FR2805430A1 - METHOD FOR MANAGING MOBILITY IN A TELECOMMUNICATIONS NETWORK, AND MOBILITY SERVER FOR IMPLEMENTING THE METHOD - Google Patents

Abstract

The invention concerns a network comprising several sub-networks (11-13) equipped with mobility servers (8) operating in accordance with an application layer signalling protocol. For a specific mobile user, one of the sub-networks (11) constitutes a nominal sub-network (9) containing a first information locating the user. When the mobile user is located with an access point (7) connected to a visited sub-network (12), the first locating information designates the mobility server of said visited sub-network, the sub-network mobility server holds in a register of visitors (10) a second information locating the mobile user, enabling him to transmit data through said access point, and each flow of information addressed to or originating from the mobile user, in the context of communication with a remote correspondent, is addressed to the visited sub-network mobility server.

Description

 <U> MOBILITY <U> MANAGEMENT METHOD IN A <U> TELECOMMUNICATIONS <U> NETWORK, <U> AND <U> M <U> MOBILITY SERVER FOR IMPLEMENTING </ U> The present invention relates to user mobility services, and finds particular application in telecommunication networks comprising several interconnected subnetworks.

The user mobility in question here is the ability of the user to communicate over the network by connecting to the network at different access points. The user can carry his terminal when he moves. This is the typical case of a radio network with mobiles, the access points being constituted by the fixed relays. Without moving equipment, the user can also connect at different points by a registration procedure comprising the transmission, to a mobility management member, of identification data provided by the user. Another possibility is that the user has a removable data medium that he presents to the visited access point equipped with a suitable reader to automate the registration procedure.

We can distinguish two types of mobility, one called here micromobility consisting of allowing the user to change access point within a sub-network, and the other called here macromobility by which the user can connect through access points belonging to separate subnets.

On the other hand, there is the ability of a user to change the access point outside a session, that is to say in the absence of an active data stream (roaming), and the ability to a user equipped with a wireless terminal to change access point during the session (handover).

For network administration purposes, each mobile user generally has a nominal subnet to which is connected an organ referred to herein as a home location register, to which the user's requests originating from sources that are not familiar with the home address are initially addressed. location. When moving into another subnet, another organ connected to this visited subnet, called here visitor manager, can cooperate with the nominal location register for the implementation of mobility services. The development of applications on networks operating according to the IP protocol (J. Postel, Internet Protocol, Request For Comments (RFC) 791, IETF, September 1981), especially those of voice and data transport, contemporary to the development communication systems with mobiles, has naturally led to consider the issue of mobility in IP networks, be they wide area networks (WANs), such as the Internet, or networks Local (LAN, Local Area Network).

The Internet Engineering Task Force (IETF) has standardized a network-layer protocol that supports mobility in an IP network (C. Perkins, IP Mobility Support, RFC 2002, IETF, October 1996). After the mobile user has left his nominal subnet, his equipment is directed to an organ called foreign agent, connected to the visited subnet, to send registration information to an organ called the nominal agent. (home agent), connected to the nominal subnet. The identity of the visited subnet is also provided to the home agent. One of the main functions of the nominal agent is then to intercept the data intended for the mobile user to retransmit, by an encapsulation mechanism, to a temporary address of the mobile user (called address c / o or care-of address) provided by the foreign agent. Once at this temporary address, the data can be delivered to the recipient.

This mobility scheme thus imposes a triangular path to the IP packets intended for the mobile user. On the other hand, data coming from the mobile user is delivered to their destination via the usual IP routing mechanisms.

In addition, it assumes the use of a fixed address for the mobile user. Covering it thus poses difficulties in networks where the address management is dynamic, which is the case for example when a DHCP server is used in the nominal subnetwork (see for example EP-A-0 938 217). The DHCP protocol is specified in RFC 2131 published by the IETF (R. Droms, Dynamic Host Configuration Protocol, March 1997).

A path optimization method, designed as an extension of the RFC 2002 protocol has been proposed (C. Perkins, D. Johnson, Route Optimization in Mobile IP, Internet Draft, IETF, February 25, 1999) to overcome the disadvantages Triangular routing mode. This method provides in particular the sending of update messages to the possible correspondents of a mobile user to keep in memory the temporary addresses c / o of the latter. These can then send directly to the temporary address of the mobile user, thus avoiding the operations performed on the data by the nominal agent.

