Correlating machine learning models related to a vertical federated learning operation

Abstract

A method is provided that includes assigning a vertical federated learning (VFL) session identifier to a VFL model training session. The method includes selecting network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with different feature spaces. And the method includes sending a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces. The request includes the VFL session identifier for storage of the machine learning model in association with the VFL session identifier. The method may also include training a federated machine learning model to compute a prediction based on intermediate output data from the network functions, and storing the federated machine learning model in association with the VFL session identifier.

Description

CORRELATING MACHINE LEARNING MODELS RELATED TO A VERTICAL FEDERATED LEARNING OPERATIONTECHNOLOGICAL FIELD

[0001] The present disclosure relates generally to telecommunications and, in particular, to federated machine learning among multiple network functions in a telecommunications system.BACKGROUND

[0002] A telecommunications system can be seen as a facility that enables communication sessions between two or more entities such as user terminals, base stations and / or other nodes by providing carriers between the various entities involved in the communications path. A telecommunications system can be provided for example by means of a communication network and one or more compatible communication devices. The communication sessions may comprise, for example, communication of data for carrying communications such as voice, video, electronic mail (email), text message, multimedia and / or content data and so on. Nonlimiting examples of services provided comprise two-way or multi-way calls, data communication or multimedia services and access to a data network system, such as the Internet.

[0003] In a wireless telecommunications system at least a part of a communication session between at least two stations occurs over a wireless link. Examples of wireless systems comprise public land mobile networks (PLMN), satellite based communication systems and different wireless local networks, for example wireless local area networks (WLAN). Some wireless systems can be divided into cells, and are therefore often referred to as cellular systems.

[0004] A user can access the telecommunications system by means of an appropriate communication device or terminal. A communication device of a user may be referred to as user equipment (UE) or user device. A communication device is provided with an appropriate signal receiving and transmitting apparatus for enabling communications, for example enabling access to a communication network or communications directly with other users. The communication device may access a carrier provided by a station, for example a base station of a cell, and transmit and / or receive communications on the carrier.

[0005] The telecommunications system and associated devices typically operate in accordance with a given standard or specification which sets out what the various entities associated with the system are permitted to do and how that should be achieved. Communication protocols and / or parameters which shall be used for the connection are also typically defined. One example of a telecommunications system is the Universal Mobile Telecommunications System (UMTS). Other examples of telecommunications systems are Long-Term Evolution (LTE), LTE Advanced and the so-called 5G or New Radio (NR) networks. NR is being standardized by the 3rd Generation Partnership Project (3GPP).BRIEF SUMMARY

[0006] Example implementations of the present disclosure are directed to telecommunications and, in particular, to federated machine learning among multiple network functions in a telecommunications system. More specifically, example implementations are directed to vertical federated learning among multiple network functions. The present disclosure includes, without limitation, the following example implementations.

[0007] Some example implementations provide an apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: assign a vertical federated learning session identifier to a vertical federated learning model training session; select network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with different feature spaces; and send a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces, wherein the request comprises the vertical federated learning session identifier for storage of the machine learning model in association with the vertical federated learning session identifier.

[0008] Some example implementations provide an apparatus comprising: means for assigning a vertical federated learning session identifier to a vertical federated learning model training session; means for selecting network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with differentfeature spaces; and means for sending a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces, wherein the request comprises the vertical federated learning session identifier for storage of the machine learning model in association with the vertical federated learning session identifier.

[0009] Some example implementations provide a method comprising: assigning a vertical federated learning session identifier to a vertical federated learning model training session; selecting network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with different feature spaces; and sending a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces, wherein the request comprises the vertical federated learning session identifier for storage of the machine learning model in association with the vertical federated learning session identifier.

[0010] Some example implementations provide a computer-readable storage medium that is non-transitory and has instructions stored therein that, in response to execution by at least one processing circuitry, causes an apparatus to at least: assign a vertical federated learning session identifier to a vertical federated learning model training session; select network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with different feature spaces; and send a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces, wherein the request comprises the vertical federated learning session identifier for storage of the machine learning model in association with the vertical federated learning session identifier.

[0011] Some example implementations provide an apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: receive a request from a network function of a telecommunications system to perform model training of a machine learning model that is to be one of machine learning models trained in acommon sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; train the machine learning model in the common sample space with a respective feature space of the different feature spaces, the machine learning model trained to determine intermediate output data; send the intermediate data to the network function for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and store the machine learning model in association with the vertical federated learning session identifier.

[0012] Some example implementations provide an apparatus comprising: means for receiving a request from a network function of a telecommunications system to perform model training of a machine learning model that is to be one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; means for training the machine learning model in the common sample space with a respective feature space of the different feature spaces, the machine learning model trained to determine intermediate output data; means for sending the intermediate data to the network function for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and means for storing the machine learning model in association with the vertical federated learning session identifier.

[0013] Some example implementations provide a method comprising: receiving a request from a network function of a telecommunications system to perform model training of a machine learning model that is to be one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; training the machine learning model in the common sample space with a respective feature space of the different feature spaces, the machine learning model trained to determine intermediate output data; sending the intermediate data to the network function for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and storing the machine learning model in association with the vertical federated learning session identifier.

[0014] Some example implementations provide a computer-readable storage medium that is non-transitory and has instructions stored therein that, in response to execution by at least one processing circuitry, causes an apparatus to at least: receive a request from a network function ofa telecommunications system to perform model training of a machine learning model that is to be one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; train the machine learning model in the common sample space with a respective feature space of the different feature spaces, the machine learning model trained to determine intermediate output data; send the intermediate data to the network function for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and store the machine learning model in association with the vertical federated learning session identifier.

[0015] Some example implementations provide an apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: receive a request to provide at least one of analytics or one or more predictions for a data analytics procedure associated with an analytics identifier; select a network function to provide information about a vertical federated learning session associated with the analytics identifier; send a request comprising the analytics identifier to the network function, the request sent to the network function for a federated machine learning model; receive a response from the network function comprising a vertical federated learning session identifier and information about the federated machine learning model; access the federated machine learning model associated with the vertical federated learning session identifier; send requests to network functions of a telecommunications system to perform inferencing using the machine learning models trained in a common sample space but with different feature spaces, the requests comprising the vertical federated learning session identifier and the analytics identifier for retrieval of the machine learning models based on the vertical federated learning session identifier and the analytics identifier; receive intermediate output data from the network functions based on the inferencing; and perform an inferencing, based on the intermediate output data and using the federated machine learning model, to compute a prediction.

[0016] Some example implementations provide an apparatus comprising: means for receiving a request to provide at least one of analytics or one or more predictions for a data analytics procedure associated with an analytics identifier; means for selecting a network function to provide information about a vertical federated learning session associated with theanalytics identifier; means for sending a request comprising the analytics identifier to the network function, the request sent to the network function for a federated machine learning model; means for receiving a response from the network function comprising a vertical federated learning session identifier and information about the federated machine learning model; means for accessing the federated machine learning model associated with the vertical federated learning session identifier; means for sending requests to network functions of a telecommunications system to perform inferencing using the machine learning models trained in a common sample space but with different feature spaces, the requests comprising the vertical federated learning session identifier and the analytics identifier for retrieval of the machine learning models based on the vertical federated learning session identifier and the analytics identifier; means for receiving intermediate output data from the network functions based on the inferencing; and means for performing an inferencing, based on the intermediate output data and using the federated machine learning model, to compute a prediction.

[0017] Some example implementations provide a method comprising: receiving a request to provide at least one of analytics or one or more predictions for a data analytics procedure associated with an analytics identifier; selecting a network function to provide information about a vertical federated learning session associated with the analytics identifier; sending a request comprising the analytics identifier to the network function, the request sent to the network function for a federated machine learning model; receiving a response from the network function comprising a vertical federated learning session identifier and information about the federated machine learning model; accessing the federated machine learning model associated with the vertical federated learning session identifier; sending requests to network functions of a telecommunications system to perform inferencing using the machine learning models trained in a common sample space but with different feature spaces, the requests comprising the vertical federated learning session identifier and the analytics identifier for retrieval of the machine learning models based on the vertical federated learning session identifier and the analytics identifier; receiving intermediate output data from the network functions based on the inferencing; and performing an inferencing, based on the intermediate output data and using the federated machine learning model, to compute a prediction.

[0018] Some example implementations provide a computer-readable storage medium that is non-transitory and has instructions stored therein that, in response to execution by at least one processing circuitry, causes an apparatus to at least: receive a request to provide at least one of analytics or one or more predictions for a data analytics procedure associated with an analytics identifier; select a network function to provide information about a vertical federated learning session associated with the analytics identifier; send a request comprising the analytics identifier to the network function, the request sent to the network function for a federated machine learning model; receive a response from the network function comprising a vertical federated learning session identifier and information about the federated machine learning model; access the federated machine learning model associated with the vertical federated learning session identifier; send requests to network functions of a telecommunications system to perform inferencing using the machine learning models trained in a common sample space but with different feature spaces, the requests comprising the vertical federated learning session identifier and the analytics identifier for retrieval of the machine learning models based on the vertical federated learning session identifier and the analytics identifier; receive intermediate output data from the network functions based on the inferencing; and perform an inferencing, based on the intermediate output data and using the federated machine learning model, to compute a prediction.

[0019] Some example implementations provide an apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: receive a request from a network function of a telecommunications system to perform inferencing using a machine learning model that is one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; access the machine learning model associated with the vertical federated learning session identifier; perform an inferencing using the machine learning model to compute intermediate output data; and send the intermediate data to the network function for inferencing using a federated machine learning model to compute a prediction from the intermediate output data.

[0020] Some example implementations provide an apparatus comprising: means for receiving a request from a network function of a telecommunications system to perform inferencing using a machine learning model that is one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; means for accessing the machine learning model associated with the vertical federated learning session identifier; means for performing an inferencing using the machine learning model to compute intermediate output data; and means for sending the intermediate data to the network function for inferencing using a federated machine learning model to compute a prediction from the intermediate output data.

[0021] Some example implementations provide a method comprising: receiving a request from a network function of a telecommunications system to perform inferencing using a machine learning model that is one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; accessing the machine learning model associated with the vertical federated learning session identifier; performing an inferencing using the machine learning model to compute intermediate output data; and sending the intermediate data to the network function for inferencing using a federated machine learning model to compute a prediction from the intermediate output data.

[0022] Some example implementations provide a computer-readable storage medium that is non-transitory and has instructions stored therein that, in response to execution by at least one processing circuitry, causes an apparatus to at least: receive a request from a network function of a telecommunications system to perform inferencing using a machine learning model that is one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; access the machine learning model associated with the vertical federated learning session identifier; perform an inferencing using the machine learning model to compute intermediate output data; and send the intermediate data to the network function for inferencing using a federated machine learning model to compute a prediction from the intermediate output data.

[0023] These and other features, aspects, and advantages of the present disclosure will be apparent from a reading of the following detailed description together with the accompanyingfigures, which are briefly described below. The present disclosure includes any combination of two, three, four or more features or elements set forth in this disclosure, regardless of whether such features or elements are expressly combined or otherwise recited in a specific example implementation described herein. This disclosure is intended to be read holistically such that any separable features or elements of the disclosure, in any of its aspects and example implementations, should be viewed as combinable unless the context of the disclosure clearly dictates otherwise.

