A method for dynamically federating multiple radio communication networks, computer program and associated network

Abstract

Method for dynamic federation of a plurality of radio communication networks, computer program and associated network. The invention relates to a method (10) for dynamically federating radio communication networks, each represented by a node, comprising at least the following successive steps: - discovery (12) between the nodes, each node broadcasting a predetermined discovery message; - election (14) of a master node from among the nodes associated respectively with each network, the nodes distinct from said master node becoming slave nodes; - implementation (16) of dynamic IP addressing by said master node; - construction (18) of a federated network by said master node, by enrolling said slave nodes after their authentication by said master node; - establishment (20) of at least one communication plan between said master node and said slave nodes; - configuration (22) of said federated network using said at least one communication plan.Figure for the abbreviation: Figure 1.

Description

Description Title of the invention: Method for dynamic federation of a plurality of radiocommunication networks, program computer and associated network

[0001] — The present invention relates to a method for dynamically federating a plurality of radiocommunication networks.

[0002] = The invention also relates to a computer program comprising ins- software instructions that, when executed by a computer, implement implements such a method of dynamic federation of a plurality of radio communication networks- communication.

[0003] The invention also relates to a federated radiocommunication network via said process.

[0004] The field of the invention is the federation of a plurality of radio communication networks- communication and, more particularly, obtaining a configuration of the plurality of networks allowing them to interconnect and implement rules and protocols governing interconnectivity between two networks.

[0005] More precisely, the federation of networks, tactics (i.e. sensitive tactical bubbles) or not (eg civil telecommunications networks), allows to connect at least two separate networks (possibly through another underlying network) and configure these two networks so that they appear as one network extended for their users.

[0006] A radiocommunication network is a network of the “private mobile radiocom- communications » - PMR (“private mobile radiocommunication network”) or a network of the type meeting the 3G, 4G, 5G, or higher standard, as defined by the 3GPP (“3rd Generation Partnership Project"). In the following, will be taken more particularly for example the case of a 4G network, but the person skilled in the art knows how to apply the teaching of this description in the case of another standard, in particular the 5G standard.

[0007] — Generally speaking, a radiocommunication network comprises a radio network home network (“Radio Access Network — RAN”), an infrastructure network (“Core Network — CN”) and advantageously a data center, preferably virtual {“Software-defined data center - SDDC”).

