Method for processing a request for statistical or predictive analysis, communication method and application entities capable of implementing these methods

Abstract

The invention relates to a method for processing a request (REQ) for statistical or predictive analysis received by a first application entity of a communications network originating from a second application entity of the network, this method comprising: - a step (F30) of determining, from at least one item of information (P-CTX, P-DEF) conveyed by the request, a context (CTX) in which the request was formulated by the second application entity; - a step (F50) of performing the requested statistical or predictive analysis by means of an analysis model (MOD(CTX)) selected (F40) according to the determined context; and - a step (F60) of providing at least one result (RES) of the analysis performed in response to the request from the second application entity.

Description

Description Title of the invention: Method for processing a statistical or predictive analysis request, communication method and application entities capable of implementing these methods Prior art

[0001] The invention belongs to the general field of telecommunications.

[0002] It concerns more particularly communication networks based on a plurality of application entities implementing various functionalities or services in the network, such as for example a 5G core network (or "5GC" for "5G Core network" in English) as defined by the 3GPP standard. Such application entities are for example devices (hardware or software) hosting network functions (or NF for "Network Functions" in English) implementing functionalities such as network access, user mobility or even the management of sessions established in the network, the storage and publication of network function profiles, etc.

[0003] In order to optimize procedures within the 5G core network, NF functions may call upon a specific NF function responsible for collecting and analyzing network data, called the NWDAF (NetWork Data Analytics Function). The NWDAF provides the network NF functions that request it with statistical and / or predictive analyses on the behavior of the network, in particular in terms of quality of service and / or the behavior of user equipment (or UE for "User Equipment" in English). The analyses carried out may be global, i.e. established at the level of the network, a server, an application or even a region (e.g. network resource load rate at a particular time of day or year, average quality of service, number of users connected to the network or active sessions, etc.), or be individual, i.e. relating to a particular UE or group of UEs (e.g.future location of a UE, volume of a future communication session of a UE, etc.). The analyses are carried out on the basis of raw data that the NWDAF function collects from other NF functions of the network and / or from nodes of the radio access network via the network management entity in charge of operations, administration and maintenance, also known as the OAM entity (for "Operations, Administration and Maintenance" in English), and to which the NWDAF function applies one or more analysis models depending on the analyses requested of it. Such analysis models are, for example, statistical analysis models such as a moving average, an exponential moving average, etc., or predictive analysis models using in particular supervised or unsupervised learning technologies.

[0004] Once established, the statistical and / or predictive analyses typically make it possible to implement corrective modifications to the network parameters in advance in order to optimize its operation. More specifically, the entities using these analyses, in other words the NF functions that are clients of the NWDAF function (which may or may not be distinct from the NF functions that collected and provided the raw data to the NWDAF function), are able, based on the statistical and / or predictive analyses received from the NWDAF function, to adapt their behavior in order to optimize the operation of the network and the quality of service delivered to each user on their UE.The client NF functions of the NWDAF function are for example an AMF (for "Access and Mobility management Function" in English) for access management, an SMF (for "Session Management Function" in English) for session management, a PCF (for "Policy Control Function" in English) for policy control, etc.

[0005] 3GPP TR 23.791, "Technical Specification Group Services and System Aspects; Study of Enablers for Network Automation for 5G (Release 16)", V16.2.0, June 2019, discusses various use cases for such statistical and / or predictive analyses in a 5G network.

[0006] Thus, for example, UE mobility predictions can be used by the AMF function to optimize UE mobility management, and in particular the determination of their registration area (or RA), this registration area making it possible to locate UEs in standby and to send them solicitation messages (or "paging" in English) when data intended for them reaches the network, etc. This use case is also described in the 3GPP TS 23.501 document, entitled "Technical Specification Group Services and System aspects; System architecture for the 5G system (5GS); Stage 2 (Release 17)", V17.4.0, March 2022, in particular in paragraph 5.3.2.

[0007] According to another example also described in 3GPP TS 23.501 in section 6.3.3.3, it may be useful for the SMF to have statistics or predictions on network traffic (e.g. load) when selecting a user plane UPF (User Plane Function) to carry data from PDU (Packet Data Unit) sessions.

[0008] According to yet another example described in particular in 3GPP TS 23.503, entitled "Technical Specification Group Services and System aspects; Policy and charging control framework for the 5G system (5GS); Stage 2 (Release 17)", V17.4.0, March 2022, in particular in sections 4.2.3 and 6.1.1.3, a PCF function may request statistical dispersion analyses to modify policies or to determine the average throughput for a slice of a network. It should also be noted that the same statistical and / or predictive analysis may be required to support different network NFs in different contexts and / or to implement various functionalities.

[0009] It is therefore clear, given the numerous use cases of statistical and / or predictive analyses delivered by the NWDAF function and their importance in the operational functioning of the network, that these analyses must be precise and relevant. Statement of the invention

[0010] The invention responds in particular to this need by proposing a method for processing a statistical or predictive analysis request received by a first application entity of a communications network from a second application entity of the network, this method comprising: a step of carrying out said required statistical or predictive analysis by means of an analysis model selected as a function of a context of formulation of said request by said second application entity, determined from at least one piece of information conveyed by the request; and a step of providing at least one result of the analysis carried out in response to the request from the second application entity.

[0011] Correlatively, the invention also relates to an application entity of a communications network, called the first application entity, comprising: a reception module configured to receive from a second application entity of the network, a statistical or predictive analysis request; a processing module, configured to carry out said analysis by means of an analysis model selected as a function of a context of formulation of said request by said second application entity, determined from at least one piece of information conveyed by said request; and a supply module, configured to supply at least one result of the analysis carried out in response to the request from the second application entity.

[0012] As mentioned above, the invention has a preferred but non-limiting application in the context of a 5G core network. Thus, for example, the first application entity and the second application entity host network functions, the network function hosted by the first application entity being a function for collecting and analyzing network data.

[0013] It should be noted, however, that while the need for statistical and / or predictive analyses has been formulated in the context of a 5G network, such a need may nevertheless arise in other situations, in particular in other communication networks (for example, proprietary networks), between application entities other than an NF function and an NWDAF function, etc. By application entity, we mean here any type of communicating device, hardware or virtual (i.e. software), configured to implement a specific processing logic, such as, for example, a device offering and / or consuming services in a network such as a network function (NF) or a network function instance of a core network compliant with the 3GPP standard, but also a router management unit or an SDN (Software Defined Network) controller in an IP (Internet Protocol) network, etc.