[0024] It will therefore be appreciated that this Brief Summary is provided merely for purposes of summarizing some example implementations so as to provide a basic understanding of some aspects of the disclosure. Accordingly, it will be appreciated that the above described example implementations are merely examples and should not be construed to narrow the scope or spirit of the disclosure in any way. Other example implementations, aspects and advantages will become apparent from the following detailed description taken in conjunction with the accompanying figures which illustrate, by way of example, the principles of some described example implementations.BRIEF DESCRIPTION OF THE FIGURE(S)

[0025] Having thus described example implementations of the disclosure in general terms, reference will now be made to the accompanying figures, which are not necessarily drawn to scale, and wherein:

[0026] FIG. 1 illustrates a telecommunications system that includes one or more public land mobile networks (PLMNs) coupled to one or more external data networks, according to some example implementations of the present disclosure;

[0027] FIG. 2 illustrates a deployment of a PLMN, according to some example implementations;

[0028] FIG. 3 more particularly depicts aspects of the deployment of FIG. 2, according to some example implementations;

[0029] FIG. 4 illustrates a trained machine learning (ML) model provisioning architecture, according to some example implementations;

[0030] FIGS. 5 A and 5B illustrate a signaling chart of vertical federated learning (VFL) training and inference, according to some example implementations;

[0031] FIG. 6 illustrates a signaling chart of VFL sample alignment for user equipment samples, according to some example implementations;

[0032] FIGS. 7A, 7B and 7C illustrate a signaling chart of VFL training and inference, according to other example implementations;

[0033] FIG. 8 illustrates a signaling chart of VFL sample alignment for user equipment samples, according to other example implementations;

[0034] FIGS. 9A, 9B, 9C, 9D, 9E, 9F and 9G are flowcharts illustrating various steps in a method according to various example implementations;

[0035] FIGS. 10 A, 10B, 10C and 10D are flowcharts illustrating various steps in a method according to various example implementations;

[0036] FIGS. 11 A, 11B, 11C and 11D are flowcharts illustrating various steps in a method according to various example implementations;

[0037] FIGS. 12A, 12B, 12C and 12D are flowcharts illustrating various steps in a method according to various example implementations; and

[0038] FIG. 13 illustrates an apparatus according to some example implementations.DETAILED DESCRIPTION

[0039] Some implementations of the present disclosure will now be described more fully hereinafter with reference to the accompanying figures, in which some, but not all implementations of the disclosure are shown. Indeed, various implementations of the disclosure may be embodied in many different forms and should not be construed as limited to the implementations set forth herein; rather, these example implementations are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art. Like reference numerals refer to like elements throughout.

[0040] Unless specified otherwise or clear from context, references to first, second or the like should not be construed to imply a particular order. A feature described as being above another feature (unless specified otherwise or clear from context) may instead be below, and vice versa; and similarly, features described as being to the left of another feature else may instead be to theright, and vice versa. Also, while reference may be made herein to quantitative measures, values, geometric relationships or the like, unless otherwise stated, any one or more if not all of these may be absolute or approximate to account for acceptable variations that may occur, such as those due to engineering tolerances or the like.

[0041] As used herein, unless specified otherwise or clear from context, the “or” of a set of operands is the “inclusive or” and thereby true if and only if one or more of the operands is true, as opposed to the “exclusive or” which is false when all of the operands are true. Thus, for example, “[A] or [B]” is true if [A] is true, or if [B] is true, or if both [A] and [B] are true. Further, the articles “a” and “an” mean “one or more,” unless specified otherwise or clear from context to be directed to a singular form. Furthermore, it should be understood that unless otherwise specified, the terms “data,” “content,” “digital content,” “information,” and similar terms may be at times used interchangeably. The term “network” may refer to a group of interconnected computers including clients and servers; and within a network, these computers may be interconnected directly or indirectly by various means including via one or more switches, routers, gateways, access points or the like.

[0042] Reference may be made herein to terms specific to a particular system, architecture or the like, but it should be understood that example implementations of the present disclosure may be equally applicable to any of a number of systems, architectures and the like. For example, reference may be made to 3 GPP technologies such as Global System for Mobile Communications (GSM), UMTS, LTE, LTE Advanced, 5GNR, 5G Advanced and 6G; however, it should be understood that example implementations of the present disclosure may be equally applicable to non-3 GPP technologies such as IEEE 802, Bluetooth and Bluetooth Low Energy.

[0043] Further, as used in this application, the term “circuitry” may refer to one or more or all of the following: (a) hardware-only circuit implementations (such as implementations in only analog and / or digital circuitry); (b) combinations of hardware circuits and software, such as (as applicable): (i) a combination of analog and / or digital hardware circuit(s) with software / firmware and (ii) any portions of hardware processor(s) with software (including digital signal processor(s)), software, and memory(ies) that work together to cause an apparatus, such as a mobile phone or server, to perform various functions); or (c) hardware circuit(s) and / or processor(s), such as a microprocessor s) or a portion of a microprocessor(s), that requiressoftware (e.g., firmware) for operation, but the software may not be present when it is not needed for operation.

[0044] The above definition of circuitry applies to all uses of this term in this application, including in any claims. As a further example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor (or multiple processors) or portion of a hardware circuit or processor and its (or their) accompanying software and / or firmware. The term circuitry also covers, for example and if applicable to the particular claim element, a baseband integrated circuit or processor integrated circuit for a mobile device or a similar integrated circuit in server, a cellular network device, or other computing or network device.

[0045] FIG. 1 illustrates a telecommunications system 100 according to various example implementations of the present disclosure. The telecommunications system generally includes one or more telecommunications networks. As shown, for example, the system includes one or more public land mobile networks (PLMNs) 102 coupled to one or more other external data networks 104 - notably including a wide area network (WAN) such as the Internet. Each of the PLMNs includes a core network (CN) 106 backbone such as the Evolved Packet Core (EPC) of LTE, the 5G core network (5GC) or the like; and each of the core networks and the Internet are coupled to one or more radio access networks (RANs) 108, air interfaces or the like that implement one or more radio access technologies (RATs). As used herein, a “network device” refers to any suitable device at a network side of a telecommunications network. Examples of suitable network devices are described in greater detail below.

[0046] In addition, the system includes one or more radio units that may be varyingly known as user equipment (UE) 110, terminal device, terminal equipment, mobile station or the like. The UE is generally a device configured to communicate with a network device or a further UE in a telecommunication network. The UE may be a portable computer (e.g., laptop, notebook, tablet computer), mobile phone (e.g., cell phone, smartphone), wearable computer (e.g., smartwatch), or the like. In other examples, the UE may be an Internet of Things (loT) device, an industrial loT (IIoT device), a vehicle equipped with a vehicle-to-everything (V2X) communication technology, or the like. In some examples, as referenced by 3 GPP, the UE may be a narrowbandloT (NB-IoT) device, an enhanced machine-type communication (eMTC) device, a reduced capability (RedCap) device, an ambient loT device, or the like.

[0047] In operation, these UEs may be configured to connect to one or more of the RANs 108 according to their particular radio access technologies to thereby access a particular CN 106 of a PLMN 102, or to access one or more of the external data networks 104 (e.g., the Internet). The external data network may be configured to provide Internet access, operator services, 3rd party services, etc. For example, the International Telecommunication Union (ITU) has classified 5G mobile network services into three categories: enhanced mobile broadband (eMBB), ultra-reliable and low-latency communications (URLLC), and massive machine type communications (mMTC) or massive internet of things (MIoT).

[0048] Examples of radio access technologies include 3 GPP radio access technologies such as GSM, UMTS, LTE, LTE Advanced, 5GNR, 5G Advanced, and 6G. Other examples of radio access technologies include IEEE 802 technologies such as IEEE 802.11 (Wi-Fi), IEEE 802.15 (including 802.15.1 (WPAN / Bluetooth), 802.15.4 (Zigbee) and 802.15.6 (WBAN)), Bluetooth, Bluetooth Low Energy (BLE), ultra wideband (UWB), and the like. Generally, a radio access technology may refer to any 2G, 3G, 4G, 5G, 6G or higher generation mobile communication technology and their different versions, as well as to any other wireless radio access technology that may be arranged to interwork with such a mobile communication technology to provide access to the CN 106 of a mobile network operator (MNO).

[0049] In various examples, a RAN 108 may be configured as one or more macrocells, microcells, picocells, femtocells or the like. The RAN may generally include one or more radio access nodes that are configured to interact with UEs 110. In various examples, a radio access node may be referred to as a base station (BS), access point (AP), base transceiver station (BTS), Node B (NB), evolved NB (eNB), macro BS, NB (MNB) or eNB (MeNB), home BS, NB (HNB) or eNB (HeNB), next generation NB (gNB), enhanced gNB (en-gNB), next generation eNB (ng- eNB), or the like. The RAN may include some type of network controlling / governing entity responsible for control of the radio access nodes. The network controlling / governing entity and radio access node may be separate or integrated into a single apparatus. The network controlling / governing entity may include processing circuity configured to carry out various management functions, etc. The processing circuity may be associated with a memory, computer-readable storage medium or database for maintaining information required in the management functions.

[0050] A RAN 108 may be centralized or distributed. In various examples, components of a RAN may be interconnected by Ethernet, Gigabit Ethernet, Asynchronous Transfer Mode (ATM), optical fiber, dark fiber, passive wavelength division multiplexing (WDM), WDM passive optical network (WDM-PON), optical transport network (OTN), time sensitive networking (TSN) and / or any other data link layer network, possibly including radio links. The RAN may be connected to a CN 106 through one or more gateways, network functions or the like.

[0051] As will be appreciated, a PLMN 102 may be deployed in a number of different manners. In a 4G LTE deployment, the EPC is the CN 106, and the evolved UMTS terrestrial radio access network (E-UTRAN) is the RAN 108; and the E-UTRAN includes one or more eNBs (radio access nodes) configured to connect UEs 110 to the E-UTRAN to thereby access the EPC. FIG. 2 illustrates a deployment 200, such as a 5G or 6G deployment. As shown, the 5GC 202 is the CN, and the next generation (NG) radio access network (NG-RAN) 204 is the RAN; and the NG-RAN includes one or more gNBs 206 (radio access nodes) configured to connect UEs 208 to the NG-RAN to thereby access the 5GC. The term ‘gNB’ in 5G may correspond to the eNB in 4G LTE.

[0052] Some 4G LTE and 5G deployments are considered standalone (SA) deployments. Other deployments combine 4G LTE and 5 G technologies, and are referred to as non-standalone (NSA) deployments. In some deployments, the E-UTRAN includes one or more ng-eNBs that are configured to communicate with the 5GC, and that may also be configured to communicate with one or more gNBs. Similarly, in another deployment, the NG-RAN may include one or more en-gNBs that are configured to communicate with the EPC, and that may also be configured to communicate with one or more eNBs. In various instances, a single UE 110, a dual-mode or multimode UE, may support multiple (two or more) RANs — thereby being configured to connect to multiple RANs, such as 4G LTE and 5G.

[0053] FIG. 3 more particularly depicts aspects of the deployment 200 for a MNO, according to some example implementations. As shown, the deployment includes the 5GC 202, and NG- RAN 204 with one or more gNBs 206 configured to connect UEs 208 to the NG-RAN to therebyaccess the 5GC. The 5GC may include a number of network functions (NFs) divided between the control plane and the user plane. In particular, the 5GC may include, for example, an access and mobility management function (AMF) 302, a session management function (SMF) 304, a user plane function (UPF) 306, a network exposure function (NEF) 308, a network data analytics function (NWDAF) 310, and / or an application function (AF) 312. Other examples of suitable NFs include a network repository function (NRF) 314, a network slice selection function (NSSF), a policy control function (PCF), a unified data management (UDM) 316, or the like.

[0054] In the control plane, the AMF 302 is configured to provide UE-based authentication, authorization, mobility management, etc. The SMF 304 is configured to provide various functionality including session management (SM), UE Internet Protocol (IP) address allocation and management, selection and control of UPF(s) 306, control part of policy enforcement and Quality of Service (QoS), lawful intercept, termination of SM parts of NAS messages, Downlink Data Notification (DNN), roaming functionality, handle local enforcement to apply QoS for Service Level Agreements (SLAs), charging data collection and charging interface, etc. If the UE 208 has multiple sessions, different SMFs may be allocated to each session to manage them individually and possibly provide different functionalities per session.

[0055] The UPF 306 supports various user plane operations and functionalities, such as packet routing and forwarding, traffic handling (e.g., QoS enforcement), an anchor point for intra-RAT / inter-RAT mobility (when applicable), packet inspection and policy rule enforcement, lawful intercept (UP collection), traffic accounting and reporting, etc. The UPF is the point of interconnect between the 5GC and at least one external data network (DN) 318 (i.e., point of ingress or egress for a DN), and routes packets to and from the DN. The DN may be configured to provide Internet access, operator services, 3rd party services, etc.

[0056] The NWDAF 310 collects and analyzes network data to provide insights into the performance, optimization, and overall health of the 5GC 202. This information may be used for network management, optimization, and decision-making processes to enhance the network’s efficiency.

[0057] The AF 312 may interact with the 5GC 202 to enable the deployment of specific services and applications. The AF communicates with other NFs to request and manage network resources, ensuring that the network adapts to the requirements of different applications andservices. The NEF 308 allows authorized third-party applications and services to access specific network functions and services in a controlled manner. The NEF enables the exposure of network capabilities to external entities, fostering innovation and the development of new services.