[0008] — Radiocommunication networks can be of different types, depending on particularly their maximum capacity. For example, in a military context, a radiocommunication network may for example fall under a first type, such as a tactical network with a maximum capacity of around ten users, a second type, such as a theater data network with a maximum capacity of a few hundred users, or a third type, such as a command network with a maximum capacity of a few thousand users. Additionally, radio communication networks can be deployable and / or mobile. These are called field networks. For example, tactical networks are essentially field networks. Radiocommunication networks (in particular field networks when their geographic radio coverage areas overlap) must be able to be federated automatically to offer, for example, flexibility of use (to adapt to the movements of field networks), simplicity of implementation (to build an extended network with advanced functions such as slicing or sharing of application resources in data centers), increased security (to enable secure interconnection), or even optimization of uses (to enable distribution of the load and application resources between networks once interconnected). Currently, there is no specification for network federation other than the "Federated Mission Networking" (FMN) standard (or "creation of coalition operations networks"). However, this standard does not take into account the integration aspects that allow for dynamic consideration of the technical constraints linked to interconnectivity between two networks. Indeed, currently, network federation generally involves an upstream planning phase and / or a static (i.e. non-reconfigurable) configuration and / or one or more static radiocommunication equipment(s). Additionally, when adding a new network to an existing federated network (i.e. when extending an existing federated network to incorporate a new network), new configuration is required to configure at least part of the existing federated network, including declaring the identities of the new incoming network. For example, there are automatic addressing protocols such as RFC 3927 which allow network addressing on a predetermined range of local network addresses (169.254.0.0 / 16) and allow via an ARP address resolution protocol (from the English Address Resolution protocol) to announce and detect IP addressing conflicts (from the English Internet Protocol) on the network, and this without restriction of obtaining an address which is not very secure. The multicast protocol for dynamic configuration of DHCP hosts (Dynamic Host Configuration Protocol) requires a static server, which means that one of the networks to be federated must permanently (i.e. always) carry this protocol to federate the other networks (i.e. bubbles, for example a tactical bubble). As an alternative to the two aforementioned addressing protocols, cluster federation is also currently implemented via manual configuration or via a multicast configuration with a static multicast source. The current network federation based on upstream planning and static configuration including the identities of the networks to be federated, is thus based, in view of the current examples mentioned above, on static addressing already carried out or implemented by static equipment. These solutions therefore do not allow modification of the network deployment after planning, except to re-plan which is time-consuming, are inflexible and / or not very secure, complex to implement, sometimes manual (Le. implemented humanly) and therefore sources of error, and this with a substantial processing time likely to impact responsiveness. The aim of this invention is to solve this problem. To this end, the invention relates to a method for dynamically federating a plurality of radiocommunication networks, each network being represented by a node, the method comprising at least the following successive steps: - discovery between the network nodes of said plurality of networks, each node broadcasting a predetermined discovery message; - election of a master node from among the nodes associated respectively with each network of said plurality, the distinct nodes of said elected master node becoming slave nodes; - implementation of dynamic IP addressing by said master node; - construction of a federated network by said master node, by enrolling said slave nodes after their authentication by said master node; - establishing at least one communication plan between said master node and said slave nodes; - configuration of said federated network using said at least one communication plan. Thus, the method according to the present invention, by the specific sequence of the aforementioned steps, makes it possible to avoid planning and static configuration including identities of the networks to be federated in current solutions. Such a solution also makes it possible to dynamically interconnect networks without planning, without requiring one or more static devices to connect them. In other words, the solution allows several networks to be federated, i.e. to set up at least one communication plan between two network nodes, free, beforehand, from knowledge between them and a predetermined IP addressing configuration (i.e. the networks do not know each other beforehand). According to other advantageous aspects of the invention, the method for dynamically federating a plurality of networks comprises one or more of the following characteristics, taken individually or in all technically possible combinations: - said discovery step is implemented using an L2 type broadcast protocol; - said election step is implemented using a weight and / or an age previously associated in a predetermined manner with each network node of said plurality of networks; - said step of implementing dynamic IP addressing by said master node is carried out by means of a DHCP protocol implemented by said master node and activated during its prior election; - said construction step and / or said establishment step is / are implemented using a secure connection for the authentication of each slave node and / or for the assembly of said at least one plan respectively; - said authentication is