base band sampling, direct conversion from f1 to DC is a
    possibility.
   Choice of the different IFs. To obtain requirements for the analogue filters that are not too
    demanding from an implementation point of view, the IFs must be selected with some care. This
    involves also the various inter-modulation products produced in the conversion processes.
   The number of IF stages. With a very high or variable input RF frequency it might be necessary to
    use a double conversion chain, by including the optional stage of Figure 6-1.
   Distribution of power gain among the stages. Linearity is important when low signal levels are
    involved. For instance, a CDMA system with a possible negative SNR value at the RF input should
    have a linear down converter to avoid a further reduction of the SNR due to non-linearity.
   Automatic Gain Control (AGC). In order to relax the requirements on the ADC, the dynamic range
    of the signal could be reduced. Adjusting the gain between one of the stages in the RF/IF unit does
    this. The actual control mechanisms should be implemented in the base band unit.
   Interface to the ADC. The signal bandwidth determines the necessary sampling rate. The
    alternatives are base band and IF sampling.

6.3.5 Connections between units
There are several reasons why S-UMTS terminals must provide connectivity to other units:

   S-UMTS terminals must either be placed outdoors or close to a window that happens to have a view
    towards the satellite. This restriction will in many cases make it inconvenient for a user to operate the
    terminal.
   Some S-UMTS terminals may be designed without user interface
   The customer may use S-UMTS to connect his portable PC or another device to Internet or other
    backbone infrastructure
   Terminals may serve as hubs for several users.
   S-UMTS terminals may provide extended coverage for T-UMTS or other terrestrial networks (GPRS,
    EDGE, GSM, or other)

Which interfaces to incorporate into an S-UMTS terminal, depend on its service and application
capabilities and on the targeted customer group. Some of the selection criteria are:

   Distance between the communicating units
   Whether or not line-of-sight between units will be available
   Maximum data rate required
                                                                                                            89
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A


   Maximum number of users connecting to the S-UMTS terminal
   Which interfaces that are incorporated in PCs, PDAs, mobile phones and so on
   International frequency regulations
   Security, immunity against eavesdropping
   Interference robustness
   Power consumption
   Size of components
   Cost of components


6.3.6 Terminal classes

6.3.6.1 Single-mode S-UMTS terminals


                     S-UMTS terminal                                          Device containing MMI


       RF front-      Up/Down          BBP       SRWI                      SRWI         BBP
         end         conversion        unit    tranceiver                tranceiver     unit          MMI




 Figure 6-2 Radio chain structure of a single mode S-UMTS unit. MMI is assumed to be a part of a
      stand-alone unit capable of interconnection with other communication standard unit.

Single-mode S-UMTS terminals have no capabilities to connect to terrestrial communication systems. A
schematic block diagram of such a terminal is shown in Figure 6-2. The transportable terminals are
examples of such terminals. Unless several S-UMTS modes are incorporated, integration of
satellite/terrestrial modules, software defined radio vs. predefined architecture etc. are not issues for this
terminal class.

The terminal is not assumed to be a stand-alone unit, so it must be connected to at least one other unit
by means of a short-range wireless interface. In principle, the only restriction on the device is that it
contains the same SRWI interface as the S-UMTS terminal. In this study only devices containing a user
interface are considered.

The S-UMTS terminal may also be part of a network, for instance a Bluetooth Pico-net, where several
devices may use the terminal as a hub to connect to the S-UMTS system. In this study the possibility of
several devices to connect to the S-UMTS terminal simultaneously is not considered. Several devices
may however have the possibility to connect to the terminal when it is not already connected to another
device. In that respect authentication is an important issue. To prevent unauthorised use of such an S-
UMTS terminal, it is required that the connecting device is a legal user. This must be verified through an
authentication handshaking procedure.
                                                                                                                90
Contracts number:
ESTEC - 13694/99/NL/US                                    Final Report
ESA S-UMTS study                                          Date                      Revision
Final Report                                              2000-11-29                A


6.3.6.2 Multi-mode terminals




                                                                                 SRWI
                              Terrestrial   Terrestrial          Terrestrial
                                                                               tranceiver

                               Satellite     Satellite            Satellite
                                                                                 MMI
                            RF front-end    Up/down                BBP



Figure 6-3 Multi mode terminal radio chain structure. The terminal may be a stand-alone unit with
    integrated MMI or it may be like the S-UMTS single mode terminal equipped with a SRWI


Table 6.3 Comparison of key parameters of S-UMTS, T-UMTS (IMT-DS), GSM-GPRS
                    S-UMTS                      T-UMTS (IMT-DS)                   GPRS
Frequency           Up: 1980-2010 MHz           Up:1920-1980 MHz                  Up: 890-915, 1710 – 1785
band                Down:2170-2200 MHz          Down: 2110-2170 MHz               (Europe)
                                                                                  1850 - 1910 (USA) MHz
                                                                                  Down: 935-960, 1805 – 1880
                                                                                  (Europe), 1930 - 1990 (USA)
                                                                                  MHz
Access
technology          DS-CDMA                     DS-CDMA                           TDMA

Duplex              FDD                         FDD
                                                                                  FDD
method
RF channel          Typically 5.0 MHz           Typically 5.0 MHz                 200 kHz
spacing


Multi-mode terminals consist of an S-UMTS module and at least one module supporting a terrestrial radio
system (an IMT2000 mode, GPRS, EDGE, GSM, IS-95 or others). The modules share the same MMI
(keypad, display and menu functions) if the terminal has one, and the same SRWI. In Figure 6-3, a
generic multi modus terminal is illustrated. Each sub-system has a terrestrial part and a satellite part. The
two parts may be completely separated in different modules, or completely integrated into one. In most
cases the situation will be somewhere in between the two limiting cases. An important factor related to
the integration is the chosen RF/IF architecture. This heavily influences the A/D and D/A interface and
further the BBP structure.


6.3.7 Technology roadmap
Traditional scaling, which has been at the basis of the semiconductor industry for the last 30 years, is
beginning to show the fundamental limits of the materials constituting the building blocks of the planar
CMOS process. However, new materials can be introduced in the basic CMOS structure to replace
and/or augment the existing ones to further extend the device scaling approach. Since the assimilation of
                                                                                                          91
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


these new materials into the modified CMOS process gives the device physicist and the circuit designer
improved electrical performance similar to the historical trends, this new regime has been often identified
as ―Equivalent Scaling.‖ It is expected that these new materials will provide a viable solution to extending
the limit of the planar CMOS process for the next 5–10 years.

Despite the use of these new materials, it will be challenging to maintain a rate of improvement in
electrical performance of about 2× every two years in the high-performance components by relying
exclusively on improvements in technology. Innovation in the techniques used in circuit and system
design will be essential to maintain the historical trends in performance improvement. To achieve this
result it is expected that the integration of multiple silicon technologies on the same chip and a closer
integration of package and silicon technology will be necessary. This emerging product category is
identified as Performance System-on-a-Chip (P-SoC).
On the other hand, cost-effective solutions will require an assessment of the silicon technology
complexity that can be afforded for a given cost. Specifically, given a system cost target, what technology
complexity can be afforded? This product category is identified as Cost-effective System-on-a-Chip (C-
SoC).

System-on-a-chip (SoC) devices, for promising use in applications such as digital communications
equipment, have been added to the scope of the International Technology Roadmap for Semiconductors:
1999. In addition, DRAM half pitches of 180, 130, 100, 70, 50 and 35 nm have been defined as
technology nodes that are general indices of technology development. Each node represents a reduction
to approximately 70 percent of the preceding node. Each step represents the creation of significant
technology progress.

In addition, specific years have been targeted for the commencement of ramp to mass production
(typically, monthly shipments of at least 10,000 units) in each of the various technology nodes. These
technology nodes, which are common to all of the Technology Working Groups (TWGs), will facilitate
understanding of the Roadmap for the path ahead.
             Table 6.4 Technology-nodes indicating the DRAM technology at year-nodes.
    Year                       1999        2002        2005        2008        2011        2014
    Technology node [nm]       180         130         100         70          50          35
.

Historically, the Roadmap has emphasised the technological limits of silicon production, leading to the
specification of the most complex chips that can be developed in the categories of memory,
microprocessor MPU), and ASIC at a particular technology node. With the growing importance of high-
volume consumer markets and the ability to integrate almost all aspects of a system design on a single
chip, the Roadmap has included an additional vehicle to capture the requirements of this important,
emerging area. We refer to this vehicle as a System-On-a-Chip (SoC). There are a number of
characteristics that distinguish a mainstream SoC, but the main consideration is that it is primarily defined
by its performance and cost rather than by technological limits. As a system-on-a-chip, these chips are
often mixed-technology designs, including such diverse combinations as embedded DRAM, high-
performance or low-power logic, analogue, RF, and even more esoteric technologies like Micro-Electro
Mechanical Systems (MEMS) and optical input/output.

