Cellular telecommunication network creating process, involves connecting controller to nodes such that messages are routed either to host operator node or to virtual operator node according to value transmitted to network

The field of the invention is that of cellular telecommunications networks. In a cellular telecommunications network, a distinction is generally made between an access network and a core network, managed by a telecommunications operator. The access network allows mobiles to send and receive messages on radio frequencies as a function of geographic areas in one of which each mobile subscribed to the operator is located at a given time. The core network makes it possible to establish mobile connections with another mobile or a fixed processing terminal equipment by passing this connection, if necessary, through other networks.

The interconnection of several operators to the same access network allows these operators to benefit from this access network, improving the economic conditions linked to infrastructure costs and the environmental conditions linked to a sometimes undesirable proliferation of radio antennas. An operator who does not have a license to operate radio frequency bands is named virtual operator (MVNO for short for Mobile Virtual Network Operator in English) with respect to a host operator (MNO for short for Mobile Network Operator in English). English) which has a license to operate an access network.

An operator with an access network will be able to increase the profitability of its access network if it offers its services to another operator who does not have an equivalent access network.

In the state of the prior art for interconnecting several operators to the same access network, solutions are known which consist in interconnecting the core network of a second operator to the core network of a first operator who as operator host, has an access network. The core network of the second operator is seen by the core network of the first operator in much the same way as any other network with which the core network of the first operator is required to establish connections. The nodes of the core network of the second operator, then the virtual operator, are generally interconnected with the nodes of the core network of the host operator.

Such an interconnection in the core network has many drawbacks. Hosting the virtual operator impacts the host operator's core network through the need to adapt the configurations of each operator as changes occur (extension of IP addressing domains, modification of protocols in networks ). Crossing a transit node induces additional delays on packet transmissions. Smart network services for the virtual operator are more difficult to implement. Routing in the core network to international destinations is not flexible for the virtual operator. The virtual operator has no control over the quality of service (QoS) in the access network.

Other solutions are also known which consist in separating each element of the access network, base stations, base station station controllers into two distinct logical parts, each assigned to an operator with specific frequencies and cells. These solutions have the drawbacks of heaviness, bulk and lack of flexibility. For example the frequencies for which an operator has a license, cannot be used by the other operator. There are in a way two neighboring but separate access networks whose independence allows only a lesser symbiosis between the operators.

An objective of the invention is to be able to interconnect several operators to the same access network in a better way than those of the prior art.

A first object of the invention is a method of producing a cellular telecommunications network allowing a so-called virtual operator who does not have a radio access network, to use a radio access network managed by a so-called operator. host, said radio access network comprising access network controllers linked to at least one core network node of the host operator.

The process is remarkable in that it comprises:

a construction step in which batteries of core network nodes are deployed by placing in each battery at least one core network node of the host operator with at least one core network node of the virtual operator, by associating one or more access network controllers to each farm and by connecting the access network controllers to all the core network nodes of the farm with which they are associated;

a configuration step in which each access network controller connected to the network core nodes of the same battery is configured so as to direct messages received from a communication mobile to a frequency of the host operator, either to a core network node of the host operator, or to a core network node of the virtual operator, depending on a value that the mobile sends to the access network to identify itself, said value being distinctive for the host operator and for the virtual operator.

The interconnection provided by the links of the controllers with the core network nodes of the two operators in each cluster of core network nodes, avoids having to establish connections between the core network nodes of the virtual operator and the core network nodes of the host operator. All of the radio transmission frequencies, available at the access network controller, can be used for the entire cluster of core network nodes because the distinction between operators is not made on the frequencies used by the network. mobile but on a distinctive value emitted by the mobile.

To identify itself, the mobile can send a value to the access network in different ways. The value can be based on a permanent identity and sent for example in a first way in a header which precedes a message body intended for a core network, in a second way in a message body intended for the core network , in a third way in a message to the access network. The value can also be based on a temporary identity allocated by the core network in comparable ways.

Particularly to take advantage of the first way, a first range of values ​​is reserved for the host operator so as to cover a permanent identity image when the mobile is subscribed to the host operator and a second range of values ​​is reserved. to the virtual operator so as to cover the permanent identity image when the mobile is subscribed to the virtual operator.

This allows a value which is a permanent identity image emitted by the mobile to identify itself, to be distinctive by the range of values.

Thus, when a mobile only has its permanent subscriber identity to be recognized by the network, the distinct value ranges for each operator allow the access network controller to recognize that of the operators with which the mobile is a subscriber to determine the appropriate core network node cluster, regardless of the frequency used by the mobile to send its message. This image can consist of all or part of the bits of the IMSI or in any other way such as a more elaborate function.

Particularly to take advantage of the second way, in the configuration step, each access network controller is configured so as to take as a distinctive value, the value of a field of a permanent identity which is extracted from a message sent by the mobile to the core network to which the mobile is subscribed.

Particularly to take advantage of the third way, in the configuration step, each access network controller is configured so as to store a correspondence between, on the one hand, a local identifier assigned by the access network controller to the mobile to establish a radio connection between the mobile and said controller, and on the other hand a permanent identity sent by the mobile to said controller to establish said radio connection, and so as to take as distinctive value, the value of a field of the permanent identity which corresponds to the local identifier for a message sent by the mobile on said radio connection, to the core network to which the mobile is subscribed .

Particularly also, in the construction step, each core network node in each farm is assigned a resource identifier that is distinctive of the host operator and the virtual operator and in the configuration step, each node is configured. core network in each battery so that any temporary identity assigned and sent by this node to a mobile communication device, includes the resource identifier so that when the value sent by the communication mobile to identify itself, is an image of the temporary identity, this value including the resource identifier, is distinctive for the host operator and for the virtual operator.

Thus, when a mobile has a temporary identity specially allocated to be recognized by the network, the resource identifiers directly allow the access network controller to determine the appropriate node of the core network of nodes. , regardless of the frequency used by the mobile to send its message.

Advantageously, each battery of core network nodes implemented in the method comprises two core network nodes and the resource identifier is a bit of first state for the host operator and of second state for the virtual operator.

This allows a simplification of the cellular telecommunications network produced by the method.

A second object of the invention is a cellular telecommunications network comprising a first core network managed by a first operator and which comprises core network nodes, organized by batteries of core network nodes and a radio access network which comprises core network nodes. access network controllers, each connected to one or more batteries.

The telecommunications network is remarkable in that it comprises a second core network managed by a second operator and which includes core network nodes, distributed in the batteries so that the access network controllers are connected to the network. at least one node of the second core network, the access network controllers being arranged to direct to a node of the first or second core network, a message sent by a communication mobile to identify itself, as a function of a value received from the communication mobile by the controller in said message, said value being distinctive from the first and from the second operator.

