Support of vertical federated learning in enablement layer

Abstract

According to one aspect, a method in a network node is described. The method includes receiving a first request from a consumer network node. The first request requests at least one artificial intelligence machine learning (AI / ML) operation. The method also includes, in response to receiving the first request, determining that VFL training between dataset domains is to be used, transmitting a second request to selected one or more AIMLE clients associated to the dataset domains, the second request being a request to evaluate a capability and availability of at least one AIMLE client to join the VFL training, and determining the one or more AIMLE clients for the VFL training based at least on a response to the second request from the selected one or more AIMLE clients. Further, the method also includes coordinating the determined one or more AIMLE clients to perform the VFL training.

Description

[0001] SUPPORT OF VERTICAL FEDERATED LEARNING IN ENABLEMENT LAYER

[0002] TECHNICAL FIELD

[0003] The present disclosure relates to wireless communications, and in particular, to support vertical federated learning (VFL) in the enablement layer.

[0004] BACKGROUND

[0005] The Third Generation Partnership Project (3 GPP) has developed and is developing standards for Fourth Generation (4G) (also referred to as Long Term Evolution (LTE)) and Fifth Generation (5G) (also referred to as New Radio (NR)) wireless communication systems. Such systems provide, among other features, broadband communication between network nodes, such as base stations, and mobile user equipments (UE), as well as communication between network nodes and between WDs. The 3GPP is also developing standards for Sixth Generation (6G) wireless communication networks.

[0006] Artificial intelligence and / or machine learning (AI / ML) relevant topics have been studied in 3GPP system aspects working group 2 (SA2) Release 18. For example, Horizontal Federated Learning (HFL) among multiple network data analytics functions (NWDAFs) within 5G Core (5GC) has been studied in SA2 Release 18 work item eNA_Ph3 (in the 3GPP Technical Report (TR) 23.700-81 V18.0.0 and in the normative work specified in 3GPP Technical Specification (TS) 23.288 V18.4.0). 5GC support of application layer AI / ML operations has been studied in SA2 Release 18 work item AIMLsys (in 3GPP TR 23.700-80 V18.0.0, the normative work in 3GPP TS 23.501 V18.4.0 and 3GPP TS 23.502 V18.4.0). Another 3GPP SA6 Release 19 key issue on supporting VFL at the enablement layer has been agreed in 3GPP TR 23.700-82 V0.2.0.

[0007] Key Issue (KI) on supporting VFL at enablement layer (e.g., corresponding to clause 5.4 in 3GPP TR 23.700-82 (V0.2.0))

[0008] There may be two types of federated learning defined, horizontal and vertical federated learning. Horizontal federated learning, or sample-based federated learning, is introduced in the scenarios that data sets share the same feature space but have different samples. Vertical federated learning or feature-based federated learning is applicable to the cases where two data sets share the same sample space but differ in feature space. In both horizontal and vertical federated learning, model parameters from each local model training function are sent to a model aggregator (or FL collaborator) to calculate an aggregate model. The model aggregator provides updated model parameters to each ML model training entity that each ML model training entity use to re-train its own model, thus allowing every local model training function to have a trained model using data from multiple sources.

[0009] FIG. 1 shows an example vertical FL (VFL) process. If VFL is supported by the enablement layer, some issues are distinct and need to be studied separately from KI #3, especially:

[0010] How to ensure that all training functions have an aligned sample range, e.g., the same users, to support VFL; and

[0011] How to discover what features are available between domains (for the same sample range) in order to support VFL.

[0012] The KI may reuse the study result of KI#3 on FL for aspects which are applicable to both horizontal and vertical FL (e.g. organizing training).

[0013] In addition, existing specifications may describe procedures for FL among multiple NWDAFs within 5GC given in 3GPP TS 23.288 V18.4.0 and procedures for 5GS assistance to application layer FL operations. Further, there may be procedures for regular ML model training process and HFL among multiple NWDAFs within 5GC and for 5GS assistance to application layer FL operation.

[0014] SUMMARY

[0015] Some embodiments advantageously provide methods, systems, and apparatuses for supporting VFL in the enablement layer, e.g., involving multiple UEs at different domains and may involve network nodes such as 5GC Network Functions (NFs). The study aspect listed in the agreed KI for 3GPP Release 19 in SA6 may be considered.

[0016] One or more embodiments address one or more of the following:

[0017] How a network node such as an AI / ML Enablement Server gets the information (datasets and features, etc.) of VFL members (e.g., UE(s) at different domains and 5GC NWDAF); and

[0018] How the network node such as an AI / ML Enablement Server utilizes the obtained information (e.g., information on datasets and features) to coordinate a VFL process between domains.

[0019] According to one aspect, a method in a network node configured as an artificial intelligence machine learning enablement (AIMLE) sever that supports vertical federated learning (VFL) and is configured to communicate with at least one user equipment (UE) supporting an AIMLE client and at least one consumer network node is described. The method includes receiving a first request from a consumer network node. The first request requests at least one artificial intelligence machine learning (AI / ML) operation. The method also includes, in response to receiving the first request, determining that VFL training between dataset domains is to be used, transmitting a second request to selected one or more AIMLE clients associated to the dataset domains, the second request being a request to evaluate a capability and availability of at least one AIMLE client to join the VFL training, and determining the one or more AIMLE clients for the VFL training based at least on a response to the second request from the selected one or more AIMLE clients. Further, the method also includes coordinating the determined one or more AIMLE clients to perform the VFL training.

[0020] In some embodiments, the method further includes obtaining information associated with discovery of VFL members in response to the first request.

[0021] In some other embodiments, the first request comprises ML model information. In some embodiments, the first request comprises a list of UEs having AIMLE clients.

[0022] In some other embodiments, the first request comprises a training type.

[0023] In some embodiments, the second request comprises one or more of: (A) a test task; (B) a data source; (C) one or more feature identifiers of a dataset; and (D) available time for supporting VFL operations.

[0024] In some other embodiments, the response to the second request includes a result of joining VFL training.

[0025] In some embodiments, the response to the second request includes a test result in response to a test task in the second request.

[0026] According to another aspect, a network node configured as an artificial intelligence machine learning enablement (AIMLE) sever, that supports vertical federated learning (VFL) and is configured to communicate with at least one user equipment (UE) supporting an AIMLE client and at least one consumer network node is described. The network node is configured to perform one or more steps corresponding to the method implemented in the network node.

[0027] According to one aspect, a method in a user equipment (UE) is described. The UE includes an artificial intelligence machine learning enablement (AIMLE) client. The method includes receiving a request from an AIMLE server to evaluate a capability and availability of the AIMLE client to join a vertical federated learning (VFL) training and transmitting a response indicating whether the AIMLE client is able to join the VFL training.

[0028] In some embodiments, the request comprises one or more of: (A) a test task; (B) a data source; (C) one or more feature identifiers of a dataset; and (D) available time for supporting VFL operations.

[0029] In some other embodiments, the method further includes transmitting a response to the request, where the response includes a test result in response to a test task in the request.