Silicon complexity is increased with the much larger numbers of interacting devices and interconnects
and the impact of new technologies, new logic families to meet performance goals, and the effects of
power and current requirements. System complexity is growing not only because of increased system
size, but also due to SoC designs with a diversity of design styles, integrated passive components, and
the increased need to incorporate embedded software. Design procedure complexity is also increasing
with the growing interaction among design levels, the difficulties of convergence and predictability of the
                                                                                                        92
Contracts number:
ESTEC - 13694/99/NL/US                            Final Report
ESA S-UMTS study                                  Date                    Revision
Final Report                                      2000-11-29              A


design process, and the growing size and dispersion of design teams— all required for quality,
productivity, and time-to-market. Verification complexity rises with the need to validate core-based and
mixed-technology designs, timing and function together, and behaviour at the system level. And the test
complexity grows greatly at higher speeds, higher levels of integration, and greater design heterogeneity,
making external test-through-pins less viable.

We see that even if technology makes SoC possible through scaling, the interaction between technology
development on one hand and system design and implementation on the other hand is a very
challenging task requiring progress and solutions in many areas.
                                                                                                        93
Contracts number:
ESTEC - 13694/99/NL/US                            Final Report
ESA S-UMTS study                                  Date                    Revision
Final Report                                      2000-11-29              A




7 Standardisation and regulatory issues
7.1    Standardisation
Today‘s existing satellite systems are more or less proprietary in design and interface structure. This is
however recognised as a problem today, and in a paper entitled ‗S-UMTS in the Wireless Information
Society: The Challenges Ahead‘ written by members of the European Commission, it is concluded: ―Lack
of standards for satellite access remain a critical issue, that will likely impose high terminal costs.
Software radio may be the answer, but is a long-term solution‖.

Immediately following the ITU-R endorsement of the 6 candidate IMT 2000 satellite air interfaces (S-
RTTs) in November 1998, there was little enthusiasm, by the proponents of the RTTs, to harmonise or to
standardise any of these systems. However, later that year, the ETSI Technical Committee for Satellite
Earth Stations and Systems (TC SES) approved 4 new work items, proposed by the European Space
Agency (ESA), Alcatel Space and France Telecom, and supported by Ericsson, which specifically
addressed the need for S-UMTS standards. It is now generally recognised that there are universal and
commercial benefits to be gained from the selective standardisation of some elements of S-UMTS.

It is envisaged that some changes may be needed to the current 3GPP specifications to enable satellite
systems to access the UMTS core network and its services. Essential changes, where appropriate, will
be submitted to the relevant 3GPP working- group responsible for the following:
 The satellite network interface to the terrestrial UMTS core network (Iu);
 The Universal Subscriber Identity Module (USIM) to Mobile Earth Station (MES) interface (Cu);
 Certain technical aspects that relate to the access of S-UMTS services;

New ETSI S-UMTS technical specifications (TS), derived from the 3GPP UTRAN specification, will be
published as open standards:
 S-UMTS common air interface(s), optimised for satellite operational characteristics;

Closer liaison with standards bodies outside Europe, for example the Telecommunications Terminal
Association (TTA) in Korean and American National Standards Institute / Telecommunications Industry
Association (ANSI / TIA-34) in the United States, should raise the level of awareness and help to
promote the S-UMTS air interface standard as a viable global solution.

Conclusion and recommendations for further work:
The definition of S-UMTS interfaces through selective standardisation would benefit both manufacturers
and end users.
Whilst consideration should be given to the voluntary standardisation of other ITU accredited satellite
IMT-2000 air interfaces (RTTs), priority should be given to promoting the ESA solution currently being
specified within the ETSI S-UMTS working group.
 Any changes to the Iu (core network) interface that are considered essential to accommodate
   satellite access should be discussed with 3GPP without delay.
 A sub-set of S-UMTS specifications should be completed within 12 months made available to other
   standards organisations (e.g. TIA-34) and published.
 Whilst not all standards are commercially successful, CT2 for example was upstaged by GSM, the
   standardisation process itself can create commercial opportunities through participating companies
   and by attracting investment.
                                                                                                            94
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A


7.2    Licensing issues
The regulatory situation of S-UMTS is rather complicated since S-UMTS in its nature is regional and
global. Therefore, licensing normally involves several national agencies as well as international
organisations co-ordinating their activities. Licensing practices vary from country to country, so a satellite
operator may face a formidable task in acquiring the licences necessary to provide services within the
footprint of a satellite.

Ideally, there should be a form of ―one-stop shopping‖ (OSS) for licences. The Conference of Postal and
Telecommunications Administrations (CEPT) has taken steps to set up OSS in Europe, where the
profusion of licensing requirements is most severe compared to the area and population involved.

A satellite system requires major investment, and the spectrum is limited. So precautions should be
taken to forestall ―paper satellites‖, systems that never will start service. Milestone notifications must be
given and for a system with no or very little progress in its preparations and funding, the licence should
be withdrawn. In Europe, CEPT has established rules for S-PCS (including S-UMTS/S-IMT-2000 band).

For satellite IMT-2000, in order to license spectrum for U.S. applicants and authorise the use of spectrum
for non-U.S.-licensed systems to operate in the United States, the FCC has set up a ―licensing window‖
or a ―cut-off date‖ within which applications must be filed. FCC has also normally a milestone schedule,
basically imposing that the applicants must show the financial ability to construct and put into service
their proposed system.

The purpose of spectrum licensing is to divide the limited frequency spectrum, with a minimal amount of
interference, among as many systems as possible (or wanted). Another possible target could be to sell
the spectrum as expensive as possible. The best solution for S-UMTS spectrum licensing is probably to
impose some minimum entrance requirements of technical, financial and legal nature. Then there can be
either a competitive or a comparative bidding to determine which systems should be granted licence if
there is not sufficient spectrum for all applicants. This is the normal situation in most countries today for
mobile licensing, although auctions are also used.

7.3    Spectrum issues and frequency sharing
The current allocations to Fixed and Mobile Services are done in Article S5 of the Radio Regulation with
regards to Resolution 212 concerning the implementation of International Mobile Telecommunications
2000. The baseline of this frequency assignment is that the bands 1885-2025MHz & 2110-2200MHz are
allocated for both mobile and fixed service. The bands 1980-2010MHz & 2170-2200MHz are dedicated
                                       st
also to Mobile Satellite Service from 1 January 2000.

The spectrum available for operators is very close to the spectrum provisions made by the ITU-R for IMT-
2000 for most of the regions and areas. However, some differences may occur especially for North
America.

Whereas 230MHz are today allocated to terrestrial IMT-2000, the UMTS-Forum foresee that about
187MHz additional spectrum will be necessary by the year 2010. The forecasts have proven that 2x145
MHz total requirements will be needed for MSS. The part dedicated to IMT-2000 is estimated to
2x66MHz by the year 2010 for global spots and 2x44MHz in EU15.

CEPT has identified an urgent need for at least 160MHz additional IMT-2000 terrestrial spectrum. CEPT
primary candidate for this additional spectrum is the band 2500 – 2690 MHz and eventually 2700 – 2900
MHz.
                                                                                                             95
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                     Revision
Final Report                                         2000-11-29               A


The frequency sharing between terminals of satellite and terrestrial UMTS/IMT-2000 is not possible. The
sharing between UMTS/IMT-2000 space stations and terrestrial systems relies on very severe
constraints i.e. pfd co-ordination limits, e.i.r.p. -limits and geostationary arc-angle restrictions and is not
feasible.

Sharing studies have been conducted to evaluate if the coexistence between MSS networks using the
same coverage and the same frequency bands is possible. The results are:
-       the sharing between MSS FDMA and TDMA networks is not feasible
-       the sharing between MSS FDMA or TDMA and CDMA networks is not feasible
-       the sharing between MSS CDMA networks is feasible

For this last item, the sharing is subject to the recommendation ITU-R M.1186 (Technical consideration
for the co-ordination between MSS networks utilising CDMA and other spread spectrum techniques in
the 1-3GHz band).