implemented by means of a cookie; - said at least one plan belongs to the group comprising at least: - a control plan; - a data plan; - an orchestration plan; - a management plan; - a user plan; - the method further comprises a step of inter-node transitive connection according to which when two nodes are connected via said federated network, the connection of a new node to one of said two connected nodes causes the automatic connection of said new node to the other of said two connected nodes; - the method further comprises an inter-node context sharing step implemented after or during said step of establishing at least one communication plan between said master node and said slave nodes, said sharing accelerating an election of a next master node in the event of loss of the current master node, and accelerating the resumption of the services ensuring the federation by said next master node. The invention also relates to a computer program comprising software instructions which, when executed by a computer, implement a method of dynamic federation of a plurality of networks, as defined above. The invention also relates to a federated network via said method. These characteristics and advantages of the invention will appear more clearly on reading the description which follows, given solely by way of example, not limiting, and made with reference to the attached drawings, in which: - [Fig.1] [Fig.1] is a flowchart of a method for dynamically federating a plurality of radiocommunication networks according to the present invention; - [Fig.2] [Fig.2] schematically illustrates each step of the process of [Fig.1]; - [Fig.3] [Fig.3] illustrates the dynamic federation of a plurality of radiocommunication networks in the event of loss of a representative node of one of the networks of the federated network by means of the method according to the present invention and in the event of addition of a new node entering said federated network: - [Fig.4] [Fig.4] illustrates the interconnection of two separate federated networks via the method according to the present invention. Referring to [Fig.1], the method 10 of dynamic federation of a plurality of radiocommunication networks, each network being represented by a node, the method comprising at least the following successive steps: - discovery 12 between the network nodes of said plurality of networks, each node broadcasting a predetermined discovery message; - election 14 of a master node from among the nodes associated respectively with each network of said plurality, the distinct nodes of said elected master node becoming slave nodes: - implementation 16 of dynamic IP addressing by said master node; - construction 18 of a federated network by said master node, by enrolling said slave nodes after their authentication by said master node; - establishment 20 of at least one communication plan between said master node and said slave nodes; - configuration 22 of said federated network using said at least one communication plan. Each of the aforementioned steps of said method 10 is subsequently detailed in relation to [Fig.2], which illustrates by way of example the dynamic federation between three distinct networks R;, Ra and R3 each represented respectively by a node N,, N., and N,. In a manner not shown, each of the three distinct networks R;, Rz and Ra includes in particular a radio access network or RAN (from the English Radio Access Network), integrating a plurality of access points or eNB (e-Node B in 4G). An eNB allows a user equipment UE (from the English User Equipment) (i.e. a terminal) to connect, by means of a radio link, to the network in which it is integrated. Each of the three separate networks R;, Rz and R3 also includes a core network or EPC (Evolved Packet Core in 4G). Schematically, the EPC typically includes a gateway with the RAN, a gateway with the IP world and a mobility management entity MME (Mobility Management Unit). suitable for managing signaling between terminals. In addition, each of the three separate networks R,, Rz and R3 further comprises a private network or SDDC (Software-defined data center) which is a secure network comprising several virtualized VM services, a routing element (i.e. router type), a security element of firewall type for communication to the outside, for example according to the IP protocol. Furthermore, each of the three distinct networks R,, R; and R3 further comprises respectively within each node N,, N,, and N,, a service orchestrator supervising a network controller (in particular of the SDN type from the English Software Defined Network), a security controller and an IT controller. The network controller controls the EPC machines to configure them according to commands received from the service orchestrator. This involves, for example, configuring the routing of a certain type of data flow from an identified user to ensure bandwidth. The IT controller controls the SDDC machines to configure them according to commands received from the service orchestrator. This involves, for example, instantiating a specific virtual application based on an identified need, reserving a certain memory and computing resource for it, and possibly decommissioning it when it is no longer in use. The security controller controls the machines on its network, including the firewall, to configure them according to commands received from the service orchestrator. For example, this involves filtering the data traffic of an authenticated user based on the privileges associated with their profile. The service orchestrator therefore allows, through the controllers, the allocation of available physical and / or virtual resources according to the needs of the users (in accordance with their respective profiles). To do this, the service orchestrator includes a predetermined and detailed orchestration logic, the predetermined orchestration logic defining the election method, the distributed mode (i.e. inter-node message), the control of the elements, the services and configuration rules. Physical resources include, for example, SDDC computing capacity, SDDC working memory capacity, network bandwidth capacity, Network Address Translation (NAT) capacity through the firewall, etc. Virtual