EP3991358A1

EP3991358A1 - Method for managing at least one communication of a terminal device in a communication network, methods for processing a communication set up with a terminal device in a communication network, corresponding appliances, terminal device, proxy device and computer programs

Info

Publication number: EP3991358A1
Application number: EP20747035.2A
Authority: EP
Inventors: Mohamed Boucadair; Christian Jacquenet
Original assignee: Orange SA
Current assignee: Orange SA
Priority date: 2019-06-28
Filing date: 2020-06-24
Publication date: 2022-05-04
Also published as: WO2020260826A1; FR3096530A1; CN114303346A; US20220239556A1

Abstract

The invention relates to a method for managing at least one communication in accordance with a transport protocol of a terminal device in a communication network, said terminal device being able to access said communication network via at least one IP resource, each IP resource comprising an IP address and a port number, comprising: - detecting (22) the presence, on at least one path for reaching said terminal device via said communication network on an IP address of said terminal device, called second IP address, of at least one state function configured so as to keep, in a table, a state associated with a communication on said path for a given service life, comprising transmitting, in the communication network, a first message from at least one first IP resource of said terminal device, said first IP resource comprising a first IP address and a first port number, to a second IP resource of said terminal device, comprising said second IP address and a second port number, and deciding on the presence of at least one state function on said at least one path on the basis of data received by the second IP resource in response to the transmission of the first message; and - triggering (23) an action for managing a communication of the terminal device on said at least one path for reaching said terminal device via said communication network on said second IP address on the basis of said detection.

Description

DESCRIPTION

TITLE: Method for managing at least one communication of terminal equipment in a communication network, methods for processing a communication established with terminal equipment in a communication network, devices, terminal equipment, proxy equipment and programs of 'computer matching

Field of the invention

The field of the invention is that of a communication network and more particularly of the management of communications of terminal equipment in such a network.

The invention finds an application in particular in communication networks implementing value-added IP services.

Prior art and its drawbacks

The communication protocol, called QUIC, was designed to reduce the latency times observed during communications based on the transport protocol TCP (for "Transport Control Protocol") and in particular the time for establishing a communication. between a first terminal equipment, called the client or the sending terminal or even simply the terminal, and a second terminal equipment, called the server or the receiving terminal, via the communication network. It is for this reason that the QUIC protocol is based on the UDP transport protocol (for “User Datagram Protocol”). Indeed, the UDP transport protocol, unlike the TCP protocol, does not use a signaling mechanism of the three-step handshake type (for "3-way handshake", in English), so that the terminals can in particular adapt the frequency of data transmission as a function of network bandwidth conditions for example. The QUIC protocol makes it possible in particular to transmit, under certain conditions, useful data as soon as the first packet of a communication is sent, without the QUIC client having to wait for the response from its correspondent. Latency and signaling times between clients are thus reduced.

In order to support any change of IP address without having to close a QUIC communication in progress, the QUIC transport protocol does not rely on transport addresses, and more particularly on the quadruplet {source IP address, source port number, destination IP address, destination port number} but on a connection identifier called CID (for “Connection Identifier”) The QUIC specification defines two types of CID: Destination CID and Source CID.

Currently, the QUIC protocol supports a connection migration mechanism (for "Connection migration") which allows a QUIC communication to be maintained in the event of a modification of one of the addresses (or port numbers) of the participants (including the changes related to the use of addresses allocated by intermediate address translation or NAT (for "Network Address Translation") equipment. Receipt of a message as part of an ongoing communication with a new source address is an indication of connection migration. Thus, a connection migration consists of going from one quadruplet {source address, source port number, destination address, destination port number} to another.

However, the presence of stateful functions such as NAT, firewall or proxy functions on a network path used by QUIC communication constitutes a source of problems likely to compromise quality. of communication. Indeed, in a known manner, such stateful functions are generally hosted by intermediate equipment, and maintain a table comprising entries associating in particular an internal source address and an external source address for an outgoing data packet. They use this table to filter incoming packets by rejecting those that do not match a valid entry in the table. An entry is generally maintained for a period, known as the predetermined lifetime, beyond which, in the absence of a new outgoing data packet or of an adequate control message to extend the lifetime, it is invalidated and deleted. Table. This lifespan can be short, of the order of a second.

To overcome this drawback, the communication protocols, QUIC included, use a mechanism to "keep alive" a communication which makes it possible, in particular, to verify that the link on which the communication is established is still active or to prevent this link from being broken. In the case where the communication is established on a link which includes an intermediate equipment of NAT type for example, such a mechanism consists for a terminal equipment, in sending at regular frequency a signaling message to the remote terminal equipment, so that the entries of the tables maintained by said intermediate equipment are not deleted unexpectedly, at the risk of breaking off the communication between the terminals while the data exchange has not yet been completed. For example, the IPsec protocol uses a message called "NAT-Keepalive" whose default sending frequency is 20 seconds.

Today, this "keepalive" mechanism, characteristic of some communication protocols including the QUIC protocol, is applied indifferently by a terminal equipment to all the paths available to this terminal equipment to access the communications network, as a state function either present or not on the way.

However, this mechanism is costly in terms of resources, both for the network that it overloads and for the terminal equipment, from which it requires the computation resources, or even increases the consumption. energy in the case of battery-operated terminal equipment, such as a mobile terminal.

In this regard, the deactivation of this “keepalive” mechanism makes it possible to multiply by 5 or 6 the operating time of a terminal equipment on battery, such as a mobile terminal for example. It will also be noted that reducing the frequency of sending “keepalive” messages significantly contributes to increasing the life of the battery of the terminal equipment. Indeed, the terminal equipment potentially embeds several applications, each sending its own “keepalive” messages. The potential gain is all the greater as the sending of the “keepalive” messages required by the application embedded in the terminal equipment is reduced.

There is therefore a need for a technique making it possible to avoid unnecessarily activating a keep-alive mechanism or to optimize its use (for example by optimizing the frequency of sending messages relating to the activation of the keep-alive mechanism. alive), in particular when this mechanism is used to maintain the characteristic entries of a communication established on a path comprising at least one intermediate device, in a table provided for this purpose and managed by this intermediate device.

Disclosure of the invention

The invention responds to this need by proposing a method for managing at least one communication established on a transport protocol of a terminal equipment in a communication network, said terminal equipment being able to access said communication network via at least an IP resource, each IP resource comprising an IP address and a port number, comprising:

the detection of a presence, on at least one path making it possible to reach via said communication network said terminal equipment on an IP address of said terminal equipment called second IP address of at least one stateful function configured to maintain a state in a table associated with a communication on said path for a predetermined lifetime, comprising the transmission in the communication network of a first message from at least a first IP resource of said terminal equipment, said first IP resource comprising a first IP address and a first port number to a second IP resource of said terminal equipment, comprising said second IP address and a second port number, and the decision of the presence of at least one stateful function on said at least one path as a function of data received by the second IP resource in response to the transmission of the first message; and

the triggering of an action for managing a communication of the terminal equipment on said at least one path making it possible to reach via said communication network said terminal equipment on said second IP address, according to said detection.

The invention relates to the management of communications of a terminal equipment in a communication network comprising intermediate equipment that embeds stateful functions, such as NATs or firewalls. These stateful functions allow the data packets exchanged by the terminal equipment via the communication to pass only for those which correspond to a valid entry in their state table, such validity being guaranteed only for a determined period of time. As a result, the management of communications of terminal equipment established on such a network can be complicated. To improve the situation, the invention is based on a completely new and inventive approach, which consists in discovering the presence of stateful functions on the paths interconnecting the various resources available to the terminal equipment to access the network. communications and taking into account the result of this discovery to decide on the communications management actions to be triggered. A principle of the invention is to send a message from at least a first IP resource of the terminal equipment to a second IP resource of this terminal equipment, and to decide on the presence of a state function on the path connecting the first and second resources as a function of data received by the terminal equipment on its second IP resource, following the transmission of the message. In other words, the invention proposes to discover the presence of possible stateful functions by relying on the local IP resources of the terminal equipment. The detection of a stateful function on one of the paths making it possible to reach the second IP address of the terminal equipment is taken into account for the management of a communication involving this second IP address of the terminal equipment.

The invention also applies when the terminal equipment has only a single IP address to access the network (the first and second IP resources are then characterized by identical IP addresses but different port numbers).

The invention is suitable for any type of transport protocol, in particular the QUIC protocol. According to one aspect of the invention, the presence of at least one stateful function on the path is decided when the data received includes an error message or when no data is received by the second IP resource in response to the transmission of the first message by at least unedite first IP resource.

According to another aspect of the invention, the first message comprises a request for establishment of a communication between said at least one first IP resource and the second IP resource of the terminal equipment and a decision of absence of a stateful function. on the path connecting said at least one first IP resource to the second IP resource via said network is taken when the data received by the second IP resource of the terminal equipment includes the first message.

One principle of the invention is to use the different network access resources available to the terminal equipment to discover the stateful functions present on the paths connecting these resources. In order to be able to detect the presence of a stateful function on a path associated with a second IP address of the terminal equipment, for example new communications are established from resources comprising IP addresses of this terminal equipment to a resource comprising this second IP adress. These new communications do not correspond to any valid entry in the tables maintained by any stateful functions present on the path of the messages exchanged.

According to yet another aspect, the detection step comprises the masking of routing information associated with said second IP resource and contained in said request for establishing said communication, prior to its transmission.

One advantage is to emulate terminal equipment having only one interface. Thus the second IP resource is not recognized as a local address of the sending terminal equipment item and the communication request is sent via the output interface of the terminal equipment item.

Alternatively, the detection step comprises, prior to the transmission of the communication establishment request, the recording of identification information of at least one routing device of the communication network associated with said to. minus a first IP resource.