The methodologies proposed by the UMTS Forum and the ITU RR to derive spectrum requirements to
the satellite component of IMT-2000, are exactly similar. The method to derive spectrum requirements is
described by the recommendation ITU-R M.1391 (Methodology for the calculation of IMT-2000 satellite
spectrum requirements). The method has been applied to selected system scenarios:

TDMA systems with the following two study cases (10 meter and 15 meter reflector antennas):

                                                             Option 1 Option2
                               Number of beam                  157     353
                               Frequency reuse pattern          7       7
                               Number of beam cluster           22      50


                     Traffic in overall busy hour    2590      2590     Mbit/s Comment
                     Number of beam cluster           22        50
                     Delay factor                      1        1
                     Capacity per carrier             391      391      kbit/s    S-EDGE
                     Number of carriers               301      131
                     Carrier bandwidth                200      200      kHz
                     Spectrum requirement            60,2      26,3     MHz

The spectrum requirement is roughly compliant with the overall estimation given by the ITU and UMTS
Forum for Europe zone.

CDMA systems
Spectrum requirements were calculated based on a 10-meter satellite antenna with 97 beams and a 15-
meter antenna where 218 beams can be shaped. However these spectrum requirement forecasts were
far beyond the spectrum prognosis obtained by the other studies. Basically, this is due to the very poor
technical characteristics of the terminals (Palmtops). But it is also the overall method, which is not really
suitable to CDMA systems.

On the basis of a theoretical performance calculation more in line with the spirit of the TDMA analysis,
result has been evaluated.
                                                                                                             96
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                      Revision
Final Report                                         2000-11-29                A


                    Traffic in overall busy hour      2590          2590           Mbit/s
                    Number of beam cluster             97            218
                    Delay factor                       1              1
                    Capacity per carrier              2048          2048           kbit/s
                    Number of carriers                 13             6
                    Carrier bandwidth                 5000          5000            kHz
                    Spectrum requirement              65,2          29,0            MHz

This evaluation is compliant with the additional spectrum requirement estimation from other studies, and
is more in line with the result of the TDMA analysis. This result indicates that the traffic prognosis of this
study is coherent with spectrum requirement studies performed independently.

7.4    CEPT / ITU action plan
The satellite world should play an increasingly important role in standardisation and regulation issues to
ensure that all necessary steps are done to promote S-UMTS. This includes reviewing strategic events,
being attentive to progress of T-UMTS and always being ready to act.

At the current time, with the failure of Iridium, doubts over the feasibility of new satellite projects, and the
explosive development of terrestrial telecommunications, the position of the satellite industry is severely
fragmented, and is coming under attack from all levels. In such conditions it is extremely difficult to
envisage extension of S-UMTS spectrum even if it is clear that this is one element blocking development
of the S-UMTS concept deployment. New bands must be found, and to do this requires real commitment
and a real project backed by many players.

ESA support to start such an initiative forms the basis of this future work.
                                                                                                             97
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                    Revision
Final Report                                        2000-11-29              A




8 European S-UMTS Development Strategy

History tells us that financing a satellite system is costly. Today, there are no obvious investors for S-
UMTS. Terrestrial operators have full focus on GPRS/T-UMTS. Service and content providers do not
believe in satellite, car industry does not even know about S-UMTS. All in all industry interest is rather
small, and it seems that without organisations like ESA pushing, S-UMTS may soon be forgotten.

Knowing this, one can conclude that any S-UMTS venture will probably not be started unless money is
provided from some non-profit organisation. There is a possibility that governments, military, EU or some
other non-profitable party see the interest in having an operative 3G satellite communication system,
possibly combined with positioning system or another satellite service.

The conclusion is that S-UMTS needs a strong driver, and this is not necessarily the industry.


8.1     Selection of strategies
Two different strategies are suggested in this document. The choice of these strategies is based on the
experiences of the study partners, and partly on the S-UMTS study result.

The S-UMTS study has focused on strategy 1 during the study: to start a European S-UMTS venture.
The focus has been in analysing market, business and technical aspects of S-UMTS and strength and
weaknesses are rather well known. To propose a strategy for a S-UMTS venture was a natural result of
the study.


The second strategy was introduced rather late in the study, although it has been mentioned as a
possibility in many discussions. There are many advantages with evolutionary advance, e.g. building on
existing experience, infrastructure and ease financing, which can be fully exploited if strategy 2 is chosen

Technical and business analyses performed for S-UMTS during the study show that:

     The capacity that can be provided with a state-of-the-art GEO communication satellite is limited, and
      this induces problems in a 3G scenario where the bit rate requirements will escalate over time.
      The conclusion is that a pre-requisite for S-UMTS in the horizontal market is satellite capacity and
      compatibility towards the terrestrial UMTS.
     A business case exists for a S-UMTS system, but some major uncertainties increases the risk of
      such a venture.
     S-UMTS should focus on both the horizontal and the niche markets, using the latter as the driving
      force for market development.
     S-UMTS needs a strong driver to initiate the establishment of a business consortium and attract
      sound investors in an S-UMTS venture.


8.1.1 A cautious approach to S-UMTS
There are two main problems identified with a S-UMTS venture: capacity and funding
.
Capacity:
                                                                                                            98
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A


The capacity of a mobile telecommunication geostationary satellite is about 200 Mbps (maximum) with a
state-of-the-art technical design. The business case indicates about 2-3 millions of subscribers during the
first year of operation increasing by a little less than 1 million subscribers pr. year. The satellite capacity
calculations indicate a maximum of 1.5 million subscribers dependent on terminal type mix, and
consequently the satellite saturates at an early stage and further investments, more satellites and ground
network expansion, will be needed to secure the growing subscriber base. The need for bandwidth is
evident and any uncertainty on availability of secured S-UMTS frequency bands increases the risk.

Funding.
 A S-UMTS system has the potential to be a self-driven business venture, however, the risks are high
and the revenue insecure. The interested investors are more or less non-existing. partly due to the
recent mobile satellite ventures history of failure

There are other major question-marks as well:

     Will S-UMTS really be able to compete with (or even complement) terrestrial systems in
      user friendliness and availability?
      Will there be terminals available in high volume for S-UMTS?
      How will the lack of indoor coverage be perceived by the user?
     Will the users/ operators / service providers be prepared to deploy of special equipment for indoor
      coverage?


8.2     The Way Forward
Two strategies are proposed by the S-UMTS Study. The two approaches to S-UMTS are different in how
they provide the services, the first strategy describes an approach where full synergies towards the
emerging T-UMTS are focused both on the network and terminal side, while the second approach
describes an system evolution based on existing S-PCN systems like Inmarsat for providing basically the
same services.

8.2.1 Strategy 1:Start a S-UMTS venture
The first strategy suggested by the S-UMTS study is to start an S-UMTS venture.

A number of aspects need to be considered when outlining an S-UMTS system business venture, and
some aspects need careful consideration.


Goal
To introduce a geostationary S-UMTS system by 2005, owned and operated by European Industries,
with a healthy business case.


8.2.1.1.1 Getting started
The satellite community in Europe has to join forces and be more driving and creative. As a comparison
T-UMTS forces throughout Europe are currently working together with UMTS system design,
standardisation, marketing, etc. even if they soon will compete between themselves. This scenario is not
occurring in the satellite community, and thus the drive for S-UMTS is virtually non-existent. This must be
changed. A driving force must be identified.
                                                                                                         99
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


The driving force can e.g. be one or several major operators, a non-profitable organisation (EU,
governments, ESA, etc.) or a consortium that lost the T-UMTS license auction. This driving force must
then take initiative to form a business consortium (see separate chapter).

The target market must be identified, and a business analysis for the target market should be made. It is
logical to start with a regional system, which then possibly could be extended to a global system.

New satellite design is required, where the main issue is increase of capacity. A 3G satellite system fully
integrated with T-UMTS will demand a major capacity increase (compared to what is possible today).
Research and development programs for this must be started as soon as possible.

ESA then must act as catalyst and encourage development of a robust business plan, promote
standardisation of the air interface, demonstrations and initiate the forming of a substantial programme.
Operators must be encouraged to deploy these elements early to capture S-UMTS users and earn
revenues using existing satellite capacity.

All regulatory powers at European level must be invoked to ensure that satellite operators' users are
allowed to roam onto terrestrial public networks on a fair and equitable basis and that the regulatory
framework allows equivalent access to private terrestrial networks.