[0030] According to another aspect, a user equipment (UE) comprising an artificial intelligence machine learning enablement (AIMLE) client is described. The UE is configured to perform one or more steps corresponding to the method implemented in the UE.

[0031] BRIEF DESCRIPTION OF THE DRAWINGS

[0032] A more complete understanding of the present embodiments, and the attendant advantages and features thereof, will be more readily understood by reference to the following detailed description when considered in conjunction with the accompanying drawings wherein:

[0033] FIG. 1 shows an example VFL process;

[0034] FIG. 2 is a schematic diagram of an exemplary network architecture illustrating a communication system connected via an intermediate network to a host computer according to the principles in the present disclosure;

[0035] FIG. 3 is a block diagram of a communication system including a network node configured to communicate with a user equipment over an at least partially wireless connection according to some embodiments of the present disclosure;

[0036] FIG. 4 is a flowchart of an exemplary process in a network node according to some embodiments of the present disclosure;

[0037] FIG. 5 is a flowchart of an exemplary process in a network node according to some embodiments of the present disclosure;

[0038] FIG. 6 is a flowchart of an exemplary process in a UE according to some embodiments of the present disclosure;

[0039] FIG. 7 according to some embodiments of the present disclosure; and

[0040] FIG. 8 according to some embodiments of the present disclosure. DETAILED DESCRIPTION

[0041] Before describing in detail exemplary embodiments, it is noted that the embodiments reside primarily in combinations of apparatus components and processing steps related to supporting VFL in the enablement layer. Accordingly, components have been represented where appropriate by conventional symbols in the drawings, showing only those specific details that are pertinent to understanding the embodiments so as not to obscure the disclosure with details that will be readily apparent to those of ordinary skill in the art having the benefit of the description herein. Like numbers refer to like elements throughout the description.

[0042] As used herein, relational terms, such as “first” and “second,” “top” and “bottom,” and the like, may be used solely to distinguish one entity or element from another entity or element without necessarily requiring or implying any physical or logical relationship or order between such entities or elements. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the concepts described herein. As used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises,” “comprising,” “includes” and / or “including” when used herein, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0043] In embodiments described herein, the joining term, “in communication with” and the like, may be used to indicate electrical or data communication, which may be accomplished by physical contact, induction, electromagnetic radiation, radio signaling, infrared signaling or optical signaling, for example. One having ordinary skill in the art will appreciate that multiple components may interoperate and modifications and variations are possible of achieving the electrical and data communication.

[0044] In some embodiments described herein, the term “coupled,” “connected,” and the like, may be used herein to indicate a connection, although not necessarily directly, and may include wired and / or wireless connections.

[0045] The term “network node” used herein can be any kind of network node comprised in a radio network which may further comprise any of base station (BS), radio base station, base transceiver station (BTS), base station controller (BSC), radio network controller (RNC), g Node B (gNB), evolved Node B (eNB or eNodeB), Node B, multi- standard radio (MSR) radio node such as MSR BS, multi-cell / multicast coordination entity (MCE), integrated access and backhaul (IAB) node, relay node, donor node controlling relay, radio access point (AP), transmission points, transmission nodes, Remote Radio Unit (RRU) Remote Radio Head (RRH), a core network node (e.g., mobile management entity (MME), self-organizing network (SON) node, a coordinating node, positioning node, MDT node, etc.), an external node (e.g., 3rd party node, a node external to the current network), nodes in distributed antenna system (DAS), a spectrum access system (SAS) node, an element management system (EMS), etc. The network node may also comprise test equipment, a core node such as network functions (NFs) (e.g., network element function (NEF), network repository function (NRF), NWDAF, etc.), one or more servers such as an enablement server, an AI / ML enablement server, etc., a registry such as an AI / ML registry, vertical application layer (VAL) server, etc. The term “radio node” used herein may be used to also denote a user equipment (UE) such as a wireless device (WD) or a radio network node.

[0046] In some embodiments, the non-limiting terms wireless device (WD) or a user equipment (UE) are used interchangeably. The UE herein can be any type of wireless device capable of communicating with a network node or another UE over radio signals, such as a wireless device (WD). The UE may also be a radio communication device, target device, device to device (D2D) UE, machine type UE or UE capable of machine to machine communication (M2M), low-cost and / or low-complexity UE, a sensor equipped with UE, Tablet, mobile terminals, smart phone, laptop embedded equipped (LEE), laptop mounted equipment (LME), USB dongles, Customer Premises Equipment (CPE), an Internet of Things (loT) device, or a Narrowband loT (NB-IOT) device, a client such as a VAL client, an Application Data Analytics Enablement Client (ADAEC), AI / ML Enablement Client (AIMLEC), etc.

[0047] Also, in some embodiments the generic term “radio network node” is used. It can be any kind of a radio network node which may comprise any of base station, radio base station, base transceiver station, base station controller, network controller, RNC, evolved Node B (eNB), Node B, gNB, Multi-cell / multicast Coordination Entity (MCE), IAB node, relay node, access point, radio access point, Remote Radio Unit (RRU) Remote Radio Head (RRH).

[0048] Note that although terminology from one particular wireless system, such as, for example, 3GPP LTE and / or New Radio (NR), may be used in this disclosure, this should not be seen as limiting the scope of the disclosure to only the aforementioned system. Other wireless systems, including without limitation Wide Band Code Division Multiple Access (WCDMA), Worldwide Interoperability for Microwave Access (WiMax), Ultra Mobile Broadband (UMB) and Global System for Mobile Communications (GSM), may also benefit from exploiting the ideas covered within this disclosure.

[0049] Note further, that functions described herein as being performed by a user equipment or a network node may be distributed over a plurality of user equipments and / or network nodes. In other words, it is contemplated that the functions of the network node and user equipment described herein are not limited to performance by a single physical device and, in fact, can be distributed among several physical devices.

[0050] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. It will be further understood that terms used herein should be interpreted as having a meaning that is consistent with their meaning in the context of this specification and the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

[0051] Referring again to the drawing figures, in which like elements are referred to by like reference numerals, there is shown in FIG. 2 a schematic diagram of a communication system 10, according to an embodiment, such as a 3GPP-type cellular network that may support standards such as LTE and / or NR (5G), which comprises an access network 12, such as a radio access network (RAN) 12, and a core network 14. The access network 12 comprises a plurality of network nodes 16a, 16b, 16c (referred to collectively as network nodes 16), such as NBs, eNBs, gNBs or other types of wireless access points, each defining a corresponding coverage area 18a, 18b, 18c (referred to collectively as coverage areas 18). Each network node 16a, 16b, 16c is connectable to the core network 14 over a wired or wireless connection 20. A first UE 22a located in coverage area 18a is configured to wirelessly connect to, or be paged by, the corresponding network node 16a. A second UE 22b in coverage area 18b is wirelessly connectable to the corresponding network node 16b. While a plurality of UEs 22a, 22b (collectively referred to as UEs 22) are illustrated in this example, the disclosed embodiments are equally applicable to a situation where a sole UE is in the coverage area or where a sole UE is connecting to the corresponding network node 16. Note that although only two UEs 22 and three network nodes 16 are shown for convenience, the communication system may include many more UEs 22 and network nodes 16. Core network 14 may include one or more core network nodes (also referred to herein as network nodes 16) having hardware and / or software that generally corresponds to the hardware and / or software in network node 16, but arranged to perform the functions of the particular core network node as opposed to a network node 16 in the access network 12.