Some authors have criticized the complexity of implementing such a method, and have suggested supporting IP mobility using an application layer signaling protocol such as the SIP protocol (M. Handley et al., SIP: Session Initiation Protocol, RFC 2543, IETF, March 1999).

 SIP describes client and server entities, as well as procedures for communicating with them (see also HG Schulzrinne and JD Rosenberg, The Session Initiation Protocol: Providing Advanced Telephony Services Across the Internet, Bell Labs Technical Journal, October-December 1998, p.144. -159). There are two types of SIP servers: proxy servers and redirection servers. On receipt of a request, a proxy server determines the next node of the path to the recipient and then transfers the request to this node, while a redirection server merely informs the client of the next node to which it must address its request. .

SIP addresses are similar to e-mail addresses, that is to say the form user (ahost, where the field user for example designates a user name or a phone number, and the field host a name The SIP protocol includes methods called INVITE, BYE, REGISTER, OPTIONS, ACK and CANCEL.The responses to the messages sent in the context of these methods are defined by code classes. INVITE is used to initiate a call session between two SIP users.

The SIP protocol provides personal mobility capabilities, and allows a user to obtain the same services regardless of its location or the terminal used, including the REGISTER methods. It has been proposed (E. Wedlund, H. Schulzrinne, Mobility Support using SIP, Proc., Of the ACM International Workshop on Wireless Mobile Multimedia, Seattle, August 20, 1999, pages 76-82) of mobility SIP protocol in which a wireless device connected to the network via an air interface can, after appropriate signaling exchanges, execute a handover, that is to say change network address (1P) by a DHCP mechanism while a communication is in progress A SIP server connected to the user's nominal subnet manages the home location register for the user The SIP servers used in mobility management can be proxy servers or Redirection servers Any mobility is registered with the nominal SIP server, which can then process requests from other SIP clients to the mobile user. wireless determines the need to execute a handover (by detecting tags issued by the wireless access points), it gets a dynamic address by a DHCP transaction, then it sends an INVITE message to its correspondent indicating the reference of the current call and its new network address so that the correspondent updates the destination IP address of the packets it sends. At the same time, the wireless equipment communicates its new address to its nominal SIP server by the REGISTER method. One option provides for decentralization of SIP servers used in mobility management: a proxy SIP server connected to the visited subnet manages micro mobility, so REGISTER methods to the nominal SIP server, redirection, may only concern macromobility (change of micromobility server).

The main disadvantage of the process is the delay that it can introduce in the execution of the handover. The INVITE method to the correspondent of the mobile user requires a non-negligible routing time if this correspondent is remote from the visited subnet. This delay can result in a noticeable break in the current call.

It should be noted that a similar disadvantage exists within the IP mobility protocol of RFC 2002. When the mobile device is connected to a remote visited subnet of the nominal subnet, it is necessary to route a message. location update to the nominal agent running the handover, which takes time. If the protocol is implemented with the aforementioned path optimization method, there is added the time necessary for the routing of a c / o address update packet to the user's correspondent. mobile.

An object of the present invention is to propose a user mobility management that makes it possible to limit the delays due to the signaling in the execution of handovers.

According to the invention, there is provided a mobility management method in a telecommunications network comprising several subnetworks respectively equipped with mobility servers operating according to an application layer signaling protocol, in which one of the subnetworks constitutes a sub-network. nominal network for a mobile user and is equipped with a nominal register containing a first location information of said user. When the mobile user is located near an access point connected to a visited subnetwork, the first location information designates a mobility server of said visited subnetwork, said mobility server of the visited subnetwork is a visitor register a second location information of the mobile user, for transmitting information thereto through said access point, and each information flow to or from the mobile user, in the context of a communication with a distant correspondent is sent to said mobility server of the visited subnet.

Another aspect of the present invention relates to a mobility server adapted to the implementation of the method above. This mobility server is to connect to a subnet belonging to a telecommunications network, and operates according to an application layer signaling protocol. It comprises means for managing a visitor register containing, for at least one mobile user located at an access point connected to said sub-network, location information of the mobile user, making it possible to transmit to him / her information through said access point, means for receiving information streams from the mobile user as part of a communication with a remote party and retransmitting them to the remote party, and means for receiving information flow from the remote party as part of said communication and retransmit to the mobile user.