8.2.1.1.2 Vital decisions and prerequisites

During the S-UMTS study, some important pre-requisites were identified, and several decisions were
argued and taken. These are all vital for success, and listed below:

   Time-to-market is crucial. To be able to ride on the ―terrestrial UMTS wave‖ the S-UMTS system
    must be operative 2005
   Frequency reserved in the S-band for MSS has to be preserved for S-UMTS
   The development cost must be minimised, as the business case is very sensitive to increased initial
    cost
   The S-UMTS architecture must be kept as simple as possible. Increased complexity lead to
    increased cost and development time, which must be avoided

        rd
    A 3 generation mobile satellite system should be based on geostationary satellites, as this will allow
    a gradual rollout of services according to the demand. One of the most attractive features of a GEO
    system is that a single satellite is sufficient to prove the concept.
   The UMTS core system architecture should be used.
   Same service networks and applications as T-UMTS will be supported.
   Voice is supported but not as a main application. The same applies for CS services in general.
   S-UMTS should be based on an open set of standards being as close as possible to the standards
    developed by 3GPP for T-UMTS
   S-UMTS terminals should be as similar to T-UMTS terminals as possible, and of well-known brand
    names.
                                                                                                                    100
Contracts number:
ESTEC - 13694/99/NL/US                                        Final Report
ESA S-UMTS study                                              Date                            Revision
Final Report                                                  2000-11-29                      A


8.2.1.1.3 Attracting operators

For S-UMTS to be a success in the horizontal market, it must be an integral part of the T-UMTS network.
The best way to ensure this is to make alliances with terrestrial operators and service providers.

To be able to attract operators to a commitment towards S-UMTS, a number of ―political‖ factors must be
met:

   Standardised system specification
   Political support of S-UMTS
   Global alliances and partnerships
   Secure frequency band
   Provide a competitive edge over other operators

Terrestrial operators have been used to think and operate locally. However, with the growing
globalisation of the industry, it becomes more and more important to operate and grow not only in the
domestic market, but also internationally.
S-UMTS might be the tool for a local terrestrial operator to compete in this environment. It enables the
operator to provide 100% coverage and services world-wide. This might also be a powerful argument in
e.g. beauty contests for T-UMTS-licenses.

8.2.1.2 Process and Schedule
Time is of key importance. If no concrete S-UMTS plans are made in year 2000 or early 2001 it will be
very difficult to defend the S-UMTS frequency band allocation. The terrestrial segment is already
lobbying for the frequency spectrum reserved for S-UMTS. Unless no consortium/partnership turns up
with a viable business plan for S-UMTS within 2002, there is a high risk that the spectrum will be lost to
terrestrial T-UMTS. Year 2005 should be met by the first S-UMTS service otherwise broad industry
support will fail.

To prevent the spectrum being lost to terrestrial T-UMTS, it is important to involve cellular operators in S-
UMTS. This can be done by focusing on the benefits for an operator to join the S-UMTS partnership By
offering to provide his subscribers access to both networks with the same subscription, the operator has
covered 100% of the area by i.e. 2005. This will be hard to match for the competitive bidders.

In the post-launch phase, the operator will benefit from a complete coverage area and (most likely) the
ability to offer lower S-UMTS charges than other T-UMTS operators not member of the partnership (but
with general roaming agreements).

Key milestones:

                 T-UMTS licensing             2nd round of T-UMTS licenses.               Full UMTS coverage with
                 processes in Europe          Start building the S-UMTS system.           S-UMTS and T-UMTS.
                 and Asia. Prepare            Use partnership in S-UMTS as an
                 for future S-UMTS,           advantage in beauty contests
                 secure S-UMTS                (ability to offer 100% coverage in
                 frequency band.              2005).




          2000                         2002                                        2005
                                                                                                       101
Contracts number:
ESTEC - 13694/99/NL/US                                        Final Report
ESA S-UMTS study                                              Date                        Revision
Final Report                                                  2000-11-29                  A



Launch after 2005 increases the risk for an entry that is ―too late‖. After 2005, the interest in S-UMTS
from terrestrial UMTS operators is expected to go down, due to the increased terrestrial coverage,
partnerships and take-overs among terrestrial operators will create ―global‖ operators and the fact that
most licenses are already awarded.

To reach this market window, the work has to be started immediately. A schedule is suggested below:


                    2000   2001        2002   2003    2004 2005       2006    2007    2008 2009
                    2010
                       Initial study
                                                              Launch 1st satellite
                                       System specification
                                                                        Operation 1st satellite
                                         Establish S-UMTS
                                         Consortium
                                                                      Launch 2 more satellites

                                         Development                             Full Operation



a) Initial study, the first phase, is this study of S-UMTS technology and business, to identify the
   possibility of realisation of a S-UMTS system.

b) System Specification is the next phase, in which a S-UMTS systems should be more detailed
   specified, both technically and economically, a business model must be chosen. Standardisation
   work must be encouraged. Spectrum strategy must be developed for the next WRC meeting.

c) Establishment of an S-UMTS consortium is necessary. Verification of the economical and political
   foundation for a European satellite venture. Financing of the system is the most important part of this
   phase.

d) The development phase includes space and ground segment development, integration and
   verification. The business model shall be consolidated.

e) Launching phases include satellite launches, delivery and installation of ground segment, as well as
   extensive marketing of the S-UMTS system in the target markets.

f)   Operation phase is the period of time the system generates revenue. End-user terminals must be
     available on the market from the start.
                                                                                                            102
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                     Revision
Final Report                                       2000-11-29               A




8.2.2 Strategy 2: Evolution of existing systems
This study has suggested a business case for S-UMTS that is not very strong. Indeed, today (mid-2000)
the enthusiasm experienced a few years ago for mobile satellite systems seems to have been replaced
by a general pessimism, and there appear to be no new satellite project start-ups being proposed to
meet the introduction of UMTS, as was the case for mobile telephony.
However, mobile satellite systems exist and will continue to exist. Furthermore, satellite systems never
lead the way in the telecommunications world, and follow terrestrial developments, and this inevitably
means that mobile satellite systems will move to UMTS. Whether the market is enormous or restricted to
small niches, the aim of European industry will be to win as large a part of the contracts resulting from
the evolution of these systems as possible.

Therefore the second suggested strategy is the evolutionary approach, building on what exists, improving
the services of today. Existing systems must be encouraged to provide 3G services and to adopt the
UMTS architecture as much as possible. The most obvious candidate is Inmarsat with its new system
Inmarsat 4 (which, according to press releases is IMT-2000 compliant), and presumably Inmarsat already
has such an evolution in mind, but other operators may also be interested. This can be achieved by
                                                                                                  rd
working with Inmarsat and other operators, to develop new applications and demonstrations of 3
generation services, which will allow to better serve the existing niches and to expand the services to
new markets. The demonstration of new services and applications will lead to a gradual expansion of
mobile satellite systems into the mainstream markets, and consequently an expansion of the mobile
satellite industry.

This is a realistic approach, which has every chance of meeting the critical date of 2005, if the work is
started now.

This strategy will not provide a ―clean‖ S-UMTS system, but a semi-proprietary 3G satellite system, which
would probably be fully acceptable for niche markets. Many of the prerequisites for the S-UMTS study
are fulfilled with this strategy, but not all.

Inmarsat is primarily offering services to the vertical market today. One of the conclusions from WP4000
is that without a major revolution in satellite capacity, the vertical markets are the targets for mobile
satellite communication even for 3G.

T-UMTS services and applications may need some adaptation for satellite. This is a drawback, but for
vertical markets this may not be too costly or impractical.

Considering that future mobile satellite design may need revolutionary concepts in order to increase
capacity, there is no way that a significantly improved satellite design can be put in service by 2005. A
3G satellite system have to be based on the current satellite design. The schedule needed to reach the
2005 market window is more or less coherent with Inmarsat 4. Given this, Inmarsat 4 can be considered
        st
as the 1 S-UMTS satellite, and the following satellites represent an upgrade.

The proposed strategy is to work with operators to develop new applications. For instance, Inmarsat can
bring together all the elements lacking in a new S-UMTS venture: terrestrial and satellite operators,
knowledge of the satellite market, belief in the satellite market, etc, and even a fallback to niche markets
if necessary.

ESA's role in this scenario is to assist in the demonstration of the new services and applications, and the
subsequent development of the satellite and ground segment technologies required to implement them.
                                                                                                           103
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                    Revision
Final Report                                         2000-11-29              A


As a conclusion, European industry can benefit from the UMTS developments by building on existing
systems. The most obvious candidate is Inmarsat, but there are others. Industry has to work closely with
the operators to develop the new and evolved products which the market requires. Only by this route will
the mobile satellite market remain healthy and grow – with the consequent benefits for the space
industry. In addition, if European industry works alongside operators to produce the products which the
market really requires, then we will be in a strong position to take significant parts of the contracts which
will be placed to develop the new systems.