When the first operator is a host operator who manages the access network and the second operator is a virtual operator, the telecommunications network is of the type produced by the method as the first subject of the invention. The two operators can also play a symmetrical role, for example by sharing the access networks that they manage to form a more extensive access network, an operator being the host for a part of the access network that it manages and virtual for part of the access network managed by the other operator. The operators who each manage a core network can all be virtual, the access network being managed by a third-party operator.

Particularly, in the telecommunications network, the value is distinctive by its belonging to a first range of values ​​reserved for the first operator, or to a second range of values ​​reserved for the second operator.

Particularly also in the telecommunications network, the access network controller comprises means for extracting a permanent identity from the mobile, within a message sent by the mobile to a core network and means for taking as distinctive value, that contained in a field of the permanent identity which identifies that of the operators to which the mobile is subscribed.

Particularly also in the telecommunications network, the access network controller comprises a correspondence table between a local identifier for radio connection with a mobile and a permanent identity of this mobile, including a field identifying that of the operators to which the mobile is subscribed. , gives said distinctive value.

Particularly again when the value is based on a temporary identity assigned by the core network, the value is distinctive by its content of a resource identifier which distinguishes, in each farm, the nodes according to whether they are from the first or the second network. heart.

In the telecommunications network, each node of the first core network organized by a cluster of core network nodes, respectively each node of the second core network distributed in a cluster of core network nodes, is then arranged to allocate to any mobile subscriber of the first operator , respectively of the second operator, a temporary mobile subscriber identity comprising the resource identifier assigned to this core network node with the distinctive value of the membership of the core network node to the first operator, respectively to the second operator

Advantageously in the telecommunications network, each battery comprises a core network node of each operator and the resource identifier comprises a bit of first state for the first operator and of second state for the second operator.

When more than two operators share the access network, the identifier consists of a sequence of bits which encodes the number of operators in base two so that there is always at least one first value bit for one. first operator and second value for a second operator.

A third object of the invention is a method of cellular telecommunications by means of controllers of a radio access network and of a mobile authorized to use the communication services of a first operator which manages a first core network or of a mobile phone. 'a second operator which manages a second core network.

The telecommunication method is remarkable in that an access network controller receiving from the mobile a message to be transmitted to that of the first or of the second core network, directs said message to be transmitted, into a battery of first and second nodes. second core networks to which said controller is connected, to a node of the first core network or to a node of the second core network depending on a value sent by the mobile to said controller, said value being distinctive for the operator whose mobile is authorized to use the communication services.

The operator whose mobile is authorized to use the services is for example the operator with which the mobile is subscribed or the operator having concluded a roaming agreement with an operator with which the mobile is subscribed.

Particularly, when the network controller detects in the message that the value contained is a permanent identity image, the core network node to which said controller directs the message is of the first core network if the contained value is covered by a first one. range of values ​​reserved for the first operator and of the second core network if the value contained is covered by a second range of values ​​reserved for the second operator.

The permanent identity image can be a part of the permanent identity, the permanent identity itself, or the result of a known transformation applied to the permanent identity.

Particularly also, when the network controller detects in the message that the value contained is a temporary identity image, the core network node to which said controller directs the message is the one whose resource identifier included in the image d temporary identity, distinguishes a membership of the node to the first core network or to the second core network.

More particularly, each battery comprising a core network node per operator with, for each core network node of the same operator, the same resource identifier in all the batteries, when a first access network controller sends a received message. from a mobile to a first node determined by a first resource identifier in a first battery, a second access network controller which receives the message, routes the message to a second node determined by the same first resource identifier in a second cluster.

The telecommunication method then facilitates the mobility of the mobile which passes to a situation under the control of a new controller connected to the second battery, from a situation under the control of an old controller connected to the first battery different from the first battery. . If both controllers are attached to the same farm, the second node simply remains the first node.

Particularly again, a message to establish the radio connection contains a permanent identity of the mobile of which a field identifies that of the operators to which the mobile is subscribed, the access network controller receiving the message to establish the radio connection, establishes the radio connection. and associates with the established radio connection, all or part of the permanent identity comprising at least the field which identifies that of the operators to which the mobile is subscribed, and the access network controller receiving the message to be transmitted by the established connection, deduces the distinctive value of all or part of the permanent identity associated with the radio connection

This telecommunication method is an advantageous use of a telecommunication network produced by a production method in accordance with the invention in a system of interconnection of several operators to the same access network.

The invention will be better understood in view of an exemplary embodiment now described with reference to the appended drawings in which:

FIG. 1 shows a telecommunications network different from that of the invention in that the interconnection of a second core network takes place at the level of a first core network;

- Figure 2 shows another telecommunications network different from that of the invention in that the interconnection of a second core network is done by allocation in the access network, transmission channels distinct from those of a first core network;

FIG. 3 shows yet another telecommunications network different from that of the invention in that a single core network is connected to the access network;

- Figure 4 shows an operation of the telecommunication network of Figure 3.

FIG. 5 shows process steps for producing a telecommunications network in accordance with the invention;

- Figure 6 shows a telecommunications network according to the invention;

- Figures 7, 8 and 9 show steps of the telecommunication method according to the invention.

FIG. 1 shows a possible architecture of the state of the art in which a host operator hosts a virtual operator (MVNO for short for Mobile Virtual Network in English). For the purposes of the invention, an MVNO is an operator that does not have frequencies and no access network. The host operator has a network which includes a core network and an access network. The virtual operator has a network which only includes a core network and no access network. With reference to FIG. 1 which, by way of illustration, uses UMTS terminology, the core network of the host operator comprises core network nodes such as switching centers 10 for mobile telephones (MSC for short for Mobile Switching Center in English ), and such as packet routers 11 (SGSN for short for Serving GPRS Support Node). The core network also includes other core network nodes that are not shown, such as, for example, a core network database (HLR for short for Home Location Register). The access network comprises radio network controllers 12 (RNC for short for Radio Network Controller in English) and generally several transmitting and receiving stations 13, 14 (Node B in the UMTS vocabulary which is simply one possible type of network among d 'others such as GSM for example).

In a known manner, each transceiver station essentially consists of an antenna which covers a geographical area for transmitting and receiving electromagnetic signals from and to mobile telephones located in this area. In GSM technology, a transceiver station is called a base station.

The RNC 12 has a function equivalent to a base station controller (BSC for short for Base Station Controller) in a GSM network. The essential function is on the one hand the routing of communications between node B and the core network and on the other hand the control and supervision of node B. Unimportant for the understanding of the invention, an RNC is distinguished from 'a BSC by relocation procedures and management of macrodiversity links.