resources include, for example, virtualized bricks, each corresponding to an application enabling specific processing to be carried out on specific data, such as, for example, a deep packet inspection (DPI) application by the firewall, an image processing application by the SDDC, or an application for compressing a data stream via the network. In 4G architecture, the service orchestrator is called the Operations Support System (OSS). The first step 12 of discovery of the method 10 of dynamic federation of a plurality of radiocommunication networks of [Fig.1] is represented in [Fig.2], by the view also identified by the reference 12. In view 12 of [Fig.2], said discovery step 12 is implemented by each node N,, N,, and N3. More precisely, the network R, represented by the node N, is for example connected via a bidirectional link 24 to a federation interface 26 corresponding for example to a star connection, the network Rz represented by the node N, is for example also connected via a bidirectional link 28 to the federation interface 26, and the network R3 represented by the node N3 is for example connected via a bidirectional link 30 to the federation interface 26. In other words, each of the nodes N,, N», and N3 to be federated (i.e. federation of at least two nodes) are then connected via a link 24, 28 or 30, of the Ethernet or LOS (Line of Sighr) type to the federation interface 26. During said neighborhood discovery step 12, the node N,, of the network R,, via the link 24 broadcasts a predetermined discovery message 32 to any node representative of another network also connected to the federation interface 26, including the nodes N, and N3. The same applies to the node N,, of the network R>, via the link 28, and to the node N,, of the network R,, via the link 30. In other words, following the establishment of the physical links 24, 28 and 30, the discovery step 12 between the service orchestrators of the nodes N,, N, and N3 consists of the exchange of a broadcast presence message between service orchestrators, through the links 24, 28 and 30 and the federation interface 26, in particular by implementing the predetermined discovery protocol, for example as a function of a mission carried out within said network. In other words, the broadcasting of such a discovery message 32 amounts to the automatic search for neighboring distinct network nodes capable of receiving said predetermined discovery message from the node in question broadcasting said message. By definition, the nodes “in the neighborhood” of the node in question are capable of receiving the discovery message from said node in question. According to an optional embodiment, said discovery step is implemented using a L2 type broadcast protocol (i.e. on network layer 2) corresponding to the network layer used to transfer data between nodes N,, N, and N; of the federated extended network). The second step 14 of election of the method 10 of dynamic federation of a plurality of radiocommunication networks of [Fig.1] is represented in [Fig.2], by the view also identified by the reference 14. Said election step 14 aims to elect from among the nodes N,, N,, and N3, the master node M of the federated network resulting from the federation of the networks R,, R», and R; (i.e. bringing together the networks R,, R:, and R3). More precisely, said election aims to choose the master service orchestrator, namely as represented in [Fig.2], the orchestrator of the node N;, representing the network R3, from among the orchestrators of each of the nodes N,, N,, and N, the nodes, N, and N, in [Fig.2], distinct from said elected master node N3 M becoming slave nodes E. Such an election is in particular of the L2 type (i.e. carried out on the network layer 2 called L2 from the English Layer 2)) It should be noted that according to a particular optional aspect, said discovery message 32, represented in view 12 associated with the discovery step 12, comprises a set of elements necessary for said subsequent election step 14 of the method according to the invention. According to an optional variant, said election step is implemented using a weight and / or an age associated beforehand in a predetermined manner with each network node of said plurality of networks, and indicated for example within said discovery message. An example of such an election mechanism is notably described in application FR 2109229 filed on September 3, 2021 in the name of the Applicant. Following said election 14, each service orchestrator therefore stores an attribute taking the value “master” represented by M in [Fig.2], or “slave” represented by E in [Fig.2], when the radiocommunication network managed by the service orchestrator in question is part of a federated network, or “undefined” when said radiocommunication network is not part of a federated network as illustrated by N 1 N>, and N; in view 12 of [Fig.2]. The master service orchestrator M (e.g., of node N3) has a global view of the federated network. It is therefore able to manage telecommunication services that extend between the three networks R,, R;, and R3. The master M therefore maintains a global state of the federated network resulting from the federation of the networks R,, Ra, and Ra. It receives requests, applies rules and generates commands whose scope concerns the federated network as a whole. If commands concern services associated with the network that this master orchestrator managed before the federation (network R; in our example), these commands are transmitted to the local controllers for execution. If these commands concern services associated with the other networks that this master orchestrator did not manage before the federation (networks R, and Ra in our example), these commands are transmitted and broadcast (from the English “broadcast”) to the slave orchestrators E of nodes R, and R respectively of nodes N, and N. Service orchestrators that become slave orchestrators E of the federated network put themselves in a mode of listening for commands from the master orchestrator M. When they receive a command from the master orchestrator M, they relay it to one or other of the local controllers (those of network R, or of network R; in our example). The slave