To download the results
Version française

ART launched the public consultation on number portability on 13 October 2000. It received some thirty responses from a variety of players, including a number of fixed network operators (local loop and long distance operators) and mobile network operators. Private and public users, industry and some professional organisations also expressed their opinions. Consumer representatives expressed frustration at the technical difficulty of some of the questions.

Regardless of the position of these players in the value chain, the consultation showed up a consensus on a number of priorities:

• For number portability, the need to distinguish between the types of numbers: geographic, non-geographic and mobile
• The desire to adopt a progressive approach in implementing solutions: creating simple and economical solutions initially, then putting in place more elaborate solutions
• The need to review the cost allocation principles associated with number portability (allocated to the recipient operator), since the current system is highly unfavourable to new entrants and could hamper the development of number portability
• The need to clearly define the role of each participant: the community of operators makes the technical choices and ART monitors the works, via a coordination system still to be defined (body, meetings, etc.)

The operators insist on the importance of geographic number portability within the context of increased competition on the local loop, emphasizing that this is an essential condition to ensure fair competition on this segment.

The operators consider that non-geographic number portability must begin with freephone numbers which must be ported as quickly as possible, before extending the mechanism to specially tariffed non-geographic numbers. They consider that a simple solution must be implemented, based on manual procedures before automated operational procedures are developed, which will be indispensable once number portability affects a large number of users.

Mobile operators consider that a specific mobile operators' group must first define the implementation of portability of mobile numbers and draw up specifications before discussions are extended to include other operators and telecommunications players.

Some contributors point out that the specificity of the overseas départements will have to be taken into account when implementing number portability.

Such a body could define and propose ways of implementing number portability, and its timetables. It seems necessary that a consensus be reached on how it is to function for portability to enjoy de facto legitimacy.

Most operators consider it necessary to create a coordination body among concerned parties.

However, the contributors emphasize that the specific nature of the different types of numbers (geographic, non-geographic and mobile numbers) must be taken into consideration in implementing number portability. These specificities are related to the retail market, to the number and diversity of players concerned, to the technical characteristics of the various networks, to existing offers and to the state of advancement of the works in progress. For this reason, operators suggest organising specific project structures.

Once common infrastructures—such as a reference database—are in place, many operators seem to favour the creation of a European economic interest grouping, which would operate autonomously (following the example of Belgian operators). One operator suggests that a specific study clearly define the legal framework of such an entity.

The primary concern of such a grouping would be non-geographic number portability.

The contributors consider that, at this stage, non-geographic and mobile number portability requires only working groups, which would be divided into subgroups based on the topics studied.

These groups would determine the technical solutions chosen, the architectures and the organisation of information exchange among operators. It would define how a reference base would be implemented, guarantee the methods used by each network are interoperable, study economic aspects and set implementation schedules.

All contributors agree on the need for a global coordination body which would monitor all the various groups and the interoperability of the techniques.

The operators propose that project groups then consortiums be created for each type of number. These groups would include a core of motivated operators, to which other operators could then be added. Some would also admit service providers and industry, although operators oppose the participation of third parties such as users, for efficiency reasons. User groups, however, want open structures, following the example of Internet discussion groups.

The three structures (geographic, non-geographic and mobile numbers) could then report to ART on a regular basis, in accordance with specifications previously approved by ART. These reports could be made to a global coordination body, with expanded representation to include users and outside persons, and led by ART.

Contributors suggest that ART publish guidelines, if necessary, based on the work of these groups and set deadlines, especially as regards mobile number portability.

In general, contributors seem to prefer operation by consensus, although some contributors suggest a vote in a plenary assembly, and others wish as much flexibility as possible.

Financing of a common database would be covered by the operators. Various possibilities are mentioned: equal contributions from all operators using the resource, or contributions on the basis of the size of the operator (based on numbers allocated or sales). Operators having participated in setting up the project would pay a reduced entrance fee.

For studies, each participant will cover its own expenses. This will be the case, for example, for mobile number portability.

Contributors generally distinguish between criteria regarding porting (deadline for porting, database query and operation, deadlock rate, synchronisation among operators, customer information, simplicity for the customer and operational procedures for operators, etc.) and criteria related to ported calls (call routing times, continuity of service, etc.).

Some contributors want these criteria to be progressively defined by the working groups, while others wish ART to draft a standard of quality.