[0058] In some deployments, such as deployment 200, operations of the gNB 206 or other radio access node may be carried out, at least partly, in a central / centralized unit (CU), such as a server, host or node, operationally coupled to a distributed unit (DU), such as a radio head / node. It is also possible that node operations may be distributed among a plurality of servers, hosts or nodes. It should also be understood that the distribution of work between 5GC 202 (or other CN) operations and gNB (or other radio access node) operations may vary depending on implementation.

[0059] A 5G network architecture may be based on a so-called CU-DU split. One gNB-CU (central node) may control one or more gNB-DUs. The gNB-CU may control a plurality of spatially separated gNB-DUs, acting at least as transmit / receive (Tx / Rx) nodes. In some example implementations, however, the gNB-DUs (also called DU) may include, for example, a radio link control (RLC), medium access control (MAC) layer and a physical (PHY) layer, whereas the gNB-CU (also called a CU) may include the layers above the RLC layer, such as a packet data convergence protocol (PDCP) layer, a radio resource control (RRC), and an internet protocol (IP) layer. Other functional splits are also possible. It is considered that a skilled person is familiar with the open systems interconnection (OSI) model and the functionalities within each layer.

[0060] In some example implementations, the server or CU may generate a virtual network through which the server communicates with the radio node. In general, virtual networking may involve a process of combining hardware and software network resources and network functionality into a single, software-based administrative entity, a virtual network. Such virtual network may provide flexible distribution of operations between the server and the radio head / node. In practice, any digital signal processing task may be performed in either the CU or the DU, and the boundary where the responsibility is shifted between the CU and the DU may be selected according to implementation.

[0061] Many applications in telecommunications involve use of a large amount of data from multiple elements like UEs 208 to be used to train a single common machine learning (ML) model (at times more simply referred to as a “model”). In this regard, federated learning (FL) is aform of machine learning where instead of model training at a single node, different versions of the model are trained at different distributed nodes. This is different from distributed machine learning, where a single ML model is trained at distributed nodes to use computation power of different nodes.

[0062] A number of federated learning architectures follow a client-server model including a server and a number of distributed clients. In federated learning, each distributed client computes parameters for its local ML model, and the server does not compute a version or part of the ML model but combines parameters of all the distributed ML models to generate a federated model. In this regard, after training a local ML model, each individual client transfers its local ML model parameters, instead of raw training dataset, to the server. The server utilizes the local ML model parameters to update the federated model which is eventually fed back to the distributed clients for their use.

[0063] Federated learning may be categorized into a number of scenarios, including horizontal federated learning (HFL) and vertical federated learning (VFL). In HFL, the distributed clients train their local ML models with a common feature space but with different sample spaces (e.g., UE IDs). As described herein, a feature space refers to a feature or a set of features. In HFL, each client keeps its training dataset local where it is generated. The clients benefit from the datasets of the other learners, but only through the federated model, shared by the server, without accessing a high volume of privacy-sensitive data.

[0064] In VFL, the distributed clients train their local ML models with a common sample space but with different feature spaces (different features or sets of features). In this regard, each client may train its local ML model in the common sample space with a respective one of the different feature space. As compared to HFL, the federated model in VFL may be trained with fewer samples that are shared among the clients. But the federated model in VFL may provide a deeper federated model by taking advantage of a larger number of features that are organized into the different sample spaces, which are then distributed among the clients for training their respective local ML models.

[0065] In 3 GPP, a key issue aims to aims to provide solutions for enabling 5GC support for VFL involving one or more NFs, such as NWDAF(s) 310 and / or AF(s) 312. FIG. 4 illustrates a trained ML model provisioning architecture 400, according to some example implementations.As shown, the trained ML model provisioning architecture includes analytics logical function (AnLF) 402 and a model training logical function (MTLF) 404, which are logical functions of an NWDAF and may be connected to each other by an Nnwdaf interface. In this regard, an NWDAF containing an AnLF may be used trained ML model provisioning services from another NWDAF containing an MTLF.

[0066] The AnLF 402 is a logical function of an NWDAF 310 configured to perform inference, derive analytics information (e.g., derive statistics and / or prediction based on an analytics consumer request), and expose an analytics service (e.g.,Nnwdaf AnalyticsSub scription, Nnwdaf Analyticsinfo). The MTLF 404 is a logical function configured of an NWDAF to train ML models and expose new training services (e.g. providing a trained ML model). An NWDAF may include an AnLF, an MTLF, or both an AnLF and an MTLF.

[0067] In the architecture of the NWDAF 310, an MTLF 404 may be configured to train a ML model, and an AnLF 402 may be configured to infer analytics based on the trained model. An ML model may also be stored in and retrieved from can be stored in an analytics data repository function (ADRF), which is a NF that may be used for storage, retrieval and removal of data or analytics by NF consumers (e.g., NWDAF) that access the data using a Nadrf service.

[0068] Example implementations of the present disclosure are directed to VFL among multiple NFs, such as NWDAFs 310 and / or AFs 312. Some example implementations assume that VFL is controlled by an NWDAF acting as server (or active participant), and that clients (or passive participants) may be either other NWDAFs or AFs. Some example implementations use the internal NWDAF architecture in which an MTLF 404 may be configured to train models, and an AnLF 402 may be configured to perform inference based on trained models. In some examples, the models may be stored at an ADRF.

[0069] According to some examples, when starting a VFL model training session, an MTLF 404 operating as an VFL training server may discover and select one or more MTLFs and / or AFs 312 acting or otherwise operable as VFL training clients for one or more specific feature spaces. This discovery may use the NRF 314, where MTLFs register their capabilities to act or otherwise operate as a VFL training server or VFL training client, along with related analytics identifiers (IDs). Similarly, VFL training clients may also register their supported feature spaces, which maybe associated with feature IDs. In some examples in which feature spaces and related intermediate results to be exchanged between training clients and training servers are not standardized, and feature IDs may include a vendor ID for a vendor, as well as an identifier assigned by the vendor.

[0070] In examples in which the training involves UE-related data as samples, the VFL training server (MTLF 404) may select UEs 208 that are registered and reachable as samples, and monitors whether the UEs remain reachable during the training. For AFs 312 operable as VFL training clients, the VFL training server may send a request to check whether the AFs can reach the UEs.

[0071] The MTLF 404 acting as the VFL training server may assign a VFL session identifier for the VFL model training session, and send requests to start the training to the selected VFL clients (MTLFs or AFs 312). Each request may indicate an analytics ID, the feature ID, the desired sample (e.g. list of UEs), and / or the VFL session ID. The MTLF acting as the VFL training server may itself train a model (a federated model) that combines information related to several features across feature spaces. Each VFL training client may train a model relating to a feature space. Those models are trained together over several cycles. After each cycle, each VFL training clients provide so-called “gradients” as intermediate results to the server that the server uses to update its federated model, and the server also provides intermediate results to each client that the clients use to update their feature-related models.

[0072] After the training completes, both the MTLF 404 acting as the VFL training server and the MTLFs acting as VFL training clients store their trained models at the ADRF along with the related analytics ID, feature ID (for clients) or server model indicator and VFL session ID.

[0073] The inference of analytics or predictions related to VFL-trained models may be performed by an AnLF 402 acting as inference server and inference clients (either AnLFs or AFs 312) for specific feature spaces. The models that have been trained together may be used together for the inference.

[0074] When the AnLF 402 acting as the inference server receives a request or subscription for analytics or predictions for a particular analytics ID, the AnLF may discover an MTLF 404 for that analytics ID, and request a related model from that MTLF. For VFL, the MTLF acting as the VFL training server may be contacted. The MTLF as the VFL training server may indicatethat a federated model relates to VFL, and provide the VFL session ID and feature IDs used to train the federated model. The AnLF acting as the VFL inference server may discover VFL inference clients for the indicated feature spaces using the NRF 314, send a request or subscription for inference output related to each feature space to each VFL inference client. The request to each VFL inference client may include the analytics ID, feature ID and training session ID. The AnLFs acting as inference clients may discover MTLFs acting as training clients for the same analytics ID and feature IDs using the NRF, and request models from those MTLFs for the analytics ID, feature IDs and VFL session ID.

[0075] Each of the AnLFs 402 as the VFL inference clients may retrieve input data and calculate output data related to a feature space and analytics ID using the retrieved model for that feature space, and provide the output data to the AnLF acting as the inference server. The AnLF 402 acting as the VFL inference server may then calculate analytics or one or more predictions using the retrieved federated model and the received feature-related output data.

[0076] In some examples, no separate controller is assumed. In these examples, the MTLF 404 acting as the VFL training server may also train the federated model, may therefore be considered an “active participant” for training. In other examples, a separate NWDAF 310 or other NF may be provided as a “VFL coordinator” in that the NWDAF may select active and passive participants for the VFL training, and request that the active participants train the federated model and passive participants train models related to the different feature spaces. The MTLF acting as the VFL training client may also be referred to as a “passive participant” for training.

[0077] In some examples, the AnLF 402 acting as the VFL inference server may also use a federated model, and may therefore be considered an “active participant” for inference. In other examples, a separate NWDAF 310 or other NF may be provided as a “VFL coordinator” in that the NWDAF may select an active participant and passive participants for the inference. This NWDAF may request that the active participant performs inference on the federated model, and that the passive participants perform inference related to feature models. The AnLF acting as the VFL inference client may also be considered a “passive participant” for inference. In some examples, the VFL coordinator may be the same for training and inference.

[0078] To further illustrate some example implementations of the present disclosure, FIGS. 5 A and 5B illustrate a signaling chart 500 of VFL training and inference. As shown in FIG. 5 A at step 501, the AnLF 402A as an VFL inference server registers its capability to act as VFL inference server in the NRF 314 together with analytics IDs it supports and a vendor ID. As shown at step 502, the AnLF acting 402B as VFL inference client registers its capability to act as VFL inference client in the NRF together with analytics ID(s) and related feature ID(s) it supports. The MTLF 404B acting as an VFL training client at step 503 registers its capability to act as VFL training client in the NRF together with analytics ID(s) and related feature ID(s) it supports. And based on configured information, an NEF 308 at step 504 registers on behalf of an AF 312 the capability of the AF to act as VFL client in the NRF together with analytics IDs and related feature ID(s) the AF supports.

[0079] The MTLF 404A as the VFL training server at step 505 receives or observes a trigger to start a VFL model training for an analytics ID. The MTLF as the VFL training server at step 506 selects the required features for the model training based on implementation.

[0080] For each feature, at step 507, the MTLF 404A as the VFL training server sends a discovery request to the NRF 314 for an MTLF 404B as the VFL training client or for an AF 312 as the VFL client and provides the analytics ID and feature ID. For each feature, the NRF at step 508 provides profiles of candidate NFs matching the discovery request and the MTLF as the VFL training server selects one such NF.

[0081] The MTLF 404A as the VFL training server may at step 509 perform sample alignment with the VFL training clients. If UEs 208 are used as samples, the procedures described below in FIG. 7 may apply.

[0082] The MTLF 404A as the VFL training server at step 510 assigns a unique VFL session ID. The MTLF as the VFL training server at step 511 sends a request to start the VFL training to the VFL training client for each feature and provides the analytics ID, feature ID, and VFL session ID.

[0083] The MTLF 404A as the VFL training server and the VFL training clients at step 512 perform VFL model training. The VFL training server trains a model which is able to combine information related to several features. Each VFL training client trains a model relating to a special feature. Those models are trained together over several cycles. After each cycle, each1VFL training clients provide so-called “gradients” as intermediate results to the VFL training server that the server uses to update its model, and the server also provides intermediate results to each VFL training client that the clients use to update their feature-related models.

[0084] The VFL training server at step 513 stores the model it trained in the ADRF together with the analytics ID, VFL session ID, and feature IDs of all features used to train the model and an indication that the model is for an VFL server. Each VFL training client at step 514 stores the model it trained in the ADRF together with the analytics ID, VFL session ID and the feature ID of the feature the model relates to and an indication that the model is for an VFL client.

[0085] An AnLF 402A acting as the VFL inference server at step 515 receives a request or subscription for analytics or statistics for an analytics ID. The AnLF as the VFL inference server at step 516 queries the NRF 314 for an VFL training server for the analytics ID and its own vendor ID. The NRF at step 517 provides profiles of candidate NFs matching the discovery request and AnLF acting as the VFL inference server selects one such NF. The AnLF acting as the VFL inference server at step 518 requests a model from the selected MTLF 404A acting as the VFL training server providing the analytics ID.

