EXECUTIVE SUMMARY

INTRODUCTION

This Briefing Report was prepared for the third ITU Regulatory Colloquium. It considers regulatory issues involved in the introduction of Global Mobile Personal Communication Systems (GMPCS) using advanced satellite communications architectures which are often (sometimes inaccurately) called "Big LEOs". The Briefing Report represents the views of its author, not those of other Colloquium participants or the ITU. The report of the Colloquium itself is a separate document, prepared by the Chairman.

Part I of the full Briefing Report provides a definition of GMPCS. The definition is "technology-independent": it is based on what a system does for its users, not how it does it. "Global personal communications" is not synonymous with "Big LEOs". Flexible personal communications will also be achieved through roaming arrangements among terrestrial "wireless" networks, and probably also in future by seamless interworking of fixed and mobile networks in which each user's personal number travels with them and can be reached via either type of network. Nevertheless, the satellite aspect of advanced personal communications involves especially challenging regulatory issues, and these are the focus of this report.

All of these issues represent one or another aspect of a single theme, best expressed as a question:

"How can the desire for rapid deployment of GMPCS to maximise its large potential benefits to users (a desire which appears to be shared by many, but not necessary all, national governments) be reconciled with other national objectives, such as maintaining sovereign control of telecommunications policy; equitably sharing the radio-spectrum resource; or maximising opportunities for national business and industry to participate in GMPCS?"

Part II considers regulatory issues in six different areas: frequency spectrum management; licensing; interconnection to existing networks; "non-discriminatory access" to GMPCS systems and services, together with related issues of competition policy; standards; and numbering.

Part III discusses how effectively these issues can be dealt with by existing national regulatory policies, institutions and processes, and by the existing machinery for international co-operation among national governments. It also considers what alternatives might be available if and when existing practices are judged to be unsatisfactory or insufficient. We present a set of six broad "scenarios": each represents one view of how regulatory policies and processes concerning GMPCS might evolve at the national level, and how international co-operation, consultations and negotiations about the issues involved might develop in future.

Part  III also presents the author's view of the "pros and cons" of each scenario. This inevitably involves subjective judgement; it is intended to provide a starting point for debate.

SUMMARY OF PART I
GMPCS AND REGULATORY POLICY - THE BACKGROUND

Defining GMPCS

Part I of the report defines a GMPCS as a system with the following attributes:

(1) It provides real-time voice communications and/or fax or data links at 9.6 Kbit/s or faster, to small mobile or portable terminals (fixed terminals may also be supported).

(2) It can provide service simultaneously anywhere on the earth's surface, or at least within an area much larger than the geographic coverage of typical geostationary systems in the Fixed Satellite Service (FSS).

(3) The capacity provided, over the whole service area, is substantial: enough to support at least 20,000 simultaneous conversations.

This technology independent definition does not even stipulate the use of satellites, although the practical realisation of GMPCS will make extensive use of satellite "airtime". Consequently the definition does not specify what orbits are used. The GMPCS systems that have reached the detailed design stage so far all use non-geostationary orbits: circular Low Earth Orbits (LEO), Intermediate Circular Orbits (ICO), or elliptical orbits. A geostationary GMPCS would be possible: such an architecture is known to have been considered (but rejected) by at least one prospective operator, INMARSAT.

The definition, and hence the scope of this report, excludes:

• The "little LEO" systems such as ORBCOMM and Starsys
• Regional geostationary systems in the Mobile Satellite Service (MSS), e.g. that of the American Mobile Satellite Corporation (AMSC)
• INMARSAT's existing geostationary MSS system, and future evolutionary extensions (as distinct from INMARSAT-P, a GMPCS system)
• Proposed wide-coverage global satellite systems such as Teledesic and Hughes' Spaceway, which operate in frequency bands most suitable for fixed rather than mobile service (note 1).

The Central Challenge of Regulatory Policy for GMPCS

The wide international equity participation in companies proposing to build GMPCS systems shows that many investors expect a favourable "verdict of the marketplace": they believe GMPCS will be substantially profitable. Many national governments are likely to judge that this points to an opportunity for substantial benefits to their national economies, implying that regulatory policy should facilitate rapid deployment of GMPCS.

At the same time it is also apparent, for example from some initiatives at the 1992 World Administrative Radio Conference (WARC 92) and from recent statements by the European Commission, that consensus about how the deployment of GMPCS should be handled cannot be taken for granted. There are evidently policy concerns in some quarters that could come into conflict with the desire for rapid deployment.

This report deals with both aspects, considering how to facilitate deployment; how to resolve policy disagreements that might delay it; and how to do so in a manner consistent with diverse national objectives.

SUMMARY OF PART II
ISSUES AND CONCLUSIONS IN SIX MAJOR AREAS OF REGULATORY POLICY

This report considers regulatory issues arising in six areas:

• Frequency allocation, assignment and co-ordination
• Licensing
• Network interconnection and revenue sharing
• "Equitable access" and competition policy
• Standards
• Numbering

In each case, the discussion follows the same sequence:

• Background
• Fundamental regulatory issues
• Recent and current regulatory developments
• Policy issues and options for the future
  

Background

Each new type of satellite system has required the solution of new issues in national and international spectrum management on at least two different levels:

• The allocation of frequencies to services such as the Fixed Satellite Service (FSS) or the Mobile Satellite Service (MSS)
• The assignment of frequencies and orbital slots to individual satellites, and the co-ordination of these assignments with other spectrum users to avoid harmful interference

In some cases the agreements adopted have also involved:

• The allotment of frequencies and orbital slots on an a-priori basis within an allocated band to particular national administrations.

The way that these three processes have worked up to now has set the stage for the current challenges of spectrum-management for GMPCS by:

• Committing more and more of the spectrum to pre-existing allocations, allotments and assignments, so that it becomes difficult to find room for new services (the new allocations for GMPCS are consequently very small)
• Increasingly requiring several different "services" to share the same allocations, which intensifies and complicates the problems of co-ordination (the new allocations for GMPCS are all shared (note 2)).
• Moving away from a-priori allotments of frequencies and orbital slots, since they are inflexible and likely to waste the orbit/spectrum resource. This means that specific assignments of the orbit/spectrum resource in the GMPCS bands will be determined by national licensing together with the international process of "advance publication", notification, co-ordination and registration of assignments administered by the ITU. This operates on the basis of date priority; essentially "first come, first served"(note 3).

Fundamental Regulatory Issues

Earlier generations of satellites, as well as GMPCS systems, needed frequency allocations, assignments, and co-ordination. However, the very small allocations for GMPCS (only 16.5 MHz in the L-band, for example); very wide international coverage; and intensive requirements for spectrum sharing, when considered together imply some new questions:

• In view of the small allocations, is it necessary to revisit the "first come, first served" method of registration and co-ordination, not to adopt an alternative (which is certainly not possible and probably not desirable), but to refine it? For example, it is likely that some planned GMPCS systems may never come into operation. The Federal Communications Commission (FCC) in the United States has "advance published" with the ITU, under the name HIBLEO, a generic spectrum requirement to accommodate five US-licensed GMPCS systems. These prospectively will occupy the whole of the 1610-1626.5 MHz allocation (in the "L-band") and the whole of the 2483.5-2500 MHz allocation (in the "S-band"). If some of this resource (legally protected by an ITU registration with date-priority) subsequently proves not to be required for the five systems (possibly because one or more of them are never deployed), should the whole assignment remain "warehoused" by the FCC for subsequent release to the remaining US licensees to accomodate growth, or to new US licensees (a process that this report calls "national reassignment")? Should these frequencies also be available to other national administrations that may notify their requirements later? Similar considerations apply to the UK administration's advance publication on behalf of INMARSAT for the 1980-2010 MHz and 2170-2200 MHz bands.
• Because non-geostationary GMPCS satellites are constantly in motion relative to the earth's surface, their signals may potentially interfere with those of numerous geostationary satellites, not just a few as in the case of conventional co-ordinations among geostationary satellites. This is a major problem for the "feeder-links" between the GMPCS satellites and gateways. Currently the international Radio Regulations (RR) are somewhat tilted in favour of geostationary systems in this situation (note 4). Whether revisions are needed, or whether common-sense co-operation and interpretation of the RR is sufficient, is an important question.
• Once GMPCS frequency assignments are internationally co-ordinated and registered, what limitations will apply to the operator's use of them? In practice, the answer in the case of geostationary satellites has been "in the space-to-earth direction, none". For example, the Astra satellites, internationally registered by Luxembourg, transmit to essentially all of Europe. (Transmitting from the ground to the satellite still requires national licensing). Will this de facto "legitimate spillover" extend to the global coverage by a GMPCS system? (While Resolutions 46 and 70 of WARC 92 affirm national governments' right to control the provision of GMPCS-based services within their national territory, and prospective GMPCS operators have indicated that they intend to comply with this (note 5), this is not the same as saying that a GMPCS space segment licensee cannot use its frequency assignment when its satellites are over countries other than the country that licensed them).
• Will the need to continually replace the relatively short-lived GMPCS satellites nullify in practice the established principle that international registration does not guarantee spectrum usage rights in perpetuity, but only for a finite satellite lifetime?
• Since all or most of the most useful frequency allocations for GMPCS may, initially at least, be occupied by licensees of those administrations who first notified a requirement, it is necessary to take a fresh look at the relationship between national and international decision-making. Will these administrations' decisions about how the GMPCS bands are to be used (for example the recent FCC decision to divide the 1610-1626.5 MHz band into separate sub-bands for systems using different architectures, TDMA and CDMA (note 6)) become in effect a de facto global orbit/spectrum plan for GMPCS? Will other national administrations find this satisfactory, or will they object? If they object, what mechanisms could be used to resolve the disagreement?