The contributors consider that operators should analyse service quality, with ART monitoring it in order to rule on disputes.

The criteria will depend on the type of solution chosen, and on how it is monitored. For example, some checks could be automated if a central database is used. Customer satisfaction surveys would also be useful.

In general, contributors want to ensure the quality of the service associated with ported numbers not be degraded. Thus, the criteria in effect for calls to non-ported numbers should also apply calls to ported numbers.

Contributors welcome the formation of an observatory using statistics submitted by portability players, according to a process created by operators.

All contributors emphasize the importance of "pricing transparency", i.e. guaranteeing the caller that he/she will not pay a surcharge, when a number's price changes because of portability. The operators want to maintain freedom in setting their own prices. The contributors distinguish between non-geographic and mobile numbers, and geographic numbers, where prices are set by the transport operator (local loop or long distance operator) and which should not change when the number called has been ported.

As for non-geographic numbers, where part of the routing cost might be charged to the person called, contributors emphasize that a maximum price per numbering range is set when ART allocates these numbers, which necessarily limits too great a variation in prices.

As for mobile numbers, contributors note a trend to even out retail prices, thanks to the convergence of the call termination prices on the mobile networks. This should prevent the introduction of pricing complexity for the consumer.

Most contributors consider clear customer information on the call price indispensable. Some contributors suggest that this information be made available to consumers at all times via voice servers or on Internet, short messages (SMS) or even via on-line information before a call is put through, especially for non-geographic or mobile numbers.

This question is generally dealt with in the previous question. Some contributors insist that the costs of number portability should be, if necessary, charged only to the ported customer, and not to the caller.

The vast majority of contributors consider that the change in the current distribution of portability costs (charged to the recipient operator) is very important, because the current system would be highly unfavourable to new entrants and would hamper the development of competition on the local market. Indeed, some users—such as government agencies, associations or companies—need to be able to keep their number.

The contributors note that the technical choices made in implementing portability are very closely linked to the principles of cost allocation which will be or have already been chosen. These principles must be based on the best interest of the consumer, encourage efficiency, be oriented to the costs of reciprocal services and minimise the number of items cross charged.

A number of proposals were made:

• Charge costs to the operator originating the call
• Charge costs to the donor operator
• Systems in which the costs would be shared by operators

For this last proposal, each operator would, for example, bear the costs of upgrading its own network. Administration and management costs would be borne by the recipient operator and additional routing costs would be charged to the donor operator or to the operator originating the call to encourage it to use a more efficient method. If a central database is used, the allocation of shared charges would have to be defined, and the costs of using this database would be borne by the operator to whom the number is being ported or by the operator originating the call.

For these last costs, some contributors suggest following the example of Belgian and Dutch operators, where each operator covers its own costs of interfacing with the database.

A progressive system, which would change with the market, was also suggested where cost allocation favouring new entrants could—initially—help them enter the market, with the distribution then progressively changing to a balanced distribution between the donor and the recipient operators.

While contributors note that the person receiving the number would not necessarily be the donor operator in mobile networks where competition is already well established, they generally agree that the cost allocation for fixed and mobile networks will likely follow the same principles.

Contributors wish reciprocal services to be cost oriented and that the methods adopted simplify the repayment process in order to make it easier for operators in the chain to monitor it.

The vast majority of concerned contributors note that the current method, while simple, will not last forever, because it requires costly and inefficient rerouting.

All contributors mention database query methods, using "off-switch" mechanisms. "All Call Query", "Call Drop Back" and "Query On De-activate" methods were rarely mentioned in the responses. One contributor suggested letting each operator decide, as long as the choices are compatible with the interface and that the "prefixing" system is kept at the inter-operator interface.

Concerned operators note that the current "manual" process, is cumbersome, a source of errors and unsuited to large volumes of ported numbers, whereas the introduction of unbundling and the wireless local loop make number portability crucial.

All operators consider standardisation and automation using secure electronic procedures and exchange processes between operators essential. Some contributors suggest that working groups establish these inter-operator processes or that ART draft guidelines or recommendations, as it has already done for preselection (order and one-stop shopping system, setting deadlines in processing requests, etc.).

Few contributors, except for users, express the desire to see operator and consumer procedures standardised. However, contributors express the desire for rapid and flexible procedures in order to simplify service to the consumer.

As a general rule, contributors approve of opening a range of dedicated non-geographic numbers (0878) to meet legislative requirements.