[0086] As shown in FIG. 5B, the MTLF 404A acting as the VFL training server may at step 519 retrieve the requested model from the ADRF. The ADRF provides together with the model the VFL session identifier and the feature IDs of all features used to train the model. The MTLF acting as the VFL training server at step 520 provides information related to the requested model it trained (either the model or information how to retrieve it). The VFL training server also provides the VFL session identifier and the feature IDs of all features used to train the model. If the MTLF acting as the VFL training server provided information how to retrieve the model, the AnLF 402A acting as the VFL inference server retrieves the model, as shown at step 521.

[0087] For each feature, the AnLF 402A acting as the VFL inference server at step 522 queries the NRF 314 for an AnLF acting 402B as an VFL inference client or an AF 312 acting as an VFL client for the analytics ID and related feature ID. For each feature, the NRF at step 523 provides profiles of candidate NFs matching the discovery request and the AnLF acting as the VFL inference server selects one such NF. For each selected AnLF acting 402B as the VFL inference client, the AnLF acting as the VFL inference server at step 524 sends a VFL inferencerequest or subscription and provides the analytics ID, VFL session ID and feature ID within the request.

[0088] Each AnLF acting 402B as the VFL inference client at step 525 queries the NRF 314 for an MTLF 404B acting as the VFL training client for the analytics ID and feature ID. Toward each AnLF acting as the VFL inference client, the NRF at step 526 provides profiles of candidate NFs matching the discovery request and the AnLF acting as the VFL inference client selects one such NF. Each AnLF acting as the VFL inference client at step 527 requests the model from the selected MTLF acting as the VFL training client providing the analytics ID, feature ID and VFL session identifier.

[0089] Each MTLF 404B acting as the VFL training client may at step 528 retrieve the requested model from the ADRF, providing the analytics ID, feature ID and VFL session identifier. Each MTLF acting as VFL training client at step 529 provides information related to the requested model it trained (either the model or information how to retrieve it). If the MTLF acting as the VFL training client provided information how to retrieve the model, the AnLF acting 402B as the VFL inference client retrieves the model, as shown at step 530.

[0090] Each AnLF acting 402B as the VFL inference client at step 531 retrieves input data and calculates output data related to the corresponding feature and analytics ID using the retrieved model for that feature. Each AnLF acting 402B as the VFL inference client at step 532 provides the output data to the AnLF 402A acting as the VFL inference server.

[0091] For each AF 312 acting as the VFL client selected in step 523, the AnLF 402A acting as the VFL inference server at step 533 sends a VFL inference request or subscription and provides the analytics ID, VFL session ID and feature ID within the request. Each AF acting as the VFL client at step 534 selects the applicable model for the inference based on the analytics ID, VFL session ID and feature ID, retrieves input data, and calculates output data using the selected model. Each AF acting as the VFL client at step 535 provides the output data to the AnLF acting as the VFL inference server.

[0092] The AnLF 402A acting as the VFL inference server at step 536 calculates analytics or predictions using the retrieved model and the received feature-related output data.

[0093] FIG. 6 illustrates a signaling chart 600 of VFL sample alignment for user equipment samples, according to some example implementations. As shown at step 601, before starting avertical federated learning, the MTLF 404A acting as the VFL training server determines an initial list of target UE to be used as sample for vertical federated learning. For each candidate UE, at step 602, the MTLF acting as the VFL training server inquiries at the UDM 316 whether an AMF is assigned to the UE. The UDM at step 603 returns the assigned AMF ID (if any). If no AMF is assigned to a UE, at step 604, the MTLF acting as the VFL training server removes this UE from the list of target UE.

[0094] The MTLF at steps 605, 606 discovers the AMF address using the NRF 314. And at steps 607, 608, the MTLF inquiries the UE location at the AMF.

[0095] If the UE is located outside the desired target area, at step 609, the MTLF 404A acting as the VFL training server removes this UE from the list of target UE.

[0096] For each AF 312 acting as the VFL training client, at step 610, the MTLF 404A acting as the VFL training server sends a request to the AF to check the availability of target UEs and provides the list of target UEs.

[0097] Each AF 312 at step 611 checks for each target UE whether it can retrieve related input data and otherwise removes the UE from the list of target UEs. Each AF at step 612 replies with a reduced list of target UEs As shown at step 613, the MTLF 404A acting as the VFL training server computes the intersection of the reduced list of target UEs provided by the AFs and uses the resulting interslection list as target UEs to be used as sample for the VFL model training.

[0098] FIGS. 7A, 7B and 7C illustrates a signaling chart 700 of VFL training and inference, according to other example implementations. As shown in FIG. 7Aat step 701, an AnLF 402A acting as a VFL inference active participant registers its capability to act as a VFL inference active participant in the NRF 314 together with analytics IDs it supports and a vendor ID. An AnLF 402B acting as a VFL inference passive participant at step 702 registers its capability to act as a VFL inference passive participant in the NRF 314 together with analytics ID(s) and related feature ID(s) it supports. An MTLF 404B acting as a VFL training passive participant at step 703 registers its capability to act as a VFL training passive participant in the NRF 314 together with analytics ID(s) and related feature ID(s) it supports. And based on configured information, an NEF 308 at step 704 registers on behalf of an AF 312 the capability of the AF toact as a VFL passive participant in the NRF 314 together with analytics IDs and related feature ID(s) the AF supports.

[0099] The MTLF 404 A acting as a VFL training active participant at step 705 registers its capability to act as a VFL training active participant in the NRF 314 together with analytics ID(s) it supports. An NWDAF 310 acting as a VFL coordinator at step 706 receives or observes a trigger to start a VFL model training for an analytics ID.

[0100] The NWDAF 310 acting as the VFL coordinator at step 707 selects the required features for the model training based on implementation. The NWDAF acting as the VFL coordinator at step 708 sends a discovery request to the NRF 314 for an MTLF 404A acting as the VFL training active participant and provides the analytics ID and a Vendor ID. The NRF at step 709 provides profiles of candidate NFs matching the discovery request and the NWDAF acting as the VFL coordinator server selects one such NF.

[0101] For each feature, the NWDAF 310 acting as the VFL coordinator at step 710 sends a discovery request to the NRF 314 for an MTLF 404B acting as VFL training passive participant or for an AF 312 acting as VFL passive participant and provides the analytics ID and feature ID. For each feature, the NRF at step 711 provides profiles of candidate NFs matching the discovery request and the NWDAF acting as the VFL coordinator server selects one such NF.

[0102] The NWDAF 310 acting as the VFL coordinator may at step 712 perform sample alignment with the VFL training clients. If UEs are used as samples, the procedures in FIG. 8 may apply.

[0103] The NWDAF 310 acting as the VFL coordinator at step 713 assigns a unique VFL session ID. The NWDAF acting as the VFL coordinator at step 714 sends a request to start the VFL training to the VFL training active participant and provides the analytics ID, feature ID, and VFL session ID. It may also provide VFL training passive participant ID(s) and / or AF 312 VFL passive participant IDs.

[0104] The NWDAF 310 acting as the VFL coordinator at step 715 sends a request to start the VFL training to the VFL training passive participant for each feature and provides the analytics ID, feature ID, and VFL session ID. In an alternative example implementation, the VFL active participant sends this request. The NWDAF acting as the VFL coordinator at step 716 sends a request to start the VFL training to the AF 312 VFL passive participant for each featureand provides the analytics ID, feature ID, and VFL session ID. In an alternative example implementation, the VFL active participant sends this request.

[0105] As shown in FIG. 7B, the MTLF 404A acting as the VFL training active participant, the MTLF(s) 404B acting as VFL training passive participants and / or the AF 312 VFL passive participant(s) at step 717 perform VFL model training. The VFL training server trains a model which is able to combine information related to several features. Each VFL training client trains a model relating to a special feature. Those models are trained together over several cycles. After each cycle, each VFL training clients provide so-called “gradients” as intermediate results to the VFL training server that the server uses to update its model, and the server also provides intermediate results to each VFL training client that the clients use to update their feature-related models.

[0106] The MTLF 404 A acting as the VFL training active participant at step 718 stores the model it trained in the ADRF together with the analytics ID, VFL session ID, and feature IDs of all features used to train the model and an indication that the model is for a VFL active participant. Each MTLF 404B acting as VFL training passive participant at step 719 stores the model it trained in the ADRF together with the analytics ID, VFL session ID and the feature ID of the feature the model relates to and an indication that the model is for a VFL client.

[0107] The NWDAF 310 acting as the VFL coordinator at step 720 receives a request or subscription for analytics or statistics for an analytics ID

[0108] The NWDAF 310 acting as the VFL coordinator at step 721 queries the NRF 314 for a VFL inference active participant for the analytics ID and its own vendor ID. The NRF at step 722 provides profiles of candidate NFs matching the discovery request and NWD AF acting as the VFL coordinator selects one such NF.

[0109] For each feature, the NWDAF 310 acting as the VFL coordinator at step 723 queries the NRF 314 for a VFL inference passive participant for the analytics ID and feature ID. The NRF at step 724 provides profiles of candidate NFs matching the discovery request and NWDAF acting as the VFL coordinator selects one such NF for each feature.

[0110] The NWDAF 310 acting as the VFL coordinator at step 725 sends a request or subscription for VFL inference to the AnLF 402A acting as the VFL inference active participantand provides the analytics ID and VFL session ID. It may also provide the VFL training active participant ID, VFL inference passive participant ID(s) and / or AF VFL passive participant IDs.

[0111] The AnLF 402A acting as the VFL inference active participant at step 726 requests a model from the MTLF 404A acting as the VFL training active participant providing the analytics ID. The MTLF acting as the VFL training active participant may at step 727 retrieve the requested model from the ADRF. The ADRF provides together with the model the VFL session identifier and the feature IDs of all features used to train the model. The MTLF acting as the VFL training active participant at step 728 provides information related to the requested model it trained (either the model or information how to retrieve it). The VFL training active participant also provides the VFL session identifier and the feature IDs of all features used to train the model.

[0112] As shown in FIG. 7C, if the MTLF 404A acting as the VFL training active participant provided information how to retrieve the model, the AnLF 402A acting as the VFL inference active participant at step 729 retrieves the model.

[0113] For each feature, the NWD AF 310 acting as the VFL coordinator at step 730 sends a request or subscription for VFL inference to the AnLF 402B acting as VFL inference passive participant and provides the analytics ID, VFL session ID and feature ID. It may also provide the related VFL training passive participant ID and / or the the VFL inference active participant ID. In an alternative example implementation, the AnLF 402A acting as the VFL inference active participant sends this request.

[0114] For each feature, if no VFL training passive participant is received in step 730, the AnLF 402B acting as VFL inference passive participant at step 731 queries the NRF 314 for an MTLF 404B acting as VFL training passive participant for the analytics ID and related feature ID. For each feature, the NRF at step 732 provides profiles of candidate NFs matching the discovery request and the AnLF 402B acting as VFL inference passive participant selects one such NF.

[0115] Each AnLF 402B acting as VFL inference passive participant at step 733 requests the model from the selected MTLF 404B acting as VFL training passive participant providing the analytics ID, feature ID and VFL session identifier. Each MTLF acting as VFL training passive participant may at step 734 retrieve the requested model from the ADRF, providing the analyticsID, feature ID and VFL session identifier. Each MTLF acting as VFL training passive participant at step 735 provides information related to the requested model it trained (either the model or information how to retrieve it).

[0116] If the MTLF 404B acting as VFL training passive participant provided information how to retrieve the model, the AnLF 402B acting as VFL inference passive participant at step 736 retrieves the model.

[0117] Each AnLF 402B acting as VFL inference passive participant at step 737 retrieves input data and calculates output data related to the corresponding feature and analytics ID using the retrieved model for that feature.

[0118] Each AnLF 402B acting as VFL inference passive participant at step 738 provides the output data to the AnLF 402A acting as the VFL inference active participant.

[0119] For each AF 312 acting as VFL passive participant, the NWDAF 310 acting as the VFL coordinator at step 739 sends a VFL inference request or subscription and provides the analytics ID, VFL session ID and feature ID within the request. It may also provide the ID of the AnLF 402A acting as the VFL inference active participant. In an alternative example implementation, the AnLF acting as the VFL inference active participant sends this request.