8.3    Discussion of an S-UMTS venture
Below is a discussion on strategic aspects in starting a new S-UMTS venture.

8.3.1 Introduction
UMTS will without doubt be a dominant and influential technology and offer integrated services to the
users beyond what any telecommunication system offers today to the public. There is an enormous effort
in many areas related to the development and implementation of the terrestrial part of UMTS. Global
expectations are also reflected in the interest attaining for terrestrial spectrum for UMTS services.
Satellites have since the UMTS concept was introduced been considered as an integrated part of the
system in order to provide global coverage and coverage where terrestrial technology was not practical
for various reasons, i.e. due to the terrain or too low expected traffic to cover the costs.

Experiences:
The expectations on mobile satellite communications have been set high in the past. Over the last
decade a number of mobile satellite systems based on GSM-like infrastructures have been introduced or
announced. These include Iridium, Globalstar, ICO, Odessey, Ellipso, Constellation, Satphone, Thuraya,
ACeS, APMT and EAST to name perhaps the most well known systems. None of these systems have
yet proven successful. Some have not yet had their chance, some have been abandoned and some are
in or have been in deep financial trouble. Satellite communications remains a niche market in year 2000.

Total revenue from satellite mobile users is currently around 700 million dollars per annum. Satellite
remains a niche market where investment must be paid from revenues (as with Inmarsat) or strategic
investment (as with ICO). Nevertheless this community is able to afford high charges for a good quality
service and is willing to pay where there is no terrestrial alternative. Attempts so far to attack horizontal
market segments have all failed. Generally the service has not been considered to be of comparable
utility to the terrestrial systems, the price has been higher than terrestrial and the distribution channels
have not been able to cope with the small volumes.

A number of inconvenient characteristics of satellite based mobile communications systems have
become clearer, like the inability to provide indoor coverage. None of the systems have offered satellite
communications with handsets as small and elegant as they can be for GSM, although this is perhaps
not a fundamental limitation. There are more issues as well, like the pace at which terrestrial GSM
systems have been deployed with, and unfortunately these issues have together overshadowed the
advantages of satellite communications. There is a general scepticism towards mobile satellite systems,
and given the recent incidents it is only logical to be careful.

The terrestrial cellular market has experienced explosive growth, which is predicted to continue as prices
continue to decline. With the deployment of GPRS and UMTS systems the price of data transfer is set to
fall further. It is expected that UMTS systems will initially take market share from the 2G and 2.5G
systems in cities and businesses. Outside of these areas GSM systems should be upgraded to GPRS in
order to compete, but again this may not reach rural areas. Expansion of these systems is being
achieved from cash flow and investment despite the present lack of profitability. Shareholders appear to
                                                                                                         104
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                    Revision
Final Report                                        2000-11-29              A


be patient and ready to wait for future dividends in order to attract users onto the network and to allow for
continued investment in infrastructure.

Urgency:
It must be understood that the focus for UMTS is on the terrestrial element, and that the level of
awareness of S-UMTS outside (and sometimes even inside) the satellite community is low. Therefore,
the development of an operational system for the satellite UMTS has an element of urgency, and it is
important that a number of technological and operational issues are resolved as soon as possible.
Otherwise, the implementation of S-UMTS may be in jeopardy.

The requirement for urgency stems from equipment manufacturing and frequency allocation reasons:
 Manufacturers are developing the terrestrial UMTS terminals, and any consideration to dual mode
   terrestrial/ satellite UMTS should be done as early as possible. An early S-UMTS development may
   be necessary to give credibility to the S-UMTS component.
 A bandwidth of 30 MHz has been allocated for the satellite component. Some pressure, in particular
   in Europe, has been reported to make parts or the whole of this band available for the terrestrial
   systems. The exorbitant total price obtained at the UK UMTS frequency auction, 22 bill. £, certainly
   has not reduced the appetite for this band

Another reason for urgency is that business analysis shows that there is a market window for a satellite
UMTS system in 2005. UMTS is still a quite new concept with a limited amount of users at this point in
time, and by offering satellite UMTS as an additional feature to the growing UMTS market at the exact
right time, a positive business case can be achieved. Only 2 years later, the market potential will be
considerably reduced.

Considering the car scenario where S-UMTS is used for offering advanced communication solutions in
cars and other vehicles there is also an element of urgency. The vehicle market is currently specifying
their communication solutions, and co-operation needs to be initiated as soon as possible.

Success factors:
There are three interdependent key factors that will be decisive for the success of S-UMTS: early system
rollout, global allocation of spectrum and willingness to make billion dollar investments in S-UMTS. The
satellite community have to defend and secure that the tentative frequency allocation next to the T-UMTS
band is not cannibalised by terrestrial service interests, and demonstrate and create enough interest
among various stakeholders, who actively will make contributions of the S-UMTS realisation.

Business Case:
The European Industry business potential is probably most important for the European satellite industry
corporations and satellite niche industries for gateways and satellite terminals. For the major
telecommunication companies the satellite part is considered a niche market as they can gain enough
revenue by focusing on T-UMTS. To summarise the total sales for European Industry may be higher if
the spectrum is utilised for T-UMTS rather than S-UMTS. On the other hand S-UMTS can fulfil important
regional service requirements regulated by domestic political decisions.

Proposed Opportunity
The challenge for the satellite community is firstly to improve services to the niche market presently
served and to try to increase revenues against a backdrop of falling price expectations. Secondly, the
satellite operators need to be shown ways in which they can reduce prices and improve quality of service
to a level at which mainstream cellular users can be captured at least in areas presently uncovered by
terrestrial systems.
                                                                                                         105
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


If revenues can be stimulated by these measures, then successful satellite operators will be able to
expand and increase shareholder value. If not, satellite will remain a niche market serving users only
where there is no alternative.

It is assumed that the satellite competitor is not only T-UMTS but also GPRS. There are two ways to
compete effectively. Firstly by offering GPRS extension where no GPRS is available. The second is to
offer full S-UMTS extension of T-UMTS in areas where GPRS cannot offer the same performance in
terms of bandwidth.

Present GSM based satellite systems have been designed to allow paging inside buildings with some
user co-operation to achieve voice calls. This has not proven acceptable to date for mainstream users. It
is proposed that an acceptable performance may be messaging and voice to handsets within buildings
with some user co-operation to effect high-speed data links.

The car industry is currently very interested in providing Internet and other wireless communications in
their cars. An S-UMTS system could make this possible. As cars are inherently well suited for satellite
communications, this may well be one of the best opportunities for S-UMTS to reach the horizontal
market. By providing a fully integrated satellite/terrestrial system, a virtually complete UMTS coverage
can be offered to all cars in a large area. As most of the services are narrowband, capacity will not be a
big problem.


8.3.2 Players and partnerships
Lessons learnt from the first generation satellite system indicate that satellite subscriber demand and
growth, which are so important for the survivability of the satellite system operator, must exist and
increase at a healthy pace. Therefore we would like to claim that the most important drivers for S-UMTS
are the T-UMTS operators, the service providers and the satellite terminal equipment manufacturers. For
commercial success several European operators and at least two out of the four leading cellular phone
manufacturers should be committed to the S-UMTS project.

Firstly, the T-UMTS operators must acknowledge S-UMTS as a complementary service, which generates
good additional revenues for them. Thereby they are willing to sign roaming agreements with the S-
UMTS operator which all together will make S-UMTS a true global service. They are going to promote
the service and many operators are willing to make equity investments in the S-UMTS company.
Secondly, some of the world‘s leading mobile manufacturers are needed behind the S-UMTS rollout in
order to secure attractive mobile phone terminals with recognised brandnames and exposed to the end-
users through their global distribution networks. Above is a prerequisite to attract the necessary
investments from venture capital companies and industry partners, mainly the satellite industry.

In a fast growing, ever-changing market it is impossible for one corporation to build big complex system
with high investments on its own. Partnership is the best solution to this on a short-term basis. An S-
UMTS system venture presents several complex problems and also requires a large initial financial base
to cover development costs. To overcome this, an S-UMTS venture must consist of a broad partnership,
including many partners with expertise in their own area. A strong and dedicated S-UMTS consortium is
needed. Such a consortium could primarily consist of e.g.:
  Global Satellite operator
  UMTS operator
  Service provider
  Contents provider
  Key Customer
                                                                                                           106
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                    Revision
Final Report                                         2000-11-29              A


    Other type of industry, e.g. car industry

It is possible to also include:
  Satellite manufacturer
  Telecom system manufacturer
  Terminal manufacturer
  Launch companies and consortia

A consortium consisting of many of the parties listed above would have a strong chance of success.
Together the S-UMTS consortium would have deep knowledge on all relevant areas. Both economical
and technical risks will be shared, and probably minimised due to the variety of focus areas. A strong
consortium is also a good ground for attracting further investors.

