Support for Federated Learning (FL)

Abstract

The present disclosure relates to a server, a client, a network node, and a method for supporting federated learning (FL), the method including at least one of sending a first message to one or more clients associated with an FL process indicating that the corresponding client has not been selected by a server for the FL process and / or that the FL process has been paused, and receiving a second message from one or more second network nodes associated with the FL process indicating that the corresponding client has left the FL process.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to PCT International Application No. PCT / CN2023 / 071204, entitled "SUPPORT FOR FEDERATED LEARNING (FL)," filed January 9, 2023, which is incorporated herein by reference in its entirety.

[0002] The present disclosure relates to the field of communications, and in particular to servers, clients, network nodes, and methods for supporting federated learning (FL). [Background technology]

[0003] With the development of electronic and communication technologies, mobile devices such as mobile phones, smartphones, laptops, tablets, and in-vehicle devices have become an important part of our daily lives. Supporting a large number of mobile devices requires highly efficient core networks, such as the 3rd Generation Partnership Project (3GPP) 5th Generation Core (5GC).

[0004] Machine learning (ML) is a key enabler for optimizing, securing, and managing core networks. This increases the collection and processing of data from network functions, which in turn may increase threats to end-user confidential information. As a result, to maximize the benefits of ML, mechanisms are needed to mitigate threats to end-user privacy.

[0005] As an emerging ML technique, federated learning (FL), also known as collaborative learning, trains algorithms across multiple distributed edge devices or servers that hold local data samples without exchanging those data samples. This approach contrasts with traditional centralized machine learning techniques, where all local datasets are uploaded to a single server, and more classical distributed approaches that often assume that local data samples are similarly distributed.

[0006] Federated learning allows multiple parties to build a common, robust machine learning model without sharing data, thereby addressing important issues such as data privacy, data security, data access rights, and access to heterogeneous data. Its applications span many industries, including defense, communications, the Internet of Things (IoT), and pharmaceuticals.

[0007] Therefore, there is increasing interest in supporting FL in 3GPP 5GC. Summary of the Invention

[0008] According to the latest technical report from 3GPP, in Section 8.8 of Technical Report (TR) 23.700-81 V2.0.0 (2022-11), the maintenance of FL processes between multiple Network Data Analysis Functions (NWDAFs) in 5GC has been added to the conclusions (mainly Principle 5) as follows: Principle 5: An NWDAF including a Model Training Logic Function (MTLF) as an FL server may determine a final list of NWDAFs including MTLFs as FL clients through an initial FL request to FL clients to determine the availability and compatibility of FL clients. During the FL procedure, an NWDAF including an MTLF as an FL server may trigger reselection, addition, or removal of FL clients based on local policy or the status of FL clients, e.g., load, availability, capacity, latency, accuracy, etc., and may issue new FL client discoveries via the Network Repository Function (NRF). FL clients can dynamically join or leave FL operations during the execution phase.

[0009] However, in the FL process and implementation maintenance, it is still unclear how the server NWDAF obtains analytical information (e.g., network function (NF) load) of the client NWDAF and how to terminate the federated learning operation in the client NWDAF.

[0010] To address or at least mitigate the above problems, some embodiments of the present disclosure provide servers, clients, network nodes, and methods for supporting federated learning (FL) in core networks such as 5GC networks.

[0011] According to a first aspect of the present disclosure, there is provided a method in a server associated with an FL process. The method includes at least one of sending a first message to one or more clients associated with the FL process indicating that the corresponding client has not been selected by the server for the FL process and / or that the FL process has been paused, and receiving a second message from one or more clients associated with the FL process indicating that the corresponding client has left the FL process. In some embodiments, the method further includes any of the steps in any of the methods of the second aspect.

[0012] According to a second aspect of the present disclosure, there is provided a method in a server associated with an FL process. The method includes receiving one or more fifth messages indicating analytical data, the analytical data associated with one or more clients in the FL process and / or one or more candidate clients selected in the FL process, and selecting at least one from the one or more candidate clients and / or one or more clients for the FL process based at least on the analytical data. In some embodiments, the method further includes any of the steps in any of the methods of the first aspect.

[0013] According to a third aspect of the present disclosure, there is provided a server comprising: a processor; and a memory storing instructions that, when executed by the processor, cause the processor to perform any of the methods of the first and / or second aspects.

[0014] According to a fourth aspect of the present disclosure, there is provided a server associated with an FL process. The server comprises at least one of a sending module configured to send a first message to one or more clients associated with the FL process indicating that the corresponding client has not been selected by the server for the FL process and / or that the FL process has been paused, and a receiving module configured to receive a second message from one or more clients associated with the FL process indicating that the corresponding client has left the FL process. In some embodiments, the server comprises one or more additional modules, each of which may perform any of the steps of any of the methods of the first aspect.

[0015] According to a fifth aspect of the present disclosure, there is provided a server associated with an FL process. The server comprises: a receiving module configured to receive one or more fifth messages indicating analytical data associated with one or more clients in the FL process and / or one or more candidate clients selected in the FL process; and a selection module configured to select at least one of the one or more candidate clients and / or one or more clients for the FL process based at least on the analytical data. In some embodiments, the server comprises one or more additional modules, each of which may perform any of the steps of any of the methods of the second aspect.

[0016] According to a sixth aspect of the present disclosure, there is provided a method in a client associated with an FL process. The method includes at least one of receiving a first message from a server associated with the FL process indicating that the client has not been selected by the server for the FL process and / or that the FL process has been suspended, and sending a second message to the server associated with the FL process indicating that the client is leaving the FL process. In some embodiments, the method further includes any of the steps in any of the methods of the seventh aspect.

[0017] According to a seventh aspect of the present disclosure, there is provided a method in a client associated with an FL process or a candidate client selected in the FL process. The method includes at least one of sending a fifth message to a server associated with the FL process indicating analytical data associated with the client or candidate client, and sending a seventh message to one or more network nodes indicating data associated with the client or candidate client, the data being used as input data in determining the analytical data associated with the client or candidate client. In some embodiments, the method further includes any of the steps in any of the methods of the sixth aspect.

[0018] According to an eighth aspect of the present disclosure, there is provided a client method, the method including: a processor; and a memory storing instructions that, when executed by the processor, cause the processor to perform any of the methods of the sixth and / or seventh aspects.

[0019] According to a ninth aspect of the present disclosure, there is provided a client associated with an FL process. The client comprises at least one of a receiving module configured to receive a first message from a server associated with the FL process indicating that the client has not been selected by the server for the FL process and / or that the FL process has been paused, and a transmitting module configured to send a second message to the server associated with the FL process indicating that the client is leaving the FL process. In some embodiments, the client comprises one or more additional modules, each of which may perform any of the steps of any of the methods of the sixth aspect.

[0020] According to a tenth aspect of the present disclosure, there is provided a client associated with an FL process or a candidate client selected in the FL process. The client or candidate client comprises at least one of a first sending module configured to send a fifth message indicating analytical data associated with the client or candidate client to a server associated with the FL process, and a second sending module configured to send a seventh message indicating data associated with the client or candidate client to one or more network nodes, the data being used as input data in determining the analytical data associated with the client or candidate client. In some embodiments, the client or candidate client comprises one or more additional modules, each of which may perform any of the steps of any of the methods of the seventh aspect.

[0021] According to an eleventh aspect of the present disclosure, there is provided a method in a network node, the method including sending, to a server associated with the FL process, a fifth message indicating analytical data associated with one or more clients associated with the FL process and / or one or more candidate clients selected by the server for the FL process.

[0022] According to a twelfth aspect of the present disclosure, there is provided a network node comprising: a processor; and a memory storing instructions that, when executed by the processor, cause the processor to perform any of the methods of the eleventh aspect.

[0023] According to a thirteenth aspect of the present disclosure, there is provided a network node comprising: a transmitting module configured to transmit, to a server associated with the FL process, a fifth message indicating analytical data associated with one or more clients associated with the FL process and / or one or more candidate clients selected by the server for the FL process. In some embodiments, the network node comprises one or more further modules, each of which may perform any of the steps of any of the methods of the eleventh aspect.

[0024] According to a fourteenth aspect of the present disclosure, there is provided a computer program comprising instructions that, when executed by at least one processor, cause the at least one processor to perform any of the methods of any of the first, second, sixth, seventh, and / or eleventh aspects.

[0025] According to a fifteenth aspect of the present disclosure, there is provided a carrier comprising the computer program of the fourteenth aspect. In some embodiments, the carrier is one of an electronic signal, an optical signal, a radio signal, or a computer-readable storage medium.

[0026] According to a sixteenth aspect of the present disclosure, there is provided a communication system for supporting FL. The communication system comprises a server of the third, fourth, and / or fifth aspects and one or more clients of the eighth, ninth, and / or tenth aspects. In some embodiments, the communication system further comprises one or more network nodes of the twelfth and / or thirteenth aspects.

[0027] According to some embodiments of the present disclosure, FL can be supported in a core network, such as a 5GC network. In this manner, the federated learning process and implementation maintenance can be completed. In one aspect, the server NWDAF can obtain analytical information of the client NWDAF through the assisting NWDAF and / or directly from the client NWDAF. Furthermore, in another aspect, the federated learning operation can be completed in the client NWDAF, taking into account two different methods of ML model information exchange in the FL execution phase: a method of reusing an existing service (or its extension) and a method of using a new service.

[0028] The above and other features of the present disclosure will become more fully apparent from the following description and appended claims, taken in conjunction with the accompanying drawings, in which: The present disclosure will be described with additional specificity and detail through the use of the accompanying drawings, with the understanding that these drawings illustrate only some embodiments in accordance with the present disclosure and therefore should not be considered limiting of its scope. [Brief explanation of the drawings]

[0029] [Figure 1] FIG. 1 illustrates an exemplary communication network in which support for FL is applicable, according to some embodiments of the present disclosure. [Figure 2] FIG. 10 illustrates an example procedure for selection of a client NWDAF in the FL preparation phase where support for FL is applicable, in accordance with some embodiments of the present disclosure. [Figure 3]FIG. 1 illustrates an example procedure for monitoring and reselection of an NWDAF during an FL execution phase where support for FL is applicable, in accordance with some embodiments of the present disclosure. [Figure 4] FIG. 10 illustrates an example procedure for dynamic discovery of a new NWDAF in the FL execution phase when information about a server NWDAF is known in a new client NWDAF for which support for FL is applicable, according to some embodiments of the present disclosure. [Figure 5] FIG. 10 illustrates an example procedure for dynamic discovery of a new NWDAF in the FL execution phase when information about a server NWDAF is unknown in a new client NWDAF for which support for FL is applicable, according to some embodiments of the present disclosure. [Figure 6] FIG. 1 illustrates an exemplary system for analytical information collection, according to some embodiments of the present disclosure. [Figure 7] FIG. 1 illustrates an example system for FL process termination in a client NWDAF, according to some embodiments of the present disclosure. [Figure 8] FIG. 1 illustrates an exemplary scenario for analytical information collection, according to some embodiments of the present disclosure. [Figure 9] FIG. 10 illustrates an example scenario for FL process termination at a client NWDAF, according to some embodiments of the present disclosure. [Figure 10] FIG. 1 illustrates an exemplary procedure for analytical information collection, according to some embodiments of the present disclosure. [Figure 11A] FIG. 10 illustrates an example procedure for FL process termination at a client NWDAF, according to some embodiments of the present disclosure. [Figure 11B] FIG. 10 illustrates an example procedure for FL process termination at a client NWDAF, according to some embodiments of the present disclosure. [Figure 12] 1 is a flowchart of an exemplary method in a server according to an embodiment of the present disclosure. [Figure 13] 10 is a flowchart of another exemplary method in a server according to another embodiment of the present disclosure. [Figure 14] 10 is a flowchart of an exemplary method in a client according to an embodiment of the present disclosure. [Figure 15] 10 is a flowchart of another exemplary method in a client according to another embodiment of the present disclosure. [Figure 16] 1 is a flowchart of an exemplary method in a network node according to one embodiment of the present disclosure. [Figure 17] FIG. 1 is a schematic diagram illustrating an embodiment of a configuration that may be used in a server, client, and / or network node, according to an embodiment of the present disclosure. [Figure 18] FIG. 2 is a block diagram of an exemplary server according to one embodiment of the present disclosure. [Figure 19] FIG. 10 is a block diagram of another exemplary server according to another embodiment of the present disclosure. [Figure 20] FIG. 2 is a block diagram of an exemplary client according to one embodiment of the present disclosure. [Figure 21] FIG. 10 is a block diagram of another exemplary client according to another embodiment of the present disclosure. [Figure 22] FIG. 2 is a block diagram of an exemplary network node according to one embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0030] The present disclosure will be described below with reference to the embodiments shown in the accompanying drawings. However, it should be understood that these descriptions are provided for illustrative purposes only and not to limit the present disclosure. Furthermore, in the following, descriptions of known structures and techniques are omitted so as not to unnecessarily obscure the concepts of the present disclosure.