Recent and Current Regulatory Developments

Several national government agencies and international institutions (most notably the FCC, INMARSAT, the ITU, and the European Commission) have been active in the field of spectrum policy for GMPCS. The WARC 92 conference produced international agreement on two major steps needed for deployment of GMPCS:

• New global frequency allocations (See Exhibit ES1)
• Procedures for notification, co-ordination and registration of assignments of frequencies and orbital positions for GMPCS, based on previous ITU practice, but with new features appropriate to large non-geostationary systems. (The procedures are provisional, and subject to refinement at future World Radiocommunication Conferences.)

National authorities in countries wishing to take the lead in GMPCS were quick to follow up WARC 92 by incorporating the allocations into national regulations, and proceeding with the "advance publication" stage of the ITU notification/co-ordination/registration process.

WARC 92 ended in March 1992. In the United States, the FCC promptly adopted the new allocations (note 7). Even before the conference, in February 1992, the US government had filed with the ITU for advance publication of a "generic" GMPCS requirement called HIBLEO. It followed this up by initiating international co-ordination for variants of HIBLEO on 28 March 1993, while the WARC 92 conference was still in session (note 8). As we noted earlier, HIBLEO, if co-ordinated without modification, would occupy the whole of two of the frequency bands newly allocated to GMPCS at WARC 92, and in fact the two considered most suitable for early deployment (note 9). The UK administration, acting on behalf of INMARSAT, "advance published" a requirement in June 1992 for all of two further frequency bands at 1.9/2.2 GHz (note 10). (For this reason, this report refers to the UK and US administrations, for brevity, as the "first movers"). What is less well-known is that many other administrations followed suit soon after, with filings by Brazil, Germany, Netherlands, Russia (note 11), Tonga and others. Nevertheless, the date-priority of the US and UK filings put the prospective US licensees and INMARSAT in a strong position (note 12).

WARC 92 ALLOCATIONS INTENDED FOR GMPCS USE

Policy Questions and Options for the Future

Where spectrum use for GMPCS is concerned (unlike some other aspects of regulation), the "die is cast": regulatory decisions are far advanced, and therefore it is of limited value to consider wide-ranging alternatives. Nevertheless, there are important unanswered questions:

• Will the prospective GMPCS space segment operators and the "first mover" regulators that license them be able to secure the far-reaching co-operation from other spectrum users and national administrations that is needed in order to satisfactorily solve the formidable co-ordination and spectrum-sharing problems inherent in the WARC 92 allocations?
• Will administrations that now have date-priority for assignments (the "first movers") subsequently resist modifying their assignments or will they (as is usual and customary) respond to other countries' requirements notified subsequently in a flexible way, "making room" for other users if new systems emerge and if some licensed systems never come into operation. The FCC's Notice of Proposed Rulemaking (NPRM (note 13) in February 1994 and the Joint Proposal and Settlement Agreement (note 14)(a proposal by four prospective US GMPCS licensees) proposed "national re-assignment" of spectrum. The FCC envisaged that spectrum within the US assignment that became surplus to the needs of the initial licensees (eg. because one goes out of business) "could be made available to [another US] licensee upon a showing of need (note 15)". However, the FCC's decisions on licensing policy in the Report and Order (note 16)of October 14 are less categorical and place more emphasis on international consultation: the FCC is apparently maintaining policy flexibility. This involves an important principle: other systems will eventually need to be accommodated, and the spirit of international co-operation on which economically successful GMPCS deployment necessarily depends will be enhanced by a flexible approach on the part of the "first mover" administrations that have date-priority on their side.

Licensing

Four types of licensing are relevant for GMPCS:

• Space segment licensing
• Gateway licensing
• Licensing of the provision of GMPCS-based services to end-users (or to resellers, which in mobile communications are often referred to as "service providers")
• Licensing or type-approval of user terminals

The distinctions are important. Space segment licensing by one country, though it includes assignment of uplink and downlink frequencies, does not give a GMPCS operator the right to set up gateway earth stations, interconnect to the Public Switched Telephone Network (PSTN) or provide services to end-users in other countries.

Space Segment Licensing

Background

Space segment licensing is carried out by national regulatory authorities such as MPT in Japan, DTI and the Radiocommunication Agency in the UK, or the FCC (note 17). What we call "licensing" is known by various names; it comprises:

• Approval of the proposed technical characteristics, orbital characteristics ("the ephemeris" (note 18)), and mode of operation of the satellite(s)
• Permission to construct and launch the satellite(s)
• Frequency assignments for uplinks and downlinks, and (where applicable) inter-satellite links.

National government, through the act of licensing a satellite, undertakes responsibility with respect to other countries for its operation. The satellite assumes a national identity: the licensing government holds the satellite owner responsible for operating the satellite within the terms of the licence. This national identity gives rise to a series of reciprocal rights and obligations under international law and practice in which the ITU plays a significant role. Chief among them are the co-ordination and registration provisions of the international Radio Regulations (RR), specifying an obligation to prevent harmful radio interference through co-ordination with the governments of countries where other spectrum users may be affected.

Any analysis of space segment licensing has to begin with a review of current procedures and a consideration of whether and how improvements can be made, short of a lengthy and complex process to develop radically new multilateral rules.

Fundamental Regulatory Issues

Administratively, there is little difference between GMPCS licensing and the licensing of other kinds of satellites. In substance, there is a difference at both the national and international levels. New regulatory problems arise at the national level, and the various possible solutions have international repercussions. For example, as we already noted, two main types of GMPCS systems, those based on CDMA (such as Globalstar) and those based on TDMA (such as Iridium) cannot in practice share the same spectrum, but require different band segments. The FCC has had to struggle to accommodate the five prospective licensees (four CDMA systems and one TDMA system) the FCC has approved from among the "first round" applicants (the initial "processing group"), even looking at the problem on a strictly national basis. It has also had to consider what to do if one or more of these applicants' GMPCs systems is not implemented.

Other novel problems arise as well. For geostationary systems, national regulators grant licences for the life of a satellite: they do not grant an automatic or perpetual right to replace one satellite with another. In a "Big-LEO" GMPCS, however, the design life of each satellite is short (typically 5 to 7 years). Significant numbers of the satellites will fail before the end of their design life, because of problems arising from the use of low orbits (e.g. the presence of orbital debris). Therefore allowance must be made for relatively frequent replacement of the satellites in a GMPCS "constellation". This requires a modified licensing approach. As we noted earlier, the same issue also arises in connection with the international registration of assignments.