[0052] Also, it is contemplated that a UE 22 can be in simultaneous communication and / or configured to separately communicate with more than one network node 16 and more than one type of network node 16. For example, a UE 22 can have dual connectivity with a network node 16 that supports LTE and the same or a different network node 16 that supports NR. As an example, UE 22 can be in communication with an eNB for LTE / E-UTRAN and a gNB for NR / NG-RAN.

[0053] A network node 16 is configured to include a node management unit 24 which is configured to perform any step and / or task and / or process and / or method and / or feature described in the present disclosure, e.g., network node functions. A UE 22 is configured to include a UE management unit 26 which is configured to perform any step and / or task and / or process and / or method and / or feature described in the present disclosure, e.g., UE functions.

[0054] Further, any of the radio access network 12 (and / or its components such as network nodes 16 and / or UEs 22) and / or core network 14 may be in communication with any other network such as a cloud network. Although not shown, core network 14 may include one or more network nodes 16 (and / or UEs 22).

[0055] Example implementations, in accordance with an embodiment, of the UE 22 and network node 16 discussed in the preceding paragraphs will now be described with reference to FIG. 3.

[0056] The communication system 10 includes a network node 16 provided in a communication system 10. Network node 16 includes hardware 28 enabling it to communicate with the UE 22. The hardware 28 may include a radio interface 30 for setting up and maintaining at least a wireless connection 32 with a UE 22 located in a coverage area 18 served by the network node 16. The radio interface 30 may be formed as or may include, for example, one or more RF transmitters, one or more RF receivers, and / or one or more RF transceivers. The radio interface 30 includes an array of antennas 34 to radiate and receive signal(s) carrying electromagnetic waves. In some embodiments, radio interface 30 may be configured for setting up and maintaining at least a wireless / wired connection with other network nodes 16. In some embodiments, network node 16 may include a communication interface configured to perform functions similar to the radio interface functions, e.g., communicate with other network nodes 16 via wired or wireless links.

[0057] In the embodiment shown, the hardware 28 of the network node 16 further includes processing circuitry 36. The processing circuitry 36 may include a processor 38 and a memory 40. In particular, in addition to or instead of a processor, such as a central processing unit, and memory, the processing circuitry 36 may comprise integrated circuitry for processing and / or control, e.g., one or more processors and / or processor cores and / or FPGAs (Field Programmable Gate Array) and / or ASICs (Application Specific Integrated Circuitry) adapted to execute instructions. The processor 38 may be configured to access (e.g., write to and / or read from) the memory 40, which may comprise any kind of volatile and / or nonvolatile memory, e.g., cache and / or buffer memory and / or RAM (Random Access Memory) and / or ROM (Read-Only Memory) and / or optical memory and / or EPROM (Erasable Programmable Read-Only Memory).

[0058] Thus, the network node 16 further has software 42 stored internally in, for example, memory 40, or stored in external memory (e.g., database, storage array, network storage device, etc.) accessible by the network node 16 via an external connection. The software 42 may include application 44 which may include software application configured to provide application functions, such as a functions associated with a service provided to UE 22.

[0059] The software 42 may be executable by the processing circuitry 36. The processing circuitry 36 may be configured to control any of the methods and / or processes described herein and / or to cause such methods, and / or processes to be performed, e.g., by network node 16. Processor 38 corresponds to one or more processors 38 for performing network node 16 functions described herein. The memory 40 is configured to store data, programmatic software code and / or other information described herein. In some embodiments, the software 42 may include instructions that, when executed by the processor 38 and / or processing circuitry 36, causes the processor 38 and / or processing circuitry 36 to perform the processes described herein with respect to network node 16. For example, processing circuitry 36 of the network node 16 may include a node management unit 24 which is configured to perform any step and / or task and / or process and / or method and / or feature described in the present disclosure, e.g., network node functions.

[0060] The communication system 10 further includes the UE 22 already referred to. The

[0061] UE 22 may have hardware 46 that may include a radio interface 48 configured to set up and maintain a wireless connection 32 with a network node 16 serving a coverage area 18 in which the UE 22 is currently located. The radio interface 48 may be formed as or may include, for example, one or more RF transmitters, one or more RF receivers, and / or one or more RF transceivers. The radio interface 48 includes an array of antennas 50 to radiate and receive signal(s) carrying electromagnetic waves. In some embodiments, UE 22 may include a communication interface configured to perform functions similar to radio interface 48, e.g., communicate with other UEs 22 via wired or wireless links.

[0062] The hardware 46 of the UE 22 further includes processing circuitry 52. The processing circuitry 52 may include a processor 54 and memory 56. In particular, in addition to or instead of a processor, such as a central processing unit, and memory, the processing circuitry 52 may comprise integrated circuitry for processing and / or control, e.g., one or more processors and / or processor cores and / or FPGAs (Field Programmable Gate Array) and / or ASICs (Application Specific Integrated Circuitry) adapted to execute instructions. The processor 54 may be configured to access (e.g., write to and / or read from) memory 56, which may comprise any kind of volatile and / or nonvolatile memory, e.g., cache and / or buffer memory and / or RAM (Random Access Memory) and / or ROM (Read-Only Memory) and / or optical memory and / or EPROM (Erasable Programmable Read-Only Memory).

[0063] Thus, the UE 22 may further comprise software 58, which is stored in, for example, memory 56 at the UE 22, or stored in external memory (e.g., database, storage array, network storage device, etc.) accessible by the UE 22. The software 58 may be executable by the processing circuitry 52. The software 58 may include an application 60. The application 60 may be operable to provide a service to a human or non-human user via the UE 22 and / or be configured to provide application client functions, e.g., associated with application 44.

[0064] The processing circuitry 52 may be configured to control any of the methods and / or processes described herein and / or to cause such methods, and / or processes to be performed, e.g., by UE 22. The processor 54 corresponds to one or more processors 54 for performing UE 22 functions described herein. The UE 22 includes memory 56 that is configured to store data, programmatic software code and / or other information described herein. In some embodiments, the software 58 and / or the application 60 may include instructions that, when executed by the processor 54 and / or processing circuitry 52, causes the processor 54 and / or processing circuitry 52 to perform the processes described herein with respect to UE 22. For example, the processing circuitry 52 of the UE 22 may include UE management unit 26 which is configured to perform any step and / or task and / or process and / or method and / or feature described in the present disclosure, e.g., UE functions.