Another option for forcing the transmission of the communication request in the network is source routing, i.e. the explicit designation of a router equipment of the communication network in charge of routing the packets of data sent by the first terminal equipment on the path corresponding to the first IP resource.

According to another aspect of the invention, a communication being established between said at least one first IP resource of the terminal equipment and an IP resource of a second terminal equipment, said first message is sent via said communication to said resource IP of the second terminal equipment, said first message comprises at least one command for sending a response to the second IP resource and a command for inserting security information into said at least one response and a decision to The absence of a stateful function on the path making it possible to reach via said communication network said terminal equipment on the second IP resource is taken when the data received on said second IP resource from the terminal equipment (T1) includes said response. An advantage of this embodiment is that it uses a communication already established by the terminal equipment from one of its network accesses, to test its other network accesses. According to yet another aspect of the invention, when the presence of at least one stateful function has been detected on said at least one path, the method comprises determining a time period for sending a hold message. alive of a state of a communication established on said at least one path by said at least one stateful function, the storage of the determined period and the taking into account of the determined period in the decision of a management action on a communication via said path.

One advantage of knowing this period is that it enables the management of communications to be optimized. For example, we can choose the communication on the path whose state function has the highest period and activate the keep-alive mechanism by configuring a sending of messages according to said period, to limit the energy consumption of the equipment. terminal.

According to yet another aspect of the invention, said determination comprises:

the establishment of a communication on said path associated with the second IP address, coming from an IP resource comprising an IP address of the terminal equipment, distinct from the second, associated with a path making it possible to reach said terminal equipment via said network, for which no presence of a stateful function has been detected;

the transmission of data via the established communication, at a succession of temporal instants, two consecutive instants of said succession of instants being separated by a temporal interval, said interval having a current value initialized at zero and doubled at each new transmission of data, as long as communication is not lost; and

determining a parameter representative of said transmission time period less than said current value.

For example, the value of the chosen parameter is equal to half of the current value of the time interval. An advantage of this determination is that it makes it possible to obtain, in a simple manner, an optimum frequency of transmission of messages for keeping a state function alive which conserves energy resources.

According to another aspect of the invention, the method comprises updating a status representative of a configuration of a mechanism for keeping a state function alive, said status being associated with said at least one path allowing to join via said communication network said terminal equipment on the IP address of the second IP resource of the terminal equipment according to said at least one decision taken for said at least one path.

For example, the status is set to an “optimized” value when no stateful function has been detected on the path, since it is not necessary in this case to activate the mechanism. keeping tables alive, or following the determination of a lifetime associated with the detected state function. On the contrary, it is set to a value "not optimized" otherwise, in particular when a state function has been detected, to indicate that the mechanism for keeping the state function tables alive must be activated and that its function must be optimized. configuration. An advantage of associating such a status with an IP resource of the terminal equipment is to facilitate the management of communications via the network and in particular to allow the terminal equipment to set up communications that are more efficient in terms of resources.

According to yet another aspect of the invention, the triggered management action comprises the establishment of a communication via said path of the terminal equipment without activating a mechanism for keeping a state associated with the communication alive when a stateful function has not been detected on said path and the establishment of the communication via said path by activating a mechanism for keeping said state associated with the communication alive when a stateful function has been detected on said path . Thus, with the invention, it is possible to activate the mechanism for keeping a state associated with a communication alive only when this is necessary, that is to say when a stateful function has been detected on a path. involved in communication, which saves terminal and network resources.

According to another aspect of the invention, during the establishment of a communication with another terminal equipment from the IP address of the second IP resource of the terminal equipment, the triggered management action comprises the transmission in the request for establishment of said communication or during said communication of the status of said at least one path associated with said IP address and / or of the determined time period.

Thus, according to the invention, a terminal equipment includes its own table of paths (typically identified by local IP addresses of the terminal equipment) and a copy of that of the other terminal equipment with which it communicates (typically identified by IP addresses of the remote terminal equipment). It is understood that the statuses and lifetimes associated with the paths on which the communication is established may be different. Their mutual knowledge enables the terminals to negotiate between themselves the best choice of path and lifetime to be applied to a mechanism for keeping the states maintained by the state functions possibly present on the communication path alive. In doing so, optimization is more efficient because it is implemented by all participants in a communication.

The invention also relates to a computer program product comprising program code instructions for implementing a method for managing at least one communication as described above, when it is executed by a processor.

The invention also relates to a recording medium readable by a computer on which is recorded a computer program comprising program code instructions for executing the steps of the management method according to the invention as described above.

Such a recording medium can be any entity or device capable of storing the program. For example, the medium may comprise a storage means, such as a ROM, for example a CD ROM or a microelectronic circuit ROM, or else a magnetic recording means, for example a USB key or a hard disk.

On the other hand, such a recording medium can be a transmissible medium such as an electrical or optical signal, which can be conveyed via an electrical or optical cable, by radio or by other means, so that the program computer it contains can be executed remotely. The program according to the invention can in particular be downloaded over a network, for example the Internet.

Alternatively, the recording medium can be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the aforementioned management method.

The invention also relates to a device for managing at least one communication of terminal equipment in a communication network, said terminal equipment being able to access said communication network via at least one IP resource, characterized in that it is configured for:

detecting a presence, on at least one path making it possible to join via said communication network said terminal equipment on an IP address of said terminal equipment called second IP address of at least one stateful function configured to maintain in a table a state associated with a communication on said path for a determined lifetime, comprising the transmission in the communication network of a first message from at least a first IP resource of said terminal equipment, said first IP resource comprising a first IP address and a first number of port to a second IP resource of said terminal equipment, comprising said second IP address and a second port number, and the decision of the presence of at least one stateful function on said at least one path as a function of data received by the second resource IP in response to the transmission of the first message; and triggering an action for managing a communication of the terminal equipment on said at least one path making it possible to reach via said communication network said terminal equipment on said second IP address, as a function of said detection.

More generally, such a device is able to implement a method for managing at least one communication as described above. Advantageously, said device is integrated into terminal equipment capable of accessing a communication network from at least one IP resource, comprising an IP address and a port number.

The device can also be integrated into proxy equipment of a communication network, able to relay the data sent by terminal equipment connected to said network.

The terminal equipment, the device for managing at least one communication and the corresponding computer program mentioned above have at least the same advantages as those conferred by the management method according to the present invention.

Correlatively, the invention also relates to a method of processing a communication established on a transport protocol between a first terminal equipment and a second terminal equipment in a communication network via at least one IP resource of the first terminal equipment, each IP resource comprising an IP address and a port number, comprising:

the receipt, from said first IP resource of a message comprising the IP address, called the first IP address and the port number, called the first port number, of said first IP resource, said message comprising a send command d 'a response to at least a second IP resource of the first terminal equipment, distinct from the first and comprising a second IP address and a second port number, and a command for inserting security information in the response ;

extracting said at least one second IP resource and said security information from said command;

sending at least one second IP address and said second port number to said at least one response comprising said security information.

This response from the second terminal equipment to the message sent by the first terminal equipment is exploited by the method for managing at least one communication implemented by said first terminal equipment item to detect the presence of a state function on the path associated with the second IP address. Advantageously, an acknowledgment message is transmitted to the first terminal equipment. Said first terminal equipment thus receives confirmation that the second terminal equipment has indeed received its order. Therefore, if it does not receive the response comprising the security information, the first terminal equipment will be able to deduce therefrom the presence of a state function on the path tested.

Correspondingly, the invention also relates to a method of processing a communication on a transport protocol between a first terminal equipment item and a second terminal equipment item in a communication network via at least one IP resource of the first terminal equipment item, each IP resource comprising an IP address and a port number, characterized in that that it includes:

the reception on an IP resource of the second terminal equipment, called the destination IP resource, of a message relating to a communication established or to be established from an IP resource of the first terminal equipment, called the source IP resource, said message comprising a status representative of a configuration of a mechanism for keeping an entry of a table maintained by a stateful function alive, said entry associating a state with a communication on a path making it possible to reach via said communication network said first device terminal on the IP address of the source IP resource for a predetermined lifetime, said status being associated with said IP address, and a period of time representative of a transmission frequency of a keep alive message of a communication status;

adjusting said status and said period based on status and period values stored in memory in association with said destination IP address;

issuing a response including the adjusted values.

Thus, with the invention, the two terminal equipments negotiate the activation / deactivation of the “keepalive” mechanism and the optimum frequency of transmission of the state keep alive messages.

The invention also relates to a computer program product comprising program code instructions for implementing a method of processing a communication as described above, when it is executed by a processor.

The invention also relates to a recording medium readable by a computer on which is recorded a computer program comprising program code instructions for the execution of the steps of the processing method according to the invention as described above. .

Alternatively, the recording medium can be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the aforementioned management method. The invention also relates to a device for processing a communication between a first terminal equipment item and a second terminal equipment item via a communication network on a transport protocol, characterized in that it is configured to implement a processing method. of the aforementioned communication.

Advantageously, said device is integrated into terminal equipment capable of accessing a communication network from at least one IP resource, comprising an IP address and a port number. It can also be integrated into the aforementioned proxy equipment.

The terminal equipment, the processing device of at least one communication and the corresponding computer program mentioned above have at least the same advantages as those conferred by the processing method according to the present invention.

The invention finally relates to a node device of a communication network capable of receiving, on at least one IP resource comprising an IP address and a port number, data from a communication between an IP resource, called the source IP resource, d. 'a first terminal equipment and an IP resource of a second terminal equipment, called the destination IP resource, and to retransmit them from said at least one IP resource. Such equipment is configured for:

detecting the presence, on at least one path making it possible to join via said communication network said node equipment on an IP resource of said terminal equipment called second IP resource, of at least one stateful function configured to maintain in a table a state associated with a communication on said path for a determined lifetime, comprising the transmission in the communication network of a first message from at least a first IP resource of said node equipment to said second IP address of said terminal equipment and the decision of the presence of at least one stateful function on said at least one path as a function of data received by the second IP resource in response to the transmission of the first message;

triggering an action for managing a communication of the terminal equipment on said communication network via said path as a function of said detection.