[0031] Those skilled in the art will appreciate that the term "exemplary" is used herein to mean "illustrative" or "serving as an example," and is not intended to suggest that a particular embodiment is preferred over another, or that a particular feature is essential. Similarly, the terms "first" and "second," and similar terms, are used merely to distinguish one particular instance of an item or feature from another, and do not dictate a particular order or configuration unless the context clearly dictates otherwise. Furthermore, as used herein, the term "step" is intended to be synonymous with "operation" or "action." The description herein of a sequence of steps does not imply that these operations must be performed in a particular order, or even that these operations be performed in any order, unless the context or details of the described operations clearly dictate otherwise.

[0032] Conditional language used herein, such as "can," "might," "may," "for example," and the like, unless expressly stated otherwise or understood otherwise within the context in which it is used, is intended to generally convey that some embodiments include certain features, elements, and / or conditions, but not others. Thus, such conditional language generally does not imply that features, elements, and / or conditions are in any way required for one or more embodiments, or that one or more embodiments necessarily include logic for determining, with or without author input or prompts, whether these features, elements, and / or conditions are included or should be implemented in any particular embodiment. Also, the term "or," when used to connect, for example, a list of elements, is used in its inclusive sense (and not its exclusive sense), such that the term "or" means one, some, or all of the elements in the list. Furthermore, the term "each," as used herein, in addition to having its ordinary meaning, can refer to any subset of the set of elements to which the term "each" applies.

[0033] The term "based on" should be read as "based at least in part on." The terms "one embodiment" and "an embodiment" should be read as "at least one embodiment." The term "another embodiment" should be read as "at least one other embodiment." Other provisions, both explicit and implicit, may be included below. Additionally, unless otherwise specified, phrases such as "at least one of X, Y, and Z" should generally be understood with the context in which they are used to convey that the item, term, etc. can be either X, Y, or Z, or a combination thereof.

[0034] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit example embodiments. As used herein, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly dictates otherwise. It will be further understood that the terms "comprises," "comprising," "has," "having," "includes," and / or "including," when used herein, specify the presence of stated features, elements, and / or components, etc., but do not exclude the presence or addition of one or more other features, elements, components, and / or combinations thereof. It will also be understood that the terms "connect(s)," "connecting," "connected," and the like, when used herein, only mean that there is an electrical or communication connection between two elements, and that they may be connected either directly or indirectly, unless expressly stated otherwise.

[0035] Of course, the present disclosure may be carried out in other specific ways than those described herein without departing from the scope and essential characteristics of the present disclosure. One or more of the specific processes described below may be carried out in any electronic device including one or more appropriately configured processing circuits, which may, in some embodiments, be incorporated into one or more application-specific integrated circuits (ASICs). In some embodiments, these processing circuits may comprise one or more microprocessors, microcontrollers, and / or digital signal processors, or variations thereof, programmed with appropriate software and / or firmware to perform one or more of the operations described above. In some embodiments, these processing circuits may comprise customized hardware to perform one or more of the functions described above. The present embodiments, therefore, are to be considered in all respects as illustrative and not restrictive.

[0036] Although several embodiments of the present disclosure are illustrated in the accompanying drawings and described in the following detailed description, it should be understood that the present disclosure is not limited to the disclosed embodiments, but instead is capable of numerous rearrangements, modifications, and substitutions without departing from the present disclosure as set forth and defined in the claims.

[0037] Furthermore, while the following description of some embodiments of the present disclosure is described in the context of a 5G system (5GS), it should be noted that the present disclosure is not limited thereto. In fact, as long as support for FL is involved, the inventive concepts of the present disclosure may be applicable to any suitable communication architecture, such as, for example, Global System for Mobile Communications (GSM) / General Packet Radio Service (GPRS), Enhanced Data Rates for GSM Evolution (EDGE), Code Division Multiple Access (CDMA), Wideband CDMA (WCDMA), Time Division Synchronous CDMA (TD-SCDMA), CDMA2000, Worldwide Interoperability for Microwave Access (WiMAX), Wireless Fidelity (Wi-Fi), Universal Terrestrial Radio Access Network (UTRAN), Enhanced UTRAN (E-UTRAN), Long Term Evolution (LTE), Evolved Packet System (EPS), etc. Therefore, those skilled in the art can readily understand that the terms used herein may also refer to their equivalents in any other infrastructure. For example, the term "user equipment" or "UE" as used herein may refer to a mobile device, a mobile terminal, a mobile station, a user device, a user terminal, a wireless device, a wireless terminal, an IoT device, a vehicle, or any other equivalent. In another example, the term "network node" as used herein may refer to or comprise a base station, a base transceiver station, an access point, a hotspot, a Node B (NB), an evolved Node B (eNB), a gNB, a network element, a network function, or any other equivalent.

[0038] Additionally, the following 3GPP documents are incorporated herein by reference in their entirety: -3GPP Technical Specification (TS) 23.288 V18.0.0 (2022-12), Technical Specification, 3rd Generation Partnership Project; Technical Specification Group Services and System Aspects; Architecture Extensions for 5G Systems (5GS) to Support Network Data Analysis Services (Release 18) -3GPP TR 23.700-81 V2.0.0(2022-11), Technical Report, 3rd Generation Partnership Project; Technical Specification Group Services and System Aspects; Network Automation Enablement Study for 5G and 5G Systems (5GS); Phase 3 (Release 18)

[0039] 1 is a block diagram illustrating an exemplary communication network 10 in which support for FL is applicable, according to one embodiment of the present disclosure. The communication network 10 is a network defined in the context of 5GS, although the present disclosure is not limited thereto.

[0040] As shown in FIG. 1, the network 10 may comprise one or more UEs 100 and a (Radio) Access Network ((R)AN) 105, which includes one or more RAN nodes, such as base stations, Node Bs, evolved Node Bs (eNBs), gNBs, or access network (AN) nodes that provide the UEs 100 with access to other parts of the network 10. Further, network 10 may include a core network portion thereof, which may include (but is not limited to) one or more User Plane Function (UPF) 115, NWDAF 120, Authentication Server Function (AUSF) 125, Access and Mobility Management Function (AMF) 130, Session Management Function (SMF) 135, Service Communication Proxy (SCP) 140, Network Slice Admission Control Function (NSACF) 145, Network Slice Selection Function (NSSF) 150, Network Publication Function (NEF) 155, Network Repository Function (NRF) 160, Policy Control Function (PCF) 165, Unified Data Management (UDM) 170, Application Function (AF) 175, and Edge Application Server Discovery Function (EASDF) 180. As shown in FIG. 1 , these entities may communicate with each other via service-based interfaces, such as Namf, Nsmf, and / or reference points, such as N1, N2, N3, N4, N6, and N9.

[0041] However, the present disclosure is not limited thereto. In some other embodiments, network 10 may include additional network functions, fewer network functions, or variations of the existing network functions shown in FIG. 1. For example, in a network with a 4G EPS architecture, the entities performing these functions (e.g., mobility management entities (MMEs)) may differ from those shown in FIG. 1 (e.g., AMF 120). In another example, in a network with a mixed 4G / 5G architecture, some of the entities may be the same as those shown in FIG. 1, and others may be different. Furthermore, the functions shown in FIG. 1 are not essential to embodiments of the present disclosure. In other words, some of them may be missing from some embodiments of the present disclosure.

[0042] As shown in FIG. 1, the UPF 115 is communicatively connected to a data network (DN) 185, which may be the Internet or may similarly be communicatively connected to the Internet, thereby allowing the UE 100 to ultimately communicate its user plane data with other devices outside the network 10, for example, via the RAN 105 and the UPF 115.

[0043] Some of the network functions shown in FIG. 1 that may be involved in some embodiments of the present disclosure are described below.

[0044] In some embodiments, the NWDAF 120 may include one or more of the following functions: -Support for data collection from NF and AF -Support for data collection from Operations, Administration and Maintenance (OAM) - Registration and Metadata Publication of NWDAF Services to NFs and AFs -Support for providing analytical information to NF and AF - Support for training and serving machine learning (ML) models to the NWDAF (including analytical logic functions)

[0045] As mentioned above, the maintenance of the FL process among multiple NWDAFs in 5GC has been added to the conclusions of Section 8.8 of TR 23.700-81, v2.0.0 (mainly Principle 5). The content of Principle 5 is as follows: Principle 5: An NWDAF including an MTLF as an FL server may determine the final list of NWDAFs including MTLFs as FL clients through an initial FL request to FL clients to determine the availability and compatibility of FL clients. During the FL procedure, an NWDAF including an MTLF as an FL server may trigger reselection, addition, or removal of FL clients based on local policy or the status of FL clients, e.g., load, availability, capability, latency, accuracy, etc., and may issue new FL client discoveries via the NRF. FL clients can dynamically join or leave FL operations during the execution phase.

[0046] A solution for the maintenance of FL processes among multiple NWDAFs in 5GC is presented in TR 23.700-81, 2.0.0 (i.e., Solution #51). Solution #51: Selecting, Monitoring, and Maintaining NWDAFs for Federated Learning in 5GC

[0047] explanation This solution is proposed to address Key Problem #8: "Supporting Federated Learning in 5GC." Bullet points for consideration regarding this key problem include: - Selection of participating NWDAF instances in the federated learning group, including consideration of how multiple NWDAFs will be coordinated, including, for example, determining the supporting information (if any) for making the selection and the roles of the participating NWDAFs. -Consider whether and how to monitor the performance of federated learning operations (e.g., network performance and model performance) by NWDAF.

[0048] To address the challenges in the above bullet points regarding supporting federated learning in 5GC, this solution focuses on the selection of NWDAFs in the preparation phase of federated learning, and the monitoring and maintenance of NWDAFs in the execution phase of federated learning.

[0049] Many factors influence the selection of a client NWDAF during the preparation phase of federated learning, including the capabilities of the NWDAF, the interoperability and availability of the client NWDAF to participate in federated learning.