[0014] Thus, the invention proposes to improve the quality and precision of the statistical and predictive analyses required by an application entity (second application entity within the meaning of the invention) from another application entity (first application entity within the meaning of the invention) by taking into account the context in which this request is formulated, i.e. the context of use of the requested predictive and / or statistical analysis. Such a formulation context (or of use of the required analysis) can typically comprise the reason for this request, i.e. the destination of the analysis result or the use case for which the statistical and / or predictive analysis is required. Thus, in a particular embodiment, the formulation context identifies a procedure intended to be executed by the second application entity in the network using the analysis result provided by the first application entity (e.g.registration of a UE, opening of a data session, etc.).

[0015] The inventors have indeed noted that, in a 5G core network, each network procedure that calls for statistical and / or predictive analyses to be performed (e.g., registration, login, optimization of a user profile, etc.) constitutes a specific use case for which it is desirable that the NWDAF function (first application entity within the meaning of the invention) delivers the most relevant result in order to optimize this procedure. Furthermore, as mentioned previously, the same statistical and / or predictive analysis may be required by the same application entity or by separate application entities during two different procedures, in other words for different uses (e.g., long-term needs vs. short-term needs). However, to date, to perform the analysis requested of it, the NWDAF function relies solely on the parameters provided in the request that strictly define the analysis that is required (e.g.,type of analysis expected and its level of precision, the target of the analysis, etc.) and on the raw data it collects from various entities in the network. The analysis model used by the NWDAF function may lose efficiency by having to respond to a plurality of different use cases.

[0016] Knowledge of the query formulation context and in particular the motivation for which this query was made, constitutes valuable information for the first application entity to improve the accuracy and relevance of the statistical and / or predictive analysis that it performs. Indeed, this information offers the first application entity the possibility of selecting in a perfectly adapted manner the analysis model that will deliver the best statistical and / or predictive analysis for a given application situation, in other words for a given use case. For example, for an AMF function, knowing that an analysis is used for an update of a recording zone or to locate a UE (also more commonly referred to by the English term “paging”) can lead to choosing very different analysis models in terms of mathematical technique and / or adjustment parameters.To better understand the contribution of the invention, we can make an analogy with everyday life and the situation of a doctor who has to carry out analyses on a patient without having information about the patient (e.g. their age, whether they are a smoker or not, etc.) or the reason why these analyses are necessary (e.g. the presence of pain).

[0017] The formulation context obtained therefore constitutes external assistance which helps guide choices. made by the first application entity to carry out the analysis requested of it and to take adequate measures to obtain a relevant result for this analysis.

[0018] Different ways can be considered to obtain the formulation context.

[0019] Thus, in a particular embodiment, the formulation context is indicated by a parameter inserted by the second application entity in a dedicated field of the request.

[0020] In other words, the formulation context is explicitly provided by the second application entity, in a specific parameter, when it formulates its request. Such a dedicated parameter is for example named UseCaseContext. Different formats can be considered for this dedicated parameter, such as an integer, a string, a list of integers or strings, etc. The parameter values ​​can be predefined or not.

[0021] This embodiment is particularly simple to implement and allows the first application entity to obtain reliable and precise information on the context of the formulation of the request.

[0022] Alternatively, the formulation context may be determined by the first application entity from a plurality of parameters defining the analysis required by the second application entity and provided in the request. Such parameters are taken for example from: an identity of the second application entity; a parameter representative of a type of analysis required (e.g. type and quantity of results obtained, etc.); at least one piece of equipment on which the analysis relates; a parameter representative of a condition for carrying out the analysis (e.g. geographical area, time period, network slice concerned, etc.); an expected level of precision for the analysis; a parameter representative of a method of notification of said at least one result of the analysis (e.g. periodicity, subscription, etc.).

[0023] Certain groups of parameters provided in the request and the values ​​associated with these groups of parameters can indeed be considered as signatures representing distinct contexts. Different possibilities can then be envisaged: the first application entity is configured with a certain number of signatures to which contexts are associated. Upon receiving the request, the first application entity determines from the parameters provided in the request whether they correspond to one of the signatures available to it. If so, the context for formulating the request is that which is associated with the signature; the first application entity implements a learning phase during which it analyzes the parameters present in the requests received and determines, for example by means of a classification algorithm, which groups of parameters and associated values ​​are significant and representative of a particular context.The different groups of parameters and associated values ​​thus obtained define a set of signatures corresponding to distinct formulation contexts. It should be noted that in this case, the first application entity is able to distinguish between distinct formulation contexts, and to associate a given formulation context with a query, but it does not necessarily have the information necessary to associate with this formulation context a specific operation executed or intended to be executed by the second application entity (e.g. determination of a recording zone).

[0024] It should also be noted that some of the parameters provided in the query may present a certain variability for a given use case; therefore, it is important to carefully select the parameters allowing the first application entity to deduce the context of formulation of the query according to the different use cases.

[0025] In another variant embodiment, it can be envisaged that the formulation context provided by the second application entity in a specific field of the query or refined by the first application entity by exploiting the parameters conveyed by the query defining the analysis required by the second application entity. The formulation context determined by the first application entity then includes the context provided by the second application entity (e.g. the procedure within the framework of which the query is formulated and the analysis result will be exploited) possibly supplemented by other information from the analysis definition parameters (e.g. UE or a group of UEs or a cell, etc.).

[0026] As mentioned previously, in accordance with the invention, the context of formulation of the request is used by the first application entity to select a relevant analysis model to carry out the statistical and / or predictive analysis requested of it.

[0027] The selection of the analysis model can be carried out at different levels. Thus, it can include, for example, a selection of at least one element from among: a statistical or predictive analysis technique to be used to carry out the analysis; at least one characteristic of a data set to be used to carry out the analysis and / or train an analysis technique used to carry out the analysis; a parameterization of an analysis technique used to carry out the analysis; etc.

[0028] Adjusting all or part of the aforementioned elements allows the first application entity to obtain analysis results that are highly relevant to the second application entity, which allows the latter to optimize the procedures that it implements within the framework of the functions entrusted to it in the network.

[0029] To enable the selection by the first application entity of an analysis model adapted to the query formulation context, different ways of proceeding can be considered.

[0030] Thus, in a particular embodiment, the processing method comprises a preliminary step of configuring the first application entity with a list of analysis models in which each analysis model is associated with at least one context for formulating a query, the analysis model applied during the implementation step being an analysis model associated in this list with the context determined during the obtaining step.

[0031] This embodiment consists of a pre-configuration of the first application entity, for example by the network operator, with a list of possible models adapted to different contexts. It is relatively simple to implement. It should be noted that an update of the list thus preconfigured can be carried out at any time via an appropriate configuration message sent to the first application entity.