service orchestrator E thus operates under the supervision of the master orchestrator M. Furthermore, the slave orchestrator E sends local information to the master orchestrator M, allowing the master orchestrator M to maintain the part of the federated network state relating to this other network. The third step 16 of implementing IP addressing of the method 10 for dynamic federation of a plurality of radiocommunication networks of [Fig. 1] is represented in [Fig. 2], by the view also identified by the reference 16. The sequence of the discovery 12, election 14 and dynamic addressing 16 steps is specific to the present invention. The dynamic IP addressing 16 is notably implemented by the node N, previously elected master M. In other words, each node orchestrator carries with it a dynamic IP addressing service which is activated only once said node orchestrator is elected master, which makes it possible to avoid prior IP addressing configuration on one or more federation interface(s) 26, most often static and / or requiring a static addressing server (i.e. no IP addressing configuration is necessary on the federation interface(s) 26). In other words, via the present invention, the service corresponding to the IP addressing becomes virtual with activation conditioned on the prior election as master of the orchestrator implementing it, so that the element implementing IP addressing is no longer a static element of the network but a dynamic element capable of being modified by prior election of a master orchestrator of the network. Such virtual dynamic IP addressing, with activation conditioned by the prior election, as master, of the orchestrator implementing it, makes it possible to free oneself from the planning and static configuration including identities of the networks to be federated in current solutions. For example, such virtual dynamic IP addressing amounts to a dynamic allocation of IP addresses on the network layer 3 (i.e. L3 from the English layer 3). According to an additional optional variant, said step 16 of implementing dynamic IP addressing by said master node is carried out by means of a DHCP protocol implemented by said master node M and activated following its prior election 14. In other words, as illustrated by view 16 in the example of [Fig.2], the node N, master M, and more generally the network R; (represented by node N°), hosts (i.e. integrates) a DHCP addressing module 38. Via this DHCP addressing module 38, node N, master M grants itself an IP address on a federation interface and regularly broadcasts (i.e. broadcasts) its master identity 34. The other nodes N, and N; respectively representative of the networks R, and R; as slaves E broadcast according to the DHCP protocol (i.e. broadcast) according to arrow 36 a request for IP addressing to the node N, master M, and thereby each recover their IP addressing. The fourth step 18 of the method 10 for dynamically federating a plurality of radiocommunication networks of [Fig.1], relates to the construction of a federated network by said master node, and is represented in [Fig.2], by the view also identified by the reference 18. According to this step 18, the construction of the federated network is carried out by enrolling said slave nodes E after their authentication 40 with the master node M. In other words, the arrow 40 of the view 18 illustrates the fact that according to this step 18, each of the slave nodes E joins or forms the network cluster federated by the master node M in a secure manner via mutual authentication of the slave nodes E and master nodes M, in particular of type L2 (i.e. on the network layer 2). According to an additional optional variant, said step 18 of construction of the federated network uses a secure connection for the authentication of each slave node E. According to an additional optional variant, the authentication, necessary for said step 18 of construction of the federated network, is implemented by means of a cookie, also called a magic cookie, corresponding by definition to an identifier without comprehensible meaning transmitted to a program by another program, to allow it to accomplish one or more tasks in a secure manner, in particular by means of a secure connection via, for example, the SSL encryption security protocol (from the English Secure Sockets Layer). The fifth step 20 of the method 10 for dynamically federating a plurality of radiocommunication networks of [Fig. 1], relates to the establishment 20 of at least one communication plan between said master node M and said slave nodes E, and is represented in [Fig. 2], by the view also identified by the reference 20. According to an additional optional variant, said at least one plan belongs to the group comprising at least: - a control plan, corresponding in particular to an IP plan for signaling and inter-system messages intended for controlling equipment and the federated network; - a data plan corresponding in particular to a routing plan indicating where all the network data packets pass (i.e. between users, servers, etc.); - an orchestration plan, corresponding in particular to a control plan dedicated to orchestration (i.e. inter-orchestrators between the nodes N,, N,, and N3 of the networks R,, R3 and Rs respectively, such as for example an East-West orchestration plan); - a management plan corresponding in particular to an IP plan for network administration (i.e. for administrator connection, for example, using the secure communication protocol SSH (Secure Shell); - a user plan, corresponding in particular to a aforementioned data plan dedicated to users. According to an additional optional variant, during this fifth step 20 of establishing at least one communication plan between the master node M and said slave nodes E, the master node M ensures the assembly of the aforementioned inter-node plan(s) of the federated network using a secure connection via, for example, the SSL encryption security protocol (Secure Sockets Layer). According to an additional optional variant, the node N, master M of the network R3 as illustrated by [Fig.2] transmits the federation context 42 to each slave node N, and N3 of the networks R, and R3 parties to the federated network resulting from