[0065] In some embodiments, network node 16 includes an AUML Enablement Server 62 configured to perform AI / ML Enablement Server functions described herein. Further, UE 22 may include VAL Client 64 configured to perform VAL client functions described herein and an AUML Enablement Client (AIMLEC) 66 configured to perform AIMLEC functions described herein. In some embodiments, AI / ML Enablement Server 62 may be at least in part comprised in Application 44 and be configured to perform AI / ML Enablement Server functions based on and / or using the software and / or hardware corresponding at least to network node 16. VAL Client 64 may be at least in part comprised in Application 60 and be configured to perform VAL client functions based on the software and / or hardware corresponding at least to UE 22. Similarly, AIMLEC 66 may be at least in part comprised in Application 44 and be configured to perform AIMLEC functions based on the software and / or hardware corresponding at least to UE 22.

[0066] In some embodiments, the inner workings of the network node 16 and UE 22 may be as shown in FIG. 3 and independently, the surrounding network topology may be that of FIG. 2.

[0067] The wireless connection 32 between the UE 22 and the network node 16 is in accordance with the teachings of the embodiments described throughout this disclosure. More precisely, the teachings of some of these embodiments may improve the data rate, latency, and / or power consumption and thereby provide benefits such as reduced user waiting time, relaxed restriction on file size, better responsiveness, extended battery lifetime, etc. In some embodiments, a measurement procedure may be provided for the purpose of monitoring data rate, latency and other factors on which the one or more embodiments improve.

[0068] Although FIGS. 2 and 3 show various “units” such as node management unit 24 and UE management unit 26 as being within a respective processor, it is contemplated that these units may be implemented such that a portion of the unit is stored in a corresponding memory within the processing circuitry. In other words, the units may be implemented in hardware or in a combination of hardware and software within the processing circuitry.

[0069] FIG. 4 is a flowchart of an exemplary process in a network node 16. The network node 16 is configured as an artificial intelligence and / or machine learning (AI / ML) enablement server 62 that supports vertical federated learning (VFL) in at least one enablement layer and configured to communicate with at least one user equipment (UE) 22 and / or at least one other network node 16. The VFL is associated with the at least one UE 22 and / or the at least one other network node 16. One or more blocks described herein may be performed by one or more elements of network node 16 such as by one or more of processing circuitry 36 (including the node management unit 24), processor 38, and / or radio interface 30.

[0070] Network node 16 such as via processing circuitry 36 and / or processor 38 and / or radio interface 30is configured to receive (Block S 116) a first request from a consumer network node 16, where the first request requests at least one AUML operation, obtain (Block SI 18) information associated with discovery of VFL members in response to the first request, and transmit (Block S120) a second request to the VFL members based on the obtained information. The second request is a preparation request for the VFL members to evaluate capability and availability to join a VFL training process. The network node 16 is also configured to receive (Block S122) a first response from the VFL members, where the first response includes an indication indicating a decision on whether to join the VFL training process, determine (Block S124) the VFL members for the VFL training process based on at least one of the information associated with discovery of VFL members, the indication, and predetermined criteria, and coordinate (Block S126) with the determined VFL members to perform at least one VFL training operation for the VFL process.

[0071] In some embodiments, a VFL profile associated with the at least one UE 22 and / or the at least one other network node 16 is registered with another network node 16 configured as an AI / ML registry.

[0072] In some other embodiments, the obtaining the information includes transmitting a third request to the other network node 16 configured as the AI / ML registry, where the third request requests the information, and receiving a second response to the third request including the requested information.

[0073] In some embodiments, the VFL members include at least one of: (A) at least one AI / ML enablement client deployed on the at least one UE 22; and (B) at least one network data analytics function (NWDAF).

[0074] In some embodiments, the method further includes reporting to the consumer network node 16 an intermediate training status associated with the VFL process.

[0075] FIG. 5 is a flowchart of an exemplary process in a network node 16. Network node 16 is configured as an artificial intelligence machine learning enablement (AIMLE) sever 62 that supports vertical federated learning (VFL) and is configured to communicate with at least one UE 22 supporting an AIMLE client 66 and at least one consumer network node. One or more blocks described herein may be performed by one or more elements of network node 16 such as by one or more of processing circuitry 36 (including the node management unit 24), processor 38, and / or radio interface 30. Network node 16 such as via processing circuitry 36 and / or processor 38 and / or radio interface 30 is configured to receive (Block S128) a first request from a consumer network node, where the first request requests at least one artificial intelligence machine learning (AI / ML) operation. Network node 16 is also configured to, in response to receiving the first request, determine (Block S130) that VFL training between dataset domains is to be used, transmit (Block S132) a second request to selected one or more AIMLE clients associated to the dataset domains, where the second request is a request to evaluate a capability and availability of at least one AIMLE client to join the VFL training, and determine (Block S134) the one or more AIMLE clients 66 for the VFL training based at least on a response to the second request from the selected one or more AIMLE clients 66. Further, network node 16 is configured to coordinate (Block S136) the determined one or more AIMLE clients 66 to perform the VFL training.

[0076] In some embodiments, the method further includes obtaining information associated with discovery of VFL members in response to the first request.

[0077] In some other embodiments, the first request comprises ML model information. In some embodiments, the first request comprises a list of UEs having AIMLE clients 66.

[0078] In some other embodiments, the first request comprises a training type.

[0079] In some embodiments, the second request comprises one or more of: (A) a test task; (B) a data source; (C) one or more feature identifiers of a dataset; and (D) available time for supporting VFL operations.

[0080] In some other embodiments, the response to the second request includes a result of joining VFL training.

[0081] In some embodiments, the response to the second request includes a test result in response to a test task in the second request.

[0082] FIG. 6 is a flowchart of an exemplary process in a UE 22. The UE includes an artificial intelligence machine learning enablement (AIMLE) client 66. One or more blocks described herein may be performed by one or more elements of UE 22 such as by one or more of processing circuitry 52 (including the UE management unit 26), processor 54, and / or radio interface 48. UE 22 such as via processing circuitry 52 and / or processor 54 and / or radio interface 48 is configured to receive (Block S138) a request from an AIMLE server 62 to evaluate a capability and availability of the AIMLE client 66 to join a vertical federated learning (VFL) training and transmit (Block S140) a response indicating whether the AIMLE client 66 is able to join the VFL training.

[0083] In some embodiments, the request comprises one or more of: (A) a test task; (B) a data source; (C) one or more feature identifiers of a dataset; and (D) available time for supporting VFL operations.

[0084] In some other embodiments, the method further includes transmitting a response to the request, where the response includes a test result in response to a test task in the request.

[0085] Having described the general process flow of arrangements of the disclosure and having provided examples of hardware and software arrangements for implementing the processes and functions of the disclosure, the sections below provide details and examples of arrangements for supporting VFL in the enablement layer.

[0086] In some embodiments, the term “AI / ML Enablement Server” is used and may refer to a network node 16. In some other embodiments, the term “AI / ML Enablement Clients” is used and may refer to UEs 22.