[0032] In another embodiment, the processing method comprises a learning step during which the first application entity: establishes correlations between analysis models, query formulation contexts, and quality indicators of the analyses carried out from said models in said contexts; and associates, from the established correlations, with at least one query formulation context an analysis model optimizing said quality, the analysis model applied during the execution step being selected from the associations established during the learning step.

[0033] This embodiment has a preferred application when the first application entity does not have a pre-configuration for a given formulation context, or in general, does not have any pre-configuration. In this embodiment, it is the first application entity which optimizes, by learning, the selection of an analysis model according to the formulation context of the analysis request addressed to it. This embodiment is advantageously scalable and allows the first application entity to take into account the quality of the analysis results it obtains to optimize the choice of analysis models.

[0034] In yet another embodiment, the treatment method comprises: - a preliminary step of configuring the first application entity with a list comprising at least one analysis model associated with at least one query formulation context; a learning step during which the first application entity: o establishes correlations between analysis models, query formulation contexts, and quality indicators of the statistical analyses carried out from said models in said contexts; and o updates and / or completes said list from said established correlations; the analysis model applied during the production step being an analysis model associated with the context determined during the obtaining step in the list updated or completed during the learning step.

[0035] This embodiment advantageously combines the two previous ones, and thus benefits from the advantages associated with each of them.

[0036] In view of the above, the invention relies on the first application entity which obtains a formulation context of the analysis request addressed to it and exploits this context to optimize the analysis results which it provides in response to this request, but also, in a particular embodiment, on the second application entity at the origin of the request and which provides the first application entity with said formulation context.

[0037] Thus, according to another aspect, the invention relates to a method of communication with a first application entity of a communications network, this communication method being implemented by a second application entity of the network and comprising: a step of sending to the first application entity a statistical or predictive analysis request comprising, in a dedicated field of the request, a parameter indicating a context for formulating the request; and a step of receiving, in response to the request, at least one result of the required analysis carried out by the first application entity by means of an analysis model selected as a function of this context.

[0038] Correlatively, the invention also relates to an application entity of a communications network, called a second application entity, comprising: a sending module, configured to send to a first application entity of the network a statistical or predictive analysis request comprising, in a dedicated field of the request, a parameter indicating a context for formulating the request; and a receiving module, configured to receive in response to said request, at least one result of the required analysis carried out by the first application entity by means of an analysis model selected according to this context.

[0039] According to yet another aspect, the invention also relates to a system in a communications network comprising at least one first application entity and at least one second application entity in accordance with the invention.

[0040] The communication method, the second application entity and the system according to the invention benefit from the same advantages cited previously as the processing method and the first application entity.

[0041] In a particular embodiment, the processing and communication methods are implemented by a computer.

[0042] The invention also relates to a computer program on a recording medium, this program being capable of being implemented in a computer or more generally in a first application entity in accordance with the invention and comprising instructions adapted to the implementation of a processing method as described above.

[0043] The invention also relates to a computer program on a recording medium, this program being capable of being implemented in a computer or more generally in a second application entity in accordance with the invention and comprising instructions adapted to the implementation of a communication method as described above.

[0044] Each of these programs may use any programming language, and may be in the form of source code, object code, or code intermediate between source code and object code, such as in a partially compiled form, or in any other desirable form.

[0045] The invention also relates to an information medium or a recording medium readable by a computer, and comprising instructions of a computer program as mentioned above.

[0046] The information or recording medium may be any entity or device capable of storing programs. For example, the medium may include a storage medium, such as a ROM, for example a CD ROM or a microelectronic circuit ROM, or a magnetic recording medium, for example a hard disk, or a flash memory.

[0047] On the other hand, the information or recording medium may be a transmissible medium such as an electrical or optical signal, which may be conveyed via an electrical or optical cable, by radio link, by wireless optical link or by other means.

[0048] The program according to the invention can in particular be downloaded from an Internet-type network.

[0049] Alternatively, the information or recording medium may be an integrated circuit in which a program is incorporated, the circuit being adapted to execute or to be used in the execution of the management, recording and communication methods according to the invention.

[0050] It is also possible to envisage, in other embodiments, that the processing and communication methods, the first and second application entities and the system according to the invention have in combination all or part of the aforementioned characteristics. Brief description of the drawings

[0051] Other characteristics and advantages of the present invention will emerge from the description given below, with reference to the appended drawings which illustrate an exemplary embodiment thereof without any limiting character. In the figures: [Fig. 1] Figure 1 represents, in its environment, a system in a network according to the invention, in a particular embodiment; [Fig. 2] Figure 2 schematically represents the hardware architecture of a computer capable of hosting any of the entities according to the invention belonging to the system of Figure 1; [Fig. 3] Figure 3 represents the functional modules of the application entities in accordance with the invention of the system of Figure 1; [Fig. 4] Figure 4 represents the main steps of a processing method according to the invention as implemented by a first application entity of the system of Figure 1; [Fig. 5] Figure 5 represents the main steps of a communication method according to the invention as implemented by a second application entity of the system of Figure 1. Description of the invention

[0052] Figure 1 represents, in its environment, a system 1 in a communications network CN, in accordance with the invention, in a particular embodiment.

[0053] In the example envisaged in Figure 1, the CN network is a 5GC core network of a 5G communications NW network as defined by the 3GPP standard, relying on a plurality of application entities hosting network functions (or NF functions) implementing various functionalities or services in the core network, such as network access, user mobility or even the management of sessions established in the network, the storage and publication of network function profiles, etc.

[0054] As mentioned above, in order to optimize the procedures within the CN core network, the NF functions may call upon a specific NF function responsible for collecting and analyzing network data, called the NWDAF function (for “NetWork Data Analytics Function”). This NWDAF function offers to the NF functions of the network that request it, statistical and / or predictive analyses on the behavior of the CN network, or more generally of the NW network, in particular in terms of quality of service, and / or on the behavior of the UEs of the NW network. These analyses may be global (e.g. established at the level of the network, a server, an application or even a region), or be individual (i.e. relating to a particular UE or group of UEs). They are carried out from raw data collected by the NWDAF function and statistical and / or predictive analysis models applied to this raw data.