 To trigger the interest of major financial institutions, it is of utmost importance to attract large companies
 like global network operators and large satellite manufacturers to bring credibility to the project. Large
 network operators have the size, capacity and competence to provide its services to the international
 market.

 By having operators, both fixed and mobile, as strategic partners, one will obtain several advantages:

 -    An established customer base. Established operators have a customer base, and it might ease the
      rollout of an S-UMTS system.
 -    Access to the market. Established operators can provide an easy access to the targeted market.
      Operators also have existing sales and marketing structures, and can also provide operation and
      maintenance functions. This can be used to further speed up the S-UMTS rollout.
 -    They have an existing network infrastructure. To simplify the roaming with terrestrial mobile
      networks and routing into fixed terrestrial networks to provide global access, it would be beneficial to
      have operators as partners. By allowing roaming into their networks, mobile operators will gain
      revenues due to the increased traffic and they will be able to deliver value added services to their
      customers. For the same reason, fixed operators will open their network in such a way that the
      traffic can be routed from a gateway and into the terrestrial network.

 In addition, each operator will be responsible for obtaining the required national licences in each country
 of interest.

 Satellite operators are important partners because of their experience and knowledge considering the
 planning, development and operation of such systems. In addition they might have terrestrial
 infrastructure that can be used for S-UMTS purposes.

 Car manufacturers. Considering the importance of the car scenario, it is important to have
 representatives from this industry as partners. If car manufacturers install S-UMTS terminals in the new
 cars, a potential customer base will be generated. By co-operating with content providers, creative and
 useful applications and services should be developed. The success will depend on which services and
 applications that will be offered, on the quality of these services and on the price.

 Content providers. In S-UMTS it is vital to develop good applications and services, which the users are
 willing to pay for.

 Other important strategic partners are the producers of terminals. One needs to have strong and reliable
 terminal producers to ensure the delivery of high quality terminals, which is appealing to the customers.
 Also important to attract manufacturers of T-UMTS equipment to get a common platform for the
 terminals, and thereby exploit the benefits from large scale production (e.g. cheaper satellite terminals).
                                                                                                          107
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A


 The terminal design is crucial considering the success of S-UMTS. The terminal design must appeal to
 the customers, and it must be adapted to the services and applications that are planned to offer.

 By partnering with manufacturers of satellites, you will be likely to get ―state of art‖ equipment. A close
 dialog is needed between those specifying the system requirements and the manufacturers of the
 satellite segment, so that the requirements conform to what the satellite manufacturers are capable of
 delivering.

To successfully establish a S-UMTS consortium, the business model must show profit potential for all
parties. The business analysis in this study indicates the possibilities, but a more extensive and detailed
business study is needed to find business cases for all involved parties.

Once the profitability is verified, clear goals must be established for the S-UMTS consortium. All parties
must be aware of their role and importance in this complex program, and it is important that all pull in the
same direction. This may prove to be more difficult than expected due to the variety in culture, language,
political trends, etc. that is found in Europe.


8.3.3 Standardisation
Standardisation is really of key importance, probably more so for a common standard air-interface than
anything else. One air-interface in all S-UMTS systems facilitates volume production and feature
transparency with T-UMTS. Without volume potential for terminals only specialist satellite terminal
vendors will develop and manufacture terminals and both price and size will become unattractive. In
general, the cellular phone industry is too busy developing new models for the terrestrial mass market to
consider using critical design resources for odd satellite products. Standardisation is probably not so
critical for the satellite manufacturers, however, lack of standardisation will lead to higher satellite
terminal unit prices, which of course may be critical for the overall business case.

An S-UMTS system should be based on an open set of standards being as close as possible to the
standards developed by 3GPP for T-UMTS. The T-UMTS infrastructure should be reused to the extent
possible. Specific satellite gateways taking the role as a combined RNC/Node B should interface
standard core network nodes through the Iu interface specified by 3GPP. Although the original intention
of the 3GPP was to make the Iu interface specifications take into account satellite specific requirements,
modifications to this interface may be required.

It is envisaged that some changes may be needed to the current 3GPP specifications to enable satellite
systems to access the UMTS core network and its services. Essential changes, where appropriate, will
be submitted to the relevant 3GPP working group responsible for the following:
The satellite network interface to the terrestrial UMTS core network (Iu);
The Universal Subscriber Identity Module (USIM) to Mobile Earth Station (MES) interface (Cu);
Certain technical aspects that relate to the access of S-UMTS services;


New ETSI S-UMTS technical specifications (TS), derived from the 3GPP UTRAN specification, will be
published as open standards:
S-UMTS common air interface(s), optimised for satellite operational characteristics;

Closer liaison with standards bodies outside Europe, for example the Telecommunications Terminal
Association (TTA) in Korean and American National Standards Institute / Telecommunications Industry
Association (ANSI / TIA-34) in the United States, should raise the level of awareness and help to
promote the S-UMTS air interface standard as a viable global solution.
                                                                                                            108
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                     Revision
Final Report                                         2000-11-29               A




8.3.4 Licensing
If governments can earn huge amounts through spectrum auctions that recently was the case in United
Kingdom when 3G licenses were auctioned it may be very difficult to reach a global consensus and
agreement among all nations on how to award licenses. This fact can delay the S-UMTS process, restrict
global coverage and even become cost prohibitive for any potential S-UMTS venture consortium.

An auction for S-UMTS spectrum between satellite operators is possible, but not quite likely to be
adopted. It assumes that the frequencies are auctioned by ITU (or another organisation) to the company
willing to pay the highest price and then freely sold and bought. This could maybe be a good solution to
large companies operating on a global basis. However, it might also mean that a large company could
prevent smaller companies commanding lesser resources to compete. It will increase the initial cost of a
system and then limit the competition, which seems to be rather limited anyhow. An auction of frequency
spectrum should therefore be avoided for S-UMTS.

The best solution for S-UMTS spectrum licensing is probably to impose some minimum entrance
requirements of technical, financial and legal nature. Then there can be either a competitive or a
comparative bidding to determine which systems should be granted license if there is not sufficient
spectrum for all applicants.

The regulatory situation of S-UMTS is rather complicated since a satellite operator may face a formidable
task in acquiring the licenses necessary to provide services within the footprint of a satellite. Ideally, there
should be a form of ―one-stop shopping‖ (OSS) for licenses. The Conference of Postal and
Telecommunications Administrations (CEPT) has taken steps to set up OSS in Europe, where the
profusion of licensing requirements is most severe compared to the area and population involved.

So far no FCC license has been awarded despite that the request for license applications in US was
closed in September 1997. None of the applications is an S-UMTS system.

A satellite system requires major investment, and the spectrum is limited. So precautions should be
taken to forestall ―paper satellites‖, systems that never will start service. Milestone notifications must be
given and for a system with no or very little progress in its preparations and funding, the licence should
be withdrawn. In Europe, CEPT has already established rules for S-PCS (including S-UMTS/S-IMT-2000
band).


8.3.5 Technical development
There are technological key issues in implementing an S-UMTS system. Listing four of them, presumably
in increasing order of difficulty:
1. Development of a large S-band multi beam satellite antenna. This work has been going on for some
     time, and a solution should be within reach.
2. Development of a flexible IF/baseband processor which allows different number of S-band frequency
     blocks from different beams to be connected to the feeder link antenna system in a flexible way by
     ground command. This component is essential to obtain sufficiently high frequency utilisation.
3. Development of an on-board processor capable of demodulating and routing W-CDMA carriers, in
     the first place from S-band beams to other S-band beams.
4. Development of a full fledged on board processor for W-CDMA signals.
5. Development of a RBS switching system for the Satellite to enable routing of signals from all GW s to
     all beams.
                                                                                                           109
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A


6. Development of a flexible Digital Beam forming payload to enable dynamical sizing and moving of
   the beams.


8.3.6 Demonstrations
Phase 1: Experimental S-UMTS satellite
System verification and the first implementation of S-UMTS could be made by means of an operator
supported EUTS, European UMTS Test Satellite, or by a payload on another operational geostationary
satellite that performs similar functions.
The satellite must as a minimum include the following functions to demonstrate feasibility of an S-UMTS
system:

1    Large S-band multi beam satellite antenna.
2    Flexible IF/baseband processor that allows different number of S-band frequency blocks from
     different beams to be connected to the feeder link antenna system by ground command. This
     component is essential to obtain sufficiently high frequency utilisation.
3    On-board processor capable of routing W-CDMA carriers.