Thus, the invention is implemented recursively by the different node devices involved on the path tested. List of Figures

Other aims, characteristics and advantages of the invention will emerge more clearly on reading the following description, given by way of simple illustrative example, and not limiting, in relation to the figures, among which:

[fig. 1]: this figure schematically represents an example of terminal equipment having several paths for connecting to a communication network;

[fig. 2]: this figure represents in the form of a logic diagram the various steps of the method for managing at least one communication of a terminal equipment item in a communication network according to one embodiment of the invention;

[fig. 3A]: this figure diagrammatically represents a first example of construction of a list of candidate communications between two distinct addresses of the same terminal equipment;

[fig. 3B]: this figure schematically represents a second example of construction of a list of candidate communications between two distinct addresses of the same terminal equipment;

[fig. 4]: this figure details in the form of a flowchart the different sub-steps of the step of discovering at least one stateful function on a path of the terminal equipment according to a first embodiment of the method for managing at least one. less one communication according to the invention;

[fig. 5]: this figure schematically represents a first example of a stateful function present on a path of the terminal equipment in the communication network;

[fig. 6]: this figure schematically represents a second example of a stateful function present on a path of the terminal equipment in the communication network;

[fig. 7]: this figure schematically represents a third example of a state function present on a path of the terminal equipment in the communication network;

[fig. 8A]: this figure schematically represents a first option of the invention for forcing the sending of a communication request intended for an address of the transmitting terminal equipment in the communication network;

[fig. 8B]: this figure schematically illustrates the reception of the communication request to the address of the sending terminal equipment according to this first option;

[fig. 9]: this figure schematically represents a second option of the invention for forcing the sending of a communication request intended for an address of the transmitting terminal equipment in the communication network; [fig. 10]: this figure schematically represents the direct communications established by a terminal equipment between two of its own interfaces according to a first exemplary embodiment of the invention;

[fig. 11]: this figure schematically represents the direct communications established by a terminal equipment item between two of its own interfaces according to a second exemplary embodiment of the invention;

[fig. 12]: this figure schematically represents the direct communications established by terminal equipment between two of its own interfaces according to a third exemplary embodiment of the invention, when the terminal equipment is connected to the network via access equipment;

[fig. 13]: this figure represents an example of implementation of the invention when the terminal equipment has been configured with a proxy equipment located in the communication network;

[fig. 14]: this figure details in the form of a flowchart the various sub-steps of the step of discovering at least one stateful function on a path of the terminal equipment according to a second embodiment of the method for managing at least one. less one communication according to the invention;

[fig. 15]: this figure schematically represents an example of the structure of a control message, called mirror, sent by a terminal equipment to a second terminal equipment via a communication already established according to an embodiment of the invention. ;

[fig. 16]: this figure schematically represents a first example of implementation of a detection of the presence of stateful functions on a path of a first terminal equipment item by using a communication already established with a second terminal equipment item according to the second mode realization of the invention;

[fig. 17]: this figure schematically represents a second example of implementation of a detection of the presence of stateful functions on a path of a first terminal equipment item by using a communication already established with a second terminal equipment item according to the second mode realization of the invention;

[fig. 18]: this figure schematically represents the determination of a lifetime of an input of a state function detected on a path of a terminal equipment item according to a third embodiment of the invention;

[fig. 19A]: this figure schematically represents a first example of the structure of a control message of a lifetime of a state function detected on a path of a terminal equipment, according to the third embodiment of the invention; [fig. 19B]: this figure schematically represents a second example of the structure of a control message of a lifetime of a state function detected on a path of a terminal equipment, according to the third embodiment of the invention;

[fig. 20]: this figure diagrammatically represents an example of transmission of a control message by a terminal equipment item to a second terminal equipment item according to this third embodiment of the invention;

[fig. 21]: this figure represents a block diagram of a device for managing at least one communication of terminal equipment in a communication network according to the invention; and

[fig. 22]: this figure represents a block diagram of a device for processing a communication of a terminal equipment item in a communication network according to the invention.

Detailed description of embodiments of the invention

General principle

The general principle of the invention is based on the discovery of stateful functions present on one or more paths making it possible to join an interface of a terminal equipment via a communication network and on the decision to trigger an action for managing a communication. with another terminal equipment on one of these paths according to said discovery. For example, one possible action is the establishment of a communication by choosing a path which does not have an intermediate state function and without activating a mechanism for keeping alive an entry of a table maintained by such a state function. . The invention thus makes it possible to manage more effectively the communications of a terminal equipment via a communications network and to optimize the energy resources brought into play by these communications. The invention applies to any type of terminal equipment: a single interface terminal equipment or a multi-interface and multi-use terminal equipment capable of establishing communications on multiple paths while preserving the continuity of the service (s) ( s) correspondent (s) when the terminal equipment is in a mobile situation.

In the remainder of the description, terminal equipment (or more simply sometimes “terminal”) denotes any entity capable of establishing or accepting the establishment of a communication based on the use of one or more communication protocols. transport, such as TCP, UDP, or QUIC. It can be a physical entity, a virtual entity, or a software application embedded in the terminal equipment.

By "QUIC protocol", or abbreviated "QUIC", is meant any protocol conforming to a version of the specification of the QUIC protocol or draft specification, such as the draft specification of the IETF entitled "QUIC: A UDP -Based Multiplexed and Secure Transport ”, or the specification of the “Quick UDP Internet Connections” protocol, known as the “QUIC” protocol, including the existing versions of these specifications or draft specifications and their evolutions. More generally, QUIC here denotes any transport protocol encapsulated on another UDP or UDP-lite transport protocol (standing for “Lightweight User Datagram Protocol”) but whose primitives and payload are encrypted.

The invention applies to any type of terminal equipment, fixed or mobile, comprising one or more communication interfaces with a communication network. These interfaces can be wired, such as for example an interface of ADSL or fiber type, or non-wired, such as for example an interface of WLAN, BlueTooth, Zigbee, or other type.

The term “communication network” is used to denote a communication network of the Internet type which can be accessed by the terminal equipment via one or more networks for accessing this network.

The terminal equipment which initiates the establishment of a communication is generally referred to as the sending or client terminal, while that for which the establishment request is intended is called the remote terminal or the server. A single terminal can therefore act both as client and server.

It is assumed that one or more paths of a terminal can be used to establish a communication and that a communication can be established via one or more paths.

Stateful functions, for example of the network address translation or NAT or firewall type, can be present on all or part of the available paths that a terminal equipment can use to access the network, both in the access network of the sending terminal, and in that of the remote terminal, or even both. NAT functions can be of different types, such as for example NAT44, NAT64, DS-Lite, NPTv6, L2NAT, NAT 66, etc. Several NAT functions can be present on a given path, as for example in the case of a so-called "double NAT" deployment.

A network path can involve both NAT functions and firewalls, without any restrictive assumption as to the number of functions present or the order in which they operate.

The term “incoming state function” denotes a state function located in the access network to which a terminal to which a communication is addressed is connected. An outgoing stateful function denotes a stateful function located on the contrary in the access network to which the terminal equipment sending the request to set up a communication is connected. Thus, a stateful function can act as an incoming state function or an outgoing state function depending on the direction of traffic.

One or more addresses or prefixes can be allocated to terminal equipment by an access network. The term IP resource denotes a pair comprising an IP address and a port number.

In the remainder of the description, an attempt is made to describe in more detail the case of communications established with the QUIC protocol and in particular signaling message structures adapted to the QUIC protocol, but the invention is not limited to this example. and relates more broadly to any other transport protocol used by a transmitting terminal equipment item to establish a communication with a remote terminal equipment item via a communication network, such as for example the TLS protocol (for “Transport Layer Security”). In the examples described below, the case of a state function of NAT type is considered in particular, but the invention applies more generally to any type of state function. In relation to FIG. 1, we present a terminal equipment Tl having several paths to access, via interfaces which allow the terminal Tl to be connected to different access networks to a communication network RC, such as the Internet network: a first path C1 associated with an IP address @ T11 of the terminal equipment Tl via an access network Nil to the network RC, a second path C2 associated with an address @ T12 of the terminal equipment Tl via an access network N12 , an i ^th path Ci associated with an address @Tli of the terminal equipment T1 via an access network Nli, with i non-zero integer.

FIG. 2 illustrates the main steps implemented by a method for managing at least one communication of the terminal equipment in the communication network RC.

During a step 20, it is checked, for example when starting the terminal T1, whether it has several network paths Cli to access the network RC and an ACAL list is obtained (for “Address Candidate List”). addresses associated with these paths. Preferably, such an ACAL list comprises, for an IP address @Tli of the terminal T1, an entry associating at least one status value STATUS with the address @Tli. By default, this parameter is set to OKOFF, which means that for this address, no optimization of management of a volume of messages sent by a keepalive mechanism for stateful functions is implemented.

In 21, a list of candidate communications CC, called the PCL list (standing for “Path Candidate List”), is constructed from the ACAL list between two distinct resources of the terminal equipment T1, that is to say. say between a first pair (first IP address, first port number) and a second pair (second IP address, second port number). It will be understood that these candidate communications are intended to connect two distinct resources of the same terminal equipment T1. Optionally, the PCL list can exclude private IPv4 addresses as described in RFC 1918, published by the IETF in February 1996 and accessible via the following URL: https://tools.ietf.org/html/rfcl918, IANA-Reserved IPv4 Prefix for Shared Address Space IPv4 addresses, as specified in RFC 6598 published by the IETF in April 2012 and accessible via the following URL: https://tools.ietf.org / html / rfc6598, IPv6 addresses of LLA (“Link Local Address”) or ULA (“Unique Local Address”) type. It will be noted that the exclusive allocation of such addresses to a terminal equipment item is an indication that an address translation mechanism is implemented in the communication network to allow global connectivity.