[0055] In accordance with the invention, the system 1 comprises: at least one first application entity 2, in accordance with the invention. In the embodiment described here, the application entity 2 hosts a network function NWDAF for collecting and analyzing data from the NW network as described above, and is designated in the remainder of the description by entity NWDAF 2; and at least one second application entity 3, in accordance with the invention, configured to consume the services offered by the entity NWDAF 2, in other words, to request the entity NWDAF 2 to obtain statistical and / or predictive analyses on the behavior of the network and / or of a UE or a group of UEs. In the embodiment described here, the application entity(ies) 3 also host NF functions of the core network CN. No limitation is attached to the NF functions hosted by the application entities 3.Thus, in the example shown in Figure 1, the system 1 comprises a plurality of application entities 3 in accordance with the invention, including in particular an application entity 3-1 hosting an AMF network access function (also referred to hereinafter as AMF entity 3-1), an application entity 3-2 hosting a session management SMF function (or SMF entity 3-2), an application entity 3-3 hosting a policy control PCF function (or PCF entity 3-3), and an application entity 3-4 hosting an application AF function (or AF entity 3-4). This list is not limiting and is given for illustrative purposes only.

[0056] Each application entity of the system 1 is a communicating device, hardware or virtual (i.e. software), configured to implement a specific processing logic, corresponding to the network function that it hosts. Thus, the invention applies equally to software or virtual application entities and to hardware application entities hosting network functions.

[0057] When the application entities considered are hardware devices (for example servers of an infrastructure of the CN core network), they then have the hardware architecture of a computer 4, as illustrated in Figure 2. This hardware architecture is based on a processor PROC, a random access memory MEM, a read only memory ROM, a non-volatile memory NVM, and means COM for communications with other entities, for example, other application entities of the CN core network, entities of an access network to the CN core network such as for example an OAM entity for managing the NW network (not shown in Figure 1), or even with UEs of the NW network.These COM means of communication may in particular rely on a wired or wireless communication interface, known per se and not described in more detail here, but also on one or more software interfaces such as an application programming interface (or API for "Application Programming Interface" in English) or a point-to-point communication interface.

[0058] When the application entities considered are software, they are themselves hosted by a hardware device having the hardware architecture of the computer 4, and can then rely on the hardware resources of the computer 4 mentioned above (PROC, MEM, ROM, NVM, COM). In the remainder of the description, for the sake of simplification, reference is made indifferently to the PROC, MEM, ROM, NVM, COM resources of the application entity considered, whether it is hardware or software.

[0059] The non-volatile memory NVM of the computer 4 constitutes a recording medium in accordance with the invention, readable by the processor PROC and on which a computer program in accordance with the invention is recorded. In the case of an application entity 2, this computer program is referenced by PROG2 and comprises instructions defining the main steps of a processing method according to the invention. For an application entity 3, the computer program in question is referenced by PROG3 and comprises instructions defining the main steps of a communication method according to the invention.

[0060] The program PROG2 defines the functional modules of an application entity 2 and in particular of the entity NWDAF 2 of the system 1, which rely on or control the elements PROC, MEM, ROM, NVM, and COM of the computer 4, cited previously. These functional modules comprise in particular, in the embodiment described here, as illustrated in FIG. 3: a reception module 2A configured to receive from an application entity 3 of the system 1 a statistical or predictive analysis request REQ; a determination module 2B, configured to obtain from at least one piece of information conveyed by this request REQ, a context CTX for formulating the request by the application entity 3.Such a formulation CTX context is typically representative of the use case of the required statistical or predictive analysis, it identifies for example a procedure that the application entity 3 will execute using the result of the requested statistical or predictive analysis; a processing module 2C, configured to carry out the required analysis by means of an analysis model selected according to the CTX context determined by the determination module 2B; and a supply module 2D, configured to supply at least one result of the analysis carried out by the processing module 2C in response to the request REQ of the application entity 3.

[0061] The configuration and operation of the modules 2A to 2D of the application entity 2 and more particularly of the NWDAF entity 2 are described in more detail later with reference to the steps of the processing method according to the invention.

[0062] Similarly, the program PROG3 defines the functional modules of an application entity 3, and more particularly of each of the entities AMF 3-1, SMF 3-2, PCF 3-3 and AF 3-4 of the system 1, these functional modules relying on or controlling the PROC, MEM, ROM, NVM, and COM elements of the computer 4, cited previously. In the embodiment described here, the functional modules of an application entity 3 according to the invention comprise in particular, as illustrated in FIG. 3: a sending module 3A, configured to send to the application entity 2, and more particularly here to the entity NWDAF 2, a statistical or predictive analysis request REQ. In the embodiment described here, the sending module 3A is configured to use, to send the request REQ, an API of the entity NWDAF 2, as detailed further later.It is further configured to insert into the REQ request, in a dedicated field of the request named for illustration purposes UseCaseContext, a parameter indicating the CTX context of formulation of the request, in other words the context of use of the result(s) of the required analysis; a reception module 3B, configured to receive in response to the REQ request addressed to the NWDAF entity 2, at least one result of the required analysis carried out by the NWDAF entity 2 by means of an analysis model selected by the latter according to the CTX context; and an execution module 3C, configured to execute a procedure associated with the network function hosted by the application entity 3 using an analysis result received by the reception module 3B. The procedure executed of course depends on the network function hosted by the application entity 3.This may be, for example, a registration procedure or determination of a registration area for the AMF entity 3-1, a login procedure or selection of a UPF function of the user plane of the CN core network for the SMF entity 3-2. of a procedure for modifying a policy or determining an average throughput for a given network slice for a PCF 3-3 entity, of a procedure for optimizing a setting for a UE or detecting fraud for an AF 3-4 entity, etc.

[0063] The configuration and operation of the modules 3A to 3C of the application entities 3 are described in more detail later with reference to the steps of the communication method according to the invention.

[0064] We will now describe, with reference to figures 4 and 5 respectively, the main steps of a processing method and a communication method according to the invention, as implemented in a particular embodiment by the NWDAF entity 2 and by one of the application entities 3 of the system 1. We will consider more particularly here, by way of illustration, the AMF entity 3-1; however, the steps of the communication method are implemented in a similar or identical manner by any other application entity 3 of the system 1.

[0065] It is therefore assumed here that the AMF entity 3-1 needs, in order to execute one of the PROC procedures for which it is responsible in the CN core network, statistical and / or predictive analyses. More particularly, in the illustrative example envisaged here, it is assumed that the AMF entity 3-1 needs a prediction of the mobility of a given UE, for example of the UE 5 shown in FIG. 1, in order to determine a registration zone for this UE 5 (PROC procedure = determination of the registration zone of the UE 5).

[0066] Of course, these assumptions are not limiting in themselves and are formulated solely for illustrative purposes. Other procedures may be implemented by the AMF 3-1 entity within the framework of the network function implemented by this entity, and require or rely on statistical and / or predictive analyses of the NWDAF 2 entity, which may also relate to other types of analyses than a prediction of mobility of a UE, such as, for example, a prediction concerning the nature of communications of a UE, etc.). Thus, the 3GPP TR 23.791 document mentioned above discusses different use cases of statistical and / or predictive analyses carried out by an NWDAF network function in a 5G network.