[0050] During the execution phase of federated learning, dynamic changes in the federated network may cause the current client NWDAF to leave or join, and dynamic participation and withdrawal of client NWDAFs in the multi-round learning / training process of federated learning in 5GC should be considered. Furthermore, a method may be applied for the server NWDAF to monitor status changes (e.g., changes in capability and availability) of the client NWDAFs.

[0051] procedure FIG. 2 illustrates an example procedure for selection of a client NWDAF in the FL preparation phase where support for FL is applicable, in accordance with some embodiments of the present disclosure.

[0052] In the preparation phase of federated learning, a server NWDAF and (potential) client NWDAFs (e.g., server NWDAF 120-S and one or more client NWDAFs 120-C-1 through 120-CN, or collectively 120-C, shown in FIG. 2) are discovered via an NRF (e.g., NRF 160 shown in FIG. 2), and a client NWDAF is selected by a handshake pattern method. The selection of a client NWDAF is based on availability, capabilities, etc.

[0053] In some embodiments, an exemplary procedure for selecting an NWDAF is shown in FIG. 2 and is described as follows.

[0054] In step S205, an NWDAF (e.g., server NWDAF 120-S and / or client NWDAFs 120-C-1 through 120-CN) may register with NRF 160 as having federated learning capabilities. In some embodiments, server NWDAF 120-S may discover client NWDAFs 120-C based on, for example, federated learning capabilities, analysis identifiers (IDs), etc.

[0055] In step S210, the server NWDAF 120-S may send a federated learning preparation request to the client NWDAF 120-C, for example, by invoking the Nnwdaf_MLPreparation_Request service operation using the interoperability information. The preparation request may include an indication of the NWDAF's role, i.e., to act as a client NWDAF.

[0056] In some embodiments, the interoperability information may indicate the capabilities required for the client NWDAF 120-C to support this FL procedure (e.g., the ability to execute a particular model), and may indicate, for example, whether and how the server NWDAF 120-S and the client NWDAF 120-C can share models.

[0057] In step S215, the client NWDAF 120-C may determine whether to participate in the federated learning process based on its availability, capabilities, and interoperability information.

[0058] In step S220, the client NWDAF 120-C may send a response to the server NWDAF 120-S indicating whether it wishes to participate in the FL procedure.

[0059] In step S225, the server NWDAF 120-S may send a test task to the client NWDAF 120-C that wishes to participate in the FL procedure. The client NWDAF 120-C may execute the test task and send the results to the server NWDAF 120-S.

[0060] In some embodiments, the test task may be a micro-computation task or a micro-training task, and the requirements for completing the micro-task are the same or similar to the requirements of the main task. In some embodiments, the test task may be a small task that causes the client NWDAF 120-C to collect local data and send local model weights back to the server 120-S, or a test to verify that the server NWDAF 120-S and the client NWDAF 120-C can communicate if they use the same FL framework or library.

[0061] In step S230, the server NWDAF 120-S may select a client NWDAF 120-C, and the results of the test task may be considered by the server NWDAF 120-S for the selection of the client NWDAF 120-C.

[0062] FIG. 3 illustrates an example procedure for monitoring and reselection of an NWDAF during an FL execution phase where support for FL is applicable, in accordance with some embodiments of the present disclosure.

[0063] During the federated learning execution phase, the server NWDAF (server NWDAF 120-S) monitors status changes of client NWDAFs (e.g., one or more client NWDAFs 120-C-1 through 120-CN, or collectively 120-C, as shown in FIG. 3). In some embodiments, a client NWDAF 120-C may be reselected for an FL task based on the updated status, availability, and / or capabilities of the client NWDAF 120-C, etc.

[0064] In some embodiments, an exemplary procedure for monitoring and reselection of a client NWDAF 120-C is shown in FIG. 3 and described as follows.

[0065] In step S305, the server NWDAF 120-S, which monitors the status of the client NWDAF 120-C during the federated learning execution process, may receive the updated status of the client NWDAF 120-C.

[0066] In some embodiments, the server NWDAF 120-S may perform monitoring and obtain updated status of the client NWDAF 120-C directly and / or via the NRF 160.

[0067] In some embodiments, the status of the client NWDAF 120-C may be the load of the NF, the availability of the NF, a change in its capabilities, for example, not supporting the FL anymore.

[0068] In step S310, the server NWDAF 120-S may check the status of the client NWDAF based on the received information and determine whether reselection of the client NWDAF 120-C is necessary in the next round of federated learning. In some embodiments, the determination may be based on the updated status of the client NWDAF 120-C, including availability, capabilities, etc.

[0069] In step S315, [if reselection is determined to be necessary in step S310], the server NWDAF 120-S may reselect a client NWDAF 120-C as in steps S210 through S230 of Figure 2. In some embodiments, the procedure for discovery of a new client NWDAF 120-C in the federated learning execution phase is described with reference to Figure 4.

[0070] In step S320, the client NWDAF 120-C may terminate operations for federated learning if it receives a termination request from the server NWDAF 120-S.

[0071] There are two possible cases for the server NWDAF to obtain the information of the new client NWDAF: directly from the new client NWDAF or via the NRF.

[0072] In some embodiments, client NWDAF #1 120-C-1 through client NWDAF #N 120-CN may be selected by server NWDAF 120-S to participate in the current round of federated learning. In some embodiments, new client NWDAFs, client NWDAF #N+1 120-C-N+1 through client NWDAF #N+X 120-C-N+X, have the ability to participate in the next round of the training process.

[0073] 4 illustrates an example procedure for dynamic discovery of new NWDAFs in the FL execution phase when information about the server NWDAF is known at the new client NWDAF for which support for FL is applicable, in accordance with some embodiments of the present disclosure. In some embodiments, new client NWDAFs (e.g., client NWDAF #N+1 120-C-N+1 through client NWDAF #N+X 120-C-N+X) that are able and / or capable of participating in the federated learning process know information about the server NWDAF 120-S and directly inform the server NWDAF 120-S.

[0074] In some embodiments, the procedure is shown in FIG. 4 and described as follows.

[0075] In step S405, the server NWDAF 120-S may register with the NRF 160 for the federated learning procedure with the following parameters: - Federated Learning (FL) Correlation ID -Analysis ID

[0076] In some embodiments, the FL correlation ID may be used to identify a particular FL procedure. For example, a server NWDAF or a client NWDAF may participate in different FL procedures simultaneously and, when receiving a message or data from another NWDAF, need to know which FL procedure the message or data is for.

[0077] In some embodiments, when a server NWDAF initiates an FL procedure, it may register the FL procedure with the NRF along with an FL correlation ID and an analysis ID. Subsequently, when a client NWDAF wishes to dynamically participate in an FL, for example, to update its local model using global information, it may query the NRF to see if there is an ongoing FL for the analysis ID. The NRF then provides the client NWDAF with the server NWDAF's ID and the FL correlation ID, so that the client NWDAF can contact the server NWDAF to participate in the FL procedure. Using the FL correlation ID, the server NWDAF knows which FL procedure the client NWDAF wishes to participate in and which model to provide the client.

[0078] In step S410, if information about the server NWDAF 120-S and the corresponding FL procedures is known via the NRF 160, the new client NWDAFs 120-C-N+1 to 120-C-N+X may inform the server NWDAF 120-S of their interoperability and availability by invoking the Nnwdaf_MLPreparation_Request service operation.

[0079] In step S415, before starting the next round of training, the server NWDAF 120-S may select a client NWDAF from among NWDAF #1 120-C-1 through NWDAF #N+X 120-C-N+X based on the updated information of the client NWDAF 120-C. In some embodiments, the procedure may be the same as steps S210 through S230 of FIG. 2.

[0080] 5 illustrates an example procedure for dynamic discovery of new NWDAFs in the FL execution phase when information about the server NWDAF is unknown at the new client NWDAF for which support for FL is applicable, according to some embodiments of the present disclosure. In some embodiments, client NWDAF #1 120-C-1 through client NWDAF #N 120-CN may be selected by the server NWDAF 120-S to participate in the current round of federated learning. In some embodiments, new client NWDAFs, client NWDAF #N+1 120-C-N+1 through client NWDAF #N+X 120-C-N+X, may have the ability to participate in the next round of training process.

[0081] In some embodiments, the procedure is shown in FIG. 5 and described as follows:

[0082] Similar to step S405, in step S505, the server NWDAF 120-S may register with the NRF 160 for the federated learning procedure with the following parameters: - Federated Learning (FL) Correlation ID -Analysis ID

[0083] In step S510, the server NWDAF 120-S may obtain information about the new client NWDAF by either subscribing to an event in which the new client NWDAF is registered or by discovering the client NWDAF via the NRF 160.

[0084] Similar to step S415, in step S515, before starting the next round of training, the server NWDAF 120-S may select a client NWDAF from among NWDAF #1 120-C-1 through NWDAF #N+X 120-C-N+X based on the updated information of the client NWDAF 120-C. In some embodiments, the procedure may be the same as steps S210 through S230 of FIG. 2.

[0085] In some embodiments, the server NWDAF 120-S may dynamically obtain information about new client NWDAFs via the NRF 160 when reselection of a client NWDAF becomes necessary, either by subscribing to an event to which the new client NWDAF is registered or by discovering the NRF 160. However, details of how the server NWDAF 120-S obtains analytical information (e.g., NF load, etc.) 120-C of the client NWDAF and how it completes federated learning operations at the client NWDAF 120-C are not shown in solution #51.

[0086] As mentioned above, the maintenance of the FL process among multiple NWDAFs in the 5GC is added to the conclusions of Section 8.8 of TR 23.700-81 (mainly Principle 5). Principle 5: An NWDAF including an MTLF as an FL server may determine the final list of NWDAFs including MTLFs as FL clients through an initial FL request to FL clients to determine the availability and compatibility of FL clients. During the FL procedure, an NWDAF including an MTLF as an FL server may trigger reselection, addition, or removal of FL clients based on local policy or the status of FL clients, e.g., load, availability, capability, latency, accuracy, etc., and may perform new FL client discovery via the NRF. FL clients can dynamically join or leave FL operations during the execution phase.

[0087] However, in the FL process and implementation maintenance, it is still unclear how the server NWDAF obtains the analytical information of the client NWDAF (e.g., the load of the NFs) and how to terminate the federated learning operation in the client NWDAF.

[0088] In some embodiments of the present disclosure, it is proposed to complete the maintenance of the FL process and implementation. In some embodiments, a procedure for a server NWDAF to obtain analytical information of a client NWDAF and a procedure for completing federated learning operations in the client NWDAF are shown.

[0089] In some embodiments, the following situations may be considered for the server NWDAF to obtain analysis information of the client NWDAF, such as NF load: The server NWDAF subscribes to some other NWDAF (referred to as a "supporting NWDAF") for analysis information of the client NWDAF. In some embodiments, the supporting NWDAF may notify the server NWDAF of the analysis results. The server NWDAF subscribes to the client NWDAF for analysis information. The client NWDAF may perform analysis on itself and notify the server NWDAF of the analysis results.

[0090] In some embodiments, the analytical information may be sent to the server NWDAF as a notification, either periodically or dynamically, when certain predetermined statuses are achieved.