The MSC 10 is a data and signaling switch for managing the establishment of a communication with a mobile telephone by passing this communication in circuit mode by the RNC which controls the node B which covers the area corresponding to a location of the telephone mobile. To manage the establishment of communication, the MSC has a temporary location database of mobile subscribers (VLR for short for Visitor Location Register in English). The VLR is a database that allocates temporary identities (TMSI for short for Temporary Mobile Subscriber Identity) to mobile telephones of subscribers who are located in a geographical area supervised by the VLR, so as to be able to manage each communication. . Although the same VLR can be associated with one or more MSCs, a VLR is generally associated with an MSC which resides in the same physical equipment. The VLR stores both locations and data of subscribers of the network itself and of subscribers of other networks currently supported. The subscribers of other networks currently supported are those roaming (roaming in English) and for which the operators of other networks have concluded agreements with the operator of the network itself who then authorizes both these subscribers of other networks as its own subscribers, to use its communication services.

The SGSN 11 is a data router with functions specific to mobile networks, to manage a communication with a mobile telephone by passing this communication in packet mode by the RNC which controls the node B which covers the zone corresponding to a location of the telephone. mobile. To manage the communication, the SGSN performs a temporary identity allocation (P-TMSI) which has a role similar to that of the TMSI, adapted to the transmission of data in packet mode (GPRS for short for General Packet Radio Service in English) . The SGSN performs packet mode specific functions equivalent to circuit mode functions performed by the MSC and VLR.

The virtual operator's core network comprises core network nodes such as MSC 20, one or more HLR 26, one or more platforms 27 called RI, acronym for intelligent network and one or more GPRS gateways 25 (GGSN for short for Gateway GPRS Support Node in English).

Unlike the MSC 10, the MSC 20 is not associated with a VLR but it is connected on the one hand to the HLR 26 of the virtual operator and on the other hand to the MSC 10 of the host operator. As explained below, the MSC 10 performs a gateway function for the circuit mode between the network of the virtual operator and the MSC (s) of the host operator.

The HLR 26 contains the data relating to the subscribers of the virtual operator such as subscription information and for each subscriber, information on the MSC 10 and the SGSN 11 which lists this subscriber.

The function of the intelligent network RI is to introduce additional functionalities and services into the general network substantially without modifying the existing structure, in particular the switches. An additional layer of intelligence is implemented in the network between the switches of the general network and the applications supporting the services. The concept of IR is based on two basic principles. The first principle is in a separation between on the one hand, the signaling through which the service information passes and on the other hand the transmission of voice or data. The second principle is in a distinction between on the one hand, the common switching functions and on the other hand, what is specific to the service. This concept makes it possible to reduce the costs of introducing and developing additional services in the network.

The GGSN 25 connects in the usual way to the SGSN 11 which is suitable for transmitting in packet mode, data coming from or going to external data networks. The establishment of the data session and the choice of the GGSN are made by means of a domain name (APN abbreviated for Access Point Name in English).

There are two possible alternatives for routing in circuit mode in the upstream direction. In a first alternative, the routing of calls from subscribers of the virtual operator is systematically done to the switches of the virtual operator (home routing in English), based on the international identity of mobile subscriber (IMSI for short for International Mobile Subscriber Identity in English). In a known manner, H MSI is a permanent identity of the mobile which consists of a country code (MCC), an operator code (MNC) and a network subscriber identifier (MSIN). This identity is sent by the mobile when it does not yet have a temporary identifier, for example when the mobile is switched on. In a second alternative, the outgoing call is routed according to the requested number, with a treatment similar to that of the subscribers of the host operator.

It should be noted that according to the state of the art shown in FIG. 1, the interconnection takes place in the core network of the host operator.

FIG. 2 shows another possible architecture of the state of the art in which two operators Op1 and Op2 share the same access network with specific frequencies and cells for each.

The access network includes at least one RNC 2 to control node B 3, 4.

The core network of the operator Op1 comprises its own HLR 36, its own MSC (s) VLR 130, its own SGSN (s) 131, its own GGSN (s) 35 and its own MSC gateway (s) 32 (GMSC for short for Gateway MSC in English). A GMSC is an MSC with a gateway function to route incoming calls from the landline network or other mobile phone networks. Most MSCs have the gateway function. An MSC 130 and an SGSN 131 are arranged to connect to a part of the RNC 2 specially allocated in logical form to the core network Op1. This part of the RNC 2 is designed to control, in each node B 3, 4, a part called node B1 with frequencies reserved for the core operator network Op1.

The core network of the operator Op2 includes its own HLR 46, its own MSC VLR 140, its own SGSN 141 (s), its own GGSN 35 and its own GMSC 42. One MSC 140 and one SGSN 141 are arranged to connect to a part of the RNC 2 specially allocated in logical form to the core network Op2. This part of the RNC 2 is arranged to control, in each node B 3, 4, a part called node B2 with frequencies reserved for the core operator network Op2.

Routing in the upstream direction, mobile to network, is done according to the cell where the mobile is connected, without requiring any particular function or routing table.

Figure 3 shows an architecture that implements the 3GPP TS 23.236 recommendation entitled Intra-domain connection of Radio Access Network (RAN) nodes to multiple Core Network (CN) nodes.

Radio access controllers of the access network are represented here by RNCs 7, 8, 9, 15, 80, 81. A set of core network nodes of the same operator, includes switching centers and packet routers represented respectively by MSCs 16, 17,18,19, 82, 83 and by SGSNs 21, 22, 23, 24, 28. The explanations which follow with the UMTS terminology can be transposed to the GSM terminology by replacing the RNC by BSCs.

3GPP recommendation TS 23.236 defines a function called luFlex to connect several core network nodes to a single access network controller. The function exists for both UMTS and

GSM / GPRS / EDGE with minor adaptations for each. Similar adaptations are possible for any other type of cellular network. The objective is in particular to share the load between the core network nodes. All RNCs, MSCs and SGSNs belong to a single operator according to 3GPP TS 23.236 recommendation.