[0067] With reference to Figure 5, the AMF entity 3-1, via its sending module 3A, sends a request REQ to the NWDAF entity 2, and more particularly to its receiving module 2A, requesting a prediction of the mobility of the UE 5 (step E10).

[0068] In the embodiment described here, the sending module 3A of the AMF entity 3-1 uses for this purpose the Nnwdaf_AnalyticsInfo service of the API of the NWDAF entity 2, described in the documents 3GPP TS 23.288 entitled “Architecture Enhancements for 5G system (5GS) to support network data analytics services (Release 17)”, V17.4.0, March 2022, and TS 28.520 entitled “Technical Specification Group Core Network and Terminals; 5G System; Network Data Analytics Services; Stage 3; (Release 17)”, V17.6.0, March 2022. The REQ request is here a simple request of type Nnwdaf_AnalyticsInfo Request. Alternatively, it can take the form of a subscription (Nnwdaf_AnalyticsSubscription service and Nnwdaf_AnalyticsSubscription Subscribe subscription), as described in 3GPP TS 23.288 and TS 29.520.It should also be noted that the sending of the REQ request by the sending module 3A, although linked to a procedure implemented by the AMF 3-1 entity (and more particularly executed or intended to be executed by its execution module 3C), is not necessarily carried out synchronously with this procedure. It can be carried out asynchronously, typically upstream of the procedure in question, in advance, in order to have the result of the analysis available to execute the procedure concerned without delay.

[0069] As defined by the 3GPP standard, the sending module 3A of the AMF 3-1 entity provides in the REQ request various P-DEF parameters defining the statistical and / or predictive analysis requested, and in particular: the identity of the application entity at the origin of the request, i.e. here the identity of the AMF 3-1 entity; one or more parameters representative of the type of analysis requested. This or these parameters may in particular define the nature of the analysis or equivalently the type of expected result (e.g. mobility prediction here), the quantity of results expected, etc.; the target of the analysis, i.e. the equipment(s) on which the analysis is carried out (e.g. EU 5 here); one or more parameters representative of the conditions for carrying out the analysis, for example the time period targeted, the geographical area targeted, the network slice concerned, etc.; a level of precision expected for the analysis; one or more parameters representative of the notification methods of the analysis network (e.g. periodicity, subscription, etc.).

[0070] Of course, this list is not exhaustive, and the person skilled in the art is invited to refer to the aforementioned 3GPP TS 23.288 and TS 29.520 documents to obtain the list of mandatory P-DEF parameters, and where applicable optional ones, which must or may be included by the sending module 3A in the REQ request addressed to the NWDAF 2 entity.

[0071] In accordance with the invention, the sending module 3A of the AMF entity 3-1 further inserts into the REQ request, in a dedicated field and more particularly in the UseCaseContext field introduced previously, a P-CTX parameter indicating the CTX context in which the REQ request is formulated, that is to say the context of use of the result(s) of the analysis requested in the REQ request. The P-CTX parameter here identifies the PROC procedure that the AMF entity 3-1 is about to execute using the result of the prediction requested in the REQ request. Thus, in the illustrative example envisaged, the P-CTX parameter contained in the UseCaseContext field of the REQ request identifies as a procedure, the determination of the registration zone of the UE 5.

[0072] Thus, in the embodiment described here, a new UseCaseContent field is added in the messages exchanged using the Nnwdaf_AnalyticsInfo and Nnwdaf_AnalyticsSubscription services.

[0073] There is no limitation on the format of the P-CTX parameter. It can be an integer uniquely designating the CTX context, a character string defining this CTX context, or a list of integers or character strings, etc. In addition, it is possible to envisage that the parameter in question takes a predefined value from a set of predefined values, each uniquely associated with a particular context (e.g. 1 = "registration of a UE by an AMF function", 2 = "determination of a registration zone of a UE by an AMF function", etc.), or conversely, that it takes any value as long as the NWDAF 2 entity is capable of associating this value with a particular context (there must be no ambiguity for the NWDAF 2 entity to identify from the value in question which context is designated, in other words the same value must not identify different formulation contexts).

[0074] Annex 1 provides, for illustrative purposes, a non-exhaustive list established by the inventors, of contexts and associated predictive or statistical analyses that may be requested from an NWDAF function in a 5G network, and therefore incidentally in the CN network. In the examples given in Annex 1, the same CTX context may be associated with different statistical / predictive analyses: for example, the context “Modification by a PCF of RFSP policies” may be based on analyses relating to the network slice load or on analyses relating to UE communications and the observed service experience (or OSE for “Observed Service Experience” in English).The corresponding requests addressed in this example by the PCF entity to the NWDAF entity include the same P-CTX parameter designating the context "Modification by a PCF of RFSP policies", while the P-DEF parameters provided in the requests designate for one of the requests, a statistical / predictive analysis relating to the load of the network slices, and for the other, statistical / predictive analyses relating to the UE communications and to the OSE.

[0075] On the other hand, it is possible to associate a different context with the same modification operation by a PCF of RFSP policies as long as it is carried out for a specific purpose, for example to preserving the quality of a vehicular communication service called V2X (for "Vehicle-to-everything" in English). It should also be noted that the X in V2X can designate any kind of entity, such as another vehicle, a drone, an infrastructure, a pedestrian, etc., so that a different context can be associated with each different X considered for the V2X service, the type of entity X considered being able to influence the choice of the analysis model.

[0076] In an alternative embodiment, it is possible to envisage including in the context information the statistical / predictive analyses on which the operation designated by the context is based, where appropriate, or more generally the use case of these analyses. Thus, by way of illustration, in the example mentioned above, it is possible to envisage a first context “Modification by a PCF of RFSP policies on the basis of statistical / predictive analysis of the network slice load” and a second context “Modification by a PCF of RFSP policies on the basis of statistical / predictive analyses of UE and OSE communications”. It should be noted that this precision provided in the context does not necessarily imply a change in the content of the P-DEF parameters provided in the request.Indeed, in order to limit the impact of the invention on existing messages already defined by the 3GPP standard (in other words to limit the modifications to be made to these messages), it is possible to envisage specifying both in the context and in the P-DEF parameters, the statistical / predictive analyses on which the operation implemented by the application entity at the origin of the request is based and which are required from the NWDAF entity.