The wide geographic coverage of GMPCS, and the fact that any individual GMPCS system will occupy a substantial proportion of the available spectrum allocations, means that the international implications of national space-segment licensing are qualitatively different from previous experience:

• Consequences of global coverage: where "spillover" of the satellite "footprint" on the earth's surface beyond the boundaries of the licensing country is concerned, GMPCS operators must deal realistically with the new situation created by their own technology. They must either live with the problem of national sensitivities in other countries concerning downlinks (which could culminate in a requirement to "switch off" the satellites' downlink beams, or bar user access to the services the beams provide, over some countries), or seek to overcome these sensitivities through international consultations on regulatory issues. Some governments might welcome this; others might oppose it as representing too great an intrusion into national decision-making.
• Consequences of very limited allocations and high spectrum demand: We already noted that practical circumstances mean space-segment licensing decisions by a national regulator could amount to de facto global orbit/spectrum plans for certain frequency bands. Whether this is how events unfold in practice will depend on the flexibility the these administrations show within the international co-ordination process.

Recent and Current Developments

GMPCS space segment licensing has so far moved ahead furthest in the United States and the UK; significant developments are also occurring in Brazil and Russia.

In the US, the FCC is at an advanced stage of a complex rulemaking proceeding (note 19)(described in detail in Chapter 4) which has recently resulted in major decisions on licensing policy (note 20). Licenses will soon be issued. Key points of the decisions were:

Policy decisions (October 1994):

1. Assigning the whole of the international allocation at 1610-1626.5 MHz and 2483.5-2500 MHz to US licensees
2. Dividing the 1610-1626.5 MHz band into sub-bands, one for CDMA and one for the TDMA (note 21)
3. Imposing minimum requirements for US national coverage
4. Classifying GMPCS space segment operators as not being "common carriers" or operators of "Commercial Mobile Radio Service" (CMRS) in the US regulatory system, so that they will be minimally regulated and will not be subjected to the foreign ownership restrictions of Section 310 of the Communications Act. This averted a potentially troublesome situation in view of international sensitivities and the GMPCS operators' plans for wide international equity participation
5. Authorising on a blanket basis handsets approved by US­licensed GMPCS operators, even if they are brought in by users from abroad
6. Requiring US licensees "to meet... any national requirements imposed by other licensing administrations regarding operations within their territories".

Licensing decisions (expected in early 1995):

1. Licensing four CDMA systems:
• Aries (Constellation)
• Ellipso (Mobile Communications Holdings Inc.)
• Globalstar (Loral/Qualcomm)
• Odyssey (TRW)
2. Licensing one TDMA system, Iridium.

Licences will be issued in early 1995 only if the applicants can agree with the FCC about how to "squeeze" their spectrum requirements down to 100% of the available worldwide L-band and S-band allocations. They appear to be close to achieving this. If they do not, the FCC will auction the frequency assignments.

In addition, the FCC's recent Report and Order also:

• Requires US-licensed GMPCS operators to provide a minimum level of global coverage (note 22). This provision ensures that US licensing does not pre-empt a major global orbit/spectrum resource without making provision for its use to provide service in other countries.
• Rejects a recent proposal by four prospective US GMPCS licensees (note 23)that would have incorporated in the FCC rules and licence conditions a provision forbidding any US-licensed GMPCS operator to accept frequency assignments from other national administrations, for operations over other parts of the world, that are more favourable to that operator than the FCC assignments. The provision appears rational from a technical point of view, but it would have raised serious international jurisdictional issues; its rejection shows that the FCC is sensitive to these issues.

In parallel with the licensing process in the US, the UK administration has similarly moved to begin the necessary regulatory steps for the establishment of the INMARSAT-P system.

Other governments that have taken action include Brazil, Germany, the Netherlands and Russia. However, none of these situations, by a margin of several years, is as close to the implementation of a real GMPCS system as are INMARSAT and the US licensees, though the Brazilian project (ECO 8) and possibly the Russian Project (Ellikom) appear to have made significant progress.

Policy Decisions and Options for the Future

Since space-segment licensing has progressed so far already, it is very late to consider fundamentally different approaches. Nevertheless, a debate is taking place:

• The European Commission has objected to decisions emerging from the US licensing process, even though they are fully compatible with the RR, because they effectively assert a national claim to 100% of two important frequency bands and also purport to settle "de facto" many matters about how these bands will be used. In a diplomatic Note to the United States, the Commission argues that:

"In view of the global nature of the satellite personal communications services proposed, the European Commission considers that it would be inappropriate for the FCC to proceed to license all or some of the proposed operators prior to an international agreement as to the principal conditions of operation of services"  (note 24)

• Some observers have suggested that a new international licensing status for private-sector GMPCS operators might be considered: this would give such operators an international standing to provide services worldwide analogous in some ways to that enjoyed today by the public-sector international satellite organisations (ISOs) such as INMARSAT (note 25).

While there may well be a longer-term role for radical concepts such as this, the immediate policy choices available are incremental. They concern:

• Whether the FCC will in fact assert in practice a right of "national re-assignment" of unused spectrum (note 26).
• How flexible the regulatory administrations of the "first mover" countries (the UK and the US) will be in the international co-ordination process.

Thus, while the licensing process will remain within the traditional framework, important choices still remain to be made.

Gateway Licensing

Background

Gateways will connect the GMPCS space segment and the Public Switched Telephone Network (PSTN). Gateway licensing is a national matter: national regulatory authorities' consent is needed for each gateway and its feeder links (note 27). An international dimension is involved, however, because feeder link frequency assignments need to be internationally co-ordinated.

Some GMPCS operators may grant "gateway regions" to companies which invest in their space segment. Such arrangements may assign to the investor exclusive rights (and obligations) to establish a gateway in one country, or gateways in several countries, possibly with the right to serve additional countries from there via the international PSTN, or by other methods such as leased-line resale.

Recent and Current Developments

So far as the author is aware, only two national administrations have so far moved far towards licensing gateways and/or their feeder links: those of the UK and US. The FCC's Report and Order (note 28)initiates a process to make feeder link frequency assignments, which would also authorise the operation of gateways. It also explores possible solutions to a variety of national and international frequency co-ordination problems involving feeder links.

While authorities in EU countries other than the UK have not yet moved far towards licensing gateways, several companies (including telecommunications operators) within the EU have agreed to invest and participate in GMPCS space-segment operations and establish one or more gateways. The European Commission has begun to consider the regulatory issues involved. It engaged the consulting firm KPMG Peat Marwick to prepare a study of these issues (note 29), among other aspects of GMPCS. This was completed in March 1994: the recommendations envisage using gateway licensing as a lever to protect European interests in negotiations with the GMPCS operators and the US government.

Policy Questions and Options for the Future

Gateway licensing can be approached in future in three different ways:

• Treating the gateway licensing decision as a "stand alone" decision
• Treating the decision essentially in that way, but taking into account the policies of neighbouring countries likely to be served from the same gateway
• Treating the gateway licensing decision more strategically, linking the decision to a broader policy for GMPCS and hence a range of other issues.

In the first and second cases, gateway licensing may be (but will not necessarily be) fairly speedy and straightforward in countries where the national regulatory authority favours competition. In countries where a regulatory regime of monopoly or limited entry still prevails, as in India or Argentina, the outcome will depend on whether the monopoly operator regards GMPCS as a threat, or an opportunity for collaboration.

In the third case, the future of gateway licensing is inseparable from a comprehensive national GMPCS policy, which may also involve international consultations or negotiations. The possibilities are explored, in scenario form, in Part III.

Handset Licensing / Type Approval

Background

The licensing of GMPCS handsets and other user terminals is obviously crucial. Fully exploiting the geographic coverage of GMPCS obviously requires some arrangement allowing handsets to be carried freely across national boundaries and used in a wide variety of jurisdictions. While handsets will be small and unobtrusive, prospective GMPCS operators have stated that they will not allow handsets to be used where they are not licensed (note 30), and in any case users will want reassurance that their use is lawful. What is needed is a generic approval of the terminals and a generic right to operate them (a "class licence" in UK terminology) rather than a license for every terminal: preferably these approvals would be valid in a wide variety of countries.

Licensing of user terminals is undoubtedly a national prerogative, but licensing by each individual country would be extremely cumbersome. Rather than abandon the principle of licensing altogether (which is undesirable for reasons of control of radio interference, and perhaps also for health and safety reasons), this suggests that some international arrangement should be made to provide for the recognition by each country participating in GMPCS of "class licenses" granted by other countries.