The mobility server of the visited subnet acts as a proxy agent for both the signaling flows, in particular concerning the mobility of the user, and for the traffic flows (voice, images, data, etc.). ). Thus, during a handover, only the local streams between the mobility server and the user are affected. The mobile user sees the proxy server as his correspondent, so he is the one he is addressing to signal his change of access point. On the other hand, nothing is changed for the exchanges between the server and the remote correspondent. It is not necessary that signaling information be sent remotely during the handover, to the nominal subnet or to the correspondent. As in general the routing of messages within a sub-network is very fast, it results in an optimization of the execution time of the handover.

This mobility management method does not imply in itself any particular adaptation at the level of the terminals communicating with the mobile equipment.

The method can be applied with various application layer signaling protocols. Currently, SIP is a preferred protocol because it constitutes a good compromise between functional richness and complexity of implementation, while having a good capacity for extension. Other examples, well known to those skilled in the art, are the protocols H.225.0 (Call signaling protocols and media stream packetization for packet-based multimedia communication systems) and H.245 (Control protocol for multimedia communication) specified by the International Telecommunication Union (ITU-T) in the framework of Recommendation H.323 (Packet-based Multimedia Communications Systems, February 1998), and the MGCP protocol (M. Arango et al., Media Gateway Controller Protocol (MGCP)) , RFC 2705, IETF, October 1999) or its Megaco variant (F. Cuervo et al., Megaco Protocol, Internet Draft, IETF, February 8, 2000).

The first location information stored in the nominal register of the user will typically be the IP address of the mobility server of the visited subnet, updated only when the user's move causes it to change the visited subnet ( and possibly when the visited subnet has multiple mobility servers and the mobile user changes mobility server without changing the visited subnet). The mobility server of the nominal subnet can then be a redirection server to that of the visited subnet.

The second location information stored in the visitor register depends on the architecture of the subnet visited. It can consist of a physical MAC address (Medium Access Control), to reach the mobile user through the access point. It can also be an IP address of the mobile user in the visited subnet, for example a dynamic address obtained from a DHCP server of the visited subnet or a composite IP address (subnet + equipment address ) version 6 of the IP protocol (IPv6). If the subnetwork visited is a LAN itself subdivided into subnetworks interconnected through routers, it is also possible to consider using a network layer mobility protocol within the visited subnetwork. such as that of RFC 2002, the mobility server acting as a nominal agent for retransmitting packets intended for the mobile user, by an encapsulation mechanism, to a c / o address provided by a foreign agent connected to the sub-subnet comprising the current access point of the mobile equipment, this c / o address of the mobile user then being recorded as second location information.

The above method makes it possible to minimize the execution time of a handover not involving a change of the proxy server locally managing the mobility of the user. This is largely the most common case since we can often have only one mobility server per site (subnet) for visitors and handovers normally occur within the same site. However, there may be cases in which the same site has multiple mobility servers, either because there are multiple subnets on the site, or because multiple servers are provided for load-sharing reasons. . In such a case, when mobile equipment detects during the session a new access point associated with a mobility server different from that which played the role of proxy before the handover, it remains necessary to exchange messages between the new mobility server and remote correspondent, so that it takes into account a change of IP address where to send its packages. In the latter case, additional means must be employed if it is desired to minimize the execution time of the handover.

One of these means consists in including, in a beacon signal transmitted on an air interface by at least one access point to which the mobile user is likely to be located, and / or in at least one transmitted message. on the air interface by said access point to the mobile user as part of an air interface-specific protocol registration procedure, a network address of a mobility server associated with said point access.

Thus, the first action of the mobile equipment after registering with the access point may be to send, to the mobility server from which it obtained the network address (1P), the message allowing it to update its location. The progress of the handover will not be delayed by additional exchanges to obtain this IP address.

Other features and advantages of the present invention will appear in the following description of nonlimiting exemplary embodiments, with reference to the accompanying drawings, in which - Figure 1 is a diagram of a telecommunications network to which the invention is applied; FIGS. 2 to 4 are diagrams illustrating message exchanges involved in mobility management according to one embodiment of the invention; FIG. 5 is a diagram of a sub-network to which an embodiment of the invention may apply; FIG. 6 is a diagram illustrating message exchanges that can intervene in the management of mobility within a subnetwork according to FIG. 5; and FIGS. 7 and 8 are diagrams showing examples of information blocks that can be transmitted by wireless access points in one embodiment of the invention.