[0077] It is noted that the determination by the sending module 3A of the AMF entity 3-1 of the CTX context for formulating the REQ request does not pose any difficulty in itself since this reflects the PROC procedure currently being executed by the AMF entity 3-1 (and more particularly by its processing module 3C) or a PROC procedure that this AMF entity 3-1 is preparing to execute (the REQ request not necessarily being sent synchronously with respect to the execution of the procedure in question).

[0078] With reference to Figure 4, upon receipt of the request REQ by its reception module 2A (step F10), the NWDAF entity 2, via its processing module 2C, determines from the P-DEF parameters provided in the request REQ, the statistical and / or predictive analysis requested of it by the AMF entity 3-1 (step F20), namely in the illustrative example envisaged here, a prediction of the mobility of the UE 5 under the conditions established by the P-DEF parameters.

[0079] The NWDAF entity 2, via its determination module 2B, also determines, from at least one piece of information conveyed by the request REQ, the CTX context in which the request REQ was formulated (step F30). This CTX context for formulating the request REQ gives the NWDAF entity 2 an indication of how the prediction requested by the AMF entity 3-1 will be used.

[0080] For this purpose, in the embodiment described here, the determination module 2B extracts the P-CTX parameter contained in the UseCaseContext field of the REQ request which identifies the PROC procedure requesting the statistical or predictive analysis requested in the REQ request.

[0081] It should be noted that in the embodiment described here, the REQ request includes the UseCaseContext field, and this field is supplied by the AMF entity 3-1 with the P-CTX parameter identifying the context in which the REQ request is formulated. Thus, the determination module 2B determines the CTX context from the content of the P-CTX parameter included in the UseCaseContext field of the REQ request. Alternatively, it may be envisaged that the UseCaseContext field is optional, and that in the absence of such a field in the request, the determination module 2B is configured to determine the CTX context for formulating the REQ request from the information conveyed by the latter, and in particular from all or part of the P-DEF parameters mentioned previously, namely the P-DEF parameters provided by the AMF entity 3-1 in the REQ request representative of its identity, the type of analysis requested, the target of this analysis (e.g. a UE or a group of UEs, a cell, etc.), the conditions for carrying out this analysis, the expected level of precision,. methods of notification of the results of the analysis, etc. For this purpose, the determination module 2B can rely on a supervised or unsupervised artificial intelligence (or AI) technique, or be configured with pre-established rules, defined for example by the operator of the CN core network, establishing a correspondence between the values ​​of all or part of the aforementioned P-DEF parameters conveyed by the REQ request and different contexts of formulation of this REQ request.

[0082] In yet another variant, we can consider a hybrid mode in which the determination module 2B determines the context from the content of the P-CTX parameter included in the UseCaseContext field and other information conveyed by the REQ request, typically the P-DEF parameters. This hybrid mode makes it possible to obtain a more precise context (and a finer granularity in the choice of the analysis model).

[0083] For example, certain P-DEF parameter groups provided in the REQ request and the values ​​associated with these parameter groups can be used to define signatures representative of distinct contexts.

[0084] It is then possible to envisage, according to an alternative embodiment, that the NWDAF entity 2, and more particularly its determination module 2B, is configured with a certain number of signatures with which contexts are associated, and that upon receipt of the request REQ, the determination module 2B determines from the P-DEF parameters provided in the request REQ, whether they correspond to one of the signatures with which it has been configured. Where appropriate, the context for formulating the request determined by the determination module 2B is that which is associated with the signature.

[0085] The configuration of the NWDAF 2 entity can be carried out in particular by the CN network operator by means of signatures determined by the latter, for example during a configuration phase as conventionally implemented during the introduction of new equipment or services into a network.

[0086] According to another embodiment variant, the module 2B for determining the NWDAF entity 2 can implement a learning phase during which it analyzes the P-DEF parameters present in a set of REQ requests previously received by the NWDAF entity 2, and determines, for example by means of a classification algorithm, which groups of parameters and associated values ​​are significant and representative of a particular context. The different groups of parameters and associated values ​​thus obtained define a set of signatures corresponding to distinct formulation contexts.

[0087] It should be noted that in this embodiment variant, the determination module 2B is able to distinguish between distinct formulation contexts, and to associate a given formulation context with a request; however, it does not necessarily have the information necessary to associate with this formulation context, a specific operation executed or intended to be executed by the application entity at the origin of the request REQ (e.g. determination of a recording zone). As an illustration, let us assume that the P-DEF parameters comprise three parameters X, Y, Z, and that the determination module 2B determines during the learning phase that: - if X=0 and Y=1, whatever the value of the parameter Z, we are always in a context to which the determination module 2B associates a label CTX1; and - if X=1 and Y=1, whatever the value of the parameter Z, we are always in another context to which the determination module 2B assigns the label CTX2. Upon receipt of a new request REQ, the determination module 2B is capable, from these signatures, of determining whether the context for formulating the new request REQ is the context CTX1 or the context CTX2, but in the absence of additional information, it is not capable of identifying the procedure associated with the context thus determined, i.e., for example, that CTX1 identifies a procedure for determining a recording area and CTX2 identifies a procedure for optimizing paging.

[0088] To remedy this, in yet another variant, it may be envisaged that the determination module 2B is configured, for example by the operator of the CN network, with rules for matching the labels determined during the learning phase by the determination module 2B and procedures likely to be implemented by the application entities corresponding to these labels. In other words, in the illustrative example envisaged above, the determination module 2B is configured with the following associations: CTX1 = Determination of a registration area of ​​a UE CTX2 = Optimization of paging

[0089] Whichever variant is chosen, it should be noted that the group of signatures on which the 2B determination module is based may evolve over time; in particular, it may be supplemented with new signatures, or these may be modified when new learning phases are carried out.

[0090] In accordance with the invention, the NWDAF entity 2 exploits (i.e. uses) the CTX context of formulation of the REQ request to select a relevant analysis model to carry out the analysis requested by the AMF entity 3-1, and more particularly here, a prediction of the mobility of the UE 5 (step F40).

[0091] Indeed, in a manner known per se, to carry out the statistical and / or predictive analyses requested of it, an NWDAF network function (and a fortiori the NWDAF entity 2 and its processing module 2C) has at its disposal a plurality of analysis techniques that it can use depending on the nature of the analyses requested of it. For example, for a statistical analysis relating to past behaviors, models conventionally used by an NWDAF network function include in particular: a model carrying out a moving average on all or part of the data collected by the NWDAF network function; a model carrying out an exponential moving average on all or part of the data collected by the NWDAF network function; etc. For predictive analysis of future behaviors, the prediction models traditionally used by a NWDAF network function rely in particular on time series modeling techniques using machine learning techniques, supervised or unsupervised, which can use a wide variety of algorithms, such as Markov chains, neural networks (e.g., LSTM neural networks for Long Short Term Memory), exponential smoothing, ARCH (AutoRegressive Conditional Heteroskedasticity) models, or ARMA or ARIMA models which combine autoregressive processes (AR for AutoRegressive), moving average (MA for Moving Average) and possibly integration (I for Integrated), etc.