[0120] Each AF 312 acting as VFL passive participant at step 740 selects the applicable model for the inference based on the analytics ID, VFL session ID and feature ID, retrieves input data, and calculates output data using the selected model. Each AF acting as VFL passive participant at step 741 provides the output data to the AnLF acting as VFL inference server.

[0121] The AnLF acting as VFL inference server at step 742 calculates analytics or predictions using the retrieved model and the received feature-related output data. And the AnLF acting as VFL inference server at step 743 provides the calculated analytics or predictions to the NWDAF 310 acting as the VFL coordinator.

[0122] FIG. 8 illustrates a signaling chart 800 of VFL sample alignment for user equipment samples, according to other example implementations. Before starting a vertical federated learning, the NWDAF 310 acting as the VFL coordinator at step 801 determines an initial list of target UE 208 to be used as sample for vertical federated learning.

[0123] For each candidate UE, the NWDAF 310 acting as the VFL coordinator at step 802 inquiries at the UDM 316 whether an AMF 302 is assigned to the UE. The UDM at step 803returns the assigned AMF ID (if any). If no AMF is assigned to a UE 208, the NWDAF acting as the VFL coordinator at step 804 removes this UE from the list of target UE.

[0124] The NWDAF 310 acting as the VFL coordinator at steps 805, 806 discovers the AMF address using the NRF 314. The NWDAF acting as the VFL coordinator at step 807, 808 inquiries the UE 208 location at the AMF 302. If the UE is located outside the desired target area, the NWDAF acting as the VFL coordinator at step 809 removes this UE from the list of target UE.

[0125] For each AF 312 acting as VFL training client, the NWDAF 310 acting as the VFL coordinator at step 810 sends a request to the AF to check the availability of target UEs 208 and provides the list of target UEs. Each AF at step 811 checks for each target UE whether it can retrieve related input data and otherwise removes the UE from the list of target UEs. Each AF at step 812 replies with a reduced list of target UEs.

[0126] The NWDAF 310 acting as VFL coordinator at step 813 computes the intersection of the reduced list of target UEs 208 provided by the AFs 312 and uses the resulting intersection list as target UEs to be used as sample for the VFL model training.

[0127] FIGS. 9 A - 9G are flowcharts illustrating various steps in a method 900 according to various example implementations. The method includes assigning a vertical federated learning session identifier to a vertical federated learning model training session, as shown at block 902 of FIG. 9A. The method includes selecting network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with different feature spaces, as shown at block 904. And the method includes sending a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces, In some of these examples, the request comprises the vertical federated learning session identifier for storage of the machine learning model in association with the vertical federated learning session identifier, as shown at block 906.

[0128] In some examples, the method 900 further includes receiving intermediate output data from the network functions based on the model training, as shown at block 908 of FIG. 9B. In some of these examples, the method also includes training, based on the intermediate output data, a federated machine learning model to compute a prediction, as shown at block 910. Andthe method includes storing the federated machine learning model in association with the vertical federated learning session identifier, as shown at block 912.

[0129] In some examples, the different feature spaces are associated with respective feature identifiers. In some of these examples, the federated machine learning model is stored at block 912 in further association with the respective feature identifiers associated with the different feature spaces.

[0130] In some examples, storing the federated machine learning model at block 912 comprises sending a request to an analytics data repository function for storage of the federated machine learning model at the analytics data repository function, as shown at block 914 of FIG. 9C.

[0131] In some examples, the method 900 further includes receiving a request for the federated machine learning model, as shown at block 916 of FIG. 9D. In some of these examples, the method also includes sending at least one of the federated machine learning model, information for retrieval of the federated machine learning model, the vertical federated learning session identifier, an analytics identifier, or respective feature identifiers with which the different feature spaces are associated, as shown at block 918.

[0132] In some examples, the different feature spaces are associated with respective feature identifiers; and in some of these examples, the request sent to each of the network functions at block 906 further comprises a feature identifier of the respective feature identifiers.

[0133] In some examples, the federated learning session identifier is associated with an analytics identifier; and in some of these examples, the request sent to each of the network functions at block 906further comprises the analytics identifier.

[0134] In some examples, the method 900 further includes sending a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and an analytics identifier, as shown at block 920 of FIG. 9E. In some of these examples, the method also includes receiving the information from the network repository function, wherein a network function of the network functions is selected based on the information received from the network repository function, as shown at block 922.

[0135] In some examples, the method 900 is performed by a network function. In some of these examples, the method further includes sending a request to a network repository function tostore information related to the network function, wherein the request comprises at least one of an indication that the network function is operable as a vertical federated learning training server, an indication that the network function is operable as a vertical federated learning training active participant, an analytics identifier, or an identifier of a vendor of the network function, as shown at block 924 of FIG. 9F.

[0136] In some examples, the method 900 further includes selecting a network function to operate as an active participant for the model training of the machine learning models, as shown at block 926 of FIG. 9G. In some of these examples, the method also includes sending a request to the network function to carry out the model training with the network functions as active participants, the request comprising the vertical federated learning session identifier and an analytics identifier, as shown at block 928.

[0137] In some examples, the method is performed by a network data analytics function, a model training logical function or an application function of the telecommunications system.

[0138] FIGS. 10A - 10D are flowcharts illustrating various steps in a method 1000 according to various example implementations. The method includes receiving a request from a network function of a telecommunications system to perform model training of a machine learning model that is to be one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier, as shown at block 1002 of FIG. 10A. The method includes training the machine learning model in the common sample space with a respective feature space of the different feature spaces, the machine learning model trained to determine intermediate output data, as shown at block 1004. The method includes sending the intermediate data to the network function for training, based on the intermediate output data, a federated machine learning model to compute a prediction, as shown at block 1006. And the method includes storing the machine learning model in association with the vertical federated learning session identifier, as shown at block 1008.

[0139] In some examples, the request received from the network function at block 1002 comprises a feature identifier associated with the respective feature space with which the machine learning model is trained.

[0140] In some examples, the federated machine learning model is stored at block 1008 in further association with the feature identifier.

[0141] In some examples, storing the machine learning model at block 1008 comprises a request to an analytics data repository function for storage of the machine learning model at the analytics data repository function, as shown sending at block 1010 of FIG. 10B.

[0142] In some examples, the federated learning session identifier is associated with an analytics identifier, and In some of these examples, the request further comprises the analytics identifier.

[0143] In some examples, the method 1000 further includes receiving a request for the machine learning model comprising the vertical federated learning session identifier and a feature identifier with which the respective feature space is associated, as shown at block 1012 of FIG. 10C. In some of these examples, the method also includes providing a response comprising at least one of the machine learning model, or information for retrieval of the machine learning model, as shown at block 1014.

[0144] In some examples, the method 1000 is performed by a second network function, and In some of these examples. In some of these examples, the method further includes sending a request to a network repository function to store information related to the second network function, wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning training client, an indication that the second network function is operable as a vertical federated learning training passive participant, an analytics identifier, or a feature identifier associated with the respective feature space with which the machine learning model is trained, as shown at block 1016 of FIG. ID.

[0145] In some examples, the method is performed by a network data analytics function, a model training logical function or an application function of the telecommunications system.

[0146] FIGS. 11 A - 1 ID are flowcharts illustrating various steps in a method 1100 according to various example implementations. The method includes receiving a request to provide at least one of analytics or one or more predictions for a data analytics procedure associated with an analytics identifier, as shown at block 1102 of FIG. 11A. The method includes selecting a network function to provide information about a vertical federated learning session associated with the analytics identifier, as shown at block 1104. The method includes sending a request comprising the analytics identifier to the network function, the request sent to the network function for a federated machine learning model, as shown at block 1106. The method includesreceiving a response from the network function comprising a vertical federated learning session identifier and information about the federated machine learning model, as shown at block 1108.

[0147] The method 1100 includes accessing the federated machine learning model associated with the vertical federated learning session identifier, as shown at block 1110. The method includes sending requests to network functions of a telecommunications system to perform inferencing using the machine learning models trained in a common sample space but with different feature spaces, the requests comprising the vertical federated learning session identifier and the analytics identifier for retrieval of the machine learning models based on the vertical federated learning session identifier and the analytics identifier, as shown at block 1112. The method includes receiving intermediate output data from the network functions based on the inferencing, as shown at block 1114. And the method includes performing an inferencing, based on the intermediate output data and using the federated machine learning model, to compute a prediction, as shown at block 1116.

[0148] In some examples, accessing the federated machine learning model at block 1110 comprises accessing the federated machine learning model and respective feature identifiers associated with the different feature spaces, as shown at block 1118 of FIG. 1 IB. In some of these examples, the method 1100 further includes selecting the network functions based on the respective feature identifiers, as shown at block 1120.

[0149] In some examples, each of the requests sent to a respective network function of the network functions at block 1122 further comprises a feature identifier of the respective feature identifiers.

[0150] In some examples, accessing the federated machine learning model at block 1110 comprises sending a request to at least one of a vertical federated learning training server, a vertical federated learning training active participant, a model training logical function, or an analytics data repository function for retrieval of the federated machine learning model, as shown at block 1124 of FIG. 11C.

[0151] In some examples, the method 1100 further includes sending a request to a network repository function for information related to at least one network function to provide the information about the vertical federated learning session, the request comprising at least one of the analytics identifier or a vendor identifier of the at least one network function, as shown atblock 1126 of FIG. 1 ID. In some of these examples, the method also includes receiving the information from the network repository function, In some of these examples, the network function to provide the information about the vertical federated learning session is selected based on the information received from the network repository function, as shown at block 1128.

[0152] In some examples, the method is performed by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

[0153] FIGS. 12A - 12D are flowcharts illustrating various steps in a method 1200 according to various example implementations. The method includes receiving a request from a network function of a telecommunications system to perform inferencing using a machine learning model that is one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier, as shown at block 1202 of FIG. 12A. The method includes accessing the machine learning model associated with the vertical federated learning session identifier, as shown at block 1204. The method includes performing an inferencing using the machine learning model to compute intermediate output data, as shown at block 1206. And the method includes sending the intermediate data to the network function for inferencing using a federated machine learning model to compute a prediction from the intermediate output data, as shown at block 1208.

[0154] In some examples, the request received from the network function at block 1202 comprises a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

[0155] In some examples, the machine learning model is accessed further based on the feature identifier.

[0156] In some examples, accessing the machine learning model at block 1204 comprises sending a request to at least one of a vertical federated learning training client, a vertical federated learning training passive participant, a model training logical function, or an analytics data repository function for storage of the machine learning model, as shown at block 1210 of FIG. 12B.

[0157] In some examples, the vertical federated learning session identifier is associated with an analytics identifier, and the request received from the network function at block 1202 further comprises the analytics identifier.

[0158] In some examples, the machine learning model is accessed from a vertical federated learning training client at block 1204. In some of these examples, the method 1200 further includes sending a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and the analytics identifier, as shown at block 1212 of FIG. 12C. The method includes receiving the information from the network repository function, as shown at block 1214. And the method includes selecting the vertical federated learning training client based on the information received from the network repository function, as shown at block 1216.

[0159] In some examples, the method 1200 is performed by a second network function. In some of these examples, the method further includes sending a request to a network repository function to store information related to the second network function, wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning inference client, the analytics identifier, or a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained, as shown at block 1218 of FIG. 12D.

[0160] In some examples, the method is performed by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

[0161] According to example implementations of the present disclosure, a telecommunications system 100 or PLMN 102, and its components such as a UE 110, CN 106, RAN 108, 5GC 202, NG-RAN 204, gNB 206, UE 208, AMF 302, SMF 304, UPF 306, NEF 308, NWDAF 310, AF 312, NRF 314, AnLF 402 and / or MTLF 404 may be implemented by various means. Means for implementing the system and its components may include hardware, firmware, software, or combinations thereof. In some examples, one or more apparatuses may be configured to function as or otherwise implement the system and its components shown and described herein. In examples involving more than one apparatus, the respective apparatuses may be connected to or otherwise in communication with one another in a number of different manners, such as directly or indirectly via a wired or wireless network or the like.