To implement the lu-Flex function, we define the MSCs, respectively the SGSNs, which belong to the same bank of switches, respectively to the same bank of routers. For example in Figure 3, a switch bank 84 includes the MSCs 16, 17 and 18, a switch bank 85 includes the MSCs 19, 82 and 83, a router bank 88 includes the SGSNs 21 and 22, a switch bank Routers 89 includes SGSN 23, 24 and 28. The quantities of MSC per bank of switches and of SGSN per bank of routers are arbitrary at the discretion of the operator. For each RNC type controller, the switch bank (s) and the router bank (s) to which the RNC is attached are defined. For example in FIG. 4, the RNCs 7, 8 are attached to the switch bank 84 and to the router bank 88, the RNCs 9, 15 are attached to the switch banks 84, 85 and to the router bank 89, the RNCs 80, 81 are attached to the bank of switches 85 and to the bank of routers 89. Associated with the bank of switches 84 is a zone 86 which includes the RNCs 7, 8, 9, 15. Associated with the bank of switches 85 is a zone 87 which includes the RNCs 9, 15, 80, 81. With the bank of routers 88, there is associated a zone 90 which includes the RNCs 7, 8. With the bank of routers 89. , is associated a zone 91 which includes the RNCs 9, 15, 80, 81. A zone in circuit switched mode can comprise the same RNCs as a zone in packet mode as is the case here for zones 87 and 91 or include RNCs different from those of a zone in packet mode as it is here the case of zones 86 and 90. Two zones can also overlap for the same circuit or packet switching mode, as is the case here with zones 86 and 87. Each RNC attached to a switch or router bank is attached to each switch in the switch bank or to each router in the router bank. For example the RNC 7 attached to the switch bank 84, is attached to the MSC 16, 17, 18. Each RNC attached to several switch or router banks, is attached to each switch of the switch banks or to each router of the banks. of routers. For example RNC 9 attached to switch banks 84 and 85, is attached to MSCs 16, 17, 18, 19, 82, 83. The lu-Flex function manages the routing of upstream messages from the RNC and the selection core network nodes in cases of mobility. There are essentially four cases.

A first case concerns routing for a mobile having only a permanent IMSI identity. Routing is based on the IMSI sent by the mobile in the initial message to the core network. More precisely, it is a derivative of the IMSI which is sent f (IMSI).

A second case concerns routing for a mobile having a temporary TMSI identity allocated by a core network node. Routing is based on the TMSI sent by the mobile in the initial message to the core network. More precisely, it is an extract from the TMSI which is sent. The temporary identity TMSI is that of the circuit mode, the explanations remain valid for the temporary identity P-TMSI which is that of the packet mode.

A third case concerns mobility which involves a change of core network node in inactive mode (idle in English). That is to say that the mobile is registered in the network but is not connected. For load sharing, an RNC freely selects a core network node according to an algorithm not defined by the standard of recommendation 3GPP TS 23.236. Routing is then performed to the selected node. Routing can also be based on an old TMSI, ie on the temporary identifier allocated by the old core network node. The choice depends on the setting of the function, for example NRI (explained later) distinct or not in adjacent banks.

A fourth case concerns mobility which involves a change of core network node in active mode. That is to say that the mobile is connected for a communication in circuit mode or in packet mode. For load sharing, the old core network node selects a new core network node according to an algorithm not defined by the standard of recommendation 3GPP TS 23.236. Routing is then performed to the selected node.

In GSM and UMTS, only the initial signaling messages need to be routed. As soon as a connection is established both to convey user data and possibly other signaling data, the routing problems as in the four cases described above no longer arise. The routing of the messages is done simply on this connection established between the mobile and the RNC and between the RNC and the core network. In GPRS, there is never a connection established strictly speaking, the routing is carried out for each frame of both signaling and user data.

By way of example, FIG. 4 illustrates the lu-Flex routing for the second case. A mobile 29 emits an electromagnetic carrier wave which is received by node 5, which is controlled by RNC 8.

In a step 37, the mobile processes a parameter (named IDNNS for Intra Domain NAS Node Selector in English). In UMTS networks, the acronym NAS refers to the English terms Non Access Stratum, ie the network layer comprising the functions independent of the access layer. When the mobile 29 registers with a core network node such as the MSC 16, the core network node allocates a TMSI to the mobile 29. The TMSI generally comprises thirty-two bits b0, b1, ..., b31 . This allocation is made so that the TMSI contains a Network Resource Identifier (NRI for short) which uniquely identifies the core network node within a farm or optionally, also within batteries that cover geographic areas adjacent to that where the mobile is located. The TMSI then contains the NRI at a predetermined position, for example from bit b23 up to potentially bit b14, as a function of a length of NRI chosen during the configuration of the core network nodes. Bit b23 is sufficient when the length of the NRI is one bit. The mobile 29 extracts bits b23 to b14 from the TMSI to form the routing parameter IDNNS as specified in standard 25-331. To send UL messages to the core network, the mobile 29 encapsulates these messages in bit frames 38 sent with the IDNNS parameter as the header.

When in a step 39, the RNC 8 receives a bit frame 38 via the node B 5 from the mobile 29, the RNC 8 extracts the NRI from the IDNNS parameter. The RNC includes a routing table which gives a correspondence between the NRI and a core network node address. This address is for example that of the MSC 16 in the illustration of FIG. 4. The VLR of the MSC 16 which includes the TMSI, also implicitly includes the NRI which constitutes a part of the TMSI. This allows a new core network node to find an old node and the information on the mobile which is referenced therein.

When the mobile 29 only has the IMSI, it generates an IDNNS value equal to f (IMSI) as specified in recommendation 3GPP TS 25.331.

For GSM in circuit mode, the base station controller (BSC for short for Base Station Controller) reads the TMSI included in the initial messages intended for the core network and extracts the NRI from it. If only the IMSI is sent, for example at registration, the BSC freely chooses an MSC by means of an algorithm specific to the implementation.

For GPRS in packet mode, the known LLC protocol (acronym for Logical Link Control) makes it possible to establish a link between the mobile and one of the SGSNs. In an LLC frame, there is a connection identifier called TLLI (acronym for Temporary Logical Link Identifier). When an SGSN has assigned a P-TMSI to the mobile, the TLLI contains a subset of typically the thirty low-order bits of the P-TMSI containing thirty two bits. The BSC reads the TLLI in the LLC frames, more exactly at the level of the RLC / MAC blocks encapsulating the LLC frame, and extracts the NRI to route to the appropriate SGSN. When the mobile has no P-TMSI allocated by an SGSN, the mobile randomly chooses the twenty-seven low-order bits of the TLLI, the other bits indicating that the TLLI is of random type. The BSC then freely chooses an SGSN according to the algorithm implemented. This is only acceptable when the choice of an SGSN is irrelevant, for example when all the SGSNs belong to the same operator.

FIG. 5 shows a method of producing a cellular telecommunications network allowing a so-called virtual operator who does not have a radio access network, to use a radio access network managed by a so-called host operator.

In a first step 62, batteries of core network nodes are defined comprising at least one core network node of the host operator and at least one core network node of the virtual operator, network controllers are associated with access to each bank and the access network controllers associated with this bank are connected to each core network node of the same farm.