Figure 1 shows schematically an intranet network distributed over three sites 1, 2, 3. On each site, a subnet 11, 12, 13 is installed to provide access to a number of users. These subnets 11-13 are for example of the Ethernet type (ISO 8802-3 standard, Local Area Networks, Part 3 - Carrier Sense Multiple Access with Collision Detection - Access Method and Physical Layer Specifications), and operate according to the network protocol IP. In the example shown, each subnet has an output router 4 connected to a backbone network 5 such as the Internet.

The network may in particular be designed to support real-time signal communications, for example telephony, between users connected to it and / or with correspondents accessible by the backbone network. In this case, the terminals may use known manner the UDP transport layer protocol (J. Postel, User Datagram Protocol, RFC 768, IETF, August 1980) and the RTP real-time protocol (H. Schulzrinne et al., RTP: A Transport Protocol for Real-Time Applications , RFC 1889, IETF, January 1996) for the transmission of coded speech or video. For call signaling, a preferred embodiment of the invention uses the SIP protocol, which also works with the UDP transport protocol. However, other signaling protocols (H.323, MGCP, ...) could be used, in addition to or instead of SIP.

At least some of the users of the network are mobile users equipped with wireless terminals 6. To enable them to access the network, it includes radio access points 7 connected to the subnets 11-13. The radio access points 7 are, for example, DECT base stations (ETSI standard, European Telecommunications Standards Institute) or according to the IEEE 802.11 standard, etc., depending on the type of wireless terminals used. Each of these radio access points 7 broadcasts a beacon signal on the air interface. Based on the beacon signals it captures, an active terminal selects the access point that provides it with the best radio link. This access point is used for communications involving the terminal.

A wireless terminal thus has the possibility of connecting to the network via several access points 7 belonging to different sub-networks (macromobility) or to the same sub-network (micromobility). When the terminal has a communication in progress and moves in a site, it is possible for it to change access point while maintaining the continuity of communication (handover). When real-time signal communication, for example telephony, is in progress, it is desirable that these handovers are executed in a minimum time to avoid noticeable breaks.

For the management of the mobility services, each subnet (at least those having wireless access points) has a mobility server 8 using the SIP protocol in the preferred embodiment of the invention.

Each mobile user has a nominal subnet. In the illustration of FIG. 1, the sub-network 11 is nominal for the holder of the wireless terminal 6. The SIP server 8 of this sub-network is associated with a location register 9 called said nominal register, which contains the IP address of the SIP server 8 of the subnet 12 visited by the mobile user. The SIP server 8 of the visited subnet 12 is associated with a location register 10 called visitor register, which contains an address for reaching the terminal 6 within the visited subnet 12, for example an IP address associated with the terminal.

Each SIP server 8 may be associated with a nominal register 9 for the users attached to its sub-network and a visitor register for the users attached to other sub-networks, only the registers relating to the terminal 6 being represented on In practice, the registers 9, 10 may be part of the same unit as the SIP server 8 with which they are associated, or they may be separate entities connected separately to the sub-network. On the other hand, it is possible to provide several SIP servers per subnet, for example in the case where the nominal SIP server and visited SIP server functions are provided by separate machines, or in the case where several SIP servers are provided to distribute the signaling load in the subnet.

FIG. 2 shows how the registration and location update of a mobile user visiting a subnet 12 different from his nominal subnet 11 can take place and the initialization of a communication with this subnet. user on the initiative of a remote correspondent.

In the illustration of FIG. 1, the remote correspondent uses a fixed terminal 15 connected to a subnet 13 different from the nominal subnet 11 and the visited subnet 12. It will however be noted that a similar procedure applies also - if this correspondent is another mobile user; - if it is connected to the subnet 11 or 12 or to the backbone network 5; - if the connection with the other party goes through another network, for example a switched telephone network, for which the intranet network is equipped with a gateway (in the latter case, it is this gateway which constitutes the SIP client relative to the correspondent remote shown in Figures 2 to 4).

With reference to FIG. 2, the wireless terminal 6 belonging to the nominal subnet 11 first catches on the air interface the beacon signal transmitted by an access point 7 belonging to the visited subnet 12. In response at this detection, it triggers a registration procedure of the MAC layer protocol used on the air interface, by addressing to the access point a registration message (Register). In response to this message, the access point 7 returns an acknowledgment (Register Ack) validating the registration of the wireless terminal with the access point.

In the example of FIG. 2, the location information of the wireless terminal registered in the visitor register 10 is a dynamic IP address obtained by the terminal after it has been registered with the access point 7. The subnet 12 is equipped with a DHCP server 16 that manages dynamic addresses locally. A typical DHCP transaction (DHCP_Discover, DHCP_Offer, DHCP_Request, DHCP Ack messages) is executed between the roaming terminal 6 and the DHCP server 16, after which the terminal 6 has an IP address.