[0087] FIG. 7 shows an example architecture of an AI / ML Enablement Server 62 (comprised in network node 16a) configured for coordinating a VFL training process among multiple UEs 22 (AI / ML Enablement Clients 66a, 66b, ... 66n, collectively referred to as AI / ML Enablement Clients 66). The AI / ML Enablement Clients 66 are deployed on UEs 22, and UEs 22 have the same user space. The AI / ML Enablement Server 62 (comprised in network node 16a) may interact with other network nodes 16 such as 5GC NFs, e.g., NWDAF via NEF (as NWDAF provides ML model training service in 5GC). The VFL process may involve multiple UEs 22 (e.g., AI / ML Enablement Clients 66) and multiple network nodes 16 (e.g., 5GC NFs (NWDAF via NEF)).

[0088] In some embodiments, a procedure of VFL in enablement layer, which may use multiple UEs at different domains and 5GC NFs, is described. FIG. 8 shows a procedure of VFL in the enablement layer. UE 22 may include a VAL client 64 and / or AI / ML Enablement Client (AIMLEC) 66. A VFL approach may be used to train ML model using distributed datasets for the user (i.e. for the same sample). The term sample may correspond to all the data for the user. The sample may be formed by multiple datasets. Each dataset may be at a VFL member. The term VFL member may refer to any UE 22 such as the AI / ML Enablement Client 66 deployed on a UE 22, or any network node 16 such as an NWDAF. Each data set may include sample instances which correspond to the data collected for the user from one of the entities (e.g., VAL Servers from bank and e- commerce company, mobile phone, 5GC NFs, etc.). These distributed datasets have different set of features, and the entities are the data sources of the distributed datasets. The corresponding process may include one or more of the following steps:

[0089] S200. VFL members (AI / ML Enablement Clients 66 which are deployed on UEs 22) register their VFL profile to the network node 16 (AI / ML registry which may be an AI / ML Enablement Server 62, Analytics Data Repository Function (A-ADRF)). VFL member Network node 16b such as (5GC NWDAF) registers its VFL profile to another network node 16 such as an NRF. The VFL profile that VFL member (AI / ML Enablement Client which are deployed on a UE 22) registered may include public land mobile network (PLMN) information (e.g. PLMN ID), VFL capability information, time interval supporting VFL, available data / sample / data source information, available feature information, interoperability information.

[0090] S202. Network node 16c such as consumer (e.g. VAL Server) sends request to the network node 16a such as AI / ML Enablement Server 62 for AI / ML operations, e.g. train a ML model. The request message may include detail requirements, e.g.'.

[0091] ' approach for training an ML model ( Centralized ML, Distributed ML / FL (HFL, VFL), etc.);

[0092] ' initial ML model information ( e.g. ML model, ML model address, ML model ID and information of repository, ML model structure);

[0093] ' intent of ML model (e.g. size of ML model);

[0094] ' ML model metric (e.g. accuracy);

[0095] ' list of UE(s)22 involved in the AI / ML operations;

[0096] ' data sources for train the ML model;

[0097] ' maximum response time; and / or

[0098] ' indication on whether intermediate report is needed or not (e.g. for report training status, which may be used by the consumer to adjust request).

[0099] S204. After receiving the request from the consumer, network node 16a (the AI / ML Enablement Server 62) determines that VFL training between domains is needed based on request from the consumer (if the request on VFL been provided in step S202) or local configuration (according to the requirements from the consumer on the AI / ML operations). If initial ML model is not provided in the request in S202, the network node 16 determines an initial ML model.

[0100] S206. Network node 16a sends request to and get response from the network node 16d for discovery VFL members (AI / ML Enablement Clients 66 which are deployed on UEs 22).

[0101] S206a. Network node 16a (AI / ML Enablement Server 62) sends request to the network node 16d (AI / ML registry) for discovery VFL members (AI / ML Enablement Clients 66 which are deployed on UEs 22 ). The request message may include, e.g.:

[0102] ' requirements on VFL capability;

[0103] ' data / sample (e.g. sample ID);

[0104] ' feature sets of the datasets (e.g. Feature IDs of one dataset, each dataset corresponds to the data / sample instances at one VFL member);

[0105] ' data sources (e.g. UE IDs, NF ID, VAL server IDs);

[0106] ' interoperability information (e.g., ML model interoperability information);

[0107] ' time period of the VFL training process;

[0108] ' UE capability;

[0109] ' area of interest; and / or

[0110] ' UE distribution.

[0111] S206b. Network node 16d (AI / ML registry) responds to the network node 16a (AI / ML Enablement Server 62). The response message from the AI / ML registry contains the VFL members information. In some embodiments, step S206 may be skipped if the AI / ML registry is the AI / ML Enablement Server 62.

[0112] S208. Network node 16a sends request to and get response from NRF (or via NEF) for discovery VFL member (5GC NWDAF). The service for application function (AF) to discovery NF from NRF (via NEF) may not available. Which service to be used and what content in the request to and response message from NRF (via NEF) may need coordination, e.g., with SA2. Steps S206 and S208 may be performed simultaneously or sequentially.

[0113] S210. After the VFL members information is obtained, network node 16a (AI / ML Enablement Server 62) sends preparation request to the VFL members. S210a. Network node 16a (AI / ML Enablement Server 62) sends preparation request to the VFL members (AI / ML Enablement Clients 66 which are deployed on UEs 22). The request message may include e.g. :

[0114] ' test task (compute task or mini training task);

[0115] ' interoperability information (e.g. ML model interoperability information);

[0116] ' requirements on available time for support VFL operations

[0117] ' available features of the dataset from the UE 22 (e.g. Feature IDs of the dataset);

[0118] ' available data sources (e.g. UE IDs, VAL server IDs)

[0119] ' domains of the datasets at the UE 22 (e.g. PLMN IDs of the UEs, VAL service provider IDs);

[0120] ' UE capability; and / or

[0121] ' ML model information (ML model or ML model address).

[0122] S210b. Network node 16a (AI / ML Enablement Server) sends preparation request to the VFL member (5GC NWDAF) via NEF. In some embodiments.

[0123] S212. The VFL members evaluate their capability and availability to join the VFL training process. The VFL members (AI / ML Enablement Clients 66 which are deployed on UEs 22) run the test task contained in the request in step S210a and judge whether it will join the VFL process. The VFL members may collect data from VAL client and other data sources.

[0124] S214. The VFL members send response to Network node 16a (AI / ML Enablement Server 62).

[0125] S214a. The VFL members (AI / ML Enablement Clients 66 which are deployed on UEs 22) send response to Network node 16a (AI / ML Enablement Server 62 ). The response message includes the decision on whether to join the VFL training process or not, and ' with the reason if not join (e.g. capability not satisfied, time not available, data not available, feature not applicable, cannot access to data source), or

[0126] ' with test result if join (e.g. time spend for complete the test task, energy consumption for the test task, memory usage for the test task, available time for support the VFL training process, feature information for the VFL training process, data source information for the VFL training process, domains of the dataset information for the VFL training process).

[0127] S214b. The VFL member (5GC NWDAF) sends a response to Network node 16a (AI / ML Enablement Server 62).