the federation of the networks R,, R>, and Ra. The federation context 42, according to the present invention, comprises all the information or technical data that the master node N; M has used to set up the federation (list of known and federated nodes (including their identifiers, their reference (i.e. node name) in the federated network (i.e. cluster), the federation state, ... the list of IP address allocations established via the DHCP addressing module 38, the security policies). As will be seen later, such sharing of the federation context 42 to all the network nodes participating in the federated network subsequently allows a slave node to become master in the event of failure of the current master, without loss of context. In other words, the method 10 according to the present invention further comprises a step of sharing inter-node context implemented after (according to a variant not shown in [Fig.2]) or during said step 20 of establishing at least one communication plan between said node N, master M and said slave nodes E, said sharing accelerating an election of a following master node in the event of loss of the current master node, and accelerating the resumption of the services ensuring the federation (in particular dynamic IP addressing by the following master node). View 20 of [Fig.2], illustrating step 20 of establishing at least one communication plan 44, also illustrates a complementary optional variant according to the present invention, namely, a step of inter-node transitive connection according to which when two nodes are connected via said federated network, the connection of a new node to one of said two connected nodes causes said new node to automatically connect to the other of said two connected nodes. Such a transitive connection is automatic, for example as illustrated in [Fig.2], between the nodes N, and N; slaves E, in particular because: - on the one hand the node N, slave E is connected to the node N, master M which exchanges with it the federation context 42 and establishes with it said at least one communication plan 44, preferably at least the secure control plan, and - on the other hand the slave node N3 E is connected to the master node N M which also exchanges with it the federation context 42 and establishes with it said at least one communication plan 44, preferably at least the secure control plan. The sixth step 22 of the method 10 for dynamically federating a plurality of radiocommunication networks of [Fig. 1], relates to the configuration 22 of said federated network using said at least one communication plan established previously in step 20, and such a configuration step 22 is represented in [Fig. 2], by the view also identified by the reference 22. This view 22 illustrates in particular the configuration 22 of the federated interconnection network between the participating networks R,, Ra and R3 by effective use of the plan(s) previously established during step 20, namely the control plan, the data plan, the orchestration plan, the management plan, the user plan. As indicated previously, the implementation of the method according to the present invention involves the establishment, upstream of the federation interface 26, of the network elements of an orchestrator 46, within each network node N,, N», and N4, namely the routing element (router type R), a security element (firewall type F), a control element 48 (i.e. network controller (in particular of the SDN type from the English Software Defined Network)) including an orchestration logic (election method, distributed mode (inter-node message), control of the elements, services and configuration rules) in order to configure the federated network resulting from the federation of the networks R,, Ra, and R; to effectively use the plan(s) previously established during step 20, in particular the data plane 50. More precisely, the configuration 22 of the federated interconnection network between the participating networks R,, R: and R; by effective use of the plan(s) previously established during step 20, consists of configuring, via the control element 48 (i.e. network controller (in particular of the SDN type from the English Software Defined Network M: - access (i.e. to configure the firewall(s) F of each network participating in the federated network), - routing (notably BGP from the English Border Gateway Protocol for the dissemination of the routes to be followed by each data packet). - setting up an encrypted connection, -etc. In other words, the step of configuring said federated network using said at least one communication plan makes it possible to finalize the establishment of secure communication “tunnels” between the nodes N,, N,, and N3 of the federated network (i.e. cluster), in particular on the network layer 3 (i.e. L3 from the English layer 3), said tunnels making it possible to guarantee all of the flows, their security, and the guarantee of inter-node services. [Fig.3] illustrates the dynamic federation of a plurality of radiocommunication networks in the event of loss of a representative node of one of the networks of the federated network by means of the method according to the present invention and in the event of addition of a new node entering said federated network. More precisely, in view 52 of [Fig.3], in comparison with views 12, 14, 16, 18 and 22 of [Fig.2], the node N, master M representative of the network R» has been lost. In this case of loss of node N, current master M associated with [Fig. 2], according to the method of the invention, the election step 14 is then immediately repeated, without reiteration of the method 10 as a whole (i.e. we go directly to the election step 14 without “re-discovery”), in particular thanks to the inter-node context sharing 42 implemented after or during said step 20 of establishing at least one communication plan illustrated by [Fig. 2]. According to the example of view 52 of [Fig. 3], using the previously shared context information, it is then the node N; of the network R, which then becomes master M of the federated network resulting from the remaining networks R, and Rs. Following this election, the DHCP addressing module 38 hosted by the master node N3 M, and more generally by