[0092] Furthermore, each of the aforementioned analysis techniques may be parameterized differently, or rely on data sets having distinct characteristics (in addition to the nature of the data collected, such as the quantity of data, etc.) for the analysis or for the training of said analysis technique, which leads to as many distinct models that can be selected by the NWDAF entity 2. The CTX context provides the latter with a guide that allows it to make an informed and relevant selection of the analysis model to be applied to respond to the request of the AMF entity 3-1, and more particularly of at least one element from among: a statistical or predictive analysis technique to be used to carry out the analysis from among a plurality of predefined techniques, such as for example, an ARIMA model, an ARCH model, exponential smoothing, a moving average, etc. The selection may be made for example from among different families of known models such as those mentioned above or combinations of such models; and / or at least one characteristic (e.g. a quantity of data) of a dataset to be used to perform the analysis and / or train an analysis technique used to perform the analysis; and / or a parameterization (e.g. in a neural network, number of layers, neurons per layer, synaptic weights, etc.) of an analysis technique used to perform the analysis. It should be noted that such parameterization is not limited to taking into account the input parameters provided by the AMF 3-1 entity at the origin of the REQ request.

[0093] In the embodiment described here, the selection of an analysis model by the processing module 2C as a function of the context CTX determined by the determination module 2B is enabled via a prior (pre)configuration of the NWDAF entity 2. More specifically, it is assumed that during a preliminary configuration step (step F00), the NWDAF entity 2 has been configured, for example by the operator of the NW network via an appropriate interface or the use of an appropriate configuration protocol, with a list LIST of statistical and predictive analysis models, in which each analysis model is associated with at least one query formulation context. This list LIST is stored for example in the non-volatile memory NVM of the NWDAF entity 2. It should be noted that a formulation context can be associated with several analysis models or with a family of analysis models.

[0094] As an illustration, for the contexts of defining a UE registration zone and optimizing paging mentioned previously, the following associations can be considered in the LIST list: Table 1]

[0095] Of course, this example is given for illustrative purposes only and is not limiting of the invention, other contexts, analysis models and associations may be envisaged as a variant or in addition.

[0096] Thus, during the selection step F40, the processing module 2C of the NWDAF entity 2 consults the preconfigured list LIST during step F00, and selects an analysis model MOD(CTX) in the list LIST, associated with the context CTX. If several analysis models are possible, the processing module 2C chooses one, for example by taking into account criteria such as the calculation time or the calculation power required with regard to all or part of the P-DEF parameters provided in the request REQ (for example, the level of precision and / or the time taken to provide the analysis results).

[0097] It should be noted that the list LIST stored in the non-volatile memory NVM of the entity NWDAF 2 is not necessarily fixed, and may evolve over time (step F00').

[0098] Thus, the LIST list can be updated and / or supplemented to reflect this evolution in a similar or identical manner to what was previously described for step F00.

[0099] Alternatively, it may be updated by the NWDAF 2 entity itself, for example by its processing module 2C, during a learning or training phase, during which the processing module 2C may be required to revise the associations included in the LIST list between analysis models and formulation contexts, and / or create new associations relating to new analysis models and / or new formulation contexts. These new analysis models and / or formulation contexts may be communicated to the NWDAF 2 entity by the network operator or by another actor, for example the designer of the NWDAF 2 entity, or may also be determined by learning. This learning phase may rely as learning data on the analyses previously carried out by the processing module 2C. It may also or alternatively exploit training data provided by a third party, such as for example by the network operator, by the designer of the NWDAF 2 entity, etc.

[0100] More specifically, for all or part of the analyses carried out by the processing module 2C, the latter can evaluate the quality of the results obtained using determined quality indicators. To this end, the processing module 2C can, for example, following the obtaining of an analysis result such as a prediction, continue to collect data and verify that the prediction obtained is correct, or re-evaluate a new prediction based on this new data and compare the re-evaluated prediction with the initial prediction. Of course, other techniques can be envisaged to evaluate the quality of the results obtained by the processing module 2C.

[0101] During the learning phase, the processing module 2C correlates the statistical analysis models available to it, the different possible values ​​of formulation contexts, and the quality indicators of the analyses carried out from these analysis models for these formulation contexts. Then it selects, for a given formulation context, the analysis model(s) leading to the best quality indicators. It then updates the LIST list with the (new) context - analysis model(s) associations thus obtained where appropriate (in other words, it updates / corrects certain associations in the LIST list and / or completes the LIST list with new associations obtained during the learning phase).It should be noted that the processing module 2C can execute several learning phases at different times depending on the learning data available to it; thus, in particular, during a learning phase, it can target the learning data that it uses by retaining only the learning data associated with a particular formulation context, to determine the most relevant analysis model(s) for this context, and to change this formulation context from one learning phase to another.

[0102] In another embodiment, no preconfiguration of the NWDAF entity 2 is implemented, and it is the NWDAF entity 2 via its processing module 2C which itself establishes the list LIST and updates it over time during at least one learning (or training) phase identical or similar to what has just been described. This learning sentence may be supervised or not (i.e. rely on learning data provided by the network operator and / or the designer of the NWDAF entity 2 or only on learning data collected by the NWDAF entity 2).

[0103] In parallel with this selection, or indifferently before or after this selection, the processing module 2C collects the INPUT_DATA network data necessary to carry out the statistical and / or predictive analysis requested of it. This collection can be carried out from different entities, depending on the type of analysis required, using the procedures described in particular in the 3GPP TS 23.288 document in paragraph 6.7.2.4. Once the INPUT_DATA data has been collected, the processing module 2C carries out the statistical and / or predictive analysis (mobility prediction of UE 5 in the example considered here) requested of it by the AMF 3.1 entity by applying the analysis model MOD(CTX) to the collected INPUT_DATA data (step F50).