[0128] S216. The network node 16a (AI / ML Enablement Server 62) determines the VFL members for this VFL training process based on the information received in steps S206, S208 and S214. The criteria used by network node 16a (AI / ML Enablement Server 62) include:

[0129] ' Available data for the same sample among the VFL members;

[0130] ' Feature alignment of the sample / datasets among the VFL members o Features needed for this VFL training operations; o Same features of the same sample / datasets among the VFL members (to guarantee sample alignment); o Different features of the same sample / datasets among the VFL members;

[0131] ' Available time of the VFL members for support the VFL training operations;

[0132] ' Capability of the VFL members for the VFL training operations; o Time spends for complete the test task; o Energy consumption for the test task; o Memory usage for the test task;

[0133] S218. The network node 16a (AI / ML Enablement Server 62) coordinates the selected VFL members to perform VFL training operations for the VFL process.

[0134] S218a. The network node 16a (AI / ML Enablement Server 62) may report to the consumer with the intermediate training status, e.g. ' the ML model metric of the intermediated ML model ( e.g. accuracy);

[0135] ' list of UEs 22 involved in the training;

[0136] The consumer may adjust its request on the ML model, e.g.:

[0137] ' increase or decrease requirements on the ML model metric (e.g. accuracy);

[0138] ' adjust intent of the ML model (e.g. size of the ML model);

[0139] ' extend or reduce the maximum response time;

[0140] ' add or remove UE(s) 22 from the list;

[0141] ' inform to terminate request on the AI / ML operations;

[0142] S220. The network node 16a (AI / ML Enablement Server 62) sends a response to the consumer. The response message may contain the result of the AI / ML operations based on the request from the consumer, e.g. a trained ML model.

[0143] The following section describes features (e.g., some of which may have been incorporated in 3GPP TS 23.482. ¥0.4.0) based on one or more embodiments described herein. Steps 1-10 in the following section may correspond to steps S200-S220 of FIG. 8, and “Figure 8.X3.2-1: Procedure for supporting VFL in Enablement layer” may correspond to FIG. 8. Further, the term UE may refer to UE 22, AI / ML Enablement Server may refer to AI / ML Enablement Server 62, VAL Client may refer to VAL Client 64, and AI / ML Enablement Client may refer to AI / ML Enablement Client 66 described herein.

[0144] 1. Introduction

[0145] Starting from an example, consider two different companies in the same city, one is a bank, and the other is an e-commerce company. Their user sets are likely to contain most of the residents of the area, so the intersection of their user space is large. However, since the bank records the user’s revenue and expenditure behavior and credit rating, and the e-commerce retains the user’s browsing and purchasing history, their feature spaces are very different. That is distributed datasets have different set of features. In addition, the user may shop at different locations when using apps on a mobile phone, e.g. home, shopping center, company. Suppose a prediction model for product purchase is required based on the user, product information, and shopping habit, etc. That is train a ML model for the prediction product purchase by using the distributed datasets with different set of features. Vertical Federated Learning (VFL) can be used to train such a ML model for product purchase prediction with the distributed datasets and different set of features. The UE at the bank and the UE for the e-commerce company (e.g. mobile phone) are the physical entities. The AI / ML Enablement Clients deployed on these UEs are the VFL members for a VFL process to train the ML model. In addition, information from 5GC on e.g., UE mobility, UE connection to network and communication at different locations, are also important for training the model. NWDAF in the NF which provide ML model training service in 5GC. In the example above, NWDAF from 5GC can be seen as one of the VFL members for train the ML model.

[0146] Supporting VFL at enablement layer is one key issue in the SA6 Release 19 study, sample and feature alignments between domains are the main study aspects. Meanwhile, cross domain VFL involving AF and NWDAF is under study in SA2. To enable an AI / ML Enablement Server coordinate a VFL process between domains e.g., involving application layer FL among multiple UE(s) and 5GC NF (i.e., NWDAF), the following aspects need be taken into consideration:

[0147] ■ How the AI / ML Enablement Server gets the information (datasets and features, etc.) of VFL members (e.g., UE(s) at different domains and 5GC NWDAF).

[0148] ■ How the AI / ML Enablement Server utilizes the obtained information (e.g., information on datasets and features) to coordinate a VFL process between domains.

[0149] Thus, the mechanism of getting information of VFL members between domains, and coordinating a VFL process involving multiple UEs and 5GC NWDAF according to the obtained information (e.g., information on sample and features) need to be studied.

[0150] One or more advantages of the embodiments described herein include enabling an AI / ML Enablement Server to coordinate a VFL process in enablement layer, involving multiple UEs and 5GC NWDAF, getting and utilizing information of VFL members between domains is necessary to be studied.

[0151] The proposal is a solution on getting information of VFL members between domains and coordinating a VFL process in enablement layer involving multiple UEs and 5GC NWDAF according to the obtained information (e.g., information on sample and features).

[0152] 4. Proposal

[0153] It is proposed to agree the following changes to 3GPP 23.700-82 V0.2.0.

[0154] The VFL approach trains the ML model using the distributed datasets for the user (i.e., for the same sample). Here, sample corresponds to all the data for the user. The sample is formed by multiple datasets, each dataset at the VFL member (e.g. the AI / ML Enablement Client deployed on an UE, NWDAF) contains sample instances which corresponds to the data collected for the user from one of the entities (e.g. VAL Servers from bank and e-commerce company, mobile phone, 5GC NFs). These distributed datasets have different set of features, and the entities are the data sources of the distributed datasets. VFL approach trains the ML model using the distributed datasets for the user (i.e. for the same sample). Here, sample corresponds to all the data for the user. The sample is formed by multiple datasets, each dataset at the VFL member (e.g. the AI / ML Enablement Client deployed on an UE, NWDAF) contains sample instances which corresponds to the data collected for the user from one of the entities (e.g. VAL Servers from bank and e- commerce company, mobile phone, 5GC NFs). These distributed datasets have different set of features, and the entities are the data sources of the distributed datasets.

[0155] Pre-conditions:

[0156] VFL member (e.g. AI / ML Enablement Client which is deployed to an UE, 5GC NWDAF) has dataset or can access to data sources or is able to collect data from data producers for the AI / ML operations.

[0157] 8.X3.2 Procedure (corresponding to Figure 8.X3.2-1: Procedure for supporting VFL in Enablement layer)

[0158] 0. VFL members (AI / ML Enablement Clients which are deployed on UEs) registered their VFL profile to the AI / ML registry (e.g. AI / ML Enablement Server, A-ADRF). VFL member (5GC NWDAF) registered its VFL profile to NRF.

[0159] The VFL profile that VFL member (AI / ML Enablement Client which are deployed on a UE) registered contain PLMN information (e.g. PLMN ID), VFL capability information, Time interval supporting VFL, Available data / sample / data source information, Available feature information, Interoperability information.

[0160] Editor’s Note: The content of VFL profile that VFL member (NWDAF) registered to NRF will depend on the definition in SA2.