Since the traffic in a satellite system will grow gradually it may not be essential that an experimental
satellite has the performance required by an operational system. Timing may be more important than
capacity.
1. Reduced antenna performance, both in number of beams and in the gain performance, will mainly
    have effect on the total system capacity. In an experimental system shortcomings in gain may be
    compensated by increased satellite power, or by reduction of the number of simultaneous carriers.
2. The development of the remotely controlled IF/baseband processing unit may be the least
    challenging one.
3. For an experimental satellite the on board signal processing may in the first place be required to
    demonstrate the functions of such a device. Capacity in terms of number of simultaneous channels
    together with mass and power consumption may be more of importance for an operational system.
The full processing device may not be essential for the experimental spacecraft, but the general interest
of on-board signal processing is such that any development effort is of interest for many other systems.

Phase 2: High capacity S-UMTS satellite with efficient feeder links
 Implications of moving radio base station and/or RNC functionality to the satellite payload.
 Satellite complexity study.
 Impact of lower RLC roundtrip delay.


8.3.7 European Industry Awareness of S-UMTS
One of the biggest challenges to make an S-UMTS program come true, is to make the European industry
aware of the strengths and opportunities of satellite communication. To achieve a successful S-UMTS
effort, a driving force is critical. This technical study point at some areas where S-UMTS may have a
strong position in the future, but the players in these segments are often not even aware of S-UMTS
today.

In general, there is a slight scepticism for satellite telecommunication today in the industry. Some of the
most well known satellite ventures have failed, and the attitude is awaiting – will the present satellite
ventures succeed? The challenge for parties with satellite interests is to turn this into interest and belief
instead.
                                                                                                              110
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                    Revision
Final Report                                         2000-11-29              A



Some means to improve the awareness of S-UMTS are presented here:

     Seminars
      All opportunities for presenting S-UMTS as a concept at different kinds of seminars must be taken. It
      is not the satellite communications seminars that shall be addressed primarily. Presentations in
      traditional satellite seminars will not reach new ears. Instead seminars on subjects connected to
      other types of industry should be addressed: telecom industry, car industry, engineering industry, etc.
      Good presentations on how S-UMTS can be beneficial to these groups of companies may well
      create the interest we need.

     ESA and EU Commission drive
      Support from ESA and the EU commission is essential, for example by providing initial finance
      support. ESA could be the driving force for organisation of presentations on the seminars mentioned
      above, providing input to the media, financing initial technical studies, etc.

     Media coverage
      Media has a large influence on the society. Presentation of S-UMTS as a concept in different form of
      magazines and newspapers may create a curiosity among the public. This can be directed to
      everyone in daily news papers, of at certain target groups, e.g. the car industry, by addressing
      relevant magazine. The aim of this is to increase the awareness of S-UMTS in the public.

     Success stories
      There have been too many failures of different satellite communication systems lately.
      There is a slight scepticism for satellite, it is considered risky. There is a big need for satellite
      success stories, and we rely on Globalstar, ACeS, Inmarsat 4, etc. for this. A success story will
      increase the belief in satellite telecommunication and hopefully convince European industry that a
      well-managed satellite venture has a large potential for success.

     Marketing directed towards major companies
      In case the target companies are identified, it is possible to direct a campaign directly towards these.
      At the moment it is too early to identify specific companies, but this may be used in a later phase.

     University and College
      University and college students can be made aware of satellite communication opportunities. This is
      a long-term improvement, with little momentary effect.


8.4     Roles in a S-UMTS venture

8.4.1 ESA Role
ESA should encourage with the help of the satellite community (operators and ground equipment
manufacturers co-ordinated by the satellite primes)
 The demonstration of a robust business plan for the operators and other actors
 The initiation of a substantial and business plan driven programme to decrease the uncertainties in
   terms of performance, cost and schedule
 The standardisation of air interfaces compatible with but not necessarily optimal for satellite use
 The development of end to end system specifications (backward compatible with GPRS) including
   single user and multi-user terminals, gateways and NCC
                                                                                                            111
Contracts number:
ESTEC - 13694/99/NL/US                               Final Report
ESA S-UMTS study                                     Date                        Revision
Final Report                                         2000-11-29                  A


   The demonstration end to end of these network elements with existing satellites (L,S,Ku and Ka
    bands) and with representative S-UMTS spacecraft elements in terms of interfaces and end to end
    performance
   The demonstration of applications, management and control including the ability to handle T-UMTS
    and GPRS terminals roaming on to the satellite network
   The definition of a next generation satellite constellation and demonstration of the key technology
    elements

A three-stage model for ESA support to S-UMTS is suggested below.




            ESA technology
                                               Subsystem - antenna, processor, platform, user terminals
            development activities


              2000
           Technology
           development
                                          Experimental                                Commercial
                                            payload                                     roll-out
         Qualification of
         demonstrators


                                         Early mission
                Lower schedule risk   with growth potential
                                                                               Advanced commercial flight
                                                                                 programme to achieve
                                                    Market development              low end user cost

Figure 8-1 ESA’s role in supporting S-UMTS development.

The first stage is where high-risk developments are underwritten by ESA in a classic technology
program. The output from this program (possibly ARTES 6 phase 1) would include qualified hardware.
Time-to-market will be a critical issue and it may be concluded that 1st generation satellite technology
has to be derived from what ESA has already supported (through TRP and ARTES 4&5). The next stage
(perhaps ARTES 6 phase 2) would concentrate on reducing the risk, to a commercial operator, by
underwriting the cost of an experimental payload. The final step of supporting full commercial systems
would primarily be financed by commercial interests, with ESA support limited to the risk management of
advanced technologies.


ESA can reduce risk for industry and help to stimulate the market for S-UMTS by co-funding the
development of enabling technologies and flight demonstrations. Risks which can be mitigated by ESA
funded activities are:
                                                                                                         112
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                    Revision
Final Report                                        2000-11-29              A


     Schedule risk. The credibility of the schedule for a flight program involving the attraction of private
      investment is crucial. New technologies and/or equipment designs which have not been flown require
      convincing proof that the development time scale will not prejudice the success of the program.
     Cost risk. The development and manufacturing costs must be quantified at the planning stage.
     Performance risk. The critical aspects of the flight equipment such as technical performance, mass,
      and power must be confirmed.

ESA has stimulated and supported wideband multimedia satellite developments through ARTES3
funding and a similar model for ARTES6 funding can be envisaged.

The early demonstration of a satellite based UMTS communication system would help to stimulate the
market for satellite based services and convince investors of the reality of the market. It would allow
potential users to experiment with different kinds of UMTS services. Consequently, the study will carefully
consider how such an ESA program can support the commercial development of S-UMTS in Europe.

ESA has a unique opportunity to promote the introduction of S-UMTS, and this task would also be in the
basic task of ESA. Key operations are:
1. Adapt the standards for terrestrial UMTS to a standard suitable for a (geostationary) satellite system.
    This task is well under way, and the results obtained so far show that both the radio interfaces and
    other elements can be made very similar.
2. Propose a European satellite project, separate spacecraft or payload, to demonstrate the feasibility
    and advantages of S-UMTS, as outlined above.
3. Initiate the necessary technology developments, in particular large S-band antenna, on-board
    IF/baseband switch and on-board processor for W-CDMA carriers.
To promote and co-ordinate the S-UMTS activities a European, or preferably global, S-UMTS forum
should be organised. ESA could play an important role in establishing this forum.

To secure European Industry involvement and technological leadership it would be valuable if ESA could
take the initiative to establish a combined test system (phase I) and commercial system (phase II) with
one geostationary satellite covering Europe, Middle East and Northern Africa. If such an initiative is not
taken by an ESA-led consortium (or other) before the end of 2000 with firm contracts signed before the
end of 2001, the realisation of S-UMTS is at great risk as T-UMTS sector will request more frequency
band.



8.5     BENEFITS OF S-UMTS FOR THE EUROPEAN INDUSTRY
The ambition of a S-UMTS system is a European initiative developing into a global industry. By starting
today, European industry can get the start and then keep its market leading position. This is a tough
ambition, but can certainly be done. Of course there are several strong benefits for European players in
such a scenario.
 Worldwide acknowledgement of satellite mobile telecommunications competence. Traditionally the
    division has been Space – USA, Ground – Europe. This tradition must be broken, the European
    space industry must show their strength world-wide.
     S-UMTS success may result in other European satellite system investments, for example broadcast,
      multicast or Ka-band-multi-media systems.
     S-UMTS may have positive influence on the T-UMTS market. By offering better UMTS coverage to
      the end users, the total interest for UMTS increases, which lead to an increase for T-UMTS.
     Path to a world-wide proliferation of UMTS through global partnerships.
                                                                                                             113
Contracts number:
ESTEC - 13694/99/NL/US                              Final Report
ESA S-UMTS study                                    Date                     Revision
Final Report                                        2000-11-29               A


   Economical return:
         -   Scenario 1: The market potential is significant and provides an attractive business
             opportunity. However, the capacity of the satellite system limits the maximum number of
             users to a marginal level. With a satellite system capable of handling 3G traffic for a large
             number of users this scenario would represent a very attractive investment.
         -   Scenario 2: The business case of scenario 2 is positive for global coverage. The return of
             14% for the global scenario might be found acceptable, particularly if there are objectives
             other than pure financial return.