[0091] In some embodiments, two possible situations may be considered to terminate the federated learning operation at the client NWDAF. - The client NWDAF leaves the federated learning process - The server NWDAF removes the client NWDAF from the FL process

[0092] In some embodiments, for both of the above two procedures, two methods for ML model information exchange between the server NWDAF and the client NWDAF in the FL execution phase may be considered, respectively. Reuse existing services (or their extensions) for ML model information exchange between server NWDAF and client NWDAF, e.g., Nnwdaf_MLModelProvision as described in TS 23.288, v18.0.0. - Use a new service for ML model information exchange between server NWDAF and client NWDAF

[0093] In some embodiments, it is proposed to complete the federated learning process and maintenance of the implementation. In some embodiments, procedures are shown for the server NWDAF to obtain analytical information of the client NWDAF through the assisting NWDAF and directly from the client NWDAF, respectively. In some embodiments, a procedure is shown for completing the federated learning operation in the client NWDAF, taking into account two different methods of ML model information exchange in the FL execution phase. In some embodiments, the two methods for ML model information exchange include: Reuse existing services (or their extensions), e.g., Nnwdaf_MLModelProvision, for ML model information exchange between server NWDAF and client NWDAF. In some embodiments, a new service is used for ML model information exchange between the server NWDAF and the client NWDAF.

[0094] Although the maintenance of the federated learning process is concluded in TR 23.700-81, some aspects are unclear, such as how a server NWDAF obtains analytical information from a client NWDAF and how to terminate the federated learning operation at the client NWDAF. In some embodiments of the present disclosure, it is proposed to complete the maintenance of the federated learning process and implementation. Procedures are provided for the server NWDAF to obtain analytical information from a client NWDAF through a supporting NWDAF or directly from the client NWDAF. In some embodiments of the present disclosure, a procedure is provided for terminating the federated learning operation at the client NWDAF, taking into account two different methods for ML model information exchange in the FL execution phase: reusing an existing service (or its extension) and using a new service.

[0095] 6 illustrates an example system for collecting analytical information, according to some embodiments of the present disclosure, for a server NWDAF 120-S to obtain analytical information of a client NWDAF 120-C through a supporting NWDAF 120-A.

[0096] 6, the server NWDAF 120-S may subscribe to the assisting NWDAF 120-A for analytical information, such as NF load, regarding the client NWDAF 120-C. The assisting NWDAF 120-A may perform an analysis regarding the client NWDAF 120-C and notify the server NWDAF 120-S of the analytical information periodically or dynamically when a certain predetermined status is achieved.

[0097] Alternatively, the server NWDAF 120-S may subscribe to the client NWDAF 120-C for analytical information, and the client NWDAF 120-C may perform analysis on itself and notify the server NWDAF 120-S of the analytical information periodically or dynamically when a certain predetermined status is achieved.

[0098] In some embodiments, the FL process may be ongoing between the server NWDAF 120-S and the client NWDAF 120-C, or the client NWDAF 120-C may be a new candidate client NWDAF that the server NWDAF 120-S selects for the FL process.

[0099] 7 illustrates an example system for terminating a federated learning operation in a client NWDAF 120-C during an FL execution phase, in accordance with some embodiments of the present disclosure.

[0100] In (a) of Figure 7, the client NWDAF 120-C may leave the federated learning operation during the FL execution phase. In (b) of Figure 7, the server NWDAF 120-S may terminate the federated learning operation for the client NWDAF 120-C that is removed from the FL process.

[0101] 8 illustrates an exemplary scenario for analytical information collection according to some embodiments of the present disclosure. Figure 8 illustrates an exemplary system for a server NWDAF 120-S to obtain analytical information of a client NWDAF 120-C through a supporting NWDAF 120-A (e.g., as shown in (a)) and directly from the client NWDAF 120-C (e.g., as shown in (b)).

[0102] 8(a), the server NWDAF 120-S may subscribe to the assisting NWDAF 120-A for analysis information regarding the client NWDAF 120-C, such as analysis ID="NF Load Information." The assisting NWDAF 120-A may collect data from other NFs (which may also be the client NWDAF 120-C), AFs, and / or OAM 810 within the 5GC, perform an analysis for the analysis ID of the client NWDAF 120-C based on the collected data, and notify the server NWDAF 120-S of the analysis result periodically or dynamically (e.g., when a certain predetermined status is achieved).

[0103] As shown in FIG. 8(b), the server NWDAF 120-S may subscribe to the client NWDAF 120-C for its own analysis information, such as analysis ID="NF Load Information." The client NWDAF 120-C may collect data from other NFs, AFs, and / or OAMs 810 in the 5GC, perform an analysis for its own analysis ID based on the collected data and possibly also on its own data, and notify the server NWDAF 120-S of the analysis results periodically or dynamically (e.g., when a certain predetermined status is achieved).

[0104] In some embodiments, the FL process may be ongoing between the server NWDAF 120-S and the client NWDAF 120-C, or the client NWDAF 120-C may be a new candidate client NWDAF that the server NWDAF 120-S selects for the FL process.

[0105] 9 illustrates an exemplary scenario for terminating a federated learning operation in a client NWDAF in accordance with some embodiments of the present disclosure. Figure 9 illustrates an exemplary system for terminating a federated learning operation in a client NWDAF 120-C during an FL execution phase by reusing an existing service (or an extension thereof) (as shown in figure (a)), using a new service for the client NWDAF 120-C to leave (as shown in figure (b)), and using a new service for the server NWDAF 120-S to terminate (as shown in figure (c)).

[0106] 9(a), the server NWDAF 120-S and the client NWDAF 120-C may unsubscribe from each other to ML model information exchange by invoking the Nnwdaf_MLModelProvision_Unsubscribe service operation during the FL execution phase. After receiving the unsubscribe request, the server NWDAF 120-S may stop sharing ML model information with the client NWDAF 120-C, and the client NWDAF 120-C may terminate the corresponding federated learning operation and stop sharing its local ML model information with the server NWDAF 120-S.

[0107] 9(b), the client NWDAF 120-C may send an Nnwdaf_MLTraining_Quit request to the server NWDAF 120-S to leave the federated learning process. After receiving the leave request from the client NWDAF 120-C, the server NWDAF 120-S may stop sharing ML model information with the client NWDAF 120-C and send an Nnwdaf_MLTraining_Quit response to the client NWDAF 120-C. After receiving the leave response, the client NWDAF 120-C may terminate the corresponding federated learning operation and stop sharing local ML model information with the server NWDAF 120-S.

[0108] 9(c), the server NWDAF 120-S may send an Nnwdaf_MLTraining_Terminate request to the client NWDAF 120-C to terminate the federated learning operation at the client NWDAF 120-C and stop sharing ML model information with the client NWDAF 120-C. After receiving the termination request from the server NWDAF 120-S, the client NWDAF 120-C may terminate the corresponding federated learning operation, stop sharing local ML model information with the server NWDAF 120-S, and send an Nnwdaf_MLTraining_Terminate response to the server NWDAF 120-S.

[0109] 10 illustrates an example procedure for collecting analytical information according to some embodiments of the present disclosure. Figure 10 illustrates two example procedures for the server NWDAF 120-S to obtain analytical information of the client NWDAF 120-C during the federated learning execution phase, corresponding to the following two cases: -Case 1: The server NWDAF 120-S obtains analytical information of the client NWDAF 120-C from the supporting NWDAF 120-A. -Case 2: The server NWDAF 120-S obtains analytical information of the client NWDAF 120-C from the client NWDAF 120-C.

[0110] The corresponding procedure is described below.

[0111] Case 1 In step S1005, the server NWDAF 120-S may subscribe to the assisting NWDAF 120-A for analytical information of the client NWDAF 120-C by invoking the Nnwdaf_AnalyticsSubscription_Subscribe service operation (e.g., analysis ID="NF Load Information"). In some embodiments, exemplary procedures for setting up subscriptions are set forth by Sections 6.1.1 and 7.2.2 (relating to Nnwdaf_AnalyticsSubscription_Subscribe) of TS 23.288, V18.0.0.

[0112] In step S1010, the supporting NWDAF 120-A may collect data for analysis from other NFs, AFs, OAM 810, etc., and may also collect data from the client NWDAF 120-C. In some embodiments, exemplary information regarding data collection for analysis is provided in Sections 6.3 through 6.16 (relating to Analysis IDs and Inputs) of TS 23.288, V18.0.0.

[0113] In step S1015, after collecting data, the supporting NWDAF 120-A may perform an analysis for the client NWDAF 120-C based on the collected data, for example, on an analysis ID, such as NF load information.

[0114] In step S1020, the supporting NWDAF 120-A may notify the server NWDAF 120-S of the analysis results periodically or dynamically (e.g., when a certain predetermined status is achieved) by invoking the Nnwdaf_AnalyticsSubscription_Notify service operation. In some embodiments, exemplary procedures for sending notifications are set forth in sections 6.1.1 and 7.2.4 (relating to Nnwdaf_AnalyticsSubscription_Notify) of TS 23.288, V18.0.0. In some embodiments, exemplary information regarding outputs is set forth in sections 6.3 through 6.16 (relating to analysis IDs and outputs) of TS 23.288, V18.0.0.

[0115] Case 2 In step S1025, the server NWDAF 120-S may subscribe to (all or part of) the client NWDAF 120-C for analytical information of the client NWDAF 120-C by invoking the Nnwdaf_AnalyticsSubscription_Subscribe service operation (e.g., analysis ID="NF Load Information"). In some embodiments, exemplary procedures for setting up subscriptions are set forth in sections 6.1.1 and 7.2.2 (relating to Nnwdaf_AnalyticsSubscription_Subscribe) of TS 23.288, V18.0.0.

[0116] In step S1030, the client NWDAF 120-C may collect data for analysis from other NFs, AFs, OAM 810, etc. In some embodiments, exemplary information regarding data collection for analysis is provided in Sections 6.3 to 6.16 (regarding Analysis IDs and Inputs) of TS 23.288, V18.0.0.

[0117] In step S1035, after collecting the data, the client NWDAF 120-C may perform an analysis ID for itself, e.g., an analysis on NF load information, based on the collected data and possibly also on its own data, and may also be its own data.

[0118] In step S1040, the client NWDAF 120-C may notify the server NWDAF 120-S of the analysis results periodically or dynamically (e.g., when a certain predetermined status is achieved) by invoking the Nnwdaf_AnalyticsSubscription_Notify service operation. In some embodiments, exemplary procedures for sending notifications are set forth in sections 6.1.1 and 7.2.4 (relating to Nnwdaf_AnalyticsSubscription_Notify) of TS 23.288, V18.0.0. In some embodiments, exemplary information regarding outputs is set forth in sections 6.3 through 6.16 (relating to analysis IDs and outputs) of TS 23.288, V18.0.0.

[0119] 11A and 11B illustrate an example procedure for terminating an FL process at a client NWDAF in accordance with some embodiments of the present disclosure. Figures 11A and 11B illustrate a procedure for terminating a federated learning operation at a client NWDAF 120-C during the FL execution phase, corresponding to the following four cases: -Case 1: An existing service (or its extension) is used for ML model information exchange and (one or more) client NWDAFs 120-C decide to leave the FL process. -Case 2: An existing service (or its extension) is used for ML model information exchange and the server NWDAF 120-S decides to remove (one or more) client NWDAF(s) 120-C from the FL process. -Case 3: A new service is used for ML model information exchange and the (one or more) client NWDAF(s) 120-C decides to leave the FL process. -Case 4: A new service is used for ML model information exchange and the server NWDAF 120-S decides to remove (one or more) client NWDAF(s) 120-C from the FL process.

[0120] The corresponding procedure is described below.

[0121] As shown in both Figures 11A and 11B, in step S1105, the federated learning process is underway, and in the FL execution phase, ML model information is exchanged between the server NWDAF 120-S and the client NWDAF 120-C by using an existing service (or an extension thereof) (e.g., the Nnwdaf_MLModelProvision service) or a new service (e.g., the Nnwdaf_MLTraining service or any other possible new service).