Still, operators doubt that such a service would meet a real need from the market today. They question the pertinence of this type of offer in the current context (explosion of mobile services meeting individuals' needs, availability of shared-cost numbers meeting companies' needs).

Only a few users think that it would be a useful service for companies.

Opinions are divided on whether this service should be considered a fixed telephony service—which would take into account the so-called "location" component—or as a fixed-mobile convergence service.

In the first case, contributors suggested using the method chosen for shared-cost and premium-rate numbers, in order to avoid delaying portability for these numbers.

In order to meet legal deadlines, some contributors suggest quickly putting in place a temporary solution.

Most contributors suggest applying an indirect interconnection method, which would allow operators to better control the service.

The vast majority of contributors prefer block number management, at least initially, although one contributor suggests using this service to test pooled number management.

It was suggested that the beneficiaries be the same as for non-geographic numbers, except a few operators which preferred that this range be reserved.

The responses to this question have been included in the responses to question 16.

The subject of pricing for these numbers is, for most contributors, a major difficulty in defining the service. Contributors generally prefer fixed pricing, for reasons of simplicity of implementation and clarity for the consumer.

Thus, it is suggested that the average price of "fixed to mobile" calls or the average of national fixed charges be used. The body responsible for setting retail prices must still be determined. Contributors state that this would require expanding France Telecom's current scale of tariffs.

Contributors lean towards no price controls, and to let the market decide, at least initially.

Contributors generally consider that the difficulties encountered in implementing the initially planned solution were caused by the complexity of the dual objective, i.e. number portability management paired with pooled number management. Concerned operators emphasize the importance of exchange processes among operators, which have not been sufficiently taken into account.

The contributors emphasize that freephone number portability must be implemented quickly, initially through bilateral agreements between France Telecom and the new operators, via a call routing process using the previously defined "prefixing" mechanism.

The use of a central database will be organised by a consortium formed by operators. It will then be necessary to define specifications then develop and operate this base. Contributors feel it important that a detailed timetable be established first. One contributor suggests that the consortium be formed in February 2001 following a feasibility study (launched in November).

The generalisation of number portability to shared-cost and premium-rate numbers is considered to pose the same technical problem as the generalisation of freephone numbers.

Contributors suggest that the consortium, which could be put in place for freephone number portability, also study this extension of portability, keeping in mind the specificities related to cross charging, invoicing and payment for this type of numbers. One contributor doubts there is a need to implement shared-cost and premium-rate number portability before the end of the migration process for the 0801, 0802, 0803 and 0836 ranges (expected for 2002) because of pricing clarity problems for number portability and professional conduct obligations which are not the same for all players.

The contributors—mostly mobile operators—suggest that a working group on mobile number portability quickly set a timetable and objectives.

Two types of work will be necessary:

• Those requiring cooperation among operators, in particular work related to modifying information systems
• Work specific to each operator for its core network and information system

Some contributors think that number portability must be operational when UMTS networks are opened in early 2002.

Operators want to set common directions together, but want to define individually the mechanisms used on their own network. They consider that the role of the working group on mobile number portability will be to encourage discussion among operators.

Contributors note that operators should ensure interoperability between fixed and mobile networks, but that they must not devise overly complex solutions in preparing for "hypothetical" fixed-mobile convergence.

Concerning implementation methods, contributors prefer—with respect to solutions based on Intelligent Network techniques—the so-called "SRF" solution (Signalling Routing Function), based on the mobile network MAP protocol, which would be the fastest to be implemented for transactions not related to calls (sending short messages, location, etc.).

Operators list the advantages and disadvantages of direct (querying a database) and indirect routing (querying the network beneficiary of the number). The direct routing method optimises routing but requires database synchronisation or the creation of a common reference database. It seems suitable when the number of ported subscribers is high, but slower to implement than indirect routing because of its increased complexity.

One contributor suggests the "prefixing" method to transport information on network interfaces (as for fixed networks).

Contributors suggest that the choice of a scenario be left up to the mobile number portability working group. It seems unreasonable to expect operators to make a decision at this point on these techniques, before cost allocation rules have been defined, which will define the structure of the architecture. Moreover, operators consider that the each operator could choose its own solution, as long as service continuity is guaranteed.

One contributor emphasizes that it will be necessary to determine the need for "manual" number portability for inter-operator information exchange in order to meet deadlines, before implementing secure electronic processing.