Fundamental Regulatory Issues

Obtaining wide generic regulatory approval of terminals may not be easy. National radio-regulatory agencies are sensitive about their sovereign prerogatives of controlling (subject to international agreements) the use of radio transmitting devices on their territory. However, some experience does exist of methods to achieve wide generic approval. For example, within the EU a "mutual recognition" (MR) regime now operates for the GSM digital cellular systems: a GSM terminal that meets specified technical standards (note 31) and is licensed or type-approved in one EU or European Economic Area (EEA) country is automatically considered to be licensed in every EU country (note 32).

Regulatory questions concerning the handsets licensing include:

• Will national administrations other than the "first mover" that authorises a particular GMPCS space segment (and the associated terminals) also license these terminals? Will they type-approve specific terminal designs, or even a broad "class" of terminals (via a "class licence"), so that individual terminals will not need to be licensed?
• Will they make conditions for such licensing beyond basic technical requirements (e.g. health and safety, avoidance of interference)? Will linkages emerge (as they may also in the case of gateways) with negotiations involving other national objectives?
• Will there be international arrangements "harmonising" licensing arrangements in multiple countries to facilitate free circulation of handsets?

Recent and Current Developments

There have been significant recent developments on this subject:

• Handset terminal licensing is one element of the GMPCS licensing process currently taking place in the United States. The FCC's Report and Order confirms a policy of type-approval of GMPCS handsets:

"rather than requiring mobile transceivers to be licensed individually, we propose a blanket licensing approach".

The FCC has effectively delegated such approval to the GMPCS operators:

"we propose to require an end-user to obtain the authorisation of the space station operator..." (note 33).

It has also contributed to international mobility by indicating that handsets in use in other countries may be brought into the United States, provided they are approved by the US-licensed GMPCS operators (note 34)

• It is important to note, however, that this is not a "mutual recognition of licences" (MR) approach like the current GSM policy in Europe. "Recognition" is extended not to the licensing decisions of other governments, but only to type-approval decisions made by the GMPCS operators themselves (and only those operators whose are US-licensed). This policy does, however, perhaps represent a beginning that could potentially be extended in the direction of a true MR regime.
• The possibility exists of achieving unrestricted movement of GMPCS handsets throughout essentially all of Western Europe (the EU plus the EEA countries). This could be done by extending the existing regulatory arrangements for mobility of GSM digital cellular terminals. A study prepared for the European Commission has proposed doing so, not unconditionally but with conditions attached to protect European interests (note 35).
• An interesting unofficial proposal has been floated in an informal paper by Stephen Cheston of Iridium Inc. (note 36) Mr Cheston proposes an ITU-administered registry mechanism. National licenses for GMPCS handsets would be registered, as would voluntary decisions made by authorities in other countries to "associate themselves" with these licenses, thereby making them effective in their own national territories. This could well be a useful device. It is not a mutual-recognition regime, because it is not reciprocal: if country A licenses a handset type and country B "associates" itself with this licence, under Mr Cheston's proposal this does not place any obligation on country A to recognise country B's handset licenses.

Policy Questions and Options for the Future

Policy in this field is still emerging. The main possibilities appear to be:

• National licensing with no systematic mechanism to foster international mobility of handsets, except for free movement into the US of handsets approved by US-licensed GMPCS operators, as provided for by the FCC
• A patchwork of bilateral arrangements between pairs of countries. These might either be "one way" arrangements (in which "country B" agrees to accept GMPCS handsets approved in the US, for example (note 37)) or mutual recognition arrangements
• Multilateral arrangements. These could be non-reciprocal, as in the ideas put forward by Stephen Cheston, or true mutual recognition arrangements.

It is an open question whether institutional mechanisms for fostering wide circulation of GMPCS handsets would be likely to gain wide international acceptance without embodying reciprocity, as in the mutual recognition of licences. Opportunities exist to create a large "free circulation area" for GMPCS handsets, based on reciprocity . One such possibility would be to combine the extension of the EU's mutual recognition regime for GSM to cover GMPCS, and an agreement between the EU and the US to associate the US with the EU mutual-recognition area. This would mean that US­licensed handsets could be used throughout the EU (and hence also in the EEA countries), and vice-versa.

Service Licensing

Background/Fundamental Issues

Service licensing in any given country is at least as important as space segment and gateway licensing, and it is a separate issue. Some observers assume that such licensing (or its absence) will be meaningless because of the mobility of pocket-sized GMPCS handsets, but in practice GMPCS operators will probably not be able to take such a cavalier view. If GMPCS operators are technically able to "switch off" service to a given country, they will feel constrained to do so if failing to do this would breach the laws of that country. Prospective operators have generally indicated that they will provide service only where this is permitted by the relevant national administration. Thus GMPCS operators need to consider how they or their local affiliates can obtain the necessary consents outside the "first mover" country.

In view of the relationships that some GMPCS consortia are establishing with local affiliates, an important related issue is whether, when licensing a GMPCS service, national regulatory administrations will allow service to be provided under an exclusive vertical relationship between the space segment operator and a local gateway operator and/or service provider.

Recent and Current Developments

Naturally neither INMARSAT-P (to be licensed by the UK) nor the US-licensed GMPCS operators will encounter significant obstacles to service licensing in their "home" countries. However, there is an important issue affecting service licensing in the US. It concerns whether GMPCS operators fall within the legal category of "common carriers", specifically providers of "Commercial Mobile Radio Services" (CMRS), and are consequently subject to obligations concerning non-discrimination among different categories of customers. Fortunately the FCC has exercised its discretion not to impose common-carrier obligations, subject to certain conditions. Designation of GMPCS as CMRS would have had the further important result of imposing foreign-ownership restrictions on GMPCS operators.

In other countries, regulatory administrations have generally not yet taken a position concerning the licensing of GMPCS services: the early evolution of EU policy is described in Chapter 13.

Policy Questions and Options for the Future

Along with gateway licensing, handset licensing, and interconnection policy, service licensing will be a key element of policy concerning GMPCS for countries other than the "first movers". In these countries, the key policy question is:

• Whether to unconditionally license GMPCS operators whose space segment is licensed in another country, so as to encourage rapid availability of services, or ...
• Whether to impose conditions. For example, in several countries (France, Germany, and the UK are examples) cellular operators are required to allow resale of airtime by resellers known as "service providers"; in the UK, cellular operators are permitted to sell their services only in that way. Will regulators in those countries insist that GMPCS operators do not restrict resale by independent service providers?

Interconnextion

Background

Interconnection, with adequate capacity, high quality, and prices that are reasonable in relation to cost, is essential for the success of GMPCS. It includes:

(1) Interconnection with the PSTN

(2) Interconnection with the Public Switched Data Network (PSDN) for those GMPCS systems whose architecture requires this

(3) Ability to interwork with terrestrial cellular systems on a "dual mode" basis, where the GMPCS handset will communicate with a terrestrial cellular base station when it can, and with a GMPCS satellite otherwise.

In countries with competitive markets for telecommunications services, requirements (1) and (2), and possibly even (3), can potentially be met by the GMPCS operator (or, depending on the GMPCS business system, an affiliated gateway operator or service provider) simply by buying telecommunications services from national carriers in the domestic resale market. This requires, however, that regulators permit the use of resold telecom services for the carriage of international calls. For reasons discussed in Chapter 4, this is often not the case today, though this may change in many countries by the time GMPCS becomes operational in the late 1990s.

Where the GMPCS operator is not able to do this, it will either have to negotiate interconnection arrangements with an established PSTN operator (possibly with the negotiations being overseen by a national regulator empowered to intervene if they break down, as in the UK), and/or on invoking mandatory interconnection policies. Such policies, like Open Network Provision (ONP) policy of the EU, would be relevant only if they are extended in future to cover GMPCS. In all but the most competitive markets, GMPCS operators have much to gain from the extension of national regulatory policies concerning interconnection to include GMPCS.

Fundamental Regulatory Issues

National regulators must make fundamental choices concerning interconnection for GMPCS. There are four main possibilities:

(1) Interconnection under essentially the same arrangements that apply to "correspondent carriers" for international traffic today:

• GMPCS operators are treated as correspondent carriers
• They make payments to the local PSTN operator ("settlements") based on the existing system of international "accounting rates"
• Intermediate ("transit") PSTN operators carrying the signal in countries between the GMPCS gateway and the calling or called party on the PSTN (when these are in a different country) receive a "transit settlement".