In another embodiment using the IPv6 protocol, the DHCP transaction is useless since the IPv6 address inherently allows a subnet / equipment hierarchy.

The next step consists in the SIP client sending the terminal 6 a SIP REGISTER message to the SIP server 8 of the visited subnet 12.

The IP address of the SIP server 8 has been provided by the access point 7 in the beacon signal broadcast on the air interface, which allows the wireless terminal to have this IP address without having to carry out transactions. for this purpose via the sub-network 12. If the beacon signal broadcast on the air interface does not have an available field large enough to contain the IP address of the SIP server 8, this can be completed, or provided in full, in the Register Ack message validating the registration of the terminal on the air interface.

The SIP REGISTER message sent by the terminal 6 enables the visited SIP server 8 to determine the IP address of the terminal, obtained in the IP header of the packet containing the SIP REGISTER message, and associate it with the SIP address ( of the form user host) contained in the SIP REGISTER message. This association is registered in register 10.

After receiving this SIP REGISTER message, the visited SIP server 8, if it does not already manage the mobile user, transmits another SIP REGISTER message to the SIP server 8 of the nominal subnet 11. This updates the input of the nominal register 9 concerning the mobile user identified in the SIP REGISTER message, by storing the IP address of the visited SIP server which sent this SIP REGISTER message. It then returns a validation message (code 200 OK) of the SIP protocol. The visited SIP server in turn sends back to the terminal 6, through the access point 7, the validation message 200 OK.

At this point, the wireless terminal 6 has registered with the access point 7, and updated its location vis-à-vis the SIP servers 8.

Upon initialization of a call from the remote party 15, the latter sends to the SIP server 8 of the nominal subnet 11 of the mobile user an SIP INVITE message requesting the mobile user. The nominal SIP server is a redirection server in the example considered, it sends back to the remote correspondent a SIP message (code 302 Move) indicating that it must direct its INVITE method to the SIP server of the subnet visited 12 which he provides the IP address. The correspondent transmits the SIP INVITE message again to this visited SIP server.

The visited SIP server then initiates another session with the mobile user, by sending him an SIP INVITE message, which the wireless terminal acknowledges by a SIP message 200 OK if it is available for the establishment of the communication. The session between the wireless terminal and the visited SIP server is opened when the SIP server has received the message 200 OK. It then transmits to the remote correspondent another SIP message 200 OK to validate the opening of the session with it. From this moment, the remote correspondent communicates with the visited SIP server as if the latter were the wireless terminal, and the wireless terminal communicates with the visited SIP server as if it were the terminal of the remote correspondent. In each open session with the visited SIP server, data is exchanged, typically according to the RTP / UDP / IP protocol stack when it represents coded speech or video, and additional signaling may be provided. according to the SIP / UDP / IP protocol stack. The visited SIP server 8 echoes to the wireless terminal 6 the RTP / UDP stream received from the remote correspondent. Similarly, the RTP / UDP data stream received in the session established with the wireless terminal is echoed to the remote party. The visited SIP server thus also acts as a proxy agent.

FIG. 3 illustrates the registration of the wireless terminal 6 with another access point 7 of the same visited subnet 12, and the location update made with the SIP server of this sub-network. The wireless terminal is registered with the new radio access point by a MAC layer transaction identical to that described with reference to FIG. 2. If necessary, a DHCP transaction (not shown) can be performed to that the terminal acquires a new IP address. The wireless terminal then sends a SIP REGISTER message to the visited SIP server, whose address it has obtained in the tag of the new access point and / or in the Register Ack message. Since the wireless terminal does not change a visited SIP server, it does not need to send a SIP REGISTER message to the SIP server of the nominal subnet 11. After modifying the visitor register entry 10 Regarding the mobile user, the visited SIP server validates its location update, returning a SIP message 200 OK to it.

To initialize a call to a remote party (lower part of Figure 3), the wireless terminal sends a SIP INVITE message to its visited SIP server and the latter, which still plays the role of proxy server, transmits another SIP message INVITES to the called SIP server of the called party. This nominal SIP server, redirection, returns a SIP message 302 Move indicating the IP address to which the correspondent can be attached. This IP address may be the address of the remote party in its nominal subnet 13 or, as in the case of FIG. 2, the address of a visited SIP server if the remote party uses a wireless terminal outside its nominal subnet.