It will also be noted that for an element, that is to say the same candidate communication of the PCL list, the source and destination addresses must be of the same type or family of addresses, for example IPv4 or IPv6, but that the list may contain candidate communications involving addresses of different types, for example candidate communications between two IPv4 type addresses and candidate communications between two IPv6 type addresses.

In 22, the presence of at least one stateful function is detected, on at least one path making it possible to reach the terminal T1 via the communication network RC on an IP address of said terminal equipment, called the second IP address. Advantageously, this step is repeated for all the paths listed in the ACAL list.

To carry out this detection, the candidate communications of the PCL list are used.

Consider for example candidate communications involving the second IP address. A first message is sent in the communication network from at least a first IP resource of said terminal equipment to a second IP resource of said terminal equipment, comprising the second IP address and a port number. The terminal T1 then decides on the presence of at least one stateful function on the path taken by this communication to connect the terminal T1 to the network RC via the second IP resource as a function of data received on this second IP resource in response to the transmission of the first message. This step is advantageously repeated for all the candidate communications involving the second IP address of the terminal T1.

It will be understood that the detection of the presence of a stateful function for one of the candidate communications is sufficient to conclude that a stateful function is present on the path connecting the terminal T1 to the network RC via the second IP address.

In 23, an action for managing a communication of the terminal equipment T1 is triggered on the path making it possible to reach via said communication network said terminal equipment on said second IP address, as a function of said detection.

For example, the management actions of step 23 include selecting a path for establishing a communication with a remote terminal as a function of the result obtained in 22. Examples of management actions will be detailed in the remainder of the description.

We now detail step 21, called LIST PCL, of obtaining a PCL list of candidate communications from the paths of the ACAL list.

LIST PCL step

This step therefore aims to create the PCL list from the ACAL list.

If the addresses in the ACAL list are not all of the same type, for example if the list includes elements of different types, for example IPv4 addresses and IPv6 addresses, then a sublist is created by address type: ACAL (v4) and ACAL (v6).

It will be noted that the ACAL list of paths available at the terminal T1 to access the network RC can include IP addresses or prefixes, but also network interface names (eg ethl, wlanO), network names, etc. etc. In what follows we will therefore refer to the elements of the ACAL list.

According to the invention, the PCL list comprises candidate communications CC between two elements of the ACAL list. It is created taking into account the following constraints:

Each entry in the PCL is made up of two items from the ACAL;

The first element of a PCL entry is called “Source”. A port number of the terminal T1, called the source port, is associated with it;

The second element of a PCL entry is called “Destination”. A port number of the terminal T1, called the destination port, is associated with it;

Each item in the ACAL list must appear in at least one PCL entry as "Destination";

Two ACAL elements can only belong to one and only one entry in the list

PCL;

An ACAL list item may appear in multiple PCL entries as "Source";

Some ACAL items may not appear in any PCL entry as "Source";

The PCL list is ordered;

An entry with an SRC1 source must not precede an entry with a SRC1 destination. Indeed, such a configuration would have the effect of creating an input in a possible intermediate state function for the element SRC1. When testing a communication with SRC1 as the destination, this stateful function could allow the data packet used to perform the test to pass and thus mask the presence of this stateful function on the path associated with

SRC1. FIGS. 3A and 3B illustrate two examples of construction of the PCL list from the ACAL list comprising the elements Nil, N12, Nli. These two examples satisfy the constraints listed above.

The example in Figure 3A generates the following PCL list:

{Nil, N12}, {N12, N 13}, {N13, N 14}, ..., {Nli-1, Nli}, {Nli, Nil}.

The example in Figure 3B generates the following PCL list:

{Nil, N12}, {Nil, N 13}, {Nil, N 14}, ..., {Nli, Nli-2}, {Nli, Nli-1}, {Nli, Nil}.

In relation to FIG. 4, the step 22 of detecting the presence of a state function according to a first embodiment of the invention is now detailed.

MULT1 embodiment or "stand-alone" mode

As illustrated by FIG. 4, a communication to be established is selected at 220 from among the candidate communications of the PCL list. During this selection, care is taken not to choose a source IP resource that has just been used as the source or destination of a previous communication, so as to prevent an entry corresponding to this resource from being still maintained as valid in a table of a stateful function present on the tested path.

The selected communication is established in 221 by forcing the transmission in the communication network of a request to establish a communication between the source address, the source port number, and the destination address and the port number destination, so that the request is not processed locally at the terminal T1. Advantageously, the source port number chosen for this candidate communication is different from those already used to test the candidate communications listed above. This condition is necessary to detect the presence of NAT functions such as EIF / EIM ("Endpoint Independent Filtering / Mapping") or firewalls.

Indeed, as illustrated by FIG. 5, a communication from a new @Tli address can be established with the @ T11 address even in the presence of a NAT if a communication from this @Tli address has been established previously (and maintained by the NAT function) to another @ T2 address. Indeed, with reference to FIG. 5, it is assumed that the NAT function supports the “EIF / EIM” mode. When the terminal Tll establishes a communication with a remote terminal Tli, the NAT function allocates to this communication an external address and an external port number (212.25.26.25:1234) and rewrites the internal source address and internal port number of the packet issued by Tll (192.168.0.2:7856) to Tli. Then, the NAT function keeps in memory an entry in its tables to associate the internal and external information of this communication. This entry does not contain information about Tli. Therefore, any incoming call to (212.25.26.25:1234) will be routed to Tll in rewriting the destination address and port number (192.168.0.2:7856) according to the instructions in the above table. In other words, the NAT function does not check the source address and port number to process an incoming packet.

Note that other types of filtering are more restrictive and can be detected more easily. This is for example ADM / ADF filtering (for “Address Dependent Filtering”), illustrated by FIG. 6, or APDM / APDF filtering (for “Address Port Dependent Filtering”, in English), illustrated by FIG. 7. In fact, unlike EIM / EIF mode, the NAT function configured in ADM / ADF mode (FIG. 6) must keep in memory the destination address of an outgoing packet, in addition to the source information. Thus, only the packets received with a source address present in an entry in the NAT function table will be routed to an internal terminal. For example, if Tll establishes a communication with a remote terminal Tli (35.26.25.25:4545), the NAT function allocates an external address and an external port number (11.11.11.11:1234) and rewrites the internal source address and number internal port of the packet sent by Tll (192.168.0.2:7856) to Tli. Then, the NAT function keeps in memory an entry in its tables to associate the internal and external information as well as the destination address characteristic of this communication (internal source: 192.168.0.2:7856, external source: 11.11.11.11:1234, destination : 35.26.25.25). The NAT function rejects all incoming communications to 1.1.1.1:1234 if the source address of a communication is not equal to 35.26.25.25.

The example of Figure 7 is similar to that of Figure 6 except that the NAT function records, in addition to the information maintained in the state tables of the example of Figure 6, the port number destination (4545) for outgoing communication sent from Tll (192.168.0.2:7856). The NAT function rejects all incoming communications to 1.1.1.1:1234 which do not have a source address and a source port number equal to 35.26.25.25:4545.

In 222, it is checked whether the request for establishing a communication sent from the source address and the source port number has been received by the destination IP resource of the terminal T1. If so, it is declared in 223 that the candidate communication (@ T11, PS, @Tli, PD) takes a path which does not include a stateful function and a PATH-STATUS status of the candidate communication is updated to the value OK.

If not, we decide that the path (@ T11, PS, @Tli, PD) is unusable in the absence of an outgoing communication and we set the status PATH_STATUS associated with NOP.

Finally, if an error message has been received, for example of the ICMP type with an indication of the destination type or unreachable protocol (for "unreachable destination or protocol", in English), the path taken by the selected candidate communication is declared unusable and the PATH-STATUS status of the communication is set to the value NOP.

It will be noted that the decisions to update the statuses of the paths taken by the candidate communications tested can be taken immediately after a single establishment attempt or alternatively after several attempts.

The sequence of steps 220 to 223 is then repeated as long as there are candidate communications to be tested.

According to a first option all the candidate communications are tested. One advantage of a systematic approach is that it allows the terminal to assess the viability of all the available paths (and to decide on a traffic routing policy accordingly).

According to a second option, only part of the communications is established in accordance with a policy local to the terminal T1. An advantage of a selective approach is that it is less expensive in computing resources and potentially better suited to particular contexts, according to which for example the choice of a path is indexed by the nature of the traffic which will pass through the communication. For example, Internet traffic goes through NAT while voice traffic is routed over a path without NAT.

Advantageously, the candidate communications are selected at 220 according to their order of appearance in the PCL list and established sequentially so as to avoid the registration of entries by a NAT function on the path that it is desired to test. However, it will be noted that several candidate communications can be established simultaneously provided that no IP resource of the terminal appears as both source and destination in candidate communications of the PCL list.

On the other hand, if the PCL list is constructed as follows: [(@ T11, @ T12), (@ T12, @ T13), ..., (@ Tli, @ Tli), ..., (@Tli , @ Tll)], then the candidate communications will necessarily be established sequentially.