One contributor wishes that ART publish guidelines regarding operational procedures, as it has done for preselection.

Contributors consider that number portability management by service companies must be organised bilaterally between the service companies and the operator(s) through the contracts they have already signed. Similar opinions are expressed for virtual mobile operators, except where they would directly receive numbering resources, in which case they would become full players in number portability.

For number portability to work, the terminal must no longer be linked to the SIM card. Generally, a terminal cannot work with another SIM card for the first six months of use unless a penalty is paid (generally around FRF 500) to the operator. De-activation is done using a specific code. Operators consider that no additional procedure is needed, since existing measures allow users to de-activate the terminal (paying a penalty if necessary). Operators could retransmit the de-activation code if necessary.

One contributor considers that an economic model needs to be defined that penalises neither the consumer nor the operator.

Contributor opinions are divided but, as a general rule, the prepaid card management system, seems to be a source of additional problems, especially the way in which cards are distributed.

For some operators the "pre-allocation" system is incompatible with number portability. For others there are no structural constraints.

Contributors emphasize that number portability will complicate the contract termination system. The mobile number portability working group will have to study how payment modes will be changed when changing operators (subscription - card).

Contributors observe that UMTS operators must have access to the same functions as GSM operators. To avoid interoperability problems, they suggest that UMTS operators be involved in the process as soon as they are chosen. Contributors consider that common numbering resources will have to be allocated for UMTS and GSM to allow number portability, although this should not affect data services, at least initially.

In general, operators consider that it is too soon to answer at this time, although fairly major changes to their information systems will be required.

Contributors consider it risky to answer because of a lack of visibility, since access to resources depends on the data management architecture.

The experiences of Belgian operators is frequently mentioned as an example we should follow for organisational aspects.

Contributors also mention the UK and the Netherlands (operators in the Netherlands use a central database) for how they implemented mobile number portability.

One contributor points out that no European country manages pooled numbers and that the general trend is to automate operator exchange processes.

Another contributor mentions the counter-example of Hong-Kong where a lack of dialogue has caused a particularly high degree of churning between operators.

Concerning mobile number portability, one contributor notes that direct routing has been chosen in Spain and in the Netherlands.

As a general rule, the inclusion of number portability for a call originating in another country is considered too complex to implement. Still, harmonisation of methods is considered useful by pan-European operators which could realise economies of scale by developing identical solutions.

Technical solutions linked to the implementation of number portability, such developing common databases shared by fixed and mobile operators or systematically introducing methods using an "Intelligent Network" architecture would help prepare for convergence.

According to most contributors, number portability between fixed and mobile numbers cannot be envisaged at this point due to different pricing methods between the two networks. Contributors consider that such a function would hurt pricing clarity and that number portability in the 0878 range, reserved for personal numbers, should be used to develop this type of service.

Contributors consider that the technical difficulties and the investments required to set up such a function are not justified by a real market need. They emphasize that at this point, operators are offering their own fixed-mobile convergence offers.

Contributors do envisage the possibility of using the location component inside a service area, i.e. the possibility of moving house within the zone while keeping the same number.

However, opinions are divided on whether to change the service area: some contributors consider that subscribers are attached to the pricing clarity offered by the service area, whereas others consider that changing will simplify pricing and that the service area could be extended to the transit zone or even the region level. One contributor points out that the geographic information contained in geographic numbers has already been distorted by the multiplication of local loop operators.

One contributor suggests letting operators control the implementation of the location component of number portability, as long as the "operator" component is not systematically linked to it (i.e. that subscribers, having kept their operator after moving house, may need to change numbers if they then change operators).

One contributor suggests making the location component mandatory whenever a customer moves house within his or her switch's area.

Contributors consider that number portability is limited to traditional telephone numbers, whenever they are used to identify an end user, regardless of the technology used.

The majority of contributors state that block number management, which is well suited to the technical constraints of conventional switching equipment, will continue to be pertinent as long as number portability does not reach a critical threshold.

Pooled number management would not be conceivable for geographic numbers unless a very large number of subscribers were ported.

However, contributors insist that blocks will have to be transferred when a large number of numbers is ported within the same block.

Contributors emphasize that potential aspects could be studied after the initial phase and that they must not delay the implementation of number portability in its current context.

Two contributors press for the examination of the problem of telephone directory listings.

Public consultation on the implementation of number portability
List of contributors