[0122] The remaining steps for cases 1-4 are described as follows:

[0123] Case 1 in Figure 11A In step S1110, the client NWDAF 120-C may decide to withdraw from the FL process.

[0124] In step S1115, the client NWDAF 120-C may unsubscribe to the server NWDAF 120-S for ML model information exchange by invoking the Nnwdaf_MLModelProvision_Unsubscribe service operation using the FL correlation ID, a cause code (e.g., the client NWDAF 120-C is leaving the FL process due to a detailed reason (e.g., a change in availability, a change in capability, etc.)), etc.

[0125] In some embodiments, the client NWDAF 120-C may send an Nnwdaf_MLModelProvision_Unsubscribe request message to the server NWDAF 120-S, and the server NWDAF 120-S may respond with an Nnwdaf_MLModelProvision_Unsubscribe response message. In some other embodiments, the client NWDAF 120-C may send an Nnwdaf_MLModelProvision_Unsubscribe request message to the server NWDAF 120-S, and the server NWDAF 120-S may respond with another Nnwdaf_MLModelProvision_Unsubscribe request message rather than an Nnwdaf_MLModelProvision_Unsubscribe response message. In either case, the client NWDAF and the server NWDAF may reach an agreement on the termination of the FL process at the client NWDAF.

[0126] In some embodiments, when the client NWDAF 120-C leaves, the context of this FL process in the client NWDAF 120-C is cleared.

[0127] In some embodiments, exemplary unsubscribing instructions are provided in sections 6.2A.1 and 7.5.3 (on Nnwdaf_MLModelProvision_Unsubscribe) of TS 23.288 V18.0.0.

[0128] In step S1120, the server NWDAF 120-S and the client NWDAF 120-C may stop FL operations relative to each other.

[0129] In step 1120a, the server NWDAF 120-S may stop federated learning operations related to the client NWDAF 120-C with respect to the FL process.

[0130] In step 1120b, the client NWDAF 120-C may stop federated learning operations associated with the FL process.

[0131] Case 2 in Figure 11A In step S1125, the server NWDAF 120-S may decide to remove the client NWDAF 120-C from the FL process.

[0132] In some embodiments, the client NWDAF 120-C may also decide to withdraw from the FL process as well.

[0133] In step S1130, the server NWDAF 120-S may unsubscribe to the client NWDAF 120-C for ML model information exchange by invoking the Nnwdaf_MLModelProvision_Unsubscribe service operation using the FL correlation ID, a cause code (e.g., the client NWDAF 120-C has been unselected by the server NWDAF 120-S for the FL process, or the FL process has been paused, etc.).

[0134] In some embodiments, the server NWDAF 120-S may send an Nnwdaf_MLModelProvision_Unsubscribe request message to the client NWDAF 120-C, and the client NWDAF 120-C may respond with an Nnwdaf_MLModelProvision_Unsubscribe response message. In some other embodiments, the server NWDAF 120-S may send an Nnwdaf_MLModelProvision_Unsubscribe request message to the client NWDAF 120-C, and the server NWDAF 120-S may respond with another Nnwdaf_MLModelProvision_Unsubscribe request message rather than an Nnwdaf_MLModelProvision_Unsubscribe response message. In either case, the client NWDAF and server NWDAF may reach an agreement on the termination of the FL process at the client NWDAF.

[0135] In some embodiments, when a client NWDAF 120-C is deselected from an FL process, the context of this FL process in the client NWDAF 120-C is cleared.

[0136] In some embodiments, an exemplary unsubscribe description is provided in sections 6.2A.1 and 7.5.3 (on Nnwdaf_MLModelProvision_Unsubscribe) of TS 23.288, V18.0.0.

[0137] In step S1135, the server NWDAF 120-S and the client NWDAF 120-C may stop FL operations relative to each other.

[0138] In step S1135a, the server NWDAF 120-S may stop federated learning operations related to the client NWDAF 120-C with respect to the FL process.

[0139] In step S1135b, the client NWDAF 120-C may stop federated learning operations related to the FL process.

[0140] Case 3 in Figure 11B In step S1140, the client NWDAF 120-C may decide to leave the federated learning process.

[0141] In step S1145, the client NWDAF 120-C may send a request to leave the FL process to the server NWDAF 120-S, for example, by calling the Nnwdaf_MLTraining_Quit request service operation (or using any other possible new service for ML model information exchange in the FL execution phase) using the FL correlation ID, a cause code (e.g., change in availability, change in capacity, etc.).

[0142] In step S1150, the server NWDAF 120-S may decide to remove the client NWDAF 120-C from the FL process and stop FL operations associated with the client NWDAF 120-C.

[0143] In step S1155, the server NWDAF 120-S may respond to the client NWDAF 120-C's request to leave by invoking the Nnwdaf_MLTraining_Quit response service operation (or using any other possible new service for ML model information exchange in the FL execution phase) with parameters (e.g., the time for the client NWDAF 120-C to leave, etc.).

[0144] In step S1160, the client NWDAF 120-C may stop the federated learning operation and clear the context associated with the FL process.

[0145] Case 4 in Figure 11B In step S1165, the server NWDAF 120-S may decide to remove the client NWDAF 120-C from the FL process.

[0146] In step S1170, the server NWDAF 120-S may send a request to the client NWDAF 120-C to terminate the FL operation of the FL process in the client NWDAF 120-C, for example, by calling the Nnwdaf_MLTraining_Terminate request service operation (or using any other possible new service for ML model information exchange in the FL execution phase) using the FL correlation ID, a cause code (e.g., the client NWDAF 120-C has been deselected by the server NWDAF 120-S for the FL process, or the FL process has been paused, etc.).

[0147] In step S1175, the client NWDAF 120-C may stop federated learning operations related to the FL process.

[0148] In some embodiments, when a client NWDAF 120-C is deselected from an FL process, the context of this FL process in the client NWDAF 120-C is cleared.

[0149] In step S1180, the client NWDAF 120-C may respond to the server NWDAF 120-S's termination request by invoking the Nnwdaf_MLTraining_Terminate response service operation (or using any other possible new service for ML model information exchange in the FL execution phase).

[0150] In step S1185, the server NWDAF 120-S may stop federated learning operations related to the client NWDAF 120-C with respect to the FL process.

[0151] According to some embodiments of the present disclosure described above, FL can be supported in a core network such as a 5GC network. In this way, the federated learning process and implementation maintenance can be completed. Procedures are provided for the server NWDAF to obtain analytical information of the client NWDAF through the assisting NWDAF and directly from the client NWDAF, respectively. Furthermore, a procedure for completing the federated learning operation at the client NWDAF is provided taking into account two different methods of ML model information exchange during the FL execution phase: a method of reusing an existing service (or its extension) and a method of using a new service.

[0152] 12 is a flowchart of an exemplary method 1200 in a server associated with an FL process, according to one embodiment of the present disclosure. Method 1200 may be implemented in a server NWDAF (e.g., server NWDAF 120-S shown in FIG. 6). Method 1200 may include at least one of steps S1210 and S1220. However, the present disclosure is not limited thereto. In some other embodiments, method 1200 may include more steps, fewer steps, different steps, or any combination thereof. Furthermore, the steps of method 1200 may be implemented in a different order than described herein. Furthermore, in some embodiments, steps in method 1200 may be split into multiple substeps and implemented by different entities, and / or multiple steps in method 1200 may be combined into a single step.

[0153] The method 1200 may begin with at least one of steps S1210 and S1220.

[0154] In step S1210, the server may send a first message to one or more clients associated with the FL process indicating that the corresponding client has not been selected by the server for the FL process and / or that the FL process has been paused.

[0155] In step S1220, the server may receive a second message from one or more clients associated with the FL process indicating that the corresponding client is leaving the FL process.

[0156] In some embodiments, at least one of the first message and the second message may further indicate at least one of an FL correlation ID and a cause code. In some embodiments, the cause code may indicate at least one of: that the corresponding client is not selected by the server for the FL process when the cause code is indicated by the first message; that the FL process is suspended when the cause code is indicated by the first message; a change in availability associated with the corresponding client when the cause code is indicated by the corresponding second message; and a change in capability associated with the corresponding client when the cause code is indicated by the corresponding second message.

[0157] In some embodiments, method 1200 may further include at least one of receiving a third message from at least one of the one or more clients in response to the corresponding first message, the third message indicating that the FL process has terminated or will terminate at the at least one client, and sending a fourth message to at least one of the one or more clients in response to the corresponding second message, the fourth message indicating that the at least one client has been removed or will be removed from the FL process. In some embodiments, the fourth message may further indicate a time at which the at least one client will leave the FL process.

[0158] In some embodiments, method 1200 may further include at least one of: stopping one or more FL operations of an FL process associated with at least one of the one or more clients in response to sending a corresponding first message; stopping one or more FL operations of an FL process associated with at least one of the one or more clients in response to receiving a corresponding second message; and stopping one or more FL operations of an FL process associated with at least one of the one or more clients in response to receiving a corresponding third message. In some embodiments, the first message may cause the client to stop one or more FL operations for the FL process. In some embodiments, at least one of the first message, the second message, the third message, and the fourth message may be a message specified in 3GPP TS 23.288, V18.0.0 and / or any of its previous releases, or an extension of that message. In some embodiments, at least one of the first message, the second message, the third message, and the fourth message may not be a message specified in 3GPP TS 23.288, V18.0.0 and / or any of its earlier releases, or an extension of that message.

[0159] In some embodiments, the first message may be one of an Nnwdaf_MLModelProvision_Unsubscribe request message and an Nnwdaf_MLTraining_Terminate request message, and / or the third message may be a corresponding one of an Nnwdaf_MLModelProvision_Unsubscribe request message, an Nnwdaf_MLModelProvision_Unsubscribe response message, and an Nnwdaf_MLTraining_Terminate response message. In some embodiments, the second message may be one of an Nnwdaf_MLModelProvision_Unsubscribe request message and an Nnwdaf_MLTraining_Quit request message, and / or the fourth message may be a corresponding one of an Nnwdaf_MLModelProvision_Unsubscribe request message, an Nnwdaf_MLModelProvision_Unsubscribe response message, and an Nnwdaf_MLTraining_Quit response message.

[0160] In some embodiments, prior to the sending of the first message and / or the receiving of the second message, method 1200 may further include at least one of sending ML model information to one or more clients and receiving ML model information from one or more clients. In some embodiments, the NWDAF may be hosted by a server. In some embodiments, the one or more NWDAFs may be hosted by one or more clients. In some embodiments, method 1200 may further include any of the steps in any of the methods described with reference to FIG. 13.

[0161] 13 is a flowchart of an example method 1300 in a server associated with an FL process, according to one embodiment of the present disclosure. Method 1300 may be implemented in a server NWDAF (e.g., server NWDAF 120-S shown in FIG. 6). Method 1300 may include steps S1310 and S1320. However, the present disclosure is not limited thereto. In some other embodiments, method 1300 may include more steps, different steps, or any combination thereof. Furthermore, the steps of method 1300 may be implemented in a different order than described herein. Furthermore, in some embodiments, steps in method 1300 may be split into multiple substeps and implemented by different entities, and / or multiple steps in method 1300 may be combined into a single step.

[0162] Method 1300 may begin at step S1310, where the server may receive one or more fifth messages indicating analytical data associated with one or more clients in the FL process and / or one or more candidate clients selected for the FL process.