(2) Interconnection arrangements similar to those used for INMARSAT services today: essentially a modification of the correspondent carrier arrangement in which operators of Land Earth Stations (gateways) buy airtime from INMARSAT, sell service to end-users, and pay settlements to the overseas PSTN operators that are involved in the completion of a call.

(3) Interconnection using an open domestic resale environment.

(4) Interconnection making use of mandatory "open network" policies.

The availability or otherwise of these options, and their "pros and cons" from the point of view of a GMPCS operator can vary widely from one country to another. The outcome matters: the price difference between interconnection on an "open resale" basis, and payment of a settlement based on the international accounting rate, can be as large as the gap between 12 US cents/minute for resold long distance service in the US (when traffic volumes are large), and 50 US cents/minute or more for settlements on many international routes.

Recent and Current Developments

So far as the author is aware, significant progress towards a policy for GMPCS interconnection has only been made in two places: the United States and the EU. In the EU, there has been discussion, for example in the recent KPMG study (note 38) for the European Commission, about the possibility of extending the ONP legislation to cover GMPCS, giving GMPCS operators the right to interconnect to the PSTN at cost-based prices. In the United States, the natural solution would be to extend the existing FCC policy favouring unrestricted resale of telecommunications carriers' services to permit GMPCS operators to purchase PSTN service at end-user prices. However, in the case of international GMPCS calls, such possibilities may fall foul of difficulties arising from wider considerations of telecommunications policy concerning international resale.

Policy Questions and Options for the Future

Interconnection policy will be most critical for those operators, like Iridium, who will focus strongly on serving international frequent travellers. For some others, such as Ellipso and Globalstar, the issue may be less pressing if they decide to concentrate mainly on selling bulk airtime to terrestrial cellular operators wanting to extend their service areas (in this case, PSTN interconnection will presumably be provided under the cellular operators' existing arrangements)

Progress can be achieved in one of two ways:

1. An ad hoc effort by the GMPCS operators and their national affiliates to obtain interconnection on a case-by-case basis
2. Efforts to develop regulatory principles multilaterally that favour unrestricted interconnection of GMPCS gateways to the PSTN at cost-based rates

Here, as in the case of handset licensing, there is a case for a multilateral initiative to "smooth the path" for GMPCS, provided the operators are willing to accept a certain amount of "give and take" in the process. For example, an agreement between the US and the EU on GMPCS might open up the possibility of applying the EU's Open Network Provision (ONP) principles to GMPCS with reciprocal opening-up of the US regime of open resale and interconnection to European-licensed operators, including INMARSAT.

Bilateral or multilateral agreements of this kind could also be established between other groups or pairs of countries. It is even possible (although probably over-ambitious) to envisage a global agreement on interconnection matters: and the principles embodied in the recently-concluded GATT Uruguay Round trade agreement may be helpful in this regard.

"Equitable access" and competition policy

Background

The principle of "equitable access" in relation to GMPCS was asserted by Resolution 70 of WARC 92. Two things can be said with confidence about this Resolution. The first is that, for many countries, its acceptance was an essential condition for their consent to the GMPCS frequency allocations: it was perceived to provide some degree of safeguard for their national interests in exchange for what was, in effect, a "go ahead" for worldwide GMPCS deployment. The second that there is no general consensus about the precise meaning of Resolution 70. Its broad intent was clearly to stake out, on behalf of the large majority of national administrations which will in practice have little influence on the deployment of the GMPCS space segment, a strong "say" in how the space segment is used within their territory.

The term "equitable access" suggests some ways this could be done, while ensuring that users receive the benefits of GMPCS. They involve applying the principles of general competition policy, and concern:

• Whether it is legitimate for GMPCS operators to grant exclusive "gateway territories" comprising several countries, or...
• Whether national (or EU) competition policy will require open competitive entry into, and participation in, the GMPCS business through resale arrangements, specifically the downstream "service provider" role (note 39).

Recent and Current Developments

It appears that GMPCS operators have become more sensitive to the issue; some may, for example, be willing to operate an open market in "service provision".

The European Commission has taken up the competition policy dimension in its diplomatic Note to the US State Department (note 40), which states that

"the conditions under which operators can provide satellite personal communications services would fall, in several important respects, under the provisions of Community law regarding the internal market, competition and trade in services and equipment".

The KPMG Peat Marwick study for the Commission explores in some detail the scope for action under EU law to combat what the Note calls "potentially anti-competitive provisions" that might be adopted by GMPCS operators. This presumably refers to "vertical" relationships such exclusive gateway franchises over extensive regions, or exclusive "service provider" rights.

Policy Questions and Future Options

The future agenda in this area of policy is inseparable from the service licensing, discussed in the previous section. It does, however, involve the additional "dimension" of competition policy: national decision-makers must decide whether to insist on wide, or even fully open, opportunities for market entry by competing service providers and perhaps even gateway operators.

Standards

Background

Technical standards for GMPCS may be needed for three main reasons: inter-operability in the interests of network connectivity and open competition; economies of scale and experience in manufacturing; and health and safety. There is as strong a case for a "zero based" approach in standardisation as there is in regulation (any particular type of standard should be considered unnecessary until it is shown to be necessary: "guilty until proved innocent"). We discuss three specific areas: handsets and other terminals, gateways, and architectural features of GMPCS networks as a whole.

Two aspects of standardisation are especially important for GMPCS deployment:

1. Establishment of standards for handsets whose acceptance is likely to facilitate free circulation of handsets over a wide area of the world
2. Establishment of interface standards that could facilitate:
• an "open network" approach to interworking with the GMPCS space segment by numerous "service providers" and/or...
• interworking by numerous cellular operators with the GMPCS space segment on a "dual mode" basis.

Some long-range efforts being carried on in the ITU and the European Telecommunications Standards Institute (ETSI) may have a large impact on the evolution of GMPCS (and personal communications generally) in the longer term. They aim at a comprehensive network architecture embracing the fixed network, terrestrial "wireless" networks, and GMPCS which could provide true Universal Personal Telecommunications (UPT), with greatly enhanced connectivity for the user. In UPT, a correspondent could be reached anywhere by calling his or her personal number.

Fundamental Regulatory Issues

The regulatory issues in this area can be summed up by three questions:

• What to standardise?
• How to standardise?
• Who should standardise?

In the area of what to standardise, the key options are:

(1) Minimal standardisation: only a few indispensable standards for terminals and interfaces that are required by other regulatory policies (e.g. mutual recognition of licences for terminals)

(2) More extensive standardisation of GMPCS system characteristics

(3) Broad public-network architectures designed to support UPT, like the long-range ITU concept, Future Personal Land-Mobile Communications Systems (FPLMTS) (now known as NMT 2000), or the corresponding ETSI concept UMTS.

Of these three, (1) is already happening, and will be especially valuable if some policy options that rely heavily on the use of standards (notably the "Mutual Recognition of Licences" scenario we describe in Part III) are adopted. Option (2) appears unlikely (and not necessarily even desirable) in a world of intense competition and proprietary technology. Option (3) is an important long range process already under way, but is likely to have only a modest impact, at most, on first-generation GMPCS.

Recent and Current Developments

Standardisation for GMPCS is proceeding in practice along two tracks:

• De facto, the existing GSM digital cellular standard, based on TDMA, is being adapted and extended to provide important elements of the standards for Iridium, also a TDMA system. In a broadly parallel way, the emerging standards for terrestrial CDMA cellular, including technology developed by Qualcomm, are being embodied in CDMA-based GMPCS systems such as Globalstar. It appears that the main owners of intellectual property specific to the GMPCS application of TDMA, Motorola, and the company with the corresponding intellectual property position for CDMA, Qualcomm, are prepared to license certain elements of their technology to others so that it becomes widely accepted as a de facto standard in each case.
• De jure formal technical standards for certain aspects of GMPCS are being developed through national standardisation bodies, ETSI and the ITU.

Long-range standardisation of comprehensive personal-communication architectures is, as we already noted, being pursued through the ITU and ETSI.