In 224, the status of the IP addresses associated with the paths listed in the ACAL list is updated as follows:

if all the candidate communications tested to an IP address @Tlk, k integer between 1 and i, have their status set to OK, it is decided that the path associated with the @Tlk address of the terminal T1 does not include stateful function and we update the status of this IP address to a value "optimized" or "OKON" (for "Optimized Keepalive On", in English), which means that for this path, the optimization procedure of the mechanism for keeping the inputs maintained by the stateful functions active;

otherwise, if at least one candidate communication tested to a resource including the IP address @Tlk has its status set to NOP, then we decide that the path associated with the IP address @Tlk of the terminal Tl includes an intermediate state function and we update the status of this STATUS resource to a value "not optimized" or "OKOFF" to signify that the optimization of the mechanism for keeping the inputs maintained by the stateful functions alive is not activated on this path and that it will be necessary to determine a message transmission frequency "Keepalive" to optimize the mechanism. After this additional processing step, the status can be modified and set to “OKON” by indicating the optimized frequency.

We now detail step 221 of sending a request to set up a selected candidate communication. As previously mentioned, because the source and destination IP addresses belong to the same terminal T1, it is necessary to force the routing of the data packet comprising said request through an output interface so that it is not processed locally by the terminal.

To do this, two options are considered:

MULT1.1 option: MP-BUND filter

According to a first variant illustrated by FIGS. 8A and 8B, the request for establishing the communication comprises a parameter MP_BUND (@ T1D) for masking the interface corresponding to the destination IP address of the candidate @ T1D communication of the terminal Tl. This MP_BLIND (lnterface Alias / Address) parameter is used to act as if the Tl terminal only had a single IP address or a single output interface (interface identifier "Interface Alias" or the the “Address” interface).

This new parameter proposed by the invention, when it is invoked to trigger the sending of a packet in the network to a given interface, has the effect of applying a filter to all the other interfaces of the terminal, which has the effect that only information relating to the unfiltered interface is retained.

In doing so, the request to establish a communication between the addresses {@ T1S, @ T1D} invoked with the MP_BLIND parameter has the consequence of sending the packet comprising this request to the default router associated with the source IP address @ T1S.

Thus the terminal T1 maintains a global routing table (table 1) which includes two routing tables associated respectively with each of its “eth0” and “ethl” interfaces.

fl; ~ # ip rule show

0: from garlic lookup local

32764: from 1.2.3.2 lookup 2

32765: from 10.20.30.2 lookup 1

32766: from garlic lookup main

32767: from ail lookup default

tl: ~ # ip route 1.2.3.0/24 dev ethO proto kernel scope link src 1.2.3.2

10.20.30.0/24 dev ethl proto kernel scope link src 10.20.30.2

default via 1.2.3.1 dev ethO

Table 1

Suppose, for example, that the terminal wants to establish a communication between the source and destination IP addresses {@ T1S = 1.2.3.2, @ T1D = 10.20.30.2}. This communication is intended to be processed locally because the destination address is that of the “ethl” interface.

In order to force the sending of the data packet comprising the request for establishing a communication via the RC communication network to the destination IP address @ T1D corresponding to the “ethl” interface, the terminal therefore emulates a behavior mono-interface using the MP_BLIND (ethl) parameter. The invoked routing table (Table 2) is shown below. The packet is then transmitted to the routing equipment RI as illustrated by FIG. 8A.

fl; ~ # ip route show table 2

1.2.3.0/24 dev ethO scope link

default via 1.2.3.1 dev ethO

Table 2

If no stateful function is present on the path connecting the terminal T1 to the network RC via the “ethl” interface, the data packet will be received by the terminal T1 via the routing equipment R2, as illustrated in figure 8B.

Option MULT1.2: Routing at source

According to a second variant illustrated in FIG. 9, the sending of the request for establishing a communication between the source @ T1S and destination @ T1D IP addresses is forced in 221 by using a source routing technique. Such a technique, known to those skilled in the art, can be implemented in different ways, such as for example the extension of routing by segment SR (for “Segment Routing”), the IPv4 option entitled “ Loose Source and Record Route or any other similar source routing function.

The option based on source routing consists of associating additional information with each of the entries in the list of candidate communications PCL to indicate, for each candidate communication, the default router associated with its source IP address @ T1S. For example, the PCL list updated by the Tl terminal is: {@ T11, PS11 @ T12, PD12, default_router = Rll}, ..., {@Tln, PSln, @ T11, PD11, default_router = Rln}.

In the following, it is assumed for example that the SR option is used.

With this option, the decision to route the data packet comprising the request to set up the selected call is based on the content of the SR option. So the terminal sends the call establishment request to the default router as identified in the corresponding entry in the PCL table. On receipt of the packet, said default router removes the SR option and proceeds to transmit the data packet to the next hop in the communication network.

The packet is thus routed step by step or rejected when no route has been found.

In relation to FIG. 10, the terminal T1 seeks to establish in 221 the candidate communications CC listed in its PCL table: {@ T11, PS11, @ T12, PD12}, {@ T12, PS12, @Tli, PDli} and { @Tli, PSli, @ T11, PD11}. In the example of FIG. 10, all these communications are established without difficulty. The ACAL table is therefore updated to set the STATUS parameter of each path associated with a tested destination IP address to the value “OKON”.

In relation to FIG. 11, the terminal T1 seeks to establish the following candidate communications on its access paths to the network: {@ T11, PS11, @ T12, PD12}, {@ T12, PS12, @Tli, PDli} and {@Tli, PSli, @ T11, PD11}. In the example of FIG. 11, all these communications are established successfully with the exception of the candidate communication {@ T12, PS12, @Tli, PDli}, because a NAT function is present on the corresponding path. The ACAL table is then updated to update the STATUS parameter of the IP addresses of the access paths to the network of the terminal T1. It sets to the value “OKON” the status of all the IP addresses available in the ACAL list at except for the @Tli address for which the STATUS parameter is kept at "OKOFF".

In relation to FIG. 12, the terminal is connected to the communication network by means of access equipment of the CPE (“Customer Premises Equipment”) type. This is for example a residential gateway. In this case, the method which has just been described can advantageously be implemented by this CPE equipment.

In relation to FIG. 13, we now describe an alternative embodiment of the embodiment of the method which has just been presented in relation to FIG. 4, according to which the terminal T1 has been configured to connect to the network via proxy equipment. P located in the RC communication network.

MULT2 embodiment or "Proxy" mode

According to the example of the QUIC protocol, this proxy equipment P implements a “Proxy QUIC” function responsible for executing operations on the QUIC communications sent by the or intended for the terminal T1. No constraint is imposed by the. invention as to the operations performed by such a proxy function. This embodiment of the invention is applicable regardless of the location of the “Proxy QUIC” function in the communication network.

The “QUIC Proxy” can typically be configured on the terminal using, for example, the DHCP protocol, the PCO option (for “PDP Configuration Options”), etc.

The requests for establishment of communication according to the QUIC protocol sent by the terminal T1 are intercepted by this “Proxy QUIC” P which then relays them to their destinations. O supposes by way of example that an IP-in-IP or GRE encapsulation scheme (for "Generic Routing Encapsulation", in English), as described in the document RFC 2784 published by the IETF in March 2000, is used between the terminal and the “Proxy QUIC” function.

If several “Proxy QUIC” functions are present in the network, the terminal can select one or more instance (s) of this function, or even use all the instances to establish communications in the network.

According to the invention, the terminal T1 builds the PCL list of candidate communications at 21 from the ACAL list according to the LIST step described above. For example, the PCL list thus constructed for the terminal T1 is {@ T11, PS11, @ T12, PD12}, ..., {@Tln, PSln, @ T11, PD11}. Then, at 221, as already described in relation to FIG. 4, the terminal T1 attempts to establish a QUIC communication for each of the elements of the PCL list, but this time via at least one “Proxy QUIC” function.

For example, as illustrated by the example of FIG. 13, the terminal T1 sends to the proxy P a request for the establishment of a communication QUIC comprising the destination IP address @Tli. To do this, it encapsulates this establishment request (formatted according to a QUIC frame) in an IP packet whose destination address is that of the proxy P. On receipt of the packet by the proxy P, the latter extracts the QUIC frame. Then, it tries to route the call establishment request to the IP address @Tli. If it has a route to reach the @Tli address, it sends the packet on that route. If no route has been found, the proxy P responds to the terminal with an ICMP message to indicate that this IP address is not reachable. If the proxy P receives the call establishment request, it retransmits it to the terminal by encapsulating it in an IP packet.

Steps 222 to 224 described above remain unchanged.

In relation to FIG. 14, the step 22 of detecting the presence of a state function according to a second embodiment of the invention is now detailed.

MULT3 embodiment or "assisted" mode

It will be recalled that in 21 a PCL list of candidate communications to be established on the network paths available at the level of the terminal equipment T1 was constructed. During a step 220 ′, the terminal selects at least one candidate communication CC from the PCL list. Note that it can select several or even all candidate communications from the PCL list.

At 220b ', it obtains the information relating to a communication CE12 which it has already established via one of its network paths with a second terminal equipment T2 in the communication network RC. For example, this communication is established between an IP resource comprising the @ T11 address allocated by the access network Nil to the terminal T1 and an IP resource comprising the @ T2 address of the terminal T2.

At 22, the terminal T1 sends to the terminal T2, via the communication CE12, a message, called a mirror frame (for “Mirror”, in English). This frame is intended to command the terminal T2 to send a response message according to the conditions specified in the fields of the mirror frame. In the case of the QUIC protocol, this is a new QUIC frame structure proposed by the invention.

This frame includes:

at least one destination IP address of the terminal T1 to which the terminal T2 must send data. Advantageously, this field comprises the destination IP addresses of the candidate communications selected at 220 ′;

at least one destination port number per destination IP address. Note that if the frame does not contain a port number, the remote terminal T2 will have to randomly generate a port number to be associated with each destination IP address;

an additional security key (or "token" or "Token") to be included in the response message to be transmitted to the address or addresses of destination indicated.