[0163] In step S1320, the server may select at least one of the one or more candidate clients and / or one or more clients for the FL process based at least on the analytical data.

[0164] In some embodiments, the one or more fifth messages may be received from at least one of at least one of the one or more clients, at least one of the one or more candidate clients, and one or more network nodes. In some embodiments, the one or more network nodes may be network nodes that assist the server in collecting analytics data associated with at least one of the one or more clients and / or at least one of the one or more candidate clients.

[0165] In some embodiments, prior to receiving the one or more fifth messages, method 1300 may further include at least one of: sending a sixth message to at least one of the one or more clients and / or at least one of the one or more candidate clients to subscribe to analytics data associated with the at least one client and / or at least one of the one or more candidate clients; and sending a sixth message to at least one of the one or more network nodes to subscribe to analytics data associated with the at least one of the one or more clients and / or at least one of the one or more candidate clients. In some embodiments, the fifth message may be a Nnwdaf_AnalyticsSubscription_Notify request message, and the sixth message may be a Nnwdaf_AnalyticsSubscription_Subscribe request message.

[0166] In some embodiments, the analytical data may include data related to at least one of load, availability, capacity, latency, and accuracy. In some embodiments, receiving the one or more fifth messages may be performed periodically and / or dynamically in response to an event. In some embodiments, the NWDAF may be hosted by a server. In some embodiments, the NWDAF may be hosted by one or more clients. In some embodiments, the NWDAF may be hosted by one or more candidate clients. In some embodiments, method 1300 may further include any of the steps in any of the methods described with reference to FIG. 12.

[0167] 14 is a flowchart of an exemplary method 1400 in a client associated with an FL process, according to one embodiment of the present disclosure. Method 1400 may be implemented in a client NWDAF (e.g., client NWDAF 120-C shown in FIG. 6). Method 1400 may include at least one of steps S1410 and S1420. However, the present disclosure is not limited thereto. In some other embodiments, method 1400 may include more steps, fewer steps, different steps, or any combination thereof. Furthermore, the steps of method 1400 may be implemented in a different order than described herein. Furthermore, in some embodiments, steps in method 1400 may be split into multiple substeps and implemented by different entities, and / or multiple steps in method 1400 may be combined into a single step.

[0168] The method 1400 may begin with at least one of steps S1410 and S1420.

[0169] In step S1410, the client may receive a first message from a server associated with the FL process indicating that the FL process has not been selected by the server and / or that the FL process has been suspended.

[0170] In step S1420, the client may send a second message to the server associated with the FL process indicating that the client is leaving the FL process.

[0171] In some embodiments, at least one of the first message and the second message may further indicate at least one of an FL correlation ID and a cause code. In some embodiments, the cause code may indicate at least one of: that the client is not selected by the server for the FL process when the cause code is indicated by the first message; that the FL process is suspended when the cause code is indicated by the first message; a change in availability associated with the client when the cause code is indicated by the second message; and a change in capability associated with the client when the cause code is indicated by the second message.

[0172] In some embodiments, method 1400 may further include at least one of: sending a third message to the server in response to the first message indicating that the FL process has been terminated or is to be terminated at the client; and receiving a fourth message from the server in response to the second message indicating that the client has been removed or is to be removed from the FL process. In some embodiments, the fourth message may further indicate a time at which the client will leave the FL process. In some embodiments, method 1400 may further include at least one of: stopping one or more FL operations of the FL process in response to receiving the first message; clearing the context of the FL process in response to receiving the first message; stopping one or more FL operations of the FL process in response to sending the second message; clearing the context of the FL process in response to sending the second message; stopping one or more FL operations of the FL process in response to receiving the fourth message; and clearing the context of the FL process in response to receiving the fourth message.

[0173] In some embodiments, at least one of the first message, the second message, the third message, and the fourth message may be a message specified in 3GPP TS 23.288, V18.0.0 and / or any of its earlier releases, or an extension of that message. In some embodiments, at least one of the first message, the second message, the third message, and the fourth message may not be a message specified in 3GPP TS 23.288, V18.0.0 and / or any of its earlier releases, or an extension of that message.

[0174] In some embodiments, the first message may be one of an Nnwdaf_MLModelProvision_Unsubscribe request message and an Nnwdaf_MLTraining_Terminate request message, and / or the third message may be a corresponding one of an Nnwdaf_MLModelProvision_Unsubscribe request message, an Nnwdaf_MLModelProvision_Unsubscribe response message, and an Nnwdaf_MLTraining_Terminate response message. In some embodiments, the second message may be one of an Nnwdaf_MLModelProvision_Unsubscribe request message and an Nnwdaf_MLTraining_Quit request message, and / or the fourth message may be a corresponding one of an Nnwdaf_MLModelProvision_Unsubscribe request message, an Nnwdaf_MLModelProvision_Unsubscribe response message, and an Nnwdaf_MLTraining_Quit response message.

[0175] In some embodiments, prior to receiving the first message and / or sending the second message, method 1400 may further include at least one of receiving ML model information from a server and sending the ML model information to the server. In some embodiments, the NWDAF may be hosted by a client. In some embodiments, the NWDAF is hosted by a server. In some embodiments, method 1400 may further include any of the steps in any of the methods described with reference to FIG. 15.

[0176] FIG. 15 is a flowchart of an exemplary method 1500 in a client associated with an FL process or a candidate client selected in an FL process, according to one embodiment of the present disclosure. Method 1500 may be implemented in a client NWDAF (e.g., client NWDAF 120-C shown in FIG. 6). Method 1500 may include at least one of steps S1510 and S1520. However, the present disclosure is not limited thereto. In some other embodiments, method 1500 may include more steps, fewer steps, different steps, or any combination thereof. Furthermore, the steps of method 1500 may be implemented in a different order than described herein. Furthermore, in some embodiments, steps in method 1500 may be split into multiple substeps and implemented by different entities, and / or multiple steps in method 1500 may be combined into a single step.

[0177] The method 1500 may begin with at least one of steps S1510 and S1520.

[0178] In step S1510, the client or candidate client may send a fifth message to a server associated with the FL process indicating analytical data associated with the client or candidate client.

[0179] In step S1520, the client or candidate client may send a seventh message to one or more network nodes indicating data associated with the client or candidate client, which data may be used as input data in determining analytical data associated with the client or candidate client.

[0180] In some embodiments, the one or more network nodes may be network nodes that assist the server in collecting analytical data associated with the client or candidate client. In some embodiments, prior to the step of sending the fifth message, method 1500 may further include at least one of receiving from the server a sixth message to subscribe to analytical data associated with the client or candidate client and receiving from at least one of the one or more network nodes an eighth message to request input data associated with the client or candidate client. In some embodiments, method 1500 may further include collecting data used to determine analytical data associated with the client or candidate client from at least one of the one or more NFs, one or more AFs, and one or more OAM nodes, and performing analysis on the collected data to determine analytical data associated with the client or candidate client.

[0181] In some embodiments, the fifth message may be a Nnwdaf_AnalyticsSubscription_Notify request message, and the sixth message may be a Nnwdaf_AnalyticsSubscription_Subscribe request message. In some embodiments, the analytics data may include data related to at least one of load, availability, capacity, latency, and accuracy. In some embodiments, sending the fifth message may be performed periodically and / or dynamically in response to an event. In some embodiments, the NWDAF may be hosted by a client or a candidate client. In some embodiments, the NWDAF may be hosted by a server. In some embodiments, method 1500 may further include any of the steps in any of the methods described with reference to FIG. 14 .

[0182] 16 is a flowchart of an example method 1600 in a network node according to one embodiment of the present disclosure. Method 1600 may be implemented in an assisting NWDAF (e.g., the assisting NWDAF 120-A shown in FIG. 6). Method 1600 may include step S1610. However, the present disclosure is not limited thereto. In some other embodiments, method 1600 may include more steps, different steps, or any combination thereof. Furthermore, the steps of method 1600 may be implemented in a different order than described herein. Furthermore, in some embodiments, steps in method 1600 may be split into multiple substeps and implemented by different entities, and / or multiple steps in method 1600 may be combined into a single step.

[0183] Method 1600 may begin at step S1610, where a network node may send a fifth message to a server associated with the FL process indicating analytical data associated with one or more clients associated with the FL process and / or one or more candidate clients selected by the server for the FL process.

[0184] In some embodiments, the network node may be a network node that assists the server in collecting analytics data associated with the one or more clients and / or one or more candidate clients. In some embodiments, prior to the step of sending the fifth message, method 1600 may further include receiving a sixth message from the server to subscribe to analytics data associated with the one or more clients and / or one or more candidate clients.

[0185] In some embodiments, the fifth message may be a Nnwdaf_AnalyticsSubscription_Notify request message and the sixth message may be a Nnwdaf_AnalyticsSubscription_Subscribe request message. In some embodiments, prior to sending the fifth message, the method 1600 may further include collecting data from at least one of the one or more clients, one or more candidate clients, one or more NFs, one or more AFs, and one or more OAM nodes, used to determine analytics data associated with the one or more clients and / or one or more candidate clients, and performing analytics on the collected data to determine analytics data associated with the one or more clients and / or one or more candidate clients.

[0186] In some embodiments, the analytical data may include data related to at least one of load, availability, capacity, latency, and accuracy. In some embodiments, the step of sending the fifth message may be performed periodically and / or dynamically in response to an event. In some embodiments, the NWDAF may be hosted by the client and / or candidate client. In some embodiments, the NWDAF may be hosted by a server.

[0187] 17 schematically illustrates an embodiment of a configuration that may be used in a server, a client, and / or a network node, according to an embodiment of the present disclosure. Included in the configuration 1700 is a processing unit 1706, e.g., with a digital signal processor (DSP) or a central processing unit (CPU). The processing unit 1706 may be a single unit or multiple units for performing different actions of the procedures described herein. The configuration 1700 may also include an input unit 1702 for receiving signals from other entities and an output unit 1704 for providing signals to other entities. The input unit 1702 and the output unit 1704 may be configured as an integrated entity or as separate entities.

[0188] Additionally, configuration 1700 may comprise at least one computer program product 1708 in the form of non-volatile or volatile memory, e.g., Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory, and / or a hard drive. Computer program product 1708 comprises a computer program 1710 including code / computer-readable instructions that, when executed by processing unit(s) 1706 in configuration 1700, cause configuration 1700 and / or servers and / or clients and / or network nodes in which configuration 1700 is included therein to perform actions, e.g., of the procedures previously described in conjunction with Figures 2 through 16 or any other variations.

[0189] Computer program 1710 may be configured as computer program code structured in computer program modules 1710A and / or 1710B. Thus, in an exemplary embodiment, when configuration 1700 is used for a server associated with an FL process, the code in the computer program of configuration 1700 includes at least one of: module 1710A configured to send a first message to one or more clients associated with the FL process indicating that the corresponding client has not been selected by the server for the FL process and / or that the FL process has been paused; and module 1710B configured to receive a second message from one or more clients associated with the FL process indicating that the corresponding client has left the FL process.

[0190] Additionally or alternatively, computer program 1710 may be configured as computer program code structured in computer program modules 1710C and 1710D. Thus, in an example embodiment, when configuration 1700 is used in a server associated with an FL process, the code in the computer program of configuration 1700 includes module 1710C configured to receive one or more fifth messages indicating analytical data associated with one or more clients in the FL process and / or one or more candidate clients selected for the FL process, and module 1710D configured to select at least one of the one or more candidate clients and / or one or more clients for the FL process based at least on the analytical data.