Policy Decisions and Future Options

Many decisions will have to be made about standardisation in the next few years. GMPCS operators will have to weigh up the desirability of ambitious standardisation efforts (which may help achieve important objectives such as free circulation of terminals) against a possible loss of competitive advantage, and certain expectations about the licensing of intellectual property, that may be involved. Probably they will select a hybrid approach, which favours extensive standardisation and even sharing of technology in certain important but circumscribed fields such as handsets, but a minimalist approach to standardisation in general.

Numbering

Background

Since many proposals for GMPCS systems envisage providing connectivity for calls to (as well as from) users who may at a particular time be located anywhere in the world, they require international numbering arrangements that meet several criteria:

• Telephone users anywhere should be able to call any GMPCS user without needing prior information about where he or she is located at the time.
• A GMPCS terminal's number should (like country codes and US area codes today) be easy to understand: new numbering arrangements should appear as a natural and easy-to-remember extension of today's numbering.
• The long range need for evolution towards a future UPT/personal number system should be taken into account.

INMARSAT today has a numbering plan, established within the ITU framework, which points to how future GMPCS numbering might be done. It treats INMARSAT's three ocean regions as "countries", each with its own code (the Atlantic has two, east and west). Similarly, future GMPCS systems could well each be assigned a single country code.

Recent and Current Developments

No major decisions have yet been taken about numbering policy for GMPCS. However, both INMARSAT (for INMARSAT-P) and several of the US-based prospective GMPCS operators are known to have a strong interest in the subject: INMARSAT, for example, has recently commissioned a study from consultants about numbering. The key arena for decision-making will in practice be the Telecommunications Standardisation Sector of the ITU (ITU-T).

Policy Decisions and Future Options

If certain GMPCS systems sell their bulk airtime to cellular operators to extend terrestrial coverage (as Ellipso and Globalstar may do), numbering may not be a major issue for them. In this situation GMPCS is likely to operate within today's numbering arrangements for terrestrial cellular. However, for GMPCS operators whose business plans require the implementation of global roaming, numbering will be a critical issue. The key choice will be between:

• Seeking a solution broadly similar to the numbering arrangement INMARSAT uses today, requiring international consensus through the ITU
• Remaining under the umbrella of one national administration. For a US-licensed GMPCS system, for example, this might involve being treated as an extra area code (strictly a Service Access Code or SAC, like 800 or 900) within the North American Numbering Plan.

SUMMARY OF PART III
FROM ISSUES TO COMPREHENSIVE POLICIES

What are the alternative ways in which the regulatory issues discussed in Part II can be resolved (or, if not resolved, at least "managed")? There are many possibilities, but this study has sought to distil their essentials into six "scenarios", as summarised in Exhibit ES2:

(1) "Business as Usual / Inflexible Approach"

(2) "Business as Usual / Flexible Consultative Approach"

(3) "Good Neighbours"

(4) "Mutual Recognition of Licences"

(5) "College of Regulators"

(6) "Global Regulatory Consultative Process".

ES. 2

The scenarios are inevitably somewhat simplified: the real course of events is likely to combine aspects of more than one. For example, it is possible that the "Mutual Recognition of Licences" scenario will apply to GMPCS handsets, while one or other of the two "Business as Usual" scenarios applies to space segment licensing.

The scenarios are not "options" among which regulators can choose freely. In practice, events are already moving with a high degree of momentum. As time goes on commitments are made and particular roads taken. For example, recent regulatory decisions in the US points strongly towards Scenario 2.

Scenario 1
"Business As Usual/Inflexible Approach"

In this scenario, national telecommunications regulators we call the "first movers" (in practice the FCC and the UK administration) license several GMPCS systems and complete the notification, co-ordination and registration process with the ITU. No agreement with other national regulatory administrations is established by the UK and US administrations prior to launch, except ITU co-ordination (which is designed specifically and solely to avoid radio interference). These "first mover" administrations prove inflexible in co-ordinations with other administrations, using date priority to retain control of a large part of the available GMPCS spectrum.

Certain other governments grant approval for installations and activities forming part of the overall GMPCS system within their jurisdictions (uplink transmissions to the GMPCS space segment; establishment of gateways; use of GMPCS handsets; and offering of GMPCS service) on a unilateral basis, i.e. without any international negotiations, at the behest of the GMPCS operator or its local partner. This is done because these regulators consider it to be in the best interests of users, and to increase competition.

Some other national governments, and the European Commission, object that unilateral approval of the space segment by the "first mover" country is in effect pre-empting telecommunications policy in other countries. Some governments may refuse consent for GMPCS gateways, PTSN interconnection arrangements, and/or the use of GMPCS handsets on their national territory; others (probably more numerous) may make approval conditional on compliance with a variety of national (or EU) policies.

Scenario 2
"Business as Usual / Flexible Consultative Approach"

In this scenario, in a formal sense national space segment licensing and international co-ordination would proceed exactly as in Scenario 1. In substance, however, there would be a large difference:

• The "first mover" administrations would act flexibly, in co-ordination discussions to accommodate other administrations requirements.
• If space segment licence terms provided for "national re-assignment" of unused spectrum, the "first mover" regulator would nevertheless ensure that the availability of such spectrum for re-assignment to other GMPCS licensees of the same country is subject to future international co-ordination (note 41).
• The "first mover" would avoid policy decisions and licence conditions tending to imply any jurisdiction outside its own country.

There would be no institutional innovations, but the GMPCS operators and the regulators in the "first mover" countries would seek to avoid the conflicts that might occur under Scenario 1, through ad hoc bilateral consultations pursuing one or more of the following goals:

• Authorisation for GMPCS operators licensed in the "first mover" countries to transmit and receive signals in other countries
• Agreements on licensing of GMPCS handsets and other terminals
• Agreements on resale of GMPCS by service providers
• Agreement on regulatory rules concerning PSTN interconnection to the and interworking with terrestrial cellular operators.

This would involve a proliferation of bilateral consultative activities, which might be simplified if one such arrangement became a widely-accepted "model".

Scenario 3
"Good Neighbours"

In this scenario, today's formal institutional arrangements still remain unchanged. As in Scenario 2, the national regulators that first license GMPCS systems (the "first movers") approach international co-ordination flexibly. As in Scenario 2 they seek understandings and agreements with other national regulators on matters of mutual interest such as free movement of GMPCS handsets.

The difference from Scenario 2, however, is that the process of mutual accommodation of interests goes considerably further:

• The "first mover" regulators respond to the concerns of other national administrations either by postponing space-segment licensing until a common view has been reached on key issues, or possibly by making licenses conditional in such a way as to allow the regulator flexibility to subsequently vary licence conditions (note 42) in light of international consultations.
• Discussions between the "first movers" and other national regulators and other relevant parties (notably the European Commission, which drafts and administers the "framework" legislation for telecommunications regulation and competition policy in Europe) establish agreed and influential (but non-binding) regulatory principles concerning, for example:
• Authorisations for GMPCS handsets and other terminals
• Conditions under which GMPCS operators or their local affiliates may install and operate gateways
• PSTN interconnection arrangements
• Management of the shared orbit/spectrum resource, e.g., its division between domains set aside for TDMA and CDMA operation.

Essentially, Scenario 3 would represent acceptance of the European Commission's position, set out in its recent diplomatic Note to the US State Department (note 43).

Scenario 4
"Mutual Recognition of Licenses"

The basic method of this scenario is already used within the EU countries to create a "single market" for telecommunications terminal equipment and allow unrestricted "roaming" by users of GSM digital cellular service. Two or more countries agree on a "mutual recognition" arrangement (which we refer to for brevity as "MR"), under which each country's regulator then undertakes to treat any products, services or activities in specified categories that have been licensed by another country participating in the MR arrangement as if it had issued the licenses itself. MR can operate bilaterally, as in the case of the existing MR agreement between France and Germany covering licenses for VSAT networks, or it can operate multilaterally, as in the existing GSM arrangements and the European Commission's proposals for extending the MR principle for VSAT licenses throughout the EU.

In the case of GMPCS, mutual recognition could also operate bilaterally or multilaterally. It could apply to only one or a few aspects of GMPCS (handset licensing, for example), or to many different aspects. The rest of this section indicates how mutual recognition might work in the case of GMPCS, with both a bilateral version of the scenario (4A) and a multilateral version (4B).