In the simplified diagram of FIG. 6, in the core network of the host operator, nodes 30, 50 of MSC type and nodes 31 of SGSN type are considered. Generally, an MSC 30, respectively an MSC 50, is normally connected to several transmission controllers 47, respectively of the access network. The same applies to SGSN 31, 51. The access network controllers 47, 52 intended to be used by the virtual operator, are for example of the BSC type in GSM technology or of the RNC type in UMTS technology. Core network node farms are defined by domain, circuit switched and packet as in the lu-Flex function of 3GPP recommendation TS 23.236 established for a single operator. Batteries 55, 57, are specially allocated to the core network nodes of the MSC type and batteries 56, 58, are specially allocated to the core network nodes of the SGSN type. In the bank 55, one connects a MSC 40 of the virtual operator's core network to all the access network controllers 47 which are connected to the MSC 30. In the bank 57, one connects a MSC 60 of the core network of the operator. 'virtual operator to all access network controllers 52 which are connected to the MSC 50. In the bank 56, an SGSN 41 of the virtual operator's core network is connected to all the access network controllers 47 which are connected to the SGSN 31. In the bank 58, an SGSN 61 of the core network of the network is connected. virtual operator to all access network controllers 52 which are linked to SGSN 51. The known lu-Flex function is not modified as such but to extend its use to several network cores managed by different operators, a particular parameterization of this function will now be described in order to make the interconnection work. in accordance with the invention. Each battery 55, 57 can include other host operator and virtual operator switches to perform load balancing or access network sharing with several virtual operators, for example. Each battery 56, 58 can include other host operator and virtual operator routers for example to perform load balancing or access network sharing with several virtual operators.

In a second step 63, each node of the same switching bank is assigned a resource identifier named NRI in recommendation 3GPP TS 23.236. The NRI identifier can be used on the uplinks (up link in English) from the mobile to the core network to allow the access network controller to direct an uplink communication to the appropriate core network node. The NRI identifier can also be used so that a core network node can find, from a former TMSI or P-TMSI and from a knowledge of a former location area of ​​a mobile, the address of a old core network node which can have information on the mobile. Each access network controller includes a routing table which includes associative correspondences between values ​​of NRI reserved for the host operator and addresses of core network nodes of the host operator on the one hand, between values ​​of NRIs reserved for the virtual operator and addresses of the core network nodes of the virtual operator on the other hand.

A simple way to assign the NRIs in step 63, is to put in step 62, an operator's core network node for each operator in a cluster and to define the NRI identifier as being a value bit for example. equal to zero for a host operator node and equal to one for a virtual operator node. As the NRIs have the same value for the host operator node, respectively for the virtual operator node in all the core network node farms, the routing by the access network controller can be done automatically to the correct core network node when the mobile moves from an area covered by a first controller to an area covered by a second controller. The load balancing for which the lu-Flex function has been provided remains possible by placing several core network nodes from the same operator in the same cluster of core network nodes. The NRI then comprises several bits, the one at zero or at one to indicate the host or virtual operator and the other bits to manage the load distribution among the nodes of the same operator.

Generally, the core network of an operator allocates to each recognized mobile a temporary TMSI or P-TMSI identity which makes it possible to follow the mobile as it moves without having to constantly communicate the IMSI subscriber identity of the mobile.

In step 63, each core network node is configured to allocate each TMSI or P-TMSI so as to contain a value of NRI, equal to the NRI of that node. While in the case of a single operator, the TMSI or P-TMSI can be freely chosen by this operator, prior consultation between operators on specific choices of NRI per operator in the case of several operators, induces a generation of TMSI or P-TMSI specific to each operator. This subsequently allows the access network controllers of the access network to extract the NRI from the TMSI or P-TMSI attached to a message sent by the mobile and then to route the message to the core network node having a value. of NRI which corresponds to the operator with which the mobile is registered because this value is contained in a TMSI or P-TMSI allocated to the mobile by a core network node of this operator.

Before registering with a core network, the mobile does not have TMSI. Generally, the mobile sends an image f (IMSI) of its subscriber identity which is the IMSI itself in the GSM case or a piece of IMSI converted into binary in the UMTS case. In the case of GPRS, the mobile sends a random number independent of the IMSI. This random nature is not a problem when the access network controllers of the access network are linked to the core network of a single operator. Any access network controller receiving a mobile message directs the message to a core network node to which it is connected. The core network node receiving this message can then assign a TMSI or P-TMSI to the mobile for subsequent monitoring.

In step 64, each access network controller provided in the access network to be shared by the host operator and by the virtual operator, is configured to distinguish the range of values ​​in which it receives an image f (IMSI). In the access network controller, the routing table giving the correspondence V = f (IMSI) -> address of the CN node, is constructed as follows: range of f (IMSI) of the MNO -> address of the MNO node and range of f (IMSI) of MVNO -> address of MVNO node. Thus, the access network controller, for example the RNC 47 or 52 of FIG. 6, is configured to direct any message received with an identity image in the header in the first range of values ​​to a core network node, for example the MSC 30 or the SGSN 31, of the host operator and any message received with an identity image in the header in the second range of values ​​towards a core network node, for example the MSC 40 or the SGSN 41 of the virtual operator. Thus, by reserving a separate range of values ​​for each operator, the mobile simply needs to send a value in the range of values ​​reserved for the operator to which it is subscribed, to allow the access network controller to direct the message to the appropriate core network. This is particularly interesting as long as the mobile does not have a TMSI which contains the NRI.

In GSM, the range of values ​​can result from the fact that the image of IMSI is the IMSI itself which is a concatenation of the country code MCC, the operator code MNC and the network subscriber identifier MSIN at the operator . Thus, the MCC with the MNC of the host operator defines for example a first range of values ​​which starts at the value defined by the bits of the MCC and of the MNC followed by bits of MSIN all at zero and which ends at the defined value by MCC and MNC bits followed by MSIN bits all one by one. Likewise, the MCC and the MNC of the virtual operator defines for example a second range of values ​​which starts at the value defined by the bits of the MCC and of the MNC followed by bits of MSIN all at zero and which ends at the value defined by MCC and MNC bits followed by MSIN bits all at one. The range of values ​​can also result from a range of values ​​defined in the MSIN independently of the MCC and the MNC.

In UMTS, the image of IMSI comprising part of the digits of the MSIN, for example three, we impose on each operator to allocate to its subscribers, the MSINs so that these three digits define numbers in base ten in a segment reserved for the operator. Each operator remains free to define at his convenience, the preceding digits and the following digit (s). Thus, for each operator, the digits which constitute the image of IMSI, define a range of values ​​reserved for the operator.