[0191] Additionally or alternatively, computer program 1710 may be configured as computer program code structured in computer program modules 1710E and / or 1710F. Thus, in an example embodiment, when configuration 1700 is used for a client associated with an FL process, the code within the computer program of configuration 1700 includes at least one of: module 1710E configured to receive a first message from a server associated with the FL process indicating that the client has not been selected by the server for the FL process and / or that the FL process has been suspended; and module 1710F configured to send a second message to the server associated with the FL process indicating that the client is leaving the FL process.

[0192] Additionally or alternatively, computer program 1710 may be configured as computer program code structured in computer program modules 1710G and / or 1710H. Thus, in an example embodiment, when configuration 1700 is used for a client associated with an FL process or a candidate client selected for an FL process, the code within the computer program of configuration 1700 includes at least one of: a module 1710G configured to send a fifth message indicating analytical data associated with the client or candidate client to a server associated with the FL process; and a module 1710H configured to send a seventh message to one or more network nodes indicating data associated with the client or candidate client, the data being used as input data in determining the analytical data associated with the client or candidate client.

[0193] Additionally or alternatively, computer program 1710 may be configured as computer program code structured in computer program module 1710I. Thus, in an example embodiment, when configuration 1700 is used in a network node, code within the computer program of configuration 1700 includes module 1710I configured to send, to a server associated with the FL process, a fifth message indicating analytical data associated with one or more clients associated with the FL process and / or one or more candidate clients selected by the server for the FL process.

[0194] The computer program modules may essentially perform the actions of the flows shown in Figures 2 through 16 to emulate a server, a client, and / or a network node. In other words, when different computer program modules are executed in the processing unit 1706, they may correspond to different modules in a server, a client, and / or a network node.

[0195] Although the code means in the embodiment disclosed above in conjunction with FIG. 17 are implemented as computer program modules that, when executed on a processing unit, cause the configuration to perform the actions described above in conjunction with the aforementioned figures, at least one of the code means may, in alternative embodiments, be implemented at least in part as a hardware circuit.

[0196] The processor may be a single CPU (Central Processing Unit), but may also comprise two or more processing units. For example, the processor may include a general-purpose microprocessor, an instruction set processor and / or related chipset, and / or a special-purpose microprocessor such as an application-specific integrated circuit (ASIC). The processor may also comprise on-board memory for caching purposes. The computer program may be carried by a computer program product connected to the processor. The computer program product may comprise a computer-readable medium on which the computer program is stored. For example, the computer program product may be a flash memory, a random access memory (RAM), a read-only memory (ROM), or an EEPROM, and the computer program modules described above may, in alternative embodiments, be distributed on different computer program products in the form of memories in servers, clients, and / or network nodes.

[0197] Corresponding to the above-described method 1200, an exemplary server associated with the FL process is provided. Figure 18 is a block diagram of a server 1800 according to one embodiment of the present disclosure. Server 1800 may be, for example, server NWDAF 120-S in some embodiments.

[0198] The server 1800 may be configured to perform the method 1200 described above with respect to Figure 12. As shown in Figure 18, the server 1800 may include at least one of a sending module 1810 configured to send a first message to one or more clients associated with the FL process indicating that the corresponding client has not been selected by the server for the FL process and / or that the FL process has been paused, and a receiving module 1820 configured to receive a second message from the one or more clients associated with the FL process indicating that the corresponding client has left the FL process.

[0199] The above modules 1810 and / or 1820 may be implemented as a pure hardware solution or as a combination of software and hardware, for example by a processor or microprocessor and appropriate software, as well as memory for storage of the software, a programmable logic device (PLD), or one or more other electronic components or processing circuits configured to perform the actions described above and shown, for example, in Figure 12. Furthermore, the server 1800 may comprise one or more further modules, each of which may perform any of the steps of the method 1200 described with reference to Figure 12.

[0200] Corresponding to the above-described method 1300, an exemplary server associated with the FL process is provided. Figure 19 is a block diagram of a server 1900 according to one embodiment of the present disclosure. Server 1900 may be, for example, server NWDAF 120-S in some embodiments.

[0201] The server 1900 may be configured to perform the method 1300 described above with respect to Figure 13. As shown in Figure 19, the server 1900 may comprise a receiving module 1910 configured to receive one or more fifth messages indicating analytical data associated with one or more clients in the FL process and / or one or more candidate clients selected for the FL process, and a selecting module 1920 configured to select at least one from the one or more candidate clients and / or one or more clients for the FL process based at least on the analytical data.

[0202] The above modules 1910 and 1920 may be implemented as a pure hardware solution or as a combination of software and hardware, for example by a processor or microprocessor and appropriate software, as well as memory for storing the software, a PLD, or one or more other electronic components or processing circuits configured to perform the actions described above and shown, for example, in Figure 13. Furthermore, the server 1900 may comprise one or more further modules, each of which may perform any of the steps of the method 1300 described with reference to Figure 13.

[0203] Corresponding to the above-described method 1400, an exemplary client associated with the FL process is provided. Figure 20 is a block diagram of a client 2000 according to one embodiment of the present disclosure. The client 2000 may be, for example, a client NWDAF 120-C in some embodiments.

[0204] The client 2000 may be configured to perform the method 1400 described above with respect to Figure 14. As shown in Figure 20, the client 2000 may include at least one of a receiving module 2010 configured to receive a first message from a server associated with the FL process indicating that the client has not been selected by the server for the FL process and / or that the FL process has been paused, and a sending module 2020 configured to send a second message to the server associated with the FL process indicating that the client is leaving the FL process.

[0205] The above modules 2010 and / or 2020 may be implemented as a pure hardware solution or as a combination of software and hardware, for example by a processor or microprocessor and appropriate software, as well as memory for storage of the software, a PLD, or one or more other electronic components or processing circuits configured to perform the actions described above and shown, for example, in Figure 14. Furthermore, the client 2000 may comprise one or more additional modules, each of which may perform any of the steps of the method 1400 described with reference to Figure 14.

[0206] Corresponding to the method 1500 described above, an exemplary client associated with the FL process, or a candidate client selected for the FL process, is provided. Figure 21 is a block diagram of a client or candidate client 2100, according to one embodiment of the present disclosure. The client or candidate client 2100 may be, for example, a client NWDAF 120-C in some embodiments.

[0207] The client or candidate client 2100 may be configured to perform the method 1500 described above with respect to Figure 15. As shown in Figure 21, the client or candidate client 2100 may comprise at least one of a first sending module 2110 configured to send a fifth message indicating analytical data associated with the client or candidate client to a server associated with the FL process, and a second sending module 2120 configured to send a seventh message indicating data associated with the client or candidate client to one or more network nodes, the data being used as input data in determining the analytical data associated with the client or candidate client.

[0208] The above modules 2110 and / or 2120 may be implemented as a pure hardware solution or as a combination of software and hardware, for example by a processor or microprocessor and appropriate software, as well as memory for storage of the software, a PLD, or one or more other electronic components or processing circuits configured to perform the actions described above and shown, for example, in Figure 15. Furthermore, the client or candidate client 2100 may comprise one or more further modules, each of which may perform any of the steps of the method 1500 described with reference to Figure 15.

[0209] An exemplary network node is provided corresponding to the above-described method 1600. Figure 22 is a block diagram of a network node 2200 according to one embodiment of the present disclosure. The network node 2200 may be, for example, the supporting NWDAF 120-A in some embodiments.

[0210] The network node 2200 may be configured to perform the method 1600 described above with respect to Figure 16. As shown in Figure 22, the network node 2200 may comprise a sending module 2210 configured to send, to a server associated with the FL process, a fifth message indicating analytical data associated with one or more clients associated with the FL process and / or one or more candidate clients selected by the server for the FL process.

[0211] The above module 2210 may be implemented as a pure hardware solution or as a combination of software and hardware, for example by a processor or microprocessor and appropriate software, as well as memory for storing the software, a PLD, or one or more other electronic components or processing circuits configured to perform the actions described above and shown, for example, in Figure 16. Furthermore, the network node 2200 may comprise one or more further modules, each of which may perform any of the steps of the method 1600 described with reference to Figure 16.

[0212] The present disclosure has been described above with reference to embodiments of the present disclosure. However, these embodiments are provided for illustrative purposes only, rather than limiting the present disclosure. The scope of the present disclosure is defined by the appended claims and their equivalents. Those skilled in the art can make various alterations and modifications without departing from the scope of the present disclosure, all of which fall within the scope of the present disclosure.

[0213] Abbreviation Explanation 5GC 5G Core Network AF Application Features AI artificial intelligence DML Distributed Machine Learning FL Associative Learning ML Machine Learning MTLF model training logic function NEF network publishing function NF Network Function NRF Network Repository Function NWDAF Network Data Analysis Function OAM Operations Administration Maintenance

Claims

1. A method (1200) in a server (120-S) associated with a federated learning (FL) process, comprising: Sending a first message to one or more clients (120-C) associated with the FL process indicating that the corresponding client has not been selected by the server (120-S) for the FL process and / or that the FL process has been suspended (S1210); and receiving a second message from one or more clients (120-C) associated with the FL process indicating that the corresponding client is leaving the FL process (S1220).

2. At least one of the first message and the second message comprises: FL correlation identifier (ID), and 12. The method of claim 1, further indicating at least one of: a cause code;

3. The cause code is: When the cause code is indicated by the first message, the corresponding client is not selected by the server (120-S) for the FL process; the FL process is suspended when the cause code is indicated by the first message; and a change in availability associated with the corresponding client when the cause code is indicated by the corresponding second message; and and a change in capabilities associated with the corresponding client when the cause code is indicated by the corresponding second message.

4. receiving a third message from at least one of the one or more clients (120-C) in response to the corresponding first message, the third message indicating that the FL process has been or will be terminated at the at least one client; and 4. The method (1200) of claim 1, further comprising at least one of: sending a fourth message to at least one of the one or more clients (120-C) in response to the corresponding second message, the fourth message indicating that the at least one client has been or is to be removed from the FL process.

5. 5. The method (1200) of claim 4, wherein the fourth message further indicates a time at which the at least one client will leave the FL process.

6. suspending one or more FL operations of the FL process associated with at least one of the one or more clients (120-C) in response to sending the corresponding first message; responsive to receiving the corresponding second message, stopping one or more FL operations of the FL process associated with at least one of the one or more clients (120-C); and The method (1200) of any one of claims 1 to 5, further comprising at least one of: stopping one or more FL operations of the FL process associated with at least one of the one or more clients (120-C) in response to receiving the corresponding third message.

7. The method (1200) of any one of claims 1 to 6, wherein the first message causes the client (120-C) to stop one or more FL operations for the FL process.

8. The first message is one of a Nnwdaf_MLModelProvision_Unsubscribe request message and a Nnwdaf_MLTraining_Terminate request message; and / or 8. The method (1200) of claim 1, wherein the third message is a corresponding one of a Nnwdaf_MLModelProvision_Unsubscribe request message, a Nnwdaf_MLModelProvision_Unsubscribe response message, and a Nnwdaf_MLTraining_Terminate response message.

9. The second message is one of a Nnwdaf_MLModelProvision_Unsubscribe request message and a Nnwdaf_MLTraining_Quit request message; and / or 9. The method (1200) of claim 1, wherein the fourth message is a corresponding one of a Nnwdaf_MLModelProvision_Unsubscribe request message, a Nnwdaf_MLModelProvision_Unsubscribe response message, and a Nnwdaf_MLTraining_Quit response message.