The establishment of the call then ends as in the case of FIG. 2. SIP messages 200 OK are sent from the remote correspondent to the SIP server of the visited subnet 12, and from this SIP server to the wireless terminal 6. after which the visited SIP server manages the two UDP streams as explained above to relay the communication.

FIG. 4 illustrates the handover procedure executed when the wireless terminal 6 has a communication in progress and changes the radio access point 7 while moving inside the visited site 2. It is assumed here that the same SIP server visited 8 is associated with the two access points 7, that is to say that the IP address of this server 8 is indicated in the beacon signal of these access points and / or in the Register Ack message. The upper part of FIG. 4 shows the data flows relayed by the visited SIP server as explained above.

After the wireless terminal 6 has detected the tag of the new access point and decided to change the access point, the registration procedure with the new access point and location update with the SIP server when the wireless terminal has received the acknowledgment of its location update (SIP message 200 OK), it sends an INVITE message to the visited SIP server, including a reference of the current call. After validation of the INVITE method (SIP message 200 OK returned to the wireless terminal by the visited SIP server), the data exchange resumes without any signaling of the visited SIP server to the nominal SIP server or the server. remote correspondent.

The execution of the handover is particularly fast since the exchanged signaling messages remain within the visited subnet 12 without having to be routed through the extended interconnection network 5.

On the other hand, the fact for the wireless terminal 6 to have directly the IP address of the visited SIP server, in the radio beacon or in the radio MAC layer signaling, saves time by avoiding a search process. this address in the IP network. FIG. 5 schematically shows another possible architecture of the subnetwork installed on the visited site 2, comprising several sub-subnetworks 120, 121 interconnected by means of one or more routers. In the example drawn, the router 4 connected to the backbone network 5 is also used to interconnect the sub-subnets 120, 121.

The SIP server 129 of the visited subnet, which locally manages the mobility of the holder of the wireless terminal 6 in the manner previously explained, is connected to one of the sub-subnets 120. It also acts as an agent. nominal in an adaptation of the Mobile IP protocol that is the subject of RFC 2002. This adapted Mobile IP protocol is implemented inside the visited subnet in order to manage in the network layer the mobility of the users between the subscribers. subnets that compose it (but not at the level of the subnetwork grouping, where an application layer protocol is used, namely SIP in the example considered). Since the agent 129 does not belong to the nominal subnet of the mobile user, it will be called a pseudo-nominal agent, and it will be said that a pseudo-mobile IP protocol is implemented.

In each sub-subnet 121 different from that provided with the pseudo-nominal agent 129 and equipped with one or more radio access points 7, a routing module is provided as a foreign agent 130 in the sense of RFC 2002.

After the wireless terminal 6 has detected the beacon of a new access point 7 belonging to such a sub-subnet 121 and decided to change access point, the registration procedure with the new point d access is performed as in the case of Figure 3 or 4 (upper part of Figure 6). The terminal 6 then sends a registration request for the Mobile IP protocol (REGISTRATION REQUEST in the terminology of RFC 2002) to the foreign agent 130, which relays it to the pseudo-nominal agent 129.

The terminal 6 places in the REGISTRATION REQUEST message, in addition to its own IP address, the IP address of the pseudo-nominal agent and the c / o address that was previously provided to it. This c / o address may have been obtained by the agent discovery mechanisms specified in RFC 2002. But in a preferred embodiment, the two IP addresses (that of the pseudo-nominal agent and the c / o address) are provided in the beacon signal broadcast by the access point 7, and / or in the Register Ack message. The clo address is then an IP address of the foreign agent 130, and can be used as the destination address by the terminal for sending the REGISTRATION REQUEST message.

After acknowledging the registration request (REGISTRATION REPLY messages in FIG. 6), the mobile terminal sends its SIP REGISTER message to update its location vis-à-vis the visited SIP server 129. The SIP REGISTER message can be sent directly to the IP address of the visited SIP server 129. As this message is seen as data by the Mobile IP protocol, it can also be relayed by the foreign agent 130 to the visited SIP server I pseudo-nominal agent 129. acknowledge the SIP REGISTER method (after an exchange with the SIP server of the nominal subnet 11 if there has been a change of subnet visited), the visited SIP server 129 uses two levels of IP header to transmit the SIP message 200 OK, the internal level having the IP address of the wireless terminal 6 and the external level having the c / o address that has been provided by the foreign agent 130. The foreign agent 130 removes the external level for to retransmit the wise 200 OK to the terminal.