In the light of the preceding explanations, it is understood that the mobile sends a value which is distinctive from the operator with which it is subscribed. This value is distinctive by the fact that it includes an NRI specific to the operator when the mobile has a TMSI or a P-TMSI and by the fact that it is in a range of values ​​reserved for the operator when the mobile not having TMSI or P-TMSI, sends an image of IMSI.

The lu-flex model for which the mobile sends the distinctive value at the header of a message to be transmitted to the core network, as seen previously, constitutes one possible means among others for implementing the invention.

The header value does not need to be distinctive if in step 64, each access network controller is configured so as to take as the distinctive value the value of an identity field which is extracted from the message sent by the mobile to the core network with which the mobile registers. When the mobile places its permanent IMSI or temporary TMSI identity in the body of the message to be transmitted to a core network node, the core network generally has protocol functions dedicated to extracting this permanent or temporary identity from the message body then to the recognition of the different fields of identity. In step 64, these dedicated functions are also installed in the access network controller. The dedicated functions installed are then organized so that the access network controller can read the distinctive value, for example in the MCC and MNC fields of the IMSI or in the NRI field of the TMSI.

. It is also not necessary for the header value of the message to be transmitted to be distinctive if in step 64, each access network controller 47 is configured so as to store a correspondence between on the one hand , a local identifier assigned by the mobile access network controller to establish a radio connection between the mobile and said controller, and on the other hand a permanent identity sent by the mobile to the controller to establish said radio connection. The access network controller is then configured so as to take as a distinctive value the value of a field of the permanent identity which corresponds to the local identifier for a message sent by the mobile on said radio connection, to the destination of the core network with which the mobile is registered.

The telecommunications network obtained by this method allows a virtual operator to use the radio access network managed by the host operator. The teaching of the invention is not limited to a single virtual operator. It is understood that the host operator can host several virtual operators on his access network. In order to host an additional virtual operator, in step 62, it is appropriate to additionally add a core network node of the additional virtual operator. In step 63, the NRI comprises as many bits as necessary to distinguish the host operator and each virtual operator, for example three bits are sufficient to distinguish a host operator and seven virtual operators in base two. The ranges of values ​​in step 64 are distributed accordingly. In this regard, provision can be made, without this being necessary, for the host operator to reserve a more restricted range for the virtual operator by reserving a larger range within which he can then reserve an additional range for a additional virtual operator without having to require the previous virtual operator (s) to go through step 63 again for each additional virtual operator. The host operator, managing the access network, can go through step 64 as many times as he wishes to configure his access network controllers accordingly.

Independently of the routing to which the invention provides a solution, it will be understood that step 65 encompasses other conventional actions such as those of configuring the mobile phones of the virtual operator to select the access network of the host operator, of the radio point of view, mobile registration is done in the core network nodes of the virtual operator. The radio transmission channels, being identical for the host operator and for the virtual operator, the vision of the localization zones by the core network nodes of the virtual operator, depends on the vision of the localization zones by the network nodes. core network of the host operator. The MVNO virtual operator must configure his CN nodes in a way adapted to the topology of the host operator in terms of cells and localization areas.

FIG. 6 shows a cellular telecommunications network illustrated with UMTS terminology, a technology for which the lu-Flex function has been particularly defined. The network of Figure 6 is produced using the method described with reference to Figure 5. With some minor differences, lu-Flex is also defined for GSM and for GPRS in an analogous fashion.

More generally, we retain from the teaching of the invention, to use the permanent and temporary identifiers of a mobile to route the messages to the correct core network in an architecture which consists in interconnecting the host operator and the virtual operator at the access network level, both for UMTS / GSM / GPRS and for other cellular networks.

With reference to FIG. 6, a mobile 59 uses a radio transmission channel on a radio frequency in a geographical area covered by a node B 48. The mobile 59 can move to other geographical areas covered by a node B 49, 53 or 54.

The access network controllers of the access network are represented here by an RNC 47 connected to Node B 48 and 49 and by an RNC 52 connected to Node B 53 and 54. The access network can include other RNCs. and other node B.

The RNC 47 passes the communications that it controls through a battery of core network nodes 55 in the circuit domain and through a battery of core network nodes 56 for the packet domain. The RNC 52 passes the communications that it controls through a battery of core network nodes 57 in the circuit domain and through a battery of core network nodes 58 for the packet domain.

An MSC 30 with its VLR in the core network node farm 55, an SGSN 31 in the core network node farm 56, an MSC 50 with its VLR in the core network node farm 57 and an SGSN 51 in the battery of core network nodes 58, constitute nodes of a first core network managed by a first operator.

An MSC 40 with its VLR in the core network node farm 55, an SGSN 41 in the core network node farm 56, an MSC 60 with its VLR in the core network node farm 57 and an SGSN 61 in the battery of core network nodes 58, constitute nodes of a second core network managed by a second operator.

In each cluster of core network nodes, a resource identifier NRI is associated with the MSCs 30, 50 and the SGSNs 31, 51, with a first value, for example equal to zero. A resource identifier NRI is associated with the MSCs 40, 60 and with the SGSNs 41, 61, with a second value for example equal to one.

When the first operator is a host operator and the second operator is a virtual operator, the telecommunications network of FIG. 6 is typically a network produced by the method described with reference to FIG. 5. The first and the second operator can as well be of identical qualification without having to fundamentally change the method of figure 5. The access network may very well be managed neither by the first nor by the second operator but by a third operator. The access network can also result from pooling the access networks of each operator. It is conceivable that, for example the RNC 47 is managed by the first operator and the RNC 52 is managed by the second operator, the first operator being the host operator for a first part of the access network and the virtual operator for another part of the network. access network, the second operator being virtual operator for the first part of the access network and host operator for the other part of the access network.

Whether the two operators play a symmetrical role or not, the configuration of RNCs does not change per se. The RNC 47 is arranged to direct, by means of its routing table, any message to the node 30 or to the node 40 as a function of a value received from the mobile 59. Different arrangements are possible such as:

An arrangement for routing a message headed by an identity image transmitted by the mobile 59, to the node 30 for an image value f (IMSI) included in a first range and to the node 40 for an image value f (IMSI) included in a second range;

An arrangement for routing a message headed by an identity image transmitted by the mobile 59, to the node 30 for an image value f (TMSI) which contains a resource identifier NRI of the node 30 and to the node 40 for an image value f (TMSI) which contains a resource identifier NRI of node 40;

An arrangement according to which the access network controller comprises means for extracting a permanent mobile identity, within a message sent by the mobile 59 to a core network and means for directing the message to node 30 if the value contained in a permanent identity field corresponds to the operator who manages node 30 and to direct the message to node 40 if the value contained in a permanent identity field, corresponds to the operator who manages node 40;

An arrangement according to which the access network controller comprises a correspondence table between a local identifier of radio connection with the mobile 59 and a permanent identity of this mobile of which a field identifying that of the operators with which the mobile is subscribed, gives a value sent by the mobile when the radio connection is established. In this arrangement, the access network controller comprises means for routing a message to node 30 if the message is received on the radio connection for which the local identifier corresponds to the permanent identity which corresponds to the operator who manages node 30 and to direct the message to node 40 if the message is received on the radio connection for which the local identifier corresponds to the permanent identity which corresponds to the operator who manages node 40.