the network Rs, is activated to resume dynamic IP addressing if necessary in the event of a new network entering the federated network, and the regular broadcast of its master identity. In addition, the master node N; M updates the context 42 (i.e. the security policies) and in a manner not shown shares this updated context to the entire federated network and in particular to the node N, of the network R,. Following this election of node N as master M, view 52 of [Fig.3] also illustrates the addition (i.e. the arrival) of a new node N, representative of the network R, which connects via a bidirectional link 54 to the federation interface 26. In other words, the new node N, representative of the network R4, will enter said federated network resulting from the remaining networks R, and R3, and to do this via the link 54 broadcasts a predetermined discovery message 32 to any node representative of another network also connected to the federation interface 26, including the nodes N, and Na. Thus, the new node N, representative of the network R, implements the aforementioned discovery step 12 by connecting via the link 54 to the federation interface 26, while the node N, master M regularly broadcasts its master identity 24. Following this broadcast, the new node N, becomes slave E as shown in view 15, and the node N; master M then implements the dynamic IP addressing 16 of the new node N, slave E, the node N, slave E broadcasting, in particular according to the DHCP protocol (i.e. broadcast), a request for IP addressing to the node N; master M, and thereby recovering its IP addressing. Then, as described previously, the orchestrator of the master node N; M implements the steps 18 of construction of the federated network by authentication, 20 of establishment of communication plans, in particular of the control plan 44 and of the data plan 50, 22 of configuration mentioned above to federate the new federated network resulting from the federation of the networks R,, R; and Ru. [Fig.4] illustrates the interconnection of two separate federated networks via the method according to the present invention. More specifically, in view 58 of [Fig.4], a first federated network is illustrated, resulting from the federation of the aforementioned Ryet R3 networks and illustrated previously in relation to figures 2 and 3, and is federated by the node N; master M. A second federated network is illustrated, resulting from the federation of the networks Rset Re, and is federated by the node N, master M of the network R. In the second federated network, resulting from the federation of the networks Rset Re, the node N, slave E is connected to the federation interface 60 via a bidirectional link 62, while the Ns master M of the network Rçy is connected via the link 64. The two connection interfaces 26 of the first federated network and 60 of the second federated network are connected (i.e. connected) by a link 66. Additionally, previously, within the second federated network the control plane 68 and data plane 70 were established and configured as previously described. View 72 illustrates the result of a new election step 14 implemented following the connection 66 of the two federated networks, in order to elect the master node from among the nodes N; of the network Rs and the node N, of the network Rs. In particular, according to the example illustrated by view 72 of [Fig.4], it is the node N, of the network R, which is elected master M of the extended federated network resulting from the networks R;, R3, Rs and Re, while the node N5 of the network Rs then becomes slave E. View 74 illustrates the federation of the extended federated network resulting from the networks R;, R3, Rs and Rs after implementation of the steps 18 of construction of the federated network by authentication, 20 of establishment of communication plans, in particular of the control plane 44 and of the data plane 50, 22 of configuration mentioned above, and application of transitive inter-node connections. Those skilled in the art will understand that the invention is not limited to the embodiments described, nor to the particular examples of the description, the embodiments and variants mentioned above being suitable for being combined with each other to generate new embodiments of the invention. The present invention thus proposes a mechanism for discovering neighborhoods and electing a master node from several network nodes to be federated. According to the present invention and in particular the specific sequence of the aforementioned steps, the master node can reposition itself on another node following the change in operational context (e.g. loss of node), and takes with it a dynamic IP addressing service which is then virtual and is no longer a static element of the network. This provides a minimum configuration while including the security elements necessary for secure and dynamic network interconnection. The solution according to the present invention thus makes it possible to federate, that is to say to set up in particular an authorized data and control plane between at least two distinct network nodes, free from prior knowledge between them and an absence of IP address configuration on their federation link. Advantageously, this solution can thus be used as a "plug and play" solution without planning and / or re-planning, based on neighborhood discovery (Ethernet or LOS) and authentication between neighbors, which in particular allows deployment of a federated network of the WAN (Wide Area Network) type, or MAN (Metropolitan Area Network) type, or LAN (Local Area Network) type, because it is easy to implement. Thus, such a dynamic network federation process facilitates the triggering in the inter-network operational exchange, for example for the sharing of telecom contexts in the ephemeral tactical bubble federation allowing mobility and offering flexibility to adapt to a tactical maneuver rhythm. In other words, the present invention makes it possible to reconcile distributed systems (i.e. cluster type) and the reactivity of programmable networks (in particular with a network controller (in particular of the SDN type from the English Software Defined Network) with security requirements.