[0161] 1. Consumer (e.g. VAL Server) sends request to the AI / ML Enablement Server for AI / ML operations, e.g. train a ML model. The request message may contain detail requirements, e.g. approach for training a ML model (Centralized ML, Distributed ML / FL (HFL, VFL), etc.), initial ML model information (e.g. ML model, ML model address, ML model ID and information of repository, ML model structure), intent of ML model (e.g. size of ML model), ML model metric (e.g. accuracy), list of UE(s) involve in the AI / ML operations, data sources for train the ML model, maximum response time, indication on whether intermediate report is needed or not (e.g. for report training status, which may be used by the consumer to adjust request).

[0162] 2. After receiving the request from the consumer, the AI / ML Enablement Server determines that VFL training between domains is needed based on request from the consumer (if the request on VFL been provided in step 1) or local configuration (according to the requirements from the consumer on the AI / ML operations). If initial ML model is not provided in the request in step 200, the AI / ML Enablement Server determines an initial ML model.

[0163] 3. AI / ML Enablement Server sends request to and get response from the AI / ML registry for discovery VFL members (AI / ML Enablement Clients which are deployed on UEs). The request message may include e.g. requirements on VFL capability, data / sample (e.g. sample ID), feature sets of the datasets (e.g. Feature IDs of one dataset, each dataset corresponds to the data / sample instances at one VFL member), data sources (e.g. UE IDs, NF ID, VAL server IDs), interoperability information (e.g., ML model interoperability information), time period of the VFL training process, UE capability, area of interest, UE distribution. The response message from the AI / ML registry contains the VFL members information.

[0164] NOTE: Step 3 may be skipped if the AI / ML registry is the AI / ML Enablement Server.

[0165] 4. AI / ML Enablement Server sends request to and get response from NRF (or via NEF) for discovery VFL member (5GC NWDAF).

[0166] Editor’s Note: The service for AF to discovery NF from NRF (via NEF) is not available now, which service to be used and what content in the request to and response message from NRF (via NEF) will need coordination with SA2.

[0167] Steps 3 and 4 can be performed simultaneously or sequentially.

[0168] 5. After obtains the VFL members information, the AI / ML Enablement Server sends preparation request to the VFL members. 5a. AI / ML Enablement Server sends preparation request to the VFL members (AI / ML Enablement Clients which are deployed on UEs). The request message may include e.g. test task (compute task or mini training task), interoperability information (e.g. ML model interoperability information), requirements on available time for support VFL operations, available features of the dataset from the UE (e.g. Feature IDs of the dataset), available data sources (e.g. UE IDs, VAL server IDs), domains of the datasets at the UE (e.g. PLMN IDs of the UEs, VAL service provider IDs), UE capability, ML model information (ML model or ML model address).

[0169] 5b. AI / ML Enablement Server sends preparation request to the VFL member (5GC NWDAF) via NEF.

[0170] Editor’s Note: The content of the request message will need coordination with SA2.

[0171] 6. The VFL members evaluate their capability and availability to join the VFL training process. The VFL members (AI / ML Enablement Clients which are deployed on UEs) run the test task contained in the request in step 5a and judge whether it will join the VFL process. The VFL members may collect data from VAL client and other data sources.

[0172] NOTE: The procedures for data collection from UE need to take user consent into account.

[0173] 7. The VFL members send response to the AI / ML Enablement Server.

[0174] 7a. The VFL members (AI / ML Enablement Clients which are deployed on

[0175] UEs) send response to the AI / ML Enablement Server. The response message includes the decision on whether to join the VFL training process or not, with the reason if not join (e.g. capability not satisfied, time not available, data not available, feature not applicable, cannot access to data source) or with test result if join (e.g. time spend for complete the test task, energy consumption for the test task, memory usage for the test task, available time for support the VFL training process, feature information for the VFL training process, data source information for the VFL training process, domains of the dataset information for the VFL training process).

[0176] 7b. The VFL member (5GC NWDAF) sends response to the AI / ML Enablement Server. Editor’s Note: The content of the response message and what is the expectation output from SA2 will need coordination with SA2.

[0177] 8. The AEML Enablement Server determines the VFL members for this VFL training process based on the information received in steps 3, 4 and 7. The criteria used by the AEML Enablement Server include:

[0178] ■ Available data for the same sample among the VFL members;

[0179] ■ Feature alignment of the sample / datasets among the VFL members ; o Features needed for this VFL training operations; o Same features of the same sample / datasets among the VFL members (to guarantee sample alignment); o Different features of the same sample / datasets among the VFL members;

[0180] ■ Available time of the VFL members for support the VFL training operations;

[0181] ■ Capability of the VFL members for the VFL training operations; o Time spends for complete the test task; o Energy consumption for the test task; o Memory usage for the test task;

[0182] 9. The ALML Enablement Server coordinates the selected VFL members to perform VFL training operations for the VFL process. The procedure for the coordination process is similar as that in the solutions to KI#3 for FL.

[0183] 9a. The AEML Enablement Server may report to the consumer with the intermediate training status, e.g. the ML model metric of the intermediated ML model (e.g. accuracy), list of UEs involved in the training. The consumer may adjust its request on the ML model, e.g. increase or decrease requirements on the ML model metric (e.g. accuracy), adjust intent of the ML model (e.g. size of the ML model), extend or reduce the maximum response time, add or remove UE(s) from the list, inform to terminate request on the AEML operations.

[0184] 10. The AEML Enablement Server sends response to the consumer. The response message may contain result of the AEML operations based on the request from the consumer, e.g. a trained ML model.

[0185] The following is nonlimiting list of example embodiments. 1. A method in a network node configured as an artificial intelligence and / or machine learning (AI / ML) enablement server that supports vertical federated learning (VFL) in at least one enablement layer and configured to communicate with at least one user equipment (UE) and / or at least one other network node, the VFL being associated with the at least one UE and / or the at least one other network node, the method comprising at least one of: receiving a first request from a consumer network node, the first request requesting at least one AI / ML operation; obtaining information associated with discovery of VFL members in response to the first request; transmitting a second request to the VFL members based on the obtained information, the second request being a preparation request for the VFL members to evaluate capability and availability to join a VFL training process; receiving a first response from the VFL members, the first response including an indication indicating a decision on whether to join the VFL training process; determining the VFL members for the VFL training process based on at least one of the information associated with discovery of VFL members, the indication, and predetermined criteria; and coordinating with the determined VFL members to perform at least one VFL training operation for the VFL process.

[0186] 2. The method of Embodiment 1, wherein a VFL profile associated with the at least one UE and / or the at least one other network node is registered with another network node configured as an AI / ML registry.

[0187] 3. The method of Embodiments 2, wherein the obtaining the information includes: transmitting a third request to the other network node configured as the AI / ML registry, the third request requesting the information; and receiving a second response to the third request including the requested information.

[0188] 4. The method of any one of Embodiments 1-3, wherein the VFL members include at least one of: at least one AI / ML enablement client deployed on the at last one UE; and at least one network data analytics function (NWDAF). 5. The method of any one of Embodiments 1-4, wherein the method further includes: reporting to the consumer network node an intermediate training status associated with the VFL process.