[0162] According to some example implementations, at least some of the method 900 described with respect to FIGS. 9A-9G may be carried out by an apparatus comprising means for performing functions corresponding steps of the method. Similarly, at least some of the method1000 described with respect to FIGS. 10A-10D may be carried out by an apparatus comprising means for performing functions corresponding steps of the method. At least some of the method 1100 described with respect to FIGS. 11 A-l ID may be carried out by an apparatus comprising means for performing functions corresponding steps of the method; and at least some of the method 1200 described with respect to FIGS. 12A-12D may be carried out by an apparatus comprising means for performing functions corresponding steps of the method. Examples of a suitable apparatus may include an NF (e.g., AMF, SMF, UPF, NEF, NWDAF, AF, AS), AnLF, MTLF or any suitable apparatus, such as a server, host or node.

[0163] FIG. 13 illustrates an apparatus 1300 in which means for performing various functions includes hardware, alone or under direction of one or more computer programs from a computer-readable storage medium or other memory, such as computer memory, according to some example implementations of the present disclosure. The apparatus may include one or more of each of a number of components such as, for example, processing circuitry 1302 connected to computer-readable storage medium or other memory 1304.

[0164] The processing circuitry 1302 may be composed of one or more processors alone or in combination with one or more computer-readable storage media. The processing circuitry is generally any piece of computer hardware that is capable of processing information such as, for example, data, computer programs and / or other suitable electronic information. The processing circuitry is composed of a collection of electronic circuits some of which may be packaged as an integrated circuit or multiple interconnected integrated circuits (an integrated circuit at times more commonly referred to as a “chip”). The processing circuitry may be configured to execute computer programs, which may be stored onboard the processing circuitry or otherwise stored in the memory 1304 (of the same or another apparatus).

[0165] The processing circuitry 1302 may be a number of processors, a multi-core processor or some other type of processor, depending on the particular implementation. Further, the processing circuitry may be implemented using a number of heterogeneous processor systems in which a main processor is present with one or more secondary processors on a single chip. As another illustrative example, the processing circuitry may be a symmetric multi-processor system containing multiple processors of the same type. In yet another example, the processing circuitry may be embodied as or otherwise include one or more ASICs, FPGAs or the like. Thus, althoughthe processing circuitry may be capable of executing a computer program to perform one or more functions, the processing circuitry of various examples may be capable of performing one or more functions without the aid of a computer program. In either instance, the processing circuitry may be appropriately programmed to perform functions or operations according to example implementations of the present disclosure.

[0166] The memory 1304 is generally any piece of computer hardware that is capable of storing information such as, for example, data, computer programs, instructions 1306 (e.g., computer-readable program code) and / or other suitable information either on a temporary basis and / or a permanent basis. The memory may include volatile and / or non-volatile memory, and may be fixed or removable. Examples of suitable memory include recording media, random access memory (RAM), read-only memory (ROM), a hard drive, a flash memory, a thumb drive, a removable computer diskette, an optical disk or some combination thereof.

[0167] The memory 1304 is a non-transitory device capable of storing information. One example of a suitable memory is a computer-readable storage medium, which is distinguishable from a computer-readable transmission medium capable of carrying information from one location to another. Examples of suitable computer-readable transmission media comprise electronic carrier signals, telecommunications signals, software distribution packages, or some combination thereof. As used herein, the term “non-transitory” is a limitation of the medium itself (i.e., tangible, not a signal) as opposed to a limitation on data storage persistency (e.g., RAM versus ROM). A computer-readable medium as described herein generally refers to a computer-readable storage medium or computer-readable transmission medium. A computer- readable medium is any entity or device capable in which information, such as one or more computer programs or portions thereof, may be stored and carried.

[0168] In addition to the memory 1304 (e.g., computer-readable storage medium), the processing circuitry 1302 may also be connected to one or more interfaces for displaying, transmitting and / or receiving information. The interfaces may include a communications interface 1308 and / or one or more user interfaces (e.g., display, user input interface). The communications interface may be configured to transmit and / or receive information, such as to and / or from other apparatus(es), network(s) or the like. The communications interface may be configured to transmit and / or receive information by physical (wired) and / or wirelesscommunications links. Examples of suitable communication interfaces include a network interface controller (NIC), wireless NIC (WNIC) or the like.

[0169] Execution of the instructions 1306 by the processing circuitry 1302, or storage of the instructions in the memory 1304, supports combinations of operations for implementing example implementations of the present disclosure. In this manner, an apparatus 1300 may comprise at least one processing circuitry and at least one memory coupled to the at least one processing circuitry, where the at least one processing circuitry is configured to execute instructions stored in the at least one memory. It will also be understood that one or more functions, and combinations of functions, may be implemented by special purpose hardware-based computer systems and / or processing circuitry which perform the specified functions, or combinations of special purpose hardware and program code instructions.

[0170] Some example implementations of the present disclosure may also be carried out in the form of a computer process defined by one or more computer programs or portions thereof. Example implementations of the present disclosure may be carried out by executing at least one portion of a computer program comprising instructions. The computer program may be in source code form, object code form, or in some intermediate form. The computer program may be stored in a computer-readable medium that is readable by a computer, processing circuitry or other suitable apparatus. As indicated above, for example, the computer program may be stored in a memory, such as a computer-readable storage medium. Additionally or alternatively, for example, the computer program may be stored in a computer-readable transmission medium. The coding of software for carrying out example implementations of the present disclosure is well within the scope of a person of ordinary skill in the art.

[0171] As will be appreciated, any suitable instructions may be loaded onto a computer, a processing circuitry or other programmable apparatus from a memory or a computer-readable medium (e.g., computer-readable storage medium, computer-readable transmission medium) to produce a particular machine, such that the particular machine becomes a means for implementing the functions specified herein. The instructions may also be stored in a computer- readable medium that can direct a computer, a processing circuitry or other programmable apparatus to function in a particular manner to thereby generate a particular machine or particular article of manufacture. In some examples, the instructions stored in the computer-readablemedium may produce an article of manufacture, where the article of manufacture becomes a means for implementing functions described herein. The instructions may be retrieved from a computer-readable medium and loaded into a computer, processing circuitry or other programmable apparatus to configure the computer, processing circuitry or other programmable apparatus to execute operations to be performed on or by the computer, processing circuitry or other programmable apparatus.

[0172] Retrieval, loading and execution of instructions comprising program code instructions may be performed sequentially such that one instruction is retrieved, loaded and executed at a time. In some example implementations, retrieval, loading and / or execution may be performed in parallel such that multiple instructions are retrieved, loaded, and / or executed together. Execution of the program code instructions may produce a computer-implemented process such that the instructions executed by the computer, processing circuitry or other programmable apparatus provide operations for implementing functions described herein.

[0173] As explained above and reiterated below, the present disclosure includes, without limitation, the following example implementations.

[0174] Clause 1. An apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: assign a vertical federated learning session identifier to a vertical federated learning model training session; select network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with different feature spaces; and send a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces, wherein the request comprises the vertical federated learning session identifier for storage of the machine learning model in association with the vertical federated learning session identifier.

[0175] Clause 2. The apparatus of clause 1, wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further at least: receive intermediate output data from the network functions based on the model training; train, based on the intermediate output data, a federated machine learning model to compute a prediction; andstore the federated machine learning model in association with the vertical federated learning session identifier.

[0176] Clause 3. The apparatus of clause 2, wherein the different feature spaces are associated with respective feature identifiers, and wherein the federated machine learning model is stored in further association with the respective feature identifiers associated with the different feature spaces.

[0177] Clause 4. The apparatus of clause 2 or clause 3, wherein the apparatus caused to store the federated machine learning model includes the apparatus caused to send a request to an analytics data repository function for storage of the federated machine learning model at the analytics data repository function.

[0178] Clause 5. The apparatus of any of clauses 2 to 4, wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further at least: receive a request for the federated machine learning model; and send at least one of the federated machine learning model, information for retrieval of the federated machine learning model, the vertical federated learning session identifier, an analytics identifier, or respective feature identifiers with which the different feature spaces are associated.

[0179] Clause 6. The apparatus of any of clauses 1 to 5, wherein the different feature spaces are associated with respective feature identifiers, and wherein the request sent to each of the network functions further comprises a feature identifier of the respective feature identifiers.

[0180] Clause 7. The apparatus of any of clauses 1 to 6, wherein the federated learning session identifier is associated with an analytics identifier, and wherein the request sent to each of the network functions further comprises the analytics identifier.

[0181] Clause 8. The apparatus of any of clauses 1 to 7, wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further at least: send a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and an analytics identifier; and receive the information from the network repository function, wherein a network function of the network functions is selected based on the information received from the network repository function.

[0182] Clause 9. The apparatus of any of clauses 1 to 8, wherein the apparatus is implemented by a network function, wherein the at least one processing circuitry is configured toexecute the instructions to cause the apparatus to further send a request to a network repository function to store information related to the network function, and wherein the request comprises at least one of an indication that the network function is operable as a vertical federated learning training server, an indication that the network function is operable as a vertical federated learning training active participant, an analytics identifier, or an identifier of a vendor of the network function.

[0183] Clause 10. The apparatus of any of clauses 1 to 9, wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further at least: select a network function to operate as an active participant for the model training of the machine learning models; and send a request to the network function to carry out the model training with the network functions as active participants, the request comprising the vertical federated learning session identifier and an analytics identifier.

[0184] Clause 11. The apparatus of any of clauses 1 to 10, wherein the apparatus is implemented by a network data analytics function, a model training logical function, or an application function of the telecommunications system.

[0185] Clause 12. An apparatus comprising: means for assigning a vertical federated learning session identifier to a vertical federated learning model training session; means for selecting network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with different feature spaces; and means for sending a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces, wherein the request comprises the vertical federated learning session identifier for storage of the machine learning model in association with the vertical federated learning session identifier.

[0186] Clause 13. The apparatus of clause 12, wherein the apparatus further comprises: means for receiving intermediate output data from the network functions based on the model training; means for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and means for storing the federated machine learning model in association with the vertical federated learning session identifier.

[0187] Clause 14. The apparatus of clause 13, wherein the different feature spaces are associated with respective feature identifiers, and wherein the federated machine learning model is stored in further association with the respective feature identifiers associated with the different feature spaces.

[0188] Clause 15. The apparatus of clause 13 or clause 14, wherein the means for storing the federated machine learning model comprises means for sending a request to an analytics data repository function for storage of the federated machine learning model at the analytics data repository function.

[0189] Clause 16. The apparatus of any of clauses 13 to 15, wherein the apparatus further comprises: means for receiving a request for the federated machine learning model; and means for sending at least one of the federated machine learning model, information for retrieval of the federated machine learning model, the vertical federated learning session identifier, an analytics identifier, or respective feature identifiers with which the different feature spaces are associated.

[0190] Clause 17. The apparatus of any of clauses 12 to 16, wherein the different feature spaces are associated with respective feature identifiers, and wherein the request sent to each of the network functions further comprises a feature identifier of the respective feature identifiers.

[0191] Clause 18. The apparatus of any of clauses 12 to 17, wherein the federated learning session identifier is associated with an analytics identifier, and wherein the request sent to each of the network functions further comprises the analytics identifier.

[0192] Clause 19. The apparatus of any of clauses 12 to 18, wherein the apparatus further comprises: means for sending a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and an analytics identifier; and means for receiving the information from the network repository function, wherein a network function of the network functions is selected based on the information received from the network repository function.

[0193] Clause 20. The apparatus of any of clauses 12 to 19, wherein the apparatus is implemented by a network function, wherein the apparatus further comprises means for sending a request to a network repository function to store information related to the network function, and wherein the request comprises at least one of an indication that the network function is operable as a vertical federated learning training server, an indication that the network function isoperable as a vertical federated learning training active participant, an analytics identifier, or an identifier of a vendor of the network function.

[0194] Clause 21. The apparatus of any of clauses 12 to 20, wherein the apparatus further comprises: means for selecting a network function to operate as an active participant for the model training of the machine learning models; and means for sending a request to the network function to carry out the model training with the network functions as active participants, the request comprising the vertical federated learning session identifier and an analytics identifier.

[0195] Clause 22. The apparatus of any of clauses 12 to 21, wherein the apparatus is implemented by a network data analytics function, a model training logical function, or an application function of the telecommunications system.

[0196] Clause 23. A method comprising: assigning a vertical federated learning session identifier to a vertical federated learning model training session; selecting network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with different feature spaces; and sending a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces, wherein the request comprises the vertical federated learning session identifier for storage of the machine learning model in association with the vertical federated learning session identifier.