An example of a mirror frame is illustrated in figure 15.

According to the invention, the terminal T2 connected to the terminal T1 via the communication CE12 implements a method of processing a communication which comprises the following steps, illustrated by FIG. 14.

It receives at 300 the mirror frame transmitted by Tl via the CEI communication. In 301, it extracts from this frame the destination IP addresses that it contains, possibly associated with destination port numbers, as well as the security key. It is recalled that if the mirror frame does not contain a destination port number, the terminal T2 generates as many as there are destination addresses extracted, for example at random. In 302, the terminal T2 sends to each pair (destination IP address, destination port number) of the terminal T1 specified in the mirror frame, a TV frame called an empty frame because it does not transport application data, comprising only the key of security. Note that the terminal T2 does not need to establish a new QUIC communication with Tl. Indeed, as mentioned previously, a QUIC communication is independent of source and destination IP resources, but relies on a security association.

It should be noted that with another transport protocol, which would depend on the source and destination IP resources, the “mirror” frame would give rise to new communications, which the terminal T1 should listen to and associate with the presence detection test in progress. .

In 222 ′, on receipt of such a TV frame on one of its addresses, the terminal T1 checks that it does understand the security key, then, if necessary, triggers the update in 223 ′ of the status ST of the corresponding CC candidate communication from its PCL list by setting it to the value “OKON”. Indeed, the fact of having received the TV frame on the path associated with the destination IP address of the terminal T1 proves that there is no state function on the path followed by the TV frame.

On the contrary, for all the destination IP addresses associated with the mirror frames which are not received after a predetermined time, the terminal T1 triggers at 223 ′ a positioning of the status of the path associated with the value OKOFF.

At 224 ′, once all the destination addresses corresponding to the candidate communications CC of the PCL list have been tested, the terminal T1 updates the status of the corresponding paths of the ACAL list, as previously described in relation to FIG. 4 .

In relation to Figures 16 and 17, we consider the example of a terminal T1 which uses a single communication CE12 established with a remote terminal T2 to control its different available addresses @ T11, @ T12, ... @ Tli, with i non-zero integer. Concretely, once the communication CE12 has been established with the remote terminal T2, the terminal T1 sends a mirror frame listing the addresses @ T11, T12, ..., @Tli as destination IP addresses and a security key CS. On receipt of the frame, the remote terminal T2 sends an acknowledgment to T1 and then proceeds to send QUIC frames of the TV frame type comprising only the security key CS, to each of the IP addresses @ T11, @ T12, etc.

In the example of FIG. 16, all these frames are successfully received by Tl. On receipt of a data packet comprising such a frame TV, the latter checks whether the packet contains the security key CS entered in the mirror frame . If so, T1 updates its ACAL table to indicate that all the addresses have a STATUS parameter equal to “OKON”.

FIG. 17 illustrates the case where the data packets comprising the TV frames sent by T2 to the address @ T12 and @Tli are not received by T1 despite the prior acknowledgment of T2, due to the presence of stateful functions on the road. At the end of a predetermined period, T1 updates the statuses of the records in its PCL list, then those of its ACAL list to set the STATUS parameter associated with the @ T12 and @Tli addresses to "OKOFF", and the associated one at @ T11 at

"OKON". Consider now the case where the terminal T1 detects a new path. It is for example a VPN tunnel (“Virtual Private Network”, in English), established from a physical interface of the terminal T1 for which it already has a network path.

This detection triggers the implementation of the step MULT for discovering stateful functions on this path according to one of the embodiments which have just been described. This new path is associated with an @ Tli + l address.

For example, the terminal T1 requests the terminal T2 with which it is already connected by sending it a mirror frame comprising the address @ Tli + 1 and a security key CS via a communication already established with it. A destination port number different from that used for the communication already established is indicated in the mirror frame. At the end of this step, the terminal updates the STATUS parameter of the path associated with the @ Tli + l address according to the result of the presence detection procedure.

Then, as a function of the responses received, the terminal T1 updates the statuses of the elements of the PCL then ACAL lists as previously described.

It can finally trigger at 23 (CONNECT) communication management actions on the paths of the ACAL list, by taking into account the value of the STATUS parameter assigned to these paths.

SOLO embodiment for a mono-interface terminal

Consider now the case of a terminal T1 having only a single interface for access to the communication network RC and therefore only a single path associated with a single IP address. The method according to the invention also applies to this terminal according to one of the embodiments described above.

For example, the terminal can implement the MULT2 embodiment or “Proxy” mode to force the sending of a message to its own IP address and to a port number different from that used to contact the proxy. :

If no response is received, then it concludes that there is a stateful function on the path.

If instead a response is received, then it concludes that there is no state function on the path.

In addition, it establishes that no address translation function is present on the path when the message sent by the proxy in response to the command from the terminal has been successfully routed and has not undergone change of IP address / destination port number. It also establishes that no filtering function is present on the path from the observation that the response packet received was sent from an address different from that used to contact the proxy.

Alternatively, the terminal T1 implements the MULT3 embodiment or “MIRROR” mode. In this case, it sends a "mirror" frame to a terminal T2 with which it is already in communication and having several IP addresses. This frame comprises a security key, the IP address of the terminal T1 and a destination port number different from that used by the terminal T1 to send the “mirror” frame as described in relation to FIGS. 15 to 17. Upon receipt of the message, the remote terminal T2 sends a response message or TV frame to the terminal address and the destination port number indicated in the mirror frame. On receipt of the data packet comprising such a TV frame, the latter checks whether the packet contains the security key CS entered in the mirror frame. If so, T1 concludes that there is no state function on the path and updates the STATUS status of this path by setting it to “OKON”.

It will be noted that if in theory the embodiment MULT1 is applicable to a single interface terminal, there is a risk that the data packets sent in the network RC by the terminal T1 are filtered on their return by the access network. Indeed, these packets sent with an address of this access network, once relayed by the intermediate nodes imposed by the routing at the source, will probably be treated as an attempt to spoof the IP address of the terminal and will therefore be blocked. .

The reception of such packets is also an indication that the network is not protected against IP address spoofing attacks (for "spoofing" in English).

In relation to FIGS. 4 and 18, the method for managing at least one communication according to a third embodiment of the invention will now be described and the case where the terminal T1 has detected the presence of at least one minus one stateful function on the path associated with the @Tli address. It therefore set the status of this path at 224 to the value “OKOFF”.

As illustrated by FIG. 4, the management method according to this third embodiment of the invention implements an optional step 225 of determining a KA_TIMER parameter representative of a lifetime of an input instantiated by the function to states present on the way. It is assumed that the method comprises a preliminary step (not shown) of extracting the elements of the ACAL list associated with an OKOFF status and a test step to verify whether at least one path has been extracted.

By way of example, we consider the case of FIG. 11, for which only the @Tli address is associated with the “OKOFF” status and therefore extracted.

Then, as illustrated by FIG. 18, the terminal T1 chooses another address from the ACAL list, for example @ T11, whose status is on the contrary equal to “OKON”.

It establishes a communication from the extracted address @Tli to the other address @ T11 and initializes the KA_TIMER parameter to a predetermined TO value, for example 15 seconds, as recommended in the document entitled “UDP Usage Guidelines”, RFC 8085, published by the IETF in March 2017.

The terminal T1 then transmits data via the established communication, at a succession of temporal instants, two consecutive instants of said succession of instants being separated by a temporal interval, initially set at T0 and doubled at each new transmission of data, both that communication is not lost or cut off. In the example of figure 18, communication is lost when the parameter KA_TIMER reaches the value KA_TIMER_LOSS equal to 960s.

From the value obtained for KA_TIMER-LOSS, we derive the lifetime of an input instantiated by the state function present on the path. For example, this lifetime is evaluated at half of the value of the time interval which induced the loss, that is to say KA_TIMER_LOSS / 2 = 480s.

Advantageously, this evaluated lifetime is used to more effectively manage the communications from the terminal T1 to the communications network RC. In particular, it is used to configure the “keepalive” mechanism and the frequency of the messages sent to maintain the entry of a communication in a state function table.

Of course, the invention is not limited to this example of an implementation of the evaluation of the lifespan, even if it offers a good compromise between the performance of a lifespan mechanism and the risk of cutting. of communication.

A detailed description will now be made in relation to FIGS. 19A, 19B and 20 of an embodiment of step 23 of triggering an action for managing at least one communication involving a network path of the terminal T1.

CONNECT step (23) It is now assumed that the ACAL table of the terminal T1 has been updated and that in particular, it informs the value of the STATUS parameter of each address associated with each path available at the level of the terminal T1 to access the communication network RC. Advantageously, it also includes the lifetime value KA_TIMER of a state function present on the path, if applicable.

If it has several paths, the terminal T1 can advantageously use the information stored in the ACAL list to select the one on which it is more efficient to establish an outgoing communication. For example, the terminal T1 can select the path which does not involve a stateful function or alternatively, a path which comprises a stateful function associated with the highest possible KA_TIMER lifetime. Indeed, the lifetime of an input maintained by a stateful function directly impacts the frequency of sending messages for keeping a “keepalive” mechanism alive and therefore the energy consumption of the terminal.

To establish a QUIC communication with a remote terminal T3, the terminal T1 includes in a communication message, such as for example the request for establishment of the communication or any other message, a new control frame called KEEPALIVE_CONTROL of a mechanism for keeping an input maintained by a stateful function alive. This frame is used to transmit to the terminal T3 information for the configuration of the mechanism for keeping an entry maintained by a state function intended to be activated on one or more paths connecting the terminal T1 to the network.