The same encapsulation mechanism will be employed for all packets sent to the terminal 6 by the visited SIP server 129 (SIP protocol or data from a remote correspondent). Similarly, the packets sent by the terminal 6 are relayed by the foreign agent 130 to their destination, namely the visited SIP server 129 since it is seen by the terminal as if it were its correspondent.

The location update performed according to FIG. 6 occurs in the same way in the case of roaming (FIGS. 2 and 3) and in the case of handover (FIG. 4). The location information associated with the terminal 6 in the visitor register 10 then consists of the c / o address that allows the SIP server to route the data to their destination.

Figure 7 illustrates a possible structure of the beacon signal broadcast by a wireless access point 7 connected to a sub-network. This signal is formed by modulating a block of digital data B, and is transmitted in a radio signal frame, or possibly in a time slot of a frame if a time division multiple access is employed on the air interface.

The digital data of the block B comprise a predetermined synchronization pattern 20, the detection of which allows the wireless terminals located within reach of the access point to synchronize in time and frequency to demodulate the signal of the frame, and certain conventional fields. 21-23 containing system information required by the physical layer and link layer protocols. This system information typically comprises - an identification of the network to which the access point belongs (field 21); system synchronization information (field 22), identifying the position of the current frame in the temporal organization of the radio signal transmitted on the carrier of the beacon (frame index in a current superframe, superframe number, etc.) ; other system information (field 23), such as a minimum received field level from which a wireless terminal is allowed to register with the access point, or indications of the frequencies carrying the tag a few nearby access points so that the terminals can monitor these carrier frequencies to determine the access point providing the best radio link.

Advantageously, the block B further comprises a field 24 in which the access point 7 places data describing its environment.

These data can represent the geographical environment where the access point is installed: location of the visited site 2, precise position of the access point (for example building, corridor, office, ...). As the network is based on the IP protocol in which there is no notion of physical connection, a communication with a wireless terminal does not suppose any knowledge, even implicit, of the geographical location of the terminal. The remote correspondents and the nominal location register 9 locate the terminal only logically, via one or more IP addresses. However, a certain number of applications may need such geographical location information, for example to render services differentiated according to the location of the terminal or to provide an indication of this location to the mobile user's correspondents. By inserting this geographical location information (explicitly or in coded form) into the beacon signal, it is made available to the wireless terminals which will then be able to communicate them to the mobility server of the nominal subnet or to correspondents in the frame. applications of this kind.

The environment data placed in the field 24 of the block B may also comprise one or more IP addresses of mobility management members, namely - the IP address of the SIP server (or H.323, MGCP, Megaco,. ..) visited in an embodiment according to Figures 1 to 4; in an embodiment according to FIGS. 5 and 6, the IP address of the SIP server (or H.323, MGCP, Megaco, etc.) also playing the role of pseudo-nominal agent 129 and the IP address of the foreign agent 130; in an embodiment where the mobility between the subnetworks is managed by a network layer protocol such as that of the RFC 2002, the IP address of the foreign agent, the environment data of the field 24 can still include a number UDP or TCP port used by the mobility server 8, 129 associated with the access point, if it is not a default port number. This allows the terminal to send the requests of the application layer signaling protocol directly to the correct port number. It is also possible to include in the environment data a MAC (Ethernet) address of the mobility manager 8, 129 connected to the same subnet or sub-subnet as the access point.

If the field 24 of the block B does not provide enough space to insert in the beacon signal all the expected environmental data, it is possible to insert some of them (if not all) in the Register Ack message. FIG. 2, 3, 4 or 6. In the example illustrated in FIG. 8, this Register Ack message is formed from a block of digital data B 'comprising a field 25 identifying the previous registration request and / or the wireless terminal that issued it; a field 26 containing codes indicating the response to the registration request (authorized, refused or conditioned access, parameters, etc.); a possible field 27 for containing complementary system information in addition to those provided in fields 21-23 of block B; a field 28 for repeating the environment data or supplying those that have not been in the field 24 of the block B.