[0104] It should be noted that for the sake of simplification, the function of the 2C processing module has been considered here as a whole, without prejudging how the 2C processing module is organized strictly speaking to perform this function. In accordance with Releases 17 and following of the 3GPP standard, the NWDAF entity in a 5G core network can be decomposed into two functions or modules, an MLTF (Model Training Logical Function) module responsible for training and providing the analysis models and an AnLF (Analytics Logical Function) module responsible for producing the analyses using the models provided by the MLTF module. The analysis models are obtained by the AnLF module using the Nnwdaf_MLModelProvision and Nnwdaf_MLModelInfo services defined by the 3GPP standard. If such an organization is considered for the NWDAF 2 entity, then the following can be added to the Nnwdaf_MLModelProvision and Nnwdaf_MLModelInfo a new UseCaseContent field, as was done previously for the Nnwdaf_AnalyticsInfo and Nnwdaf_AnalyticsSubscription services, so that the AnLF module can specify to the MLTF module the CTX context of formulation of the REQ request and the MLTF module selects the analysis model to be applied according to this CTX context of formulation. In an alternative embodiment, the AnLF module itself selects the analysis model to be used according to the CTX context and then requests the selected model from the MTLF module.

[0105] Then the NWDAF entity 2 provides, via its 2D supply module, the analysis result RES obtained by the processing module 2C (prediction of the mobility of the UE 5 in the example considered here) to the AMF entity 3-1 in response to its request REQ (step F60).

[0106] Upon receipt of the response from the NWDAF entity 2 containing the RES analysis result by the reception module 3B of the AMF-1 entity (step E20), the execution module 2C of the AMF entity 3-1 executes the PROC procedure in the context of which it requested the NWDAF entity 2 using the RES analysis result provided by the NWDAF entity 2 (step E30). Thus, in the illustrative example envisaged here, the execution module 3C determines the registration area of ​​the UE 5 using the mobility prediction of the UE 5 provided by the NWDAF entity 2, in a manner known per se.

[0107] The invention has just been described in the context of a 5G core network, and in particular of an NWDAF function of such a network, advantageously relying on the APIs defined by the 3GPP standard, modified for the purposes of the invention. The invention can however be applied in other contexts, and in particular to other networks, for example to proprietary networks, future generation networks or networks corresponding to other versions (or "Releases" in English) of a 5G network, IP networks, etc., to other application entities, such as for example routers, SDN controllers, etc. which can be requested by various application entities of the network in order to deliver statistical and / or predictive analyses, and exploit interfaces other than APIs (e.g. point-to-point interfaces) to implement the invention. Annex 1

Claims

Claims

1. Method for processing a request (REQ) for statistical or predictive analysis received by a first application entity (2) of a communications network (CN) from a second application entity (3-1) of the network, said method comprising: a step (F30) of determining, from at least a plurality of parameters (P-CTX, P-DEF) conveyed by said request and defining the analysis required by the second application entity, a context (CTX) for formulating said request by said second application entity; a step (F50) of carrying out said required statistical or predictive analysis by means of an analysis model (MOD(CTX)) selected (F40) as a function of said determined context; and a step (F60) of providing at least one result (RES) of the analysis carried out in response to said request from the second application entity.

2. Processing method according to claim 1 wherein during the determination step, said formulation context is further determined from a parameter (P-CTX) inserted by the second application entity in a dedicated field of the request.

3. Processing method according to claim 1 or 2 wherein said plurality of parameters are taken from: an identity of the second application entity; a parameter representative of a type of analysis required; at least one piece of equipment to which said analysis relates; a parameter representative of a condition for carrying out said analysis; an expected level of precision for said analysis; a parameter representative of a method of notification of said at least one result of the analysis.

4. Processing method according to any one of claims 1 to 3 wherein said formulation context (CTX) identifies a procedure (PROC) intended to be executed by the second application entity in the network using the result provided by the first application entity.

5. Processing method according to any one of claims 1 to 4 wherein the selection (F40) of the analysis model comprises a selection of at least one element from: a statistical or predictive analysis technique to be used to carry out the analysis; at least one characteristic of a data set to be used to carry out the analysis and / or train an analysis technique used to carry out the analysis; a parameterization of an analysis technique used to carry out the analysis.

6. Processing method according to any one of claims 1 to 5 comprising a preliminary step (F00) of configuring the first application entity with a list (LIST) of analysis models in which each analysis model is associated with at least one context for formulating a query, said analysis model applied during the realization step being an analysis model associated in said list with the context determined during the obtaining step.

7. Processing method according to any one of claims 1 to 5 comprising a learning step during which the first application entity: establishes correlations between analysis models, contexts for formulating a query, and quality indicators of the analyses carried out from said models in said contexts; and associates, from said established correlations, with at least one context for formulating a query an analysis model optimizing said quality, said analysis model applied during the execution step being selected from the associations established during the learning step.

8. Processing method according to any one of claims 1 to 5 comprising: a preliminary step (F00) (LIST) of configuring the first application entity with a list comprising at least one analysis model associated with at least one context for formulating a query; a learning step (F00') during which the first application entity: o establishes correlations between analysis models, contexts for formulating a query, and quality indicators of the statistical analyses carried out from said models in said contexts; and o updates and / or completes said list from said established correlations; the analysis model applied during the production step being an analysis model associated with the context determined during the obtaining step in the list updated or completed during the learning step.

9. Processing method according to any one of claims 1 to 8 wherein the first application entity (2) and the second application entity (3-1) host network functions, said network function hosted by the first application entity being a function for collecting and analyzing network data.

10. Computer program (PROG) comprising instructions for implementing a method according to any one of claims 1 to 9 when said program is executed by a computer.

11. Application entity (2) of a communications network, called first application entity, comprising: a reception module (2A) configured to receive from a second application entity of the network, a request for statistical or predictive analysis; a determination module (2B), configured to determine a context of formulation of said request by said second application entity from at least a plurality of parameters (P-DEF) defining the analysis required by the second application entity and conveyed in the analysis request; a processing module (2C), configured to carry out said analysis by means of an analysis model selected as a function of said formulation context determined by said determination module (2B); and a supply module (2D), configured to supply at least one result of the analysis carried out in response to said request from the second application entity.

12. Application entity (2) according to claim 11 in which said determination module (2B) is configured to determine said formulation context from, in addition, a parameter (P-CTX) inserted by the second application entity in a dedicated field of the request.

13. System (1) in a communications network comprising: at least one first application entity (2) according to claim 12; and at least one second application entity (3,3-1) comprising: o a sending module (3A), configured to send to said at least one first application entity of the network a statistical or predictive analysis request comprising a plurality of parameters (P-DEF) defining the analysis required by the second application entity and inserted in a dedicated field of the request, a parameter indicating a context for formulating said request;and o a receiving module (3B), configured to receive from said at least one first application entity in response to said request, at least one result of the required analysis carried out by said at least one first application entity by means of an analysis model selected according to the context of formulation of the request determined by said at least one first application entity from said plurality of parameters defining the analysis required by the second application entity and said parameter inserted in said dedicated field of the request.;

Images