Examples of the format of such a message are illustrated by Figures 19A and 19B. In relation to FIG. 19, we consider a simple TCS format intended to be used by a terminal which has a single network interface and in particular a single network path for accessing the communication network RC. It can also be used to characterize a single source address used by the terminal T1 for sending a QUIC data packet.

FIG. 19B presents a second TCM format of a control message of a keep alive mechanism intended for a multi-interface terminal, that is to say provided with several paths for accessing the RC network. The frame of this control message comprises a field representative of an element of the IP address, prefix or IP address identifier type, a STATUS status field and a KA_TIMER lifetime field. The values of the STATUS and KA_TIMER parameters are extracted from the ACAL table of the Tl terminal. The KA_TIMER field is optional. Note that if this field is absent for an element with a STATUS equal to “OKON”, this means that the procedure for keeping the inputs of a state function alive can be safely disabled. One or more KEEPALIVE_CONTROL control frames can be sent to the same remote terminal. For example, a control frame is sent per available path at the terminal T1 or else a single control frame includes all the information relating to all of the available paths.

As illustrated in FIG. 20, on receipt of the frame by the remote terminal T3, the latter extracts the information it contains and saves them in a table called ACAL_PEER. It will be understood that this ACAL_PEER table is an at least partial copy of that maintained by Tl. The remote terminal T3 advantageously uses the information in this table to decide on which path to establish a communication with Tl.

In the example of figure 20, it is assumed that for the path associated with the IP address @ T11, the status received by T3 is set to "OKON", but that no KA-TIMER lifetime value is has been positioned in the KEEPALIVE_CONTROL frame. The absence of a lifetime value when the status is set to OKON is an indication that there is no stateful function on the path. Thus, T3 knows that it can send messages to this address of Tl without having tested it beforehand.

The remote terminal T3 must in turn determine the value of the KA_TIMER parameter representative of a lifetime of an input instantiated by the stateful function present on the path between the IP address @ T11 of Tl and its own paths in the access networks N31, N32, ..., N3j with j non-zero integer, as previously described in relation to FIG. 15. Then, it transmits them to the terminals with which it wishes to communicate via the communication network, such as for example the terminal Tl.

The terminal T1 can also decide to add one or more new paths to a communication established with the terminal T3 as a function of the relative values of the parameters STATUS and KA_TIMER obtained for a new path compared to those of the paths already used. For example, a terminal can decide to add a new path to the communication even if it does not receive data from the remote terminal T3 via this path.

It is understood that for a communication established between T1 and T3, a path used for this communication can be associated with two distinct values of the parameter KA_TIMER:

KA_TIMER (Local) and KA_TIMER (PEER). In this case, the terminals must choose the smaller value between the two. In the example of Figure 20, each terminal sends KEEPALIVE_CONTROL control frames to the other. Thus, each terminal has its own ACAL table and that of the remote ACAL terminal (Peer).

It is assumed for example that the first communication has been established between T1 and T3 via the IP address @ T11 of Nil and the address @ T31 of N31. By consulting the ACAL (Peer) table, the T3 terminal decides to add a new path via the IP address @ T3i of (N3 and the IP address @ T11 of Nil) to the communication. In addition, the terminal T1 decides to add the path @Tli of Nli and @ T31 of N31 to the communication. Due to the fact that for the @ T11 address of Nil and the @ T31 address of N31, the status is set to "OKON" with no associated KA_TIMER lifetime value, the two terminals know that no stateful function has been detected on these paths and decide not to activate the “keepalive” mechanism for keeping the inputs of a stateful function alive for the path (Nil, N31). On the other hand, for the other paths used by the communication, values of KA_TIMER being entered, they activate the optimization for the other paths according to the instructions described in their ACAL tables. In particular, they configure an optimal transmission frequency of "keepalive" messages according to the KA_TIMER lifetime entered.

Thus, the two terminals can each implement the invention which has just been described.

The terminals can send new KEEPALIVE_CONTROL control frames whenever necessary, as for example in the event of modification of network parameters, such as an attachment to a new access network, obtaining a new IP address, etc.

We now present, in relation to FIG. 21, the hardware structure of a device 100 for managing at least one communication of terminal equipment in a communication network, said terminal equipment being able to access said communication network via at least two paths, one said path being associated to an IP resource, comprising at least an IP address and a port number. According to the invention, the device 100 comprises a module for detecting a presence, on at least one path making it possible to reach via said communication network said terminal equipment on an IP address of said terminal equipment called second IP address of at least one. stateful function configured to maintain in a table a state associated with a communication on said path for a determined lifetime, comprising the transmission in the communication network of a first message from at least a first IP resource of said terminal equipment, said first IP resource comprising a first IP address and a first port number to a second IP resource of said terminal equipment, comprising said second IP address and a second port number, and the decision of the presence of at least one stateful function on said at least one path as a function of data received by the second IP resource in response to the transmission of the first message, and a trigger module the execution of an action for managing a communication of the terminal equipment on said at least one path making it possible to reach via said communication network said terminal equipment on said second IP address, as a function of said detection. Advantageously, the device 100 comprises a module for obtaining a list of paths connecting the terminal to the network via an IP address of this terminal, called the second IP address, and a module for obtaining a list of candidate communications between a first resource comprising a first IP address and a first port number and a second IP resource comprising the second IP address and a second port number.

Advantageously, the device 100 also comprises a module for determining the lifetime associated with a state of a communication established on said path by said at least one state function detected on the path.

Advantageously, the device 100 further comprises a module for updating a status representative of the presence or the absence of a state function on said path as a function of the decision, where appropriate of the lifetime. associated with a state of the communication by the state function detected and the module for triggering a management action is configured to trigger said action according to the status and / or the associated lifetime.

The term “module” can correspond just as well to a software component as to a hardware component or a set of hardware and software components, a software component itself corresponding to one or more computer programs or subroutines or otherwise. more general to any element of a program capable of implementing a function or a set of functions.

More generally, such a communication management device 100 comprises a random access memory 103 (for example a RAM memory), a processing unit 102 equipped for example with a processor, and controlled by a computer program Pgl, representative the module for detecting the presence of a stateful function on said at least one path and the module for triggering at least one management action on said at least one path, and optionally modules for obtaining the list of paths , obtaining the list of candidate communications, updating a status and determining a lifetime, stored in a read only memory 101 (for example a ROM memory or a hard disk). On initialization, the code instructions of the computer program are for example loaded into the random access memory 103 before being executed by the processor of the processing unit 102. The random access memory 103 notably contains the path table ACAL. , the list of candidate communications PCL and stores for each path the associated KA_TIMER status and, if applicable, lifetime parameters. The processor of the processing unit 102 controls the obtaining of the list of paths, the detection of the presence of stateful functions on the listed paths and the triggering of a communication management action on the listed paths, in accordance with the logic diagram of FIG. 2. Advantageously, it also stores the ACAL-PEER path tables transmitted by other terminals. FIG. 21 only illustrates one particular way, among several possible, of implementing the device for managing a communication 100, so that it performs the steps of the method for managing a communication of a terminal equipment item as detailed below. above, in relation to FIG. 2 in its various embodiments. In fact, these steps can be carried out either on a reprogrammable computing machine (a PC computer, a DSP processor or a microcontroller) executing a program comprising a sequence of instructions, or on a dedicated computing machine (for example a set of logic gates such as an FPGA or ASIC, or any other hardware module).

In the case where the device 100 is produced with a reprogrammable computing machine, the corresponding program (that is to say the sequence of instructions) can be stored in a removable storage medium (such as for example a floppy disk, CD-ROM or DVD-ROM) or not, this storage medium being partially or totally readable by a computer or a processor.

The different embodiments have been described above in relation to a device for managing a communication 100 integrated into terminal equipment Tl 10, such as for example a mobile telephone of the smart phone type (for “smartphone”, in English ), a PC-type computer (for “Personal Computer” in English) or a tablet, but it can also, as described in relation to FIG. 12, be embedded in any type of CPE access equipment to a computer. communication network, such as a home automation gateway, provided that it has access to a wide area network, such as for example the Internet network.

Finally, in relation to FIG. 22, we present the hardware structure of a device 200 for processing a communication established between a first terminal equipment item and a second terminal equipment item via a communication network, comprising at least one reception module, from a first IP resource of a first terminal equipment, a message comprising at least one command to send a response to at least a second IP resource of the first terminal equipment, distinct from the first , and inserting security information in the response, a module for extracting said at least one second IP resource and said security information and a transmission module intended for said at least one second address d 'a response or TV frame comprising said security information.

Advantageously, the device 200 further comprises a module for storing an ACAL_PEER path table received from the first terminal equipment item.

The term “module” can correspond just as well to a software component as to a hardware component or a set of hardware and software components, a software component itself corresponding to one or more programs or subroutines of a computer or computer. more generally to any element of a program capable of implementing a function or a set of functions.

More generally, such device equipment 200 comprises a random access memory 203 (for example a RAM memory), a processing unit 202 equipped for example with a processor, and controlled by a computer program Pg2, representative of the reception modules, extraction and transmission, stored in a read only memory 201 (for example a ROM memory or a hard disk). On initialization, the code instructions of the computer program are for example loaded into the random access memory 203 before being executed by the processor of the processing unit 32. The random access memory 203 notably contains the second IP resources and the security information extracted from the control message received from the first terminal equipment. It can also include the ACAL_PEER path table or tables transmitted by the first terminated device and possibly other terminals. The processor of the processing unit 202 controls the reception of the message, the extraction of at least a second IP resource from the first terminal equipment and of security information included in said message, the transmission of a response comprising the security information intended for said at least one second IP resource of the first terminal equipment, in accordance with the flowchart of FIG. 14.