Claims

Claims

1. Method (10) for dynamically federating a plurality of ra- diocommunication, each network being represented by a node, the method comprising at least the following successive steps: - discovery (12) between the network nodes of said plurality of networks, each node broadcasting a predetermined discovery message; - election (14) of a master node among the associated nodes respec- tively to each network of said plurality, the distinct nodes of said elected master node becoming slave nodes: - implementation (16) of dynamic IP addressing by said node master : - construction (18) of a federated network by said master node, by en- role of said slave nodes after their authentication by said master node; - establishment (20) of at least one communication plan between said master node and said slave nodes; - configuration (22) of said federated network using said at least one communication plan.

2. Method (10) for dynamically federating a plurality of ra- diocommunication according to claim 1, wherein said step of discovery is implemented using a broadcast protocol type L2.

3. Method (10) for dynamically federating a plurality of ra- diocommunication according to claim 1 or 2, wherein said election step is implemented using weight and / or age pre-associated in a predetermined manner with each network node of said plurality of networks.

4. Method (10) for dynamically federating a plurality of ra- diocommunication according to any one of the preceding claims- preceding, wherein said step of implementing IP addressing dynamic by said master node is carried out by means of a protocol DHCP implemented by said master node and activated upon its election prior.

5. Method (10) for dynamically federating a plurality of ra- diocommunication according to any one of the preceding claims- preceding, wherein said construction step and / or said step establishment is / are implemented using a connection secure for the authentication of each slave node and / or for the assembly of said at least one plane respectively.

6. Method (10) for dynamically federating a plurality of ra- diocommunication according to any one of the preceding claims- preceding in which said authentication is implemented by means of a cookie.

7. Method (10) for dynamically federating a plurality of ra- diocommunication according to any one of the preceding claims- preceding, in which said at least one plan belongs to the group including at least: - a control plan; - a data plan; - an orchestration plan; - a management plan; - a user plan.

8. Method (10) for dynamically federating a plurality of ra- diocommunication according to any one of the preceding claims- preceding, further comprising a step of inter-transitive connection node according to which when two nodes are connected via said network federated, the connection of a new node to one of said two nodes connected causes the automatic connection of said new node to the other of said two connected nodes.

9. A method (10) of radio communication according to any one of the claims- preceding indications, further comprising a step of sharing inter-node context implemented after or during said step establishing at least one communication plan between said node master and said slave nodes, said sharing speeding up an election of a next master node in case of loss of the current master node, and ac- accelerating the resumption of services ensuring federation by said node next master.

10. A computer program comprising software instructions which, when executed by a computer, implement a radiocommunication method according to any one of the re- previous instructions.