The RNC 52 is arranged identically to the RNC 47 to direct, by means of its routing table, any message to the node 50 or to the node 60 as a function of a value received from the mobile 59.

Whether or not the RNC 52 is managed by the same operator as the RNC 47, the MSCs 30, 50 and the SGSNs 31, 51, are arranged to allocate to the mobile 59, a TMSI or P-TMSI comprising the NRI which distinguishes the nodes core network of the first operator if the RNC 47, 52 directs a message to them from the mobile 59 without TMSI. Likewise, the MSC 40, 60 and the SGSN 41, 61, are arranged to allocate to the mobile 59, a TMSI comprising the NRI which distinguishes the core network nodes of the second operator if the RNC 47, 52 directs a message to them from the second operator. mobile 59 without TMSI.

In the representation of FIG. 6, the NRI comprises a bit of first value for the first operator, and of second value for the second operator.

A sharing by two operators of their respective access networks allows them to benefit from the infrastructure of a partner operator. Normally, TMSI temporary identifiers have an internal scope and IMSI permanent identifiers have a scope at least partially internal to each operator. The allocation of identifiers is then concerted between the two operators for TMSI, P-TMSI and IMSI.

The purpose of FIG. 7 is to describe a cellular telecommunications method which uses a telecommunications network in accordance with the invention such as that of FIG. 6.

The steps and state transitions 68 to 72 are part of the steps usually executed by a mobile to route the signaling initiated by the mobile to the core network. Routing is essentially necessary as long as there is no connection established between on the one hand, the mobile and the access network (AN abbreviated for Access Network in English) and between ΓΑΝ and the network on the other hand. core (CN for short for Core Network in English). The user traffic, voice, data or additional signaling, then takes place in a known manner on the connection established when it exists. In the GPRS case where there is only one LLC connection established between the mobile and the SGSN but not between the BSC and the SGSN, the steps are repeated for each RLC / MAC frame.

In step 68, the mobile 59 generates a message to be sent in order to carry out the mobile signaling to CN. It is recalled that from the radio point of view, to transmit its message, the mobile first selects the radio access network of the host operator.

The transition 69 corresponds to an absence of temporary identity in the mobile 59 which activates the step 70 following the step 68. The transition 71 corresponds to a presence of temporary identity in the mobile 59 which activates the step. 72 following step 68. The temporary identity is for example the

TMSI in GSM or in UMTS for circuit mode. The temporary identity is for example still the P-TMSI in GPRS or in UMTS for the packet mode.

In step 70, the mobile 59 sends its signaling message with a permanent identity image f (IMSI) to the access network AN.

In the case of GSM in circuit mode, the permanent identity image is the complete international subscriber identity IMSI itself which is at the header of the signaling message to the core network while being accessible to the network. access.

In the case of GPRS in packet mode, the signaling message is sent with a header which contains a random number intended for the access network and a first message body which contains a connection establishment message intended for the core network. CN. The connection establishment message itself comprises a header which contains the full IMSI and a second message body which contains the packet to be transmitted to the core network. The permanent identity image is thus encapsulated in the message intended for the core network.

In the case of UMTS, the signaling message is sent with a header which contains the IDNNS of the lu-Flex function and a message body intended to be transmitted to the core network CN by the access network AN. The IDNNS is an f image (IMSI) which is only a deduced part of the IMSI. There is then an access network level signaling message (IDT for short for Initial Direct Transfer in English) which contains among other things the IDNNS and the message to the CN, thus respecting the independence of the user and network layers without have to read transparent part of signaling. It is recalled that from the radio point of view, to transmit its message, the mobile first selects the radio access network of the host operator.

In step 72, the mobile 59 sends its signaling message with a temporary identity image to the access network AN.

In the case of GSM in circuit mode, the temporary identity image is the temporary complete subscriber identity TMSI itself which is at the header of the signaling message in a manner similar to the permanent identity.

In the case of GPRS in packet mode, the signaling and data frame messages are sent with a header which contains the TLLI consisting of part of the temporary subscriber identity in full packet mode P-TMSI.

In the case of UMTS, the signaling message is sent with a header which contains the IDNNS of the lu-Flex function and a message body intended to be transmitted to the core network CN by the access network AN. In circuit mode, the IDNNS is an image f (TMSI) which is only a deduced part of the TMSI. In packet mode, the IDNNS is an image f (P-TMSI) which is only a deduced part of the P-TMSI.

The function of the steps and state transitions 73 to 79 is to route by the AN, the initial signaling messages in GSM and in UMTS or any message in GPRS in the mobile direction towards CN.

The message sent by the mobile by means of a radio connection allocated by the access network controller is received by the access network controller which is then in charge of directing it to the core network node of the appropriate operator. Reception of the message to be transmitted validates a state transition 105 which activates step 73.

In step 73, the access network, for example the RNC 47, reads the header of the message from the mobile 59 which is accessible to the access network. In UMTS, this is more particularly the IDNNS while in GSM, it is simply the permanent or temporary identity itself, in GPRS, it is more particularly the TLLI. The value captured in this step makes it possible to discriminate the core network managed by the operator whose mobile is authorized to use the services.

The transition 76 corresponds to an IDNNS equal to an IMSI image in UMTS or to the IMSI itself in GSM. The transition 76 activates the step 77 following the step 73. The transition 74 corresponds to an IDNNS equal to an image of TMSI in UMTS or to the TMSI itself in GSM. Transition 74 activates step 75 following step 73.

In step 77, the range of values ​​in which the header value is contained allows the RNC to directly find a core network node address of the operator to which the mobile is subscribed, among addresses associated with this. value range in its routing table.

In step 75, the access network controller directly extracts the NRI from the header value so as to find in its routing table, a core network node address of the operator with which the mobile is registered. , associated with this NRI.

The transition 78, validated by the knowledge of the address, then activates the step 79.

In step 79, the access network controller routes the message in the cluster of core network nodes to which it is connected, to the node whose address has been found in its routing table.

This communication method is particularly flexible for managing mobility.

In the event of mobility with the selection of a new CN node by the initial CN node, the initial node can only select a target node in its own core network because it is not connected to the core network nodes of the other operator (s) .