Fig. 22 only illustrates one particular way, among several possible, of implementing the processing device 200, so that it carries out the steps of the method of processing a communication detailed above, in relation to FIG. 14. Indeed, these steps can be performed indifferently on a reprogrammable computing machine (a PC computer, a DSP processor or a microcontroller) executing a program comprising a sequence of instructions, or on a dedicated computing machine (for example a set of logic gates such as a FPGA or ASIC, or any other hardware module).

In the case where the communication processing device 200 is implemented with a reprogrammable computing machine, the corresponding program (that is to say the sequence of instructions) can be stored in a removable storage medium (such as whether for example a floppy disk, a CD-ROM or a DVD-ROM) or not, this storage medium being partially or totally readable by a computer or a processor.

The various embodiments have been described above in relation to a communication processing device 200 integrated into terminal equipment T2, Tl, such as for example a mobile telephone of the smart telephone type (for “smartphone”, in English), a PC-type computer (for “Personal computer” in English) or a tablet, but it can also, as described in relation to FIG. 12, be embedded in any equipment CPE type access to a communication network, such as a residential gateway, provided that it has access to a network.

The invention which has just been described proposes to dynamically detect the presence of stateful functions in a communication network on access paths from a terminal to this network. This presence detection is used to manage the communications accordingly and thus improve their quality while optimizing the energy resources put into play in particular by the terminal equipment.

The invention allows in particular:

to optimize the battery consumption of mobile terminals;

- optimize the selection of paths / addresses for establishing communications; optimize the routing of traffic in the network;

to facilitate the establishment of new communications from an address of a remote terminal even if no data has yet been received from this address;

avoid communication establishment problems associated with the presence of stateful functions, such as NAT or firewall functions;

to optimize the value of the transmission frequency of “keepalive” messages by communication, taking into account the possible presence of stateful functions on a path taken by the communication.

Claims

1. Method for managing at least one communication according to a transport protocol of a terminal equipment in a communication network, said terminal equipment being able to access said communication network via at least one IP resource, each IP resource comprising a IP address and a port number, characterized in that it includes:

the detection (22) of a presence, on at least one path making it possible to reach via said communication network said terminal equipment on an IP address of said terminal equipment called second IP address of at least one stateful function configured to maintain in a tables a state associated with a communication on said path for a predetermined lifetime, comprising the transmission in the communication network of a first message from at least a first IP resource of said terminal equipment, said first IP resource comprising a first address IP and a first port number to a second IP resource of said terminal equipment, comprising said second IP address and a second port number, and the decision of the presence of at least one stateful function on said at least one path as a function of data received by the second IP resource in response to the transmission of the first message; and

the triggering (23) of an action for managing a communication of the terminal equipment on said at least one path making it possible to reach via said communication network said terminal equipment on said second IP address, as a function of said detection.

2. A method of managing a communication according to claim 1, characterized in that a presence of at least one stateful function on the path is decided, when the data received includes an error message or when no data. is received by the second IP resource in response to the transmission of the first message by at least one said first IP resource.

3. A method of managing a communication according to one of claims 1 to 2, characterized in that the first message comprises a request for establishment of a communication between said at least one first IP resource and the second IP resource of the terminal equipment and in that a decision of no state function on the path connecting said at least one first IP resource to the second IP resource via said network is taken when the data received by the second IP resource from the terminal equipment understand the first message.

4. A method of managing a communication according to claim 3, characterized in that it comprises the masking of routing information associated with said second IP resource and contained in said request for establishment of said communication, prior to its transmission. .

5. A method of managing a communication according to claim 3, characterized in that it comprises, prior to the transmission of the request for establishing the communication, the recording of identification information of at least one routing device of the communication network associated with said at least one first IP resource.

6. Method of managing at least one communication according to one of claims 1 to 2, characterized in that a communication (CE12) being established between said at least one first IP resource of the terminal equipment (Tl) and an IP resource of a second terminal equipment (T2), said first message is sent via said communication to said IP resource of the second terminal equipment (T2), said first message comprises at least one command for sending a response to the second IP resource and a command for inserting security information in said at least one response and in that a decision for the absence of a stateful function on the path making it possible to reach via said communication network said terminal equipment on the second IP resource is taken when the data received on said second IP resource from the terminal equipment (T1) includes said response.

7. Method for managing a communication according to one of claims 1 to 6, characterized in that, when the presence of at least one stateful function has been detected on said at least one path, the method comprises determining of a temporal period of transmission of a message to keep a state of a communication established on said at least one path alive by said at least one stateful function, the storage of the determined period and the taking in account of the period determined in the decision of a management action on a communication via said path.

8. A method of managing a communication according to claim 7, characterized in that said determination comprises:

the transmission of data via the established communication, at a succession of instants temporal, two consecutive instants of said succession of instants being separated by a temporal interval, said interval having a current value initialized at zero and doubled at each new transmission of data, as long as the communication is not lost; and updating a parameter representative of said period, by assigning it a value equal to half of the current value of the time interval;

9. A method of managing a communication according to one of claims 1 to 8, characterized in that the method comprises the updating of a status representative of a configuration of a mechanism for keeping alive a device. stateful function, said status being associated with said at least one path making it possible to reach via said communication network said terminal equipment on the second IP address of the terminal equipment as a function of said at least one decision taken for said at least one path.

10. A method of managing a communication according to any one of the preceding claims, characterized in that the triggered management action comprises establishing a communication via said path of the terminal equipment without activating a holding mechanism. alive of a state associated with the communication when a stateful function on said path has not been detected and the establishment of the communication via said path by activating a mechanism for keeping said state associated with the communication alive when a stateful function has been detected on said path.

11. A method of managing a communication according to one of claims 7 to 10, characterized in that, during the establishment of a communication with another terminal equipment from the IP address of the second IP resource of the terminal equipment, the triggered management action comprises the transmission in the request for establishment of said communication or during said communication of the status of said at least one path associated with said IP address and / or of the determined time period.

12. Device (100) for managing at least one communication of terminal equipment in a communication network, said terminal equipment being able to access said communication network via at least one IP resource, characterized in that it is configured for:

detecting a presence, on at least one path making it possible to join via said communication network said terminal equipment on an IP address of said terminal equipment called second IP address of at least one stateful function configured to maintain in a table a state associated with a communication on said path for a predetermined lifetime, comprising the transmission in the communication network of a first message from at least a first IP resource of said terminal equipment, said first IP resource comprising a first IP address and a first port number to a second IP resource of said terminal equipment, comprising said second IP address and a second port number, and the decision of the presence of at least one function stateful on said at least one path based on data received by the second IP resource in response to transmission of the first message; and

triggering an action for managing a communication of the terminal equipment on said at least one path making it possible to reach via said communication network said terminal equipment on said second IP address, as a function of said detection.

13. Process for processing a communication established according to a transport protocol between a first terminal equipment and a second terminal equipment in a communication network via at least one IP resource of the first terminal equipment, each IP resource comprising an IP address and a port number, characterized in that it includes:

the reception (300), from said first IP resource of a message comprising the IP address, called the first IP address and the port number, called the first port number, of said first IP resource, said message comprising a command d 'sending a response to at least a second IP resource of the first terminal equipment, distinct from the first and comprising a second IP address and a second port number, and a command for inserting security information in the response;

extracting (301) said at least one second IP resource and said security information from said command;

the transmission (303) to said at least one second IP address and to said second port number of at least one response (TV) comprising said security information.

14. Process for processing a communication established over a transport protocol between a first terminal equipment and a second terminal equipment in a communication network via at least one IP resource of the first terminal equipment, each IP resource comprising an IP address and a port number, characterized in that it includes

the reception on an IP resource of the second terminal equipment, called the destination IP resource, of a message relating to an established or to be established communication from an IP resource of the first terminal equipment (Tl), called the source IP resource, said message comprising a status representative of a configuration of a mechanism for keeping alive an entry of a table maintained by a stateful function, said entry associating a state with a communication on a path making it possible to reach via said communication network said first terminal equipment on the IP address of the source IP resource for a predetermined lifetime, said status being associated with said IP address, and a period of time representative of a frequency of transmission of a message to keep a communication state alive; adjusting said status and said period based on status and period values stored in memory in association with said destination IP address; and issuing a response comprising the adjusted values.

15. Device (200) for processing a communication between a first terminal equipment and a second terminal equipment via a communication network established on a transport protocol, said device being characterized in that it is configured to implement the method. for processing a communication according to claim 13 and / or claim 14.

16. Terminal equipment (Tl, 10) capable of accessing a communication network from at least one IP resource, comprising an IP address and a port number, characterized in that it comprises a device for managing at least one communication from said resource according to claim 12.

17. Terminal equipment (T2, 11, Tl, 10) according to claim 16, characterized in that it further comprises a device for processing a communication according to claim 15.

18. Proxy equipment (P) of a communication network, capable of connecting terminal equipment to said network, characterized in that it comprises a device for managing at least one communication according to claim 12.

19. Node equipment of a communication network capable of receiving, on at least one IP resource comprising an IP address and a port number, data from a communication between an IP resource, called the source IP resource, from a first equipment item. terminal and an IP resource of a second terminal equipment item, called the destination IP resource, and retransmitting them from said at least one IP resource, characterized in that it is configured for:

detecting the presence, on at least one path making it possible to join via said communication network said node equipment on an IP resource of said terminal equipment called second IP resource, of at least one stateful function configured to maintain in a table a state associated with a communication on said path for a determined lifetime, comprising transmission in the communication network a first message from at least a first IP resource of said node equipment to said second IP address of said terminal equipment and the decision of the presence of at least one stateful function on said at least one path as a function of data received by the second IP resource in response to sending the first message; and

20. A computer program product comprising program code instructions for implementing a method according to any one of claims 1 to 11 and 13 to 14 when it is executed by a processor.