10. Prior to the step of sending the first message (S1210) and / or the step of receiving the second message (S1220), the method (1200) may further comprise: Sending machine learning (ML) model information to said one or more clients (120-C); and The method (1200) of any one of claims 1 to 9, further comprising at least one of: receiving ML model information from said one or more clients (120-C).

11. a Network Data Analysis Function (NWDAF) hosted by said server (120-S); and / or The method (1200) of any one of claims 1 to 10, wherein one or more NWDAFs are hosted by said one or more clients (120-C).

12. The method (1200) of any one of claims 1 to 11, further comprising any of the steps in the method (1300) of any one of claims 13 to 21.

13. A method (1300) in a server (120-S) associated with an FL process, comprising: receiving (S1310) one or more fifth messages indicating analytical data associated with one or more clients (120-C) in the FL process and / or one or more candidate clients (120-C) selected for the FL process; and selecting (S1320) at least one of the one or more candidate clients and / or the one or more clients for the FL process based at least on the analytical data.

14. The one or more fifth messages include: At least one of said one or more clients (120-C); At least one of the one or more candidate clients (120-C), and The method (1300) of claim 13, wherein the signal is received from at least one of the one or more network nodes (120-A).

15. 15. The method of claim 14, wherein the one or more network nodes are network nodes that assist the server in collecting analytical data associated with at least one of the one or more clients and / or at least one of the one or more candidate clients.

16. Prior to receiving the one or more fifth messages (S1310), the method (1300) may further comprise: sending a sixth message to at least one of the one or more clients (120-C) and / or at least one of the one or more candidate clients to subscribe to analytical data associated with the at least one client and / or the at least one candidate client; and 16. The method (1300) of claim 13, further comprising at least one of: sending a sixth message to at least one of the one or more network nodes (120-A) to subscribe to analytics data associated with at least one of the one or more clients (120-C) and / or at least one of the one or more candidate clients (120-C).

17. The fifth message is a Nnwdaf_AnalyticsSubscription_Notify request message; 17. The method (1300) of any one of claims 13 to 16, wherein the sixth message is a Nnwdaf_AnalyticsSubscription_Subscribe request message.

18. The analytical data is load, availability, ability, Latency, and 18. The method (1300) of any one of claims 13 to 17, comprising data relating to at least one of:

19. 19. The method (1300) of any one of claims 13 to 18, wherein the step of receiving one or more fifth messages (S1310) is performed periodically and / or dynamically in response to an event.

20. a Network Data Analysis Function (NWDAF) hosted by said server (120-S); and / or The NWDAF is hosted by said one or more clients (120-C), and / or The method (1300) of any one of claims 13 to 19, wherein the NWDAF is hosted by said one or more candidate clients (120-C).

21. 21. The method (1300) of any one of claims 13 to 20, further comprising any of the steps in the method (1200) of any one of claims 1 to 12.

22. a processor (1706); A server (120-S, 1700, 1800, 1900) comprising: a memory (1708) storing instructions that, when executed by the processor (1706), cause the processor (1706) to implement a method (1200, 1300) according to any one of claims 1 to 21.

23. A method (1400) in a client (120-C) associated with an FL process, comprising: receiving a first message from a server (120-S) associated with the FL process indicating that the client (120-C) has not been selected by the server (120-S) for the FL process and / or that the FL process has been suspended (S1410); and and sending a second message to a server (120-S) associated with the FL process indicating that the client (120-C) is leaving the FL process (S1420).

24. At least one of the first message and the second message comprises: FL correlation identifier (ID), and 24. The method (1400) of claim 23, further indicating at least one of: a cause code;

25. The cause code is: When the cause code is indicated by the first message, the client (120-C) is not selected by the server (120-S) for the FL process; the FL process is suspended when the cause code is indicated by the first message; and a change in availability associated with said client (120-C) when said cause code is indicated by said second message; and a change in capabilities associated with the client (120-C) when the cause code is indicated by the second message.

26. In response to the first message, sending a third message to the server (120-S) indicating that the FL process has been or will be terminated at the client (120-C); and The method (1400) of any one of claims 23 to 25, further comprising at least one of: receiving a fourth message from the server (120-S) in response to the second message, the fourth message indicating that the client (120-C) has been or is to be removed from the FL process.

27. 27. The method (1400) of claim 26, wherein the fourth message further indicates a time at which the client (120-C) will leave the FL process.

28. suspending one or more FL operations of the FL process in response to receiving the first message; clearing the context of the FL process in response to receiving the first message; suspending one or more FL operations of the FL process in response to sending the second message; clearing the context of the FL process in response to sending the second message; suspending one or more FL operations of the FL process in response to receiving the fourth message; and clearing the context of the FL process in response to receiving the fourth message.

29. The first message is one of a Nnwdaf_MLModelProvision_Unsubscribe request message and a Nnwdaf_MLTraining_Terminate request message; and / or 29. The method (1400) of any one of claims 23 to 28, wherein the third message is a corresponding one of a Nnwdaf_MLModelProvision_Unsubscribe request message, a Nnwdaf_MLModelProvision_Unsubscribe response message, and a Nnwdaf_MLTraining_Terminate response message.

30. The second message is one of a Nnwdaf_MLModelProvision_Unsubscribe request message and a Nnwdaf_MLTraining_Quit request message; and / or 30. The method (1400) of any one of claims 23 to 29, wherein the fourth message is a corresponding one of a Nnwdaf_MLModelProvision_Unsubscribe request message, a Nnwdaf_MLModelProvision_Unsubscribe response message, and a Nnwdaf_MLTraining_Quit response message.

31. Prior to the step of receiving the first message (S1410) and / or the step of sending the second message (S1420), the method (1400) may further comprise: receiving machine learning (ML) model information from the server (120-S); and The method (1400) of any one of claims 23 to 30, further comprising at least one of: sending ML model information to said server (120-S).

32. a Network Data Analysis Function (NWDAF) hosted by said client (120-C); and / or The method (1400) of any one of claims 23 to 31, wherein the NWDAF is hosted by the server (120-S).

33. 43. The method (1400) of any one of claims 23 to 32, further comprising any of the steps in the method (1500) of any one of claims 34 to 42.

34. A method (1500) in a client (120-C) associated with a FL process or a candidate client (120-C) selected for said FL process, comprising: Sending a fifth message indicating analytical data associated with the client (120-C) or the candidate client (120-C) to a server (120-S) associated with the FL process (S1510); and transmitting (S1520) to one or more network nodes (120-A) a seventh message indicating data associated with the client (120-C) or the candidate client (120-C), the data being used as input data in determining analytical data associated with the client (120-C) or the candidate client (120-C).

35. The method (1500) of claim 34, wherein the one or more network nodes (120-A) are network nodes that assist the server (120-S) in collecting analytical data associated with the client (120-C) or the candidate client (120-C).

36. Before the step of transmitting the fifth message (S1510), the method (1500) further comprises: receiving a sixth message from the server (120-S) for subscribing to the analytical data associated with the client (120-C) or the candidate client (120-C); and The method (1500) of claim 34 or 35, further comprising at least one of: receiving an eighth message from at least one of the one or more network nodes (120-A) to request the input data associated with the client (120-C) or the candidate client (120-C).

37. collecting data used to determine analytical data associated with the client (120-C) or the candidate client (120-C) from at least one of one or more network functions (NFs), one or more application functions (AFs), and one or more operations, administration, and maintenance (OAM) nodes; 37. The method (1500) of any one of claims 34 to 36, further comprising: performing analysis on the collected data to determine the analytical data associated with the client (120-C) or the candidate client (120-C).

38. The fifth message is a Nnwdaf_AnalyticsSubscription_Notify request message; 38. The method (1500) of any one of claims 34 to 37, wherein the sixth message is a Nnwdaf_AnalyticsSubscription_Subscribe request message.

39. The analytical data is load, availability, ability, Latency, and 39. The method (1500) of any one of claims 34 to 38, comprising data relating to at least one of:

40. 40. The method (1500) of any one of claims 34 to 39, wherein the step of sending (S1510) the fifth message is performed periodically and / or dynamically in response to an event.

41. a Network Data Analysis Function (NWDAF) hosted by said client (120-C) or said candidate client (120-C); and / or The method (1500) of any one of claims 34 to 40, wherein the NWDAF is hosted by the server (120-S).

42. 42. The method (1500) of any one of claims 34 to 41, further comprising any of the steps in the method (1400) of any one of claims 23 to 33.

43. a processor (1706); A client (120-C, 1700, 2000, 2100) comprising: a memory (1708) storing instructions that, when executed by the processor (1706), cause the processor (1706) to implement a method (1400, 1500) described in any one of claims 23 to 42.

44. A method (1600) in a network node (120-A), comprising: A method (1600) including sending (S1610) to a server (120-S) associated with an FL process a fifth message indicating analytical data associated with one or more clients (120-C) associated with the FL process and / or one or more candidate clients (120-C) selected by the server (120-S) for the FL process.

45. The method (1600) of claim 44, wherein the network node (120-A) is a network node that assists the server (120-S) in collecting analytical data associated with the one or more clients (120-C) and / or the one or more candidate clients (120-C).

46. Before the step of sending the fifth message (S1610), the method (1600) further comprises: The method (1600) of claim 44 or 45, further comprising receiving a sixth message from the server (120-S) for subscribing to the analytical data associated with the one or more clients (120-C) and / or the one or more candidate clients (120-C).

47. The fifth message is a Nnwdaf_AnalyticsSubscription_Notify request message; 47. The method (1600) of any one of claims 44 to 46, wherein the sixth message is a Nnwdaf_AnalyticsSubscription_Subscribe request message.

48. Before the step of sending the fifth message (S1610), the method (1600) further comprises: collecting data from at least one of the one or more clients (120-C), the one or more candidate clients (120-C), one or more network functions (NFs), one or more application functions (AFs), and one or more operations, administration, and maintenance (OAM) nodes, used to determine analytical data associated with the one or more clients (120-C) and / or the one or more candidate clients (120-C); 48. The method (1600) of any one of claims 44 to 47, further comprising: performing analysis on the collected data to determine the analytical data associated with the one or more clients (120-C) and / or the one or more candidate clients (120-C).

49. The analytical data is load, availability, ability, Latency, and 49. The method (1600) of any one of claims 44 to 48, comprising data relating to at least one of:

50. 50. The method (1600) of any one of claims 44 to 49, wherein the step of sending (S1610) the fifth message is performed periodically and / or dynamically in response to an event.

51. The NWDAF is hosted by the client (120-C) and / or the candidate client (120-C), and / or The method (1600) of any one of claims 44 to 50, wherein the NWDAF is hosted by the server (120-S).

52. a processor (1706); A network node (120-A, 1700, 2200) comprising: a memory (1708) storing instructions that, when executed by the processor (1706), cause the processor (1706) to perform a method (1600) described in any one of claims 44 to 51.

53. 52. A computer program (1710) comprising instructions that, when executed by at least one processor (1706), cause the at least one processor (1706) to perform a method according to any one of claims 1 to 21, 23 to 42, and 44 to 51.

54. 54. A carrier (1708) containing the computer program (1710) of claim 53, the carrier (1708) being one of an electronic signal, an optical signal, a radio signal, or a computer readable storage medium.

55. A communication system (10) for supporting federated learning (FL), comprising: A server (120-S) according to claim 22; A communication system (10) comprising one or more clients (120-C) according to claim 43.

56. A communication system (10) according to claim 55, further comprising one or more network nodes (120-A) according to claim 52.

Images