A case of mobility with selection of a new CN node by the access network is typically that of a change of CN node outside of communication. Suppose for example in FIG. 6 that the mobile 59, initially in the coverage domain of the RNC 47, goes into the coverage domain of the RNC 52. The steps 73 to 79 which would have been initially executed by the RNC 47, are naturally executed by the RNC 52 which then receives the message with IDNNS originating from the old TMSI or old TMSI itself in step 73. A same resource identifier value NRI corresponding to a core network of the same operator in any battery, in step 75, this naturally leads the RNC 52 to direct the message to for example the MSC 50 or respectively the MSC 60 which belongs to to the same core network as the MSC 30 or respectively the MSC 40. The mobile 59 is then listed in the associated VLR of the appropriate core network. The behavior is the same with an SGSN.

FIG. 8 shows alternative steps of the cellular telecommunications method in accordance with the invention. Steps 68 to 72 are executed by the mobile in a manner identical to FIG. 7. However, unlike FIG. 7 where it was the header usually accessible to the access network which was used, it is the header usually accessible to the core network which is operated by the access network.

The message sent by the mobile by means of a radio connection allocated by the access network controller is received by the access network controller which is then in charge of directing it to the core network node of the appropriate operator. Reception of the message to be transmitted validates a state transition 105 which activates a step 94.

In step 94, the access network controller extracts the mobile identity from the message body. The distinctive value capture is done here by reading the identity in the core network header which is found in the message body seen by the access network. The value captured in this step makes it possible to discriminate the core network managed by the operator whose mobile is authorized to use the services. A transition 95 is validated if the identity is the TMSI temporary identity. A transition 96 is validated if the identity is the permanent identity IMSI.

Transition 95 validates step 75 in which the access network controller finds the network node address from the NRI contained in the TMSI.

The transition 96 validates a step 98 in which the access network controller finds the network node address from the operator code MNC and the country code MCC contained in the IMSI.

The process steps described with reference to FIG. 8 are especially useful when the header value accessible to the access network does not make it possible to recognize the operator whose mobile is authorized to use the services. This is particularly when the mobile does not have a temporary identity, sends in the header accessible to the access network, a random number in the case of GPRS, or when roaming, sends a permanent identity image that does not respect not the agreements fixed between operators to operate the same access network. When the mobile sends a temporary identity in the header accessible to the access network, it is not necessary to consider the transition 95 because the step 75 can very well be activated by the transition 74 following the step 73.

FIG. 9 shows other alternative steps of the cellular telecommunications method in accordance with the invention.

A transition 99 commands the mobile to establish a radio connection so as to be able to communicate with the core network with which the mobile is registered or wants to register. The transition 99 is for example activated by switching on the mobile.

The transition 99 triggers a step 100 in which the mobile uses a so-called uplink control radio channel, to send a radio connection request message. In order to identify itself, the mobile places its permanent IMSI identity in the request message that it sends in step 100.

A transition 101 is validated in the access network controller by a unique reception of a request message which contains the IMSI identifying the mobile. The transition 101 triggers a step 102.

In step 102, the access network controller establishes a radio connection. Usually, the access network controller assigns a local identifier to the mobile, for example RNTI in UMTS, so as to list the radio connection defined for each mobile.

In the execution of step 102 by the access network controller in accordance with a first alternative embodiment of the invention, at least a significant part of the IMSI, such as the MCC and the MNC, is associated with the 'radio connection identifier. In performing step 102 by the access network controller in accordance with a second alternative embodiment of the invention, at least a significant portion of the IMSI such as the MCC and the MNC is used to determine the core network of the operator who authorizes a mobile having this MCC and this MNC, to use its communication services. The radio access controller then associates the determined operator with a local identifier of the radio connection, ie the radio connection. The association which results from the first or the second alternative embodiment is stored in the access network controller at least until the mobile makes an access with identification by TMSI. It is observed that this step makes it possible to discriminate the core network by capturing the distinctive value which results from reading the permanent identity in the request message.

A transition 103 is validated in the mobile when the mobile has received the parameters to establish the radio connection. The transition 103 validates a step 104 in the usual way.

In step 104, the mobile uses the radio connection in the usual way to send a message to be transmitted to the core network.

A transition 105, validated in the access network controller by reception of the message to be transmitted, triggers a step 106.

In step 106, in the first alternative embodiment, the access network controller uses the association stored between the radio connection identifier and the significant part of the permanent identity IMSI to determine in the cluster, the address of the core network node which belongs to the operator corresponding to the mobile's IMSI. In the second alternative embodiment, the access network controller uses the association stored between the radio connection identifier and the operator determined in step 102, to determine in the battery, the address of the core network node which belongs to the operator associated with the radio connection.

Knowledge of the address of the node validates the transition 78 which makes it possible to continue the method as explained previously.

To summarize, the essential teaching of the invention is to use, in the access network, a value sent by the mobile to the access network or the core network, to direct a message to the distinguished core network. by this value. All or part of the steps of the communication method can be combined.

When the mobile has a temporary identity allocated by a core network, the temporary identity image that the mobile sends with a message to be transmitted to the core network, to the UMTS access network or to the core network while being accessible to the core network. '' access by GSM or GPRS, allows the access network controller, to direct the message to the appropriate core network by virtue of the core network node identification contained in the temporary identity image usually readable by the access network.

When the mobile does not have a temporary identity, the permanent identity image that the mobile sends with a message to be transmitted to the core network, to the UMTS access network or to the core network while being accessible to the access network in GSM, allows the access network controller to direct the message to the appropriate core network, recognizing for this core network a range of values ​​which covers the permanent identity image usually readable by the network of 'access. The permanent identity extracted from the message to be transmitted by the access network controller, or the association of a radio connection for the message to be transmitted which was established upon receipt of the permanent identity by the access network, allows the access network controller to route the message to the core network without having to recognize a range of values ​​that covers the permanent identity image. This is for example the case in GPRS when the mobile appends a random number to the message to be transmitted or in UMTS when the mobile simply has a roaming agreement with the appropriate core network.

It will be noted that the exemplary embodiments of the invention do not require any particular configuration of the mobile. The mobile usually sends a value which is an image of the temporary identity received from the core network or of the permanent identity which is on its SIM card. The distinctive value of the temporary identity image results from the prior configuration of the core network. The distinctive value of the permanent identity image results from international agreements to assign country and operator identifiers or from agreements between operators to assign ranges of values.

It will be understood, however, that other embodiments remain within the scope of the claims. This is the case, among other things, for a mobile which would be configured so as to send not a totally random number in GPRS but a pseudo-random number which would constitute a distinctive operator value, for example by a range of values ​​in which it would be understood or by an operator identifier concatenation to the generally generated random number.