[0197] Clause 24. The method of clause 23, wherein the method further comprises: receiving intermediate output data from the network functions based on the model training; training, based on the intermediate output data, a federated machine learning model to compute a prediction; and storing the federated machine learning model in association with the vertical federated learning session identifier.

[0198] Clause 25. The method of clause 24, wherein the different feature spaces are associated with respective feature identifiers, and wherein the federated machine learning model is stored in further association with the respective feature identifiers associated with the different feature spaces.

[0199] Clause 26. The method of clause 24 or clause 25, wherein storing the federated machine learning model comprises sending a request to an analytics data repository function for storage of the federated machine learning model at the analytics data repository function.

[0200] Clause 27. The method of any of clauses 24 to 26, wherein the method further comprises: receiving a request for the federated machine learning model; and sending at least one of the federated machine learning model, information for retrieval of the federated machine learning model, the vertical federated learning session identifier, an analytics identifier, or respective feature identifiers with which the different feature spaces are associated.

[0201] Clause 28. The method of any of clauses 23 to 27, wherein the different feature spaces are associated with respective feature identifiers, and wherein the request sent to each of the network functions further comprises a feature identifier of the respective feature identifiers.

[0202] Clause 29. The method of any of clauses 23 to 28, wherein the federated learning session identifier is associated with an analytics identifier, and wherein the request sent to each of the network functions further comprises the analytics identifier.

[0203] Clause 30. The method of any of clauses 23 to 29, wherein the method further comprises: sending a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and an analytics identifier; and receiving the information from the network repository function, wherein a network function of the network functions is selected based on the information received from the network repository function.

[0204] Clause 31. The method of any of clauses 23 to 30, wherein the method is performed by a network function, wherein the method further comprises sending a request to a network repository function to store information related to the network function, and wherein the request comprises at least one of an indication that the network function is operable as a vertical federated learning training server, an indication that the network function is operable as a vertical federated learning training active participant, an analytics identifier, or an identifier of a vendor of the network function.

[0205] Clause 32. The method of any of clauses 23 to 31, wherein the method further comprises: selecting a network function to operate as an active participant for the model training of the machine learning models; and sending a request to the network function to carry out themodel training with the network functions as active participants, the request comprising the vertical federated learning session identifier and an analytics identifier.

[0206] Clause 33. The method of any of clauses 23 to 32, wherein the method is performed by a network data analytics function, a model training logical function, or an application function of the telecommunications system.

[0207] Clause 34. A computer-readable storage medium that is non-transitory and has instructions stored therein that, in response to execution by at least one processing circuitry, causes an apparatus to at least: assign a vertical federated learning session identifier to a vertical federated learning model training session; select network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with different feature spaces; and send a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces, wherein the request comprises the vertical federated learning session identifier for storage of the machine learning model in association with the vertical federated learning session identifier.

[0208] Clause 35. The computer-readable storage medium of clause 34, wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further at least: receive intermediate output data from the network functions based on the model training; train, based on the intermediate output data, a federated machine learning model to compute a prediction; and store the federated machine learning model in association with the vertical federated learning session identifier.

[0209] Clause 36. The computer-readable storage medium of clause 35, wherein the different feature spaces are associated with respective feature identifiers, and wherein the federated machine learning model is stored in further association with the respective feature identifiers associated with the different feature spaces.

[0210] Clause 37. The computer-readable storage medium of clause 35 or clause 36, wherein the apparatus caused to store the federated machine learning model includes the apparatus caused to send a request to an analytics data repository function for storage of the federated machine learning model at the analytics data repository function.

[0211] Clause 38. The computer-readable storage medium of any of clauses 35 to 37, wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further at least: receive a request for the federated machine learning model; and send at least one of the federated machine learning model, information for retrieval of the federated machine learning model, the vertical federated learning session identifier, an analytics identifier, or respective feature identifiers with which the different feature spaces are associated.

[0212] Clause 39. The computer-readable storage medium of any of clauses 34 to 38, wherein the different feature spaces are associated with respective feature identifiers, and wherein the request sent to each of the network functions further comprises a feature identifier of the respective feature identifiers.

[0213] Clause 40. The computer-readable storage medium of any of clauses 34 to 39, wherein the federated learning session identifier is associated with an analytics identifier, and wherein the request sent to each of the network functions further comprises the analytics identifier.

[0214] Clause 41. The computer-readable storage medium of any of clauses 34 to 40, wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further at least: send a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and an analytics identifier; and receive the information from the network repository function, wherein a network function of the network functions is selected based on the information received from the network repository function.

[0215] Clause 42. The computer-readable storage medium of any of clauses 34 to 41, wherein the apparatus is implemented by a network function, wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further send a request to a network repository function to store information related to the network function, and wherein the request comprises at least one of an indication that the network function is operable as a vertical federated learning training server, an indication that the network function is operable as a verticalfederated learning training active participant, an analytics identifier, or an identifier of a vendor of the network function.

[0216] Clause 43. The computer-readable storage medium of any of clauses 34 to 42, wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further at least: select a network function to operate as an active participant for the model training of the machine learning models; and send a request to the network function to carry out the model training with the network functions as active participants, the request comprising the vertical federated learning session identifier and an analytics identifier.

[0217] Clause 44. The computer-readable storage medium of any of clauses 34 to 43, wherein the apparatus is implemented by a network data analytics function, a model training logical function, or an application function of the telecommunications system.

[0218] Clause 45. An apparatus comprising means for performing the method of any of clauses 23 to 33.

[0219] Clause 46. A computer-readable medium comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 23 to 33.

[0220] Clause 47. A computer-readable storage medium comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 23 to 33.

[0221] Clause 48. A computer program comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 23 to 33.

[0222] Clause 49. An apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: receive a request from a network function of a telecommunications system to perform model training of a machine learning model that is to be one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; train the machine learning model in the common sample space with a respective feature space ofthe different feature spaces, the machine learning model trained to determine intermediate output data; send the intermediate data to the network function for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and store the machine learning model in association with the vertical federated learning session identifier.

[0223] Clause 50. The apparatus of clause 49, wherein the request received from the network function comprises a feature identifier associated with the respective feature space with which the machine learning model is trained.

[0224] Clause 51. The apparatus of clause 50, wherein the federated machine learning model is stored in further association with the feature identifier.

[0225] Clause 52. The apparatus of any of clauses 49 to 51, wherein the apparatus caused to store the machine learning model includes the apparatus caused to send a request to an analytics data repository function for storage of the machine learning model at the analytics data repository function.

[0226] Clause 53. The apparatus of any of clauses 49 to 52, wherein the federated learning session identifier is associated with an analytics identifier, and wherein the request further comprises the analytics identifier.

[0227] Clause 54. The apparatus of any of clauses 49 to 53, wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further at least: receive a request for the machine learning model comprising the vertical federated learning session identifier and a feature identifier with which the respective feature space is associated; and provide a response comprising at least one of the machine learning model, or information for retrieval of the machine learning model.

[0228] Clause 55. The apparatus of any of clauses 49 to 54, wherein the apparatus is implemented by a second network function, wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further send a request to a network repository function to store information related to the second network function, and wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning training client, an indication that the second network function is operable as a vertical federated learning training passive participant, an analyticsidentifier, or a feature identifier associated with the respective feature space with which the machine learning model is trained.

[0229] Clause 56. The apparatus of any of clauses 49 to 55, wherein the apparatus is implemented by a network data analytics function, or a model training logical function, or an application function of the telecommunications system.

[0230] Clause 57. An apparatus comprising: means for receiving a request from a network function of a telecommunications system to perform model training of a machine learning model that is to be one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; means for training the machine learning model in the common sample space with a respective feature space of the different feature spaces, the machine learning model trained to determine intermediate output data; means for sending the intermediate data to the network function for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and means for storing the machine learning model in association with the vertical federated learning session identifier.

[0231] Clause 58. The apparatus of clause 57, wherein the request received from the network function comprises a feature identifier associated with the respective feature space with which the machine learning model is trained.

[0232] Clause 59. The apparatus of clause 58, wherein the federated machine learning model is stored in further association with the feature identifier.

[0233] Clause 60. The apparatus of any of clauses 57 to 59, wherein the means for storing the machine learning model comprises means for sending a request to an analytics data repository function for storage of the machine learning model at the analytics data repository function.

[0234] Clause 61. The apparatus of any of clauses 57 to 60, wherein the federated learning session identifier is associated with an analytics identifier, and wherein the request further comprises the analytics identifier.

[0235] Clause 62. The apparatus of any of clauses 57 to 61, wherein the apparatus further comprising: means for receiving a request for the machine learning model comprising the vertical federated learning session identifier and a feature identifier with which the respectivefeature space is associated; and means for providing a response comprising at least one of the machine learning model, or information for retrieval of the machine learning model.

[0236] Clause 63. The apparatus of any of clauses 57 to 62, wherein the apparatus is implemented by a second network function, wherein the apparatus further comprises means for sending a request to a network repository function to store information related to the second network function, and wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning training client, an indication that the second network function is operable as a vertical federated learning training passive participant, an analytics identifier, or a feature identifier associated with the respective feature space with which the machine learning model is trained.

[0237] Clause 64. The apparatus of any of clauses 57 to 63, wherein the apparatus is implemented by a network data analytics function, or a model training logical function, or an application function of the telecommunications system.

[0238] Clause 65. A method comprising: receiving a request from a network function of a telecommunications system to perform model training of a machine learning model that is to be one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; training the machine learning model in the common sample space with a respective feature space of the different feature spaces, the machine learning model trained to determine intermediate output data; sending the intermediate data to the network function for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and storing the machine learning model in association with the vertical federated learning session identifier.

[0239] Clause 66. The method of clause 65, wherein the request received from the network function comprises a feature identifier associated with the respective feature space with which the machine learning model is trained.

[0240] Clause 67. The method of clause 66, wherein the federated machine learning model is stored in further association with the feature identifier.

[0241] Clause 68. The method of any of clauses 65 to 67, wherein storing the machine learning model comprises sending a request to an analytics data repository function for storage of the machine learning model at the analytics data repository function.

[0242] Clause 69. The method of any of clauses 65 to 68, wherein the federated learning session identifier is associated with an analytics identifier, and wherein the request further comprises the analytics identifier.

[0243] Clause 70. The method of any of clauses 65 to 69, wherein the method further comprising: receiving a request for the machine learning model comprising the vertical federated learning session identifier and a feature identifier with which the respective feature space is associated; and providing a response comprising at least one of the machine learning model, or information for retrieval of the machine learning model.

[0244] Clause 71. The method of any of clauses 65 to 70, wherein the method is performed by a second network function, wherein the method further comprises sending a request to a network repository function to store information related to the second network function, and wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning training client, an indication that the second network function is operable as a vertical federated learning training passive participant, an analytics identifier, or a feature identifier associated with the respective feature space with which the machine learning model is trained.

[0245] Clause 72. The method of any of clauses 65 to 71, wherein the method is performed by a network data analytics function, or a model training logical function, or an application function of the telecommunications system.

[0246] Clause 73. A computer-readable storage medium that is non-transitory and has instructions stored therein that, in response to execution by at least one processing circuitry, causes an apparatus to at least: receive a request from a network function of a telecommunications system to perform model training of a machine learning model that is to be one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; train the machine learning model in the common sample space with a respective feature space of the different feature spaces, the machine learning model trained to determine intermediate output data; sendthe intermediate data to the network function for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and store the machine learning model in association with the vertical federated learning session identifier.

[0247] Clause 74. The computer-readable storage medium of clause 73, wherein the request received from the network function comprises a feature identifier associated with the respective feature space with which the machine learning model is trained.

[0248] Clause 75. The computer-readable storage medium of clause 74, wherein the federated machine learning model is stored in further association with the feature identifier.

[0249] Clause 76. The computer-readable storage medium of any of clauses 73 to 75, wherein the apparatus caused to store the machine learning model includes the apparatus caused to send a request to an analytics data repository function for storage of the machine learning model at the analytics data repository function.

[0250] Clause 77. The computer-readable storage medium of any of clauses 73 to 76, wherein the federated learning session identifier is associated with an analytics identifier, and wherein the request further comprises the analytics identifier.

[0251] Clause 78. The computer-readable storage medium of any of clauses 73 to 77, wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further at least: receive a request for the machine learning model comprising the vertical federated learning session identifier and a feature identifier with which the respective feature space is associated; and provide a response comprising at least one of the machine learning model, or information for retrieval of the machine learning model.