[0189] 6. A network node configured as an artificial intelligence and / or machine learning (AI / ML) enablement server that supports vertical federated learning (VFL) in at least one enablement layer and configured to communicate with at least one user equipment (UE) and / or at least one other network node, the VFL being associated with the at least one UE and / or the at least one other network node, the network node being configured to at least one of: receive a first request from a consumer network node, the first request requesting at least one AI / ML operation; obtain information associated with discovery of VFL members in response to the first request; transmit a second request to the VFL members based on the obtained information, the second request being a preparation request for the VFL members to evaluate capability and availability to join a VFL training process; receive a first response from the VFL members, the first response including an indication indicating a decision on whether to join the VFL training process; determine the VFL members for the VFL training process based on at least one of the information associated with discovery of VFL members, the indication, and predetermined criteria; and coordinate with the determined VFL members to perform at least one VFL training operation for the VFL process.

[0190] 7. The network node of Embodiment 6, wherein a VFL profile associated with the at least one UE and / or the at least one other network node is registered with another network node configured as an AI / ML registry.

[0191] 8. The network node of Embodiment 7, wherein the obtaining the information includes: transmitting a third request to the other network node configured as the AI / ML registry, the third request requesting the information; and receiving a second response to the third request including the requested information. 9. The network node of any one of Embodiments 6-8, wherein the VFL members include at least one of: at least one AI / ML enablement client deployed on the at last one UE; and at least one network data analytics function (NWDAF).

[0192] 10. The network node of any one of Embodiments 6-9, wherein the network node is further configured to: report to the consumer network node an intermediate training status associated with the VFL process.

[0193] As will be appreciated by one of skill in the art, the concepts described herein may be embodied as a method, data processing system, computer program product and / or computer storage media storing an executable computer program. Accordingly, the concepts described herein may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects all generally referred to herein as a “circuit” or “module.” Any process, step, action and / or functionality described herein may be performed by, and / or associated to, a corresponding module, which may be implemented in software and / or firmware and / or hardware. Furthermore, the disclosure may take the form of a computer program product on a tangible computer usable storage medium having computer program code embodied in the medium that can be executed by a computer. Any suitable tangible computer readable medium may be utilized including hard disks, CD-ROMs, electronic storage devices, optical storage devices, or magnetic storage devices.

[0194] Some embodiments are described herein with reference to flowchart illustrations and / or block diagrams of methods, systems and computer program products. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer (to thereby create a special purpose computer), special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0195] These computer program instructions may also be stored in a computer readable memory or storage medium that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instruction means which implement the function / act specified in the flowchart and / or block diagram block or blocks.

[0196] The computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0197] It is to be understood that the functions / acts noted in the blocks may occur out of the order noted in the operational illustrations. For example, two blocks shown in succession may in fact be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality / acts involved. Although some of the diagrams include arrows on communication paths to show a primary direction of communication, it is to be understood that communication may occur in the opposite direction to the depicted arrows.

[0198] Computer program code for carrying out operations of the concepts described herein may be written in an object oriented programming language such as Python, Java® or C++. However, the computer program code for carrying out operations of the disclosure may also be written in conventional procedural programming languages, such as the "C" programming language. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer. In the latter scenario, the remote computer may be connected to the user's computer through a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0199] Many different embodiments have been disclosed herein, in connection with the above description and the drawings. It will be understood that it would be unduly repetitious and obfuscating to literally describe and illustrate every combination and subcombination of these embodiments. Accordingly, all embodiments can be combined in any way and / or combination, and the present specification, including the drawings, shall be construed to constitute a complete written description of all combinations and subcombinations of the embodiments described herein, and of the manner and process of making and using them, and shall support claims to any such combination or subcombination.

[0200] Abbreviations that may be used in the preceding description include: Al Artificial Intelligence

[0201] AF Application Function

[0202] ML Machine Learning

[0203] NRF Network Repository Function

[0204] NWDAF Network Data Analytics Function VFL Vertical Federated Learning

[0205] HFL Horizontal Federated Learning

[0206] 5GC 5G Core Network

[0207] It will be appreciated by persons skilled in the art that the embodiments described herein are not limited to what has been particularly shown and described herein above. In addition, unless mention was made above to the contrary, it should be noted that all of the accompanying drawings are not to scale. A variety of modifications and variations are possible in light of the above teachings without departing from the scope of the following claims.

Claims

What is claimed is:

1. A method in a network node (16) configured as an artificial intelligence machine learning enablement, AIMLE, sever (62), that supports vertical federated learning, VFL, and is configured to communicate with at least one user equipment, UE, (22) supporting an AIMLE client (66) and at least one consumer network node (16), the method comprising: receiving (S128) a first request from a consumer network node (16), the first request requesting at least one artificial intelligence machine learning, AI / ML, operation; in response to receiving the first request, determining (S130) that VFL training between dataset domains is to be used; transmitting (S132) a second request to selected one or more AIMLE clients (66) associated to the dataset domains, the second request being a request to evaluate a capability and availability of at least one AIMLE client (66) to join the VFL training; determining (S134) the one or more AIMLE clients (66) for the VFL training based at least on a response to the second request from the selected one or more AIMLE clients (66); and coordinating (S136) the determined one or more AIMLE clients (66) to perform the VFL training.

2. The method of Claim 1, wherein the method further includes: obtaining information associated with discovery of VFL members in response to the first request.

3. The method of any one of Claims 1 and 2, wherein the first request comprises ML model information.

4. The method of any one of Claims 1-3, wherein the first request comprises a list of UEs (22) having AIMLE clients (66).

5. The method of any one of Claims 1-4, wherein the first request comprises a training type.

6. The method of any one of Claims 1-5, wherein the second request comprises one or more of:a test task; a data source; one or more feature identifiers of a dataset; and available time for supporting VFL operations.

7. The method of any one of Claims 1-6, wherein the response to the second request comprises: a result of joining VFL training.

8. The method of any one of Claims 1-7, wherein the response to the second request comprises: a test result in response to a test task in the second request.

9. A network node (16) configured as an artificial intelligence machine learning enablement, AIMLE, sever (62), that supports vertical federated learning, VFL, and is configured to communicate with at least one user equipment, UE, (22) supporting an AIMLE client (66) and at least one consumer network node (16), the network node (16) being configured to perform one or more steps corresponding to Claims 1-8.

10. A method in a user equipment, UE, (22) the UE (22) comprising an artificial intelligence machine learning enablement, AIMLE, client (66), the method comprising: receiving (S138) a request from an AIMLE server to evaluate a capability and availability of the AIMLE client (66) to join a vertical federated learning, VFL, training; and transmitting (S140) a response indicating whether the AIMLE client (66) is able to join the VFL training.

11. The method of Claim 10, wherein the request comprises one or more of: a test task; a data source; one or more feature identifiers of a dataset; and available time for supporting VFL operations.

12. The method of any one of Claims 10 and 11, wherein the method further includes: transmitting a response to the request, the response comprising a test result in response to a test task in the request.

13. A user equipment, UE, (22) comprising an artificial intelligence machine learning enablement, AIMLE, client (66), the UE (22) being configured to perform one or more steps corresponding to Claims 10-12.

Images