Claims (13)

A mobility management method in a telecommunications network comprising a plurality of subnetworks (11-13) respectively equipped with mobility servers (8) operating according to an application layer signaling protocol, in which one of the subnetworks (11) ) constitutes a nominal subnet for a mobile user and is provided with a nominal register (9) containing a first location information of said user, wherein when the mobile user is located near an access point (7). connected to a visited subnet (12), the first location information designates a mobility server of said visited subnet, said visited subnet mobility server holds in a visitor register (10) a second information of locating the mobile user to transmit information thereto through said access point, and each information flow to or from the user Mobile initiator, in the context of a communication with a remote correspondent, is addressed to said mobility server of the visited subnet. 2. Method according to claim 1, wherein the network operates according to the IP protocol. The method of claim 2, wherein said application layer signaling protocol is SIP, H.323, MGCP or Megaco. The method of claim 2 or 3, wherein the first location information is an IP address of the mobility server (8) of the visited subnet (12). The method of claim 2 or 3, wherein the second location information is an IP address of the mobile user in the visited subnet (12). The method according to any one of claims 1 to 4, wherein at least one subnetwork comprises a plurality of interconnected sub-subnetworks (120, 121) respectively equipped with mobility agents (129, 130) used in a subsystem. network layer mobility management protocol, wherein said subnetwork has a mobility server (129) connected to one of the subnetworks (120) and constituting a mobility mobility agent for visiting mobile users in said subnetwork, and wherein when a mobile user is located at an access point (7) connected to another one of the sub-subnets (121), said mobility server constituting a mobility agent nominal stores in the visitor register (10) a second location information of the mobile user, consisting of a network address, that said mobility server constituting a nominal mobility agent places in packet headers in which the information flows intended for the mobile user are retransmitted by an encapsulation mechanism. 7. Method according to any one of claims 1 to 6, wherein at least one access point (7) at which the mobile user is likely to be located is associated with one of the mobility servers (8). ), and includes a network address of the mobility server (8) to which said access point is associated in a beacon signal transmitted on an air interface by said access point (7), and / or in at least one a message transmitted on the air interface by said access point to the mobile user as part of a procedure for registering a protocol specific to the air interface. The method according to claim 7, wherein, after the registration procedure with said access point (7), a location update message of the application layer signaling protocol is issued from the mobile user to said network address. The method according to claim 7 or 8, wherein at least one subnetwork comprises a plurality of interconnected sub-subnetworks (120, 121) respectively equipped with mobility agents (129, 130) used in a management protocol. network layer mobility, wherein said subnet has a mobility server (129) connected to one of the sub-subnetworks (120) and constituting a mobility mobility agent for mobile users visiting said sub-subnet (120); a network, wherein when a mobile user is located at an access point (7) connected to another one of the sub-subnets (121), said mobility server constituting a mobility agent stores in the register of visitors (10) a second location information of the mobile user, consisting of a network address, that said mobility server constituting a nominal mobility agent places in packet headers in which the information flows for the mobile user are retransmitted by an encapsulation mechanism, and in which is further included in the beacon signal transmitted by said access point, and / or in at least one message transmitted on the air interface by said access point to the mobile user as part of the procedure for registering the protocol specific to the air interface, a network address of a foreign mobility agent (130) connected to the same sub-sub- network (121) as said access point and corresponding to the network address forming the second location information of the mobile user. A mobility server to be connected to a subnetwork (12) belonging to a telecommunications network, operating according to an application layer signaling protocol and comprising means for managing a visitor register (10) containing, for at least least one mobile user located at an access point (7) connected to said sub-network, location information of the mobile user, for transmitting information through said access point, means for receiving information flows from the mobile user in the context of a communication with a remote party and retransmitting them to the remote party, and means for receiving information streams from the remote party within the context of said communication and retransmit them to the mobile user. The mobility server of claim 10, wherein the network operates according to the IP protocol and said application layer signaling protocol is SIP, H.323, MGCP or Megaco protocol. The mobility server of claim 10 or 11, wherein the second location information is an IP address of the mobile user in the visited subnet (12). Mobility server according to claim 10 or 11, for a sub-network comprising a plurality of interconnected sub-subnetworks (120, 121) respectively equipped with mobility agents (129, 130) used in a mobility management protocol. network layer, said mobility server being connected to one of the sub-subnets (120) and constituting a mobility mobility agent for visiting mobile users in said subnetwork, wherein the register management means of visitors (10) are arranged to store, for at least one mobile user located at an access point (7) connected to another of the sub-subnetworks (121), a location information consisting of an address network, and the means for retransmitting information flows to the mobile user are arranged to encapsulate said flows in packets having a header including said network address stored in the visit register urs for the mobile user.