[0252] Clause 79. The computer-readable storage medium of any of clauses 73 to 78, wherein the apparatus is implemented by a second network function, wherein the computer- readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further send a request to a network repository function to store information related to the second network function, and wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning training client, an indication that the second network function is operable as a vertical federated learning training passive participant, an analytics identifier, or afeature identifier associated with the respective feature space with which the machine learning model is trained.

[0253] Clause 80. The computer-readable storage medium of any of clauses 73 to 79, wherein the apparatus is implemented by a network data analytics function, or a model training logical function, or an application function of the telecommunications system.

[0254] Clause 81. An apparatus comprising means for performing the method of any of clauses 65 to 72.

[0255] Clause 82. A computer-readable medium comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 65 to 72.

[0256] Clause 83. A computer-readable storage medium comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 65 to 72.

[0257] Clause 84. A computer program comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 65 to 72.

[0258] Clause 85. An apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: receive a request to provide at least one of analytics or one or more predictions for a data analytics procedure associated with an analytics identifier; select a network function to provide information about a vertical federated learning session associated with the analytics identifier; send a request comprising the analytics identifier to the network function, the request sent to the network function for a federated machine learning model; receive a response from the network function comprising a vertical federated learning session identifier and information about the federated machine learning model; access the federated machine learning model associated with the vertical federated learning session identifier; send requests to network functions of a telecommunications system to perform inferencing using the machine learning models trained in a common sample space but with different feature spaces, the requests comprising the vertical federated learning session identifier and the analytics identifier for retrieval of the machine learning models based on thevertical federated learning session identifier and the analytics identifier; receive intermediate output data from the network functions based on the inferencing; and perform an inferencing, based on the intermediate output data and using the federated machine learning model, to compute a prediction.

[0259] Clause 86. The apparatus of clause 85, wherein the apparatus caused to access the federated machine learning model includes the apparatus caused to access the federated machine learning model and respective feature identifiers associated with the different feature spaces, and wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further select the network functions based on the respective feature identifiers.

[0260] Clause 87. The apparatus of clause 86, wherein each of the requests sent to a respective network function of the network functions further comprises a feature identifier of the respective feature identifiers.

[0261] Clause 88. The apparatus of any of clauses 85 to 87, wherein the apparatus caused to access the federated machine learning model includes the apparatus caused to send a request to at least one of a vertical federated learning training server, a vertical federated learning training active participant, a model training logical function, or an analytics data repository function for retrieval of the federated machine learning model.

[0262] Clause 89. The apparatus of any of clauses 85 to 88, wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further at least: send a request to a network repository function for information related to at least one network function to provide the information about the vertical federated learning session, the request comprising at least one of the analytics identifier or a vendor identifier of the at least one network function; and receive the information from the network repository function, wherein the network function to provide the information about the vertical federated learning session is selected based on the information received from the network repository function.

[0263] Clause 90. The apparatus of any of clauses 85 to 89, wherein the apparatus is implemented by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

[0264] Clause 91. An apparatus comprising: means for receiving a request to provide at least one of analytics or one or more predictions for a data analytics procedure associated with ananalytics identifier; means for selecting a network function to provide information about a vertical federated learning session associated with the analytics identifier; means for sending a request comprising the analytics identifier to the network function, the request sent to the network function for a federated machine learning model; means for receiving a response from the network function comprising a vertical federated learning session identifier and information about the federated machine learning model; means for accessing the federated machine learning model associated with the vertical federated learning session identifier; means for sending requests to network functions of a telecommunications system to perform inferencing using the machine learning models trained in a common sample space but with different feature spaces, the requests comprising the vertical federated learning session identifier and the analytics identifier for retrieval of the machine learning models based on the vertical federated learning session identifier and the analytics identifier; means for receiving intermediate output data from the network functions based on the inferencing; and means for performing an inferencing, based on the intermediate output data and using the federated machine learning model, to compute a prediction.

[0265] Clause 92. The apparatus of clause 91, wherein the means for accessing the federated machine learning model comprises means for accessing the federated machine learning model and respective feature identifiers associated with the different feature spaces, and wherein the apparatus further comprises means for selecting the network functions based on the respective feature identifiers.

[0266] Clause 93. The apparatus of clause 92, wherein each of the requests sent to a respective network function of the network functions further comprises a feature identifier of the respective feature identifiers.

[0267] Clause 94. The apparatus of any of clauses 91 to 93, wherein the means for accessing the federated machine learning model comprises means for sending a request to at least one of a vertical federated learning training server, a vertical federated learning training active participant, a model training logical function, or an analytics data repository function for retrieval of the federated machine learning model.

[0268] Clause 95. The apparatus of any of clauses 91 to 94, wherein the apparatus further comprises: means for sending a request to a network repository function for information relatedto at least one network function to provide the information about the vertical federated learning session, the request comprising at least one of the analytics identifier or a vendor identifier of the at least one network function; and means for receiving the information from the network repository function, wherein the network function to provide the information about the vertical federated learning session is selected based on the information received from the network repository function.

[0269] Clause 96. The apparatus of any of clauses 91 to 95, wherein the apparatus is implemented by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

[0270] Clause 97. A method comprising: receiving a request to provide at least one of analytics or one or more predictions for a data analytics procedure associated with an analytics identifier; selecting a network function to provide information about a vertical federated learning session associated with the analytics identifier; sending a request comprising the analytics identifier to the network function, the request sent to the network function for a federated machine learning model; receiving a response from the network function comprising a vertical federated learning session identifier and information about the federated machine learning model; accessing the federated machine learning model associated with the vertical federated learning session identifier; sending requests to network functions of a telecommunications system to perform inferencing using the machine learning models trained in a common sample space but with different feature spaces, the requests comprising the vertical federated learning session identifier and the analytics identifier for retrieval of the machine learning models based on the vertical federated learning session identifier and the analytics identifier; receiving intermediate output data from the network functions based on the inferencing; and performing an inferencing, based on the intermediate output data and using the federated machine learning model, to compute a prediction.

[0271] Clause 98. The method of clause 97, wherein accessing the federated machine learning model comprises accessing the federated machine learning model and respective feature identifiers associated with the different feature spaces, and wherein the method further comprises selecting the network functions based on the respective feature identifiers.

[0272] Clause 99. The method of clause 98, wherein each of the requests sent to a respective network function of the network functions further comprises a feature identifier of the respective feature identifiers.

[0273] Clause 100. The method of any of clauses 97 to 99, wherein accessing the federated machine learning model comprises sending a request to at least one of a vertical federated learning training server, a vertical federated learning training active participant, a model training logical function, or an analytics data repository function for retrieval of the federated machine learning model.

[0274] Clause 101. The method of any of clauses 97 to 100, wherein the method further comprises: sending a request to a network repository function for information related to at least one network function to provide the information about the vertical federated learning session, the request comprising at least one of the analytics identifier or a vendor identifier of the at least one network function; and receiving the information from the network repository function, wherein the network function to provide the information about the vertical federated learning session is selected based on the information received from the network repository function.

[0275] Clause 102. The method of any of clauses 97 to 101, wherein the method is performed by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

[0276] Clause 103. A computer-readable storage medium that is non-transitory and has instructions stored therein that, in response to execution by at least one processing circuitry, causes an apparatus to at least: receive a request to provide at least one of analytics or one or more predictions for a data analytics procedure associated with an analytics identifier; select a network function to provide information about a vertical federated learning session associated with the analytics identifier; send a request comprising the analytics identifier to the network function, the request sent to the network function for a federated machine learning model; receive a response from the network function comprising a vertical federated learning session identifier and information about the federated machine learning model; access the federated machine learning model associated with the vertical federated learning session identifier; send requests to network functions of a telecommunications system to perform inferencing using the machine learning models trained in a common sample space but with different feature spaces, therequests comprising the vertical federated learning session identifier and the analytics identifier for retrieval of the machine learning models based on the vertical federated l earning session identifier and the analytics identifier; receive intermediate output data from the network functions based on the inferencing; and perform an inferencing, based on the intermediate output data and using the federated machine learning model, to compute a prediction.

[0277] Clause 104. The computer-readable storage medium of clause 103, wherein the apparatus caused to access the federated machine learning model includes the apparatus caused to access the federated machine learning model and respective feature identifiers associated with the different feature spaces, and wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further select the network functions based on the respective feature identifiers.

[0278] Clause 105. The computer-readable storage medium of clause 104, wherein each of the requests sent to a respective network function of the network functions further comprises a feature identifier of the respective feature identifiers.

[0279] Clause 106. The computer-readable storage medium of any of clauses 103 to 105, wherein the apparatus caused to access the federated machine learning model includes the apparatus caused to send a request to at least one of a vertical federated learning training server, a vertical federated learning training active participant, a model training logical function, or an analytics data repository function for retrieval of the federated machine learning model.

[0280] Clause 107. The computer-readable storage medium of any of clauses 103 to 106, wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further at least: send a request to a network repository function for information related to at least one network function to provide the information about the vertical federated learning session, the request comprising at least one of the analytics identifier or a vendor identifier of the at least one network function; and receive the information from the network repository function, wherein the network function to provide the information about the vertical federated learning session is selected based on the information received from the network repository function.

[0281] Clause 108. The computer-readable storage medium of any of clauses 103 to 107, wherein the apparatus is implemented by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

[0282] Clause 109. An apparatus comprising means for performing the method of any of clauses 97 to 102.

[0283] Clause 110. A computer-readable medium comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 97 to 102.

[0284] Clause 111. A computer-readable storage medium comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 97 to 102.

[0285] Clause 112. A computer program comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 97 to 102.

[0286] Clause 113. An apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: receive a request from a network function of a telecommunications system to perform inferencing using a machine learning model that is one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; access the machine learning model associated with the vertical federated learning session identifier; perform an inferencing using the machine learning model to compute intermediate output data; and send the intermediate data to the network function for inferencing using a federated machine learning model to compute a prediction from the intermediate output data.

[0287] Clause 114. The apparatus of clause 113, wherein the request received from the network function comprises a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

[0288] Clause 115. The apparatus of clause 114, wherein the machine learning model is accessed further based on the feature identifier.

[0289] Clause 116. The apparatus of any of clauses 113 to 115, wherein the apparatus caused to access the machine learning model includes the apparatus caused to send a request to at least one of a vertical federated learning training client, a vertical federated learning training passive participant, a model training logical function, or an analytics data repository function for storage of the machine learning model.

[0290] Clause 117. The apparatus of any of clauses 113 to 116, wherein the vertical federated learning session identifier is associated with an analytics identifier, and the request received from the network function further comprises the analytics identifier.

[0291] Clause 118. The apparatus of clause 117, wherein the machine learning model is accessed from a vertical federated learning training client, and wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further at least: send a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and the analytics identifier; receive the information from the network repository function; and select the vertical federated learning training client based on the information received from the network repository function.

[0292] Clause 119. The apparatus of any of clauses 113 to 118, wherein the apparatus is implemented by a second network function, wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further send a request to a network repository function to store information related to the second network function, and wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning inference client, the analytics identifier, or a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

[0293] Clause 120. The apparatus of any of clauses 113 to 119, wherein the apparatus is implemented by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

[0294] Clause 121. An apparatus comprising: means for receiving a request from a network function of a telecommunications system to perform inferencing using a machine learning model that is one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; means foraccessing the machine learning model associated with the vertical federated learning session identifier; means for performing an inferencing using the machine learning model to compute intermediate output data; and means for sending the intermediate data to the network function for inferencing using a federated machine learning model to compute a prediction from the intermediate output data.

[0295] Clause 122. The apparatus of clause 121, wherein the request received from the network function comprises a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

[0296] Clause 123. The apparatus of clause 122, wherein the machine learning model is accessed further based on the feature identifier.

[0297] Clause 124. The apparatus of any of clauses 121 to 123, wherein the means for accessing the machine learning model comprises means for sending a request to at least one of a vertical federated learning training client, a vertical federated learning training passive participant, a model training logical function, or an analytics data repository function for storage of the machine learning model.

[0298] Clause 125. The apparatus of any of clauses 121 to 124, wherein the vertical federated learning session identifier is associated with an analytics identifier, and the request received from the network function further comprises the analytics identifier.

[0299] Clause 126. The apparatus of clause 