Fair competition with American suppliers must be envisaged. For global coverage reach FCC may pose
a stumbling block as no S-UMTS system operators have filed their applications. ICO and Inmarsat are
the only European, but maybe those companies are more American today as far as equity ownership and
head office are concerned.

An interesting aspect of the benefits is the fact that the S-UMTS business potential is probably most
important for European satellite industry corporations and satellite niche industries for gateways and
satellite terminals. For the major telecommunication companies the satellite part is a niche market only.

Regarding the interest of the different parties there are great difficulties in bringing terminal
manufacturers onboard. The difficult issue is to attract the terminal vendors already having a full-fledged
distribution system. The S-UMTS volume will always be small compared with T_UMTS, GPRS and other
cellular phones, which means that the development costs for S-UMTS phones are big for the
manufacturers.

Getting operators interested in S-UMTS is also a challenge. Maybe an S-UMTS operator should target
content providers and venture capitalists as equity owners and limit the deals with T-UMTS operators to
roaming agreements, which probably most operators are willing to sign.

Some benefits and market opportunities are summarised in the table below.

End-user benefits                              Access to personal services anywhere, anytime
                                               Ideal for the word traveller, irrespective if you are
                                                a businessman or an adventurer.
                                               Increased service coverage
Global satellite operator                      First step into 3G
                                               Open satellite communication for the mass
                                                market.
                                               Increase business
                                               Roaming agreements with T-UMTS operators
UMTS operator                                  Innovative and technical advanced image
                                               Gain new business
                                               Increase customer loyalty by offering a better
                                                end-user service
                                               Provide global access & services
                                               Increased competitiveness over other local
                                                operators
                                               Increased interest as a partner in world-wide
                                                consortia
Satellite manufacturer                         Gain new business
Telecom system manufacturer                    Gain new business
                                                                                                          114
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                    Revision
Final Report                                       2000-11-29              A


                                              Offer a complete 3G communication solution
Terminal manufacturer                         Innovative and technical advanced image
                                              Increase business without too big changes in
                                               design.
Service/content provider                      Availability of the content is increasing, can be
                                               accessed from virtually anywhere on the globe.
Car industry                                  Offer an integrated world-wide car
                                               communication system
                                              Innovative and technical advanced image
                                              Increased competitiveness over other car
                                               companies



8.6     WEAKNESSES AND RISKS OF S-UMTS
It is extremely important that a S-UMTS venture consortium is aware of the risks and weaknesses of
such a system. A future consortium must have evaluated the risks and prepared contingency plans in
case of occurrence. This is probably the best way of ensuring success.

8.6.1 Capacity
The market segment addressed by S-UMTS is mainly driven by the evolution of terrestrial mobile
systems. The services are E-commerce, mobile Internet and multimedia applications plus the general
availability of the user. The capacities required are mainly governed by user expectations and based on
subjective factors. The network and server capabilities evolve according to these expectations and the
introduction of broadband multimedia services has strongly influenced this development in recent time.
The capacity of an S-UMTS satellite must be dimensioned to meet an expected similar trend for the
mobile users. The expected reasonable user requirement (comfort-zone) today for transfer capacity for
WWW mobile services is 30-40 kbps, however by 2005 this requirement is expected to rise towards 100
kbps as new terrestrial services develops and networks / servers improve their capacities. The main
parameters governing this are response -time and data-volume to be transferred.
Improved capacity is probably the single most important issue for launch of a successful S-UMTS system
planned for years 2005-2007.


8.6.2 Weak end-user belief in satellite telecommunication
The belief in use of satellite telecommunication in the mass market is very weak today. It seems like only
those with special interest believe in satellite telecommunication for the horizontal market. The belief in
S-UMTS may be just as an important issue to address as the technical design or the business case,
before moving on. The issue is increasing the awareness and belief in S-UMTS.


8.6.3 Business case pre-requisites
The business case is based on some basic statements:

     Development cost on target
     Time to market on target
     Subscriber base is as estimated.

In case later estimates show that the subscriber estimations made in this study was too optimistic, the
venture should be re-evaluated.
                                                                                                             115
Contracts number:
ESTEC - 13694/99/NL/US                                Final Report
ESA S-UMTS study                                      Date                     Revision
Final Report                                          2000-11-29               A




8.6.4 License fee for S-UMTS
During the last months, extremely high T-UMTS license costs have appeared. License costs of the same
size as the complete infrastructure costs is considered an acceptable investment for T-UMTS operators
all over Europe. This evolution may have the unlucky consequence that the governments may expect
similar income for the S-UMTS licenses, and announces auction for S-UMTS spectrum between satellite
operators. This is unlikely to happen, but will have large consequences if it does.
The initial cost will increase with a less attractive business case as result, and it will limit the competition,
which seems to be rather limited anyhow.


8.6.5 Loss of S-UMTS spectrum
A most real danger is loss of the S-UMTS spectrum to e.g. T-UMTS. Spectrum is a limited natural
resource and there are many strong and genuine interests that desire the S-UMTS spectrum for other
usage than satellite UMTS.

8.7    CONCLUSION
This study has aimed at the realisation of an S-UMTS system by European industry. The study team has
made an economical analysis, suggested a technical solution, and also pointed at some weaknesses and
risks of such a system. One important conclusion is that there are major obstacles to a successful S-
UMTS venture.

It is generally known that satellite capacity is limited. The mass-market presents an interesting
opportunity, but at the moment the technical design only marginally meets the increasing capacity
requirements set by UMTS, GPRS / EDGE and the mobile user expectations to these new services. Real
time Multimedia and broadband services for a large number of handheld mobile satellite users are thus
currently only marginally feasible. However, these services can well be delivered to the vehicular mobile
users and the fixed and nomadic satellite users. Best effort type services, WEB-browsing etc, however,
mainly depends on the user comfort level (response time and data-rate) and using ―thin client‖ type
services might prove beneficial in a satellite environment. As technical design evolves, satellite mobile
communication may reach the mass-market in the future.
                                                                           2
Certainly satellites cannot compete with the achievable capacity pr.km of terrestrial systems – the
relative size of terrestrial and satellite cells indicates this, however, the service and user availability over
waste areas inherent to satellite systems are of some importance for the customer and the service
providers. Other opportunities definitely exist and satellite systems have other strong points, which are
perhaps not best exploited by the S-UMTS market scenarios in this study. The issue to be tackled is to
match the role and services which satellites can offer to the developments in the telecommunications
world in general, which is being led by the terrestrial systems.

The S-UMTS study team suggested two possible strategies:
1. Form a business consortium that will start a European S-UMTS venture, with an operative mobile
   satellite system in 2005, focusing on reducing the associated risks.
                                                     rd
2. Evolution of existing satellite system towards a 3 generation system

The business case for S-UMTS by itself is not enough for initiating an ambitious satellite venture.
However, if an interest can be identified by the industry and system operators for developing a European
S-UMTS system with a clear goal of world coverage within a realistic time-frame, and a non-profitable
investment can be generated, strategy 1 may become reality.
                                                                                                       116
Contracts number:
ESTEC - 13694/99/NL/US                             Final Report
ESA S-UMTS study                                   Date                   Revision
Final Report                                       2000-11-29             A


Strategy 2, building on existing systems (Inmarsat 4) and encouraged them to provide 3G services, may
be the most probable and realistic approach, but the synergies with T-UMTS will be limited or non-
existing and solutions proprietary. The driving force for this approach is mainly the ambition of the
system operator, his competitive environment and the customer needs.
 The more ambitious scheme of Strategy 1 based on an ESA/ ESTEC development program and ETSI
standardisation efforts plus close co-operation
 with UMTS terminal vendors and mobile system operators, might prove fruitful by ensuring synergies
and integration with the T-UMTS evolution. The driving force for this approach is the ambition of a global
standard, open competition and market requirements to mobile service availability. If this is enough to get
past the funding hurdle in today‘s satellite investor climate depends on a number of factors like business
structure, possible untraditional consortium constellations and the evolution of the T-UMTS business
ventures which will define the business reference models.

				
DOCUMENT INFO