Scenario 4A
Bilateral Mutual Recognition

In this scenario, pairs of national governments or regulatory agencies enter into bilateral MR arrangements. Variants of Scenario 4A could range from a relatively modest (but potentially very valuable) version concerned only with the free circulation of GMPCS handsets, to a maximum version covering all aspects of GMPCS licensing. In principle, Scenario 4A involves complete reciprocity between participating countries; in this respect it is unlike the earlier Scenarios. For handset licensing, reciprocity would give all participating governments the opportunity to designate part of the "pool" of approved handset types. If space segment licensing were included, MR would mean that all participating countries would automatically recognise licenses granted in the "first mover" countries. Some governments might object that this involves voluntarily limiting the exercise of their sovereignty without obtaining in exchange reciprocal rights of equal practical value. Thus it is arguable that MR for the space segment might not work unless it were combined with MR for handsets, service providers and so on, providing market access to the new GMPCS industry for participants from numerous countries.

Scenario 4B
Multilateral Mutual Recognition and the Mutual Recognition Group (MRG)

In this scenario, national governments and regulators wishing to see rapid GMPCS deployment form an association (the "Mutual Recognition Group") as a sub-set of the ITU membership, analogous to the the existing EU arrangements for GSM. The concept of the MRG as a group formed by voluntary association is in keeping with the notion of "variable geometry" increasingly accepted in international institutions (most notably in the EU, not all of whose Member States participate in its EU's more ambitious integration initiatives).

The decision to join the MRG would not be made lightly: a national government joining a comprehensive MRG would commit itself to recognise as licensed in its own country numerous licenses issued by other member governments of the MRG. In exchange, all MRG members would agree that their national licensing would conform to agreed common policies in areas defined by a binding international agreement which established the MRG. The areas might include:

• Standards for handsets/terminals, with unrestricted provision by any vendor of handsets/terminals conforming to the standards ("open type approval")
• Rights to establish gateways and/or rights of resellers ("service providers") or cellular operators in each country to resell GMPCS service.

The MRG concept may seem ambitious, but in the author's view it may be feasible:

• Participation would be voluntary, and the MRG would be based on the principle of mutual gain to the participants: a "win/win" solution. Countries whose governments did not consider rapid deployment of GMPCS as a "win" (whether for reasons of industrial policy, social control, or otherwise) would naturally not participate, as they would find the obligations of membership unacceptable. (On the other hand, countries whose governments favoured rapid deployment also might not participate if they felt that the negotiating process in the MRG would generate common principles implying too many restrictions on their own freedom of action or that of the GMPCS space-segment operators).
• A comprehensive MRG may be more feasible than a partial one, dealing (for example) only with handset licensing. The latter may be seen by many countries as representing a concession of authority to the "first mover" country and its licensees, without sufficient reciprocity in favour of other countries.

The possibility exists of a more limited non-reciprocal arrangement for approval of handsets, as proposed by Stephen Cheston of Iridium Inc, as discussed in Part II. This has the advantage of simplicity and might be relatively easy to agree on; however, the mechanism unused might remain if governments felt it involved an insufficient degree of reciprocity.

Scenario 5
"College of Regulators"

In this scenario, governments and regulatory agencies seek to realise the benefits of the "Good Neighbours" scenario, while avoiding the web of bilateral consultations it would inevitably entail and avoiding making new binding international agreements.

In Scenario 5, national regulatory administrations would establish a multilateral process for consultation among themselves, under ITU auspices. Such a "College of Regulators" would not be a formal inter-governmental body and its work products would not have the status of treaties. Its activities would resemble those of the ITU-T and ITU-R Study Groups, but would address regulatory subject matter. Like them, it would seek consensus, embodied in formal Recommendations. The subject matter might include:

• Handset licensing/free circulation of handsets
• Regulatory principles concerning gateways and PSTN interconnection
• Principles concerning the independent service providers.

Its goal would be to facilitate the deployment of GMPCS systems worldwide, with due regard to other policy goals of participating governments.

Scenario 6
"Global Regulatory Consultative Process"

This scenario is more ambitious, and significantly less realistic, than the MRG concept of Scenario 4B, or Scenario 5. It would be more ambitious than the MRG because it would involve the entire ITU membership, and not just a self-selected sub-set of the membership, and more ambitious than the College of Regulators because it would seek to negotiate binding international agreements. It would involve three main elements:

• An inter-governmental body, possibly established as a new ITU "Sector", to undertake preparatory work for international agreements in three of the six areas of regulatory policy considered in this report; namely:
• licensing
• interconnection
• "non discriminatory access", e.g., the role of service providers

The other areas (radio spectrum, standards, and numbering) would not be included because they are already handled by existing ITU machinery.

• Regulatory Conferences at which binding international agreements would be sought. Such efforts would have to be based on a wide consensus, including the agreement of the "home" country or countries licensing the GMPCS space segment. Possibly the negotiating effort might be linked to the provisions of the GATT Uruguay Round agreement concerning free trade in telecommunications services, and considered as an effort to implement its principles.

Lest this scenario seems outlandish, it is useful to recall that the machinery of GATT in the field of international trade has performed a similar function fairly successfully for the past four decades, and the new World Trade Organisation (WTO) will expand this role.

EVALUATION: "PROS AND CONS" OF THE SCENARIOS

This section summarises the author's view of the advantages and disadvantages of each particular scenario. The material is set out in note form as "pros and cons".

While stated as "facts" in the interests of brief and clear expression, the "pros and cons" are strictly the author's own judgements. In making them, he has kept in mind that in today's climate, every new initiative for public sector action, at national or international level, must survive severe testing of its practicality and its benefits.

Scenario 1
"Business as Usual/Inflexible Approach"

Scenario 1 "works". Its major advantages lie in its relative simplicity and speed of decision making:

• Complex decisions (e.g. those concerning the co-existence of GMPCS systems with different characteristics, such as TDMA versus CDMA) are made fairly speedily.
• Regulatory approval for the GMPCS space segment is timely.
• In some other countries, at least, gateways will be established, handsets licensed, and GMPCS service offered promptly. (However, regulatory authorities in some other countries may refuse consent, or impose conditions).

• GMPCS operators may fail to get access to some national markets or may obtain it only under restrictive conditions, because certain national governments are dissatisfied with a GMPCS deployment over which they have little influence. (For example, a gateway outside a country's territory may route calls into that territory via the PSTN.)
• National regulatory authorities can to some extent use their own licensing powers to bargain with the GMPCS operators, but even this will not be able to influence certain decisions "cast in concrete" unilaterally by the "first mover" regulator, like the split of the 1610-1626.5 MHz band between CDMA and TDMA.

Governments other than the "first movers" may well divide into three camps: those who consent to the full deployment of GMPCS, without misgivings; those who do so reluctantly, concerned about their lack of influence on how GMPCS is deployed, but feeling that its benefits nevertheless outweigh the disadvantages; and those (probably relatively few) who block deployment entirely. In practice, most national administrations will probably allow GMPCS service to be offered in their territories, but in this scenario deployment may take place on a basis that is less favourable (in terms, for example, of international mobility of terminals, or PSTN interconnection) than might be negotiated in an environment where decisions are made with more international consultation and reciprocity.

Scenario 2
"Business as Usual / Flexible Consultative Approach"

• Scenario 2, like Scenario 1, would allow for speedy regulatory decision-making on space segment deployment.
• As in Scenario 1, in some countries timely national regulatory consents may then allow the establishment of gateways, the licensing of handsets, the agreement of PSTN interconnection terms, and the offering of GMPCS services might follow quickly.
• Again, as in Scenario 1, there may be countries where this does not occur, or where restrictive conditions are imposed. It seems reasonable to expect, however, that there may be fewer such countries, and therefore quicker expansion of GMPCS services, in Scenario 2 than in Scenario 1 because national regulatory authorities in the majority of countries respond favourably to the more consultative character of decision-making in Scenario 2

• The greater degree of international consultation in Scenario 2 will require additional time and effort from national regulators. This may also involve some delays, although there is nothing in Scenario 2 that would delay initial issuing of GMPSC licenses by the "first mover" regulator, except perhaps the work needed to adapt those licenses to preserve for the regulator the "room for manoeuvre" in the international orbit/spectrum co-ordination process required by the Scenario.

Scenario 3
Good Neighbours

The advantages of Scenario 3 would be similar to those of Scenario 2, except that:

• Because of the genuinely reciprocal character of the consultations in Scenario 3, with an opportunity for national administrations to influence the "first mover" regulator's licensing decisions, there are likely to be still fewer countries who withhold or delay regulatory approvals or impose restrictive conditions.
• Initial licensing of the space segment might be delayed to some degree to make it possible to embody in licence conditions any modifications agreed between the "first mover" regulator and other national regulators in the course of consultations. (However, in view of the several years still to elapse before GMPCS space segment deployment, it is not clear that this would delay the overall time-scale.)

Cons

• The extensive bilateral consultations involved in Scenario 3 would consume substantial staff time.
• It may be too late to implement this Scenario in the United States, because of the stage already reached by the licensing process.

Scenario 4
Mutual Recognition of Licenses

We discuss the "pros and cons" of Mutual Recognition (MR) in general, before turning to the specific arguments for and against Scenario 4A (bilateral MR) or Scenario 4B (the Multilateral Mutual Recognition Group or MRG).

Pros

• By definition, MR would ensure (if applied to terminals) that handsets and other terminals could circulate freely in participating countries. Similarly, MR for gateways and PSTN interconnection would break down obstacles in these areas, and MR for the space segment would ensure that the GMPCS space segment could transmit signals to, and receive signals freely and lawfully from, the territory of participating countries.
• For governments or regulators other than those of the "first mover" countries, consultations within the MR framework would give them a voice in decisions where otherwise they would have relatively little say (note 44). This "voice" might be used in bargaining for such goals as wider participation of a country's "own" businesses in the establishment and operation of gateways, the supply of equipment, or resale/"service provision".
• Why the regulatory agency in the "first mover" country should be willing to slow down its decision-making in order to engage in negotiations to establish a legally-binding MR regime may be less obvious. However, the "first mover" might expect to benefit from:
• Unrestricted circulation of GMPCS handsets and other GMPCS terminals in more countries than might otherwise be possible
• Licensing of GMPCS gateways in more countries and with more uniform conditions
• Better terms for interconnection.
  

• This scenario therefore has the potential to be a "win/win" solution, in which governments, users and GMPCS operators all gain.
• Governments are unlikely to enter MR agreements unless these agreements contain basic standards or principles with which licenses must comply before being eligible to be automatically recognised under the MR agreement in other participating countries (note 45). Reaching agreement on the applicable standards and principles may involve substantial cost or delay. (It is not clear, however, that it need take so long that it would interfere with the timing of GMPCS deployment, and there is nothing in Scenario 4 to prevent national regulators from issuing licenses unilaterally, and seeking MR treatment for them subsequently).
• MR agreements, with their associated standards and principles, involve some loss of freedom of action by national administrations.

Specific "Pros and Cons" of the Bilateral (4A) and Multilateral (4B) Approach
to Mutual Recognition of Licenses

Bilateral (Scenario 4A)

Pros

• Simplicity for any given pair of countries (but separate negotiations are needed for numerous pairs of countries)
• A possibility (but not certainty) of reaching agreement quickly
• No need to create new international institutions.

Cons

• Unlikely to result in uniform regulatory conditions for the operation for GMPCS in multiple countries. This is most significant in the case of the regulation of handsets, where uniform rules allowing unrestricted circulation of handsets are highly desirable.
• For any one country (especially the "first mover" countries that license the GMPCS space segment), negotiating and administrating bilateral MR agreements with large numbers of countries may prove cumbersome.
• If MR applies to space segment licensing, delays in deployment may result.

Multilateral (Scenario 4B)

Pros

• Would produce relatively uniform regulatory conditions for the operation of GMPCS in all the countries participating in the MRG. This would be extremely valuable in the case of free circulation of handsets.
• A multinational negotiating process, like the ITU Radiocommunications Conferences, would centralise negotiations on certain issues; and this might be less costly for participating administrations than wide-ranging bilateral negotiations.

Cons

• Putting an MRG in place through multilateral negotiations might involve substantial delay. If the "first mover" country were to agree to delay its process in order to align its space segment licensing decision with common principles agreed in the MRG, this would no doubt slow down the deployment of GMPCS. It is more plausible to assume that the "first mover" would not delay its initial licensing decision, but would instead try to build flexibility into license conditions to allow subsequent adjustments accommodating the common "standards and principles" on which consensus is needed to obtain an overall MR agreement.
• It is possible that some governments, motivated by industrial-strategy considerations might join the MR negotiations process and deliberately delay it to allow "their" own domestic industry time to catch up.
• Countries whose governments favoured rapid deployment might not be willing to participate if they felt that the MRG negotiating process might generate common principles implying too many restrictions on their freedom of action or that of the GMPCS space-segment operators.

Scenario 5
College of Regulators

Pros

• This scenario involves an informal consultative approach to build a consensus on issues involving GMPC, rather than the negotiation of binding agreements as envisaged in the Mutual Recognition scenarios. It may therefore be possible to accomplish results more quickly (but perhaps less reliably) than in the MR scenarios.
• This scenario may be more acceptable than Scenario 4 (or Scenario 6, below) to those governments which are highly sensitive about national sovereignty and are reluctant to see the formal machinery of international negotiations or binding international agreements expanded in scope, but which see value in an international consultative process on regulatory issues.

Cons

• In the absence of a mechanism for negotiating binding agreements the influence and value of a "College of Regulators" would depend on the degree of voluntary responsiveness to other administrations' issues that each national administration (especially those of the "first mover" countries) bring to the process. If confidence is not built up through voluntary initiatives, administrations other than the "first movers" may feel that this scenario does not meet their needs for influence on decision-making.
• In this scenario there could be no self-enforcing mechanism for restricting participation of any ITU member government in the "College of Regulators", as there is in the multilateral MR scenario (Scenario 4B). There is therefore a risk that national administrations unfavourably disposed to GMPCS deployment will participate and obstruct progress from the point of view of those administrations that do want rapid deployment. Timely progress towards agreement could hardly be guaranteed in a process necessarily open to all ITU member governments. The "convoy" might, in practice, steam at only the speed of the slowest ships.

Scenario 6:
Global Regulatory Consultative Process

Pros

• This approach (like Scenario 5) provides the entire ITU membership with the opportunities for consultation on regulatory matters. However, unlike Scenario 5 it would provide for negotiation of binding agreements.
• It may therefore be favoured by those (like the European Commission) who argue that a comprehensive process of international consultation is needed on the regulatory issues arising from GMPCS.

Cons

• Putting such machinery into operation, if feasible at all, would be highly controversial. It is likely to be much slower than obtaining international consensus or agreement under any of the other scenarios previously discussed.
• If national administrations delayed their own decision-making on licensing during the consultative process, this (unlike the consultation and negotiation process in the earlier scenarios) might involve a long enough time-lag to delay the physical deployment of the GMPCS space segment. In practice, this might not happen; licensing and deployment would go ahead regardless, leaving the "Global Regulatory Consultative Process" with little real influence on the course of events, at least for the first generation of GMPCS systems.
• In the author's judgement, this scenario is therefore neither feasible nor promising:
• It is politically infeasible because of the degree of perceived infringement of national sovereignty involved.
• While it appears to be a "stronger" concept than the MRG, in that it engages the machinery and tradition of voluntary compliance with rules made in the common interest that characterise the ITU as a whole, it may not in fact be more effective:
• Participation would have to be open to all ITU members, risking the same problem of "steaming as slow as the slowest ship" that arises in the "College of Regulators" scenario, but to a still greater degree, since the prospect of legally-binding agreements will engender greater caution.
• Having such machinery wrestling with licensing and other issues might well strain the tradition of consensus that has served the ITU community very well over many decades.
  

Overall, the author's personal view is that Scenarios 3, 4 and 5 would be the most beneficial from a general public interest point of view, and also for the GMPCS operators, because they balance the benefits of timely GMPCS deployment with the need for international consultation. This is desirable in its own right and will contribute to obtaining the wholehearted and timely co-operation of numerous national regulatory administrations which is a key requirement for the success of GMPCS. The author particularly favours an effort to implement the multilateral Mutual Recognition Group concept (Scenario 4B), which will, he believes, best achieve this balance and be most effective in facilitating the free movement of GMPCS handsets.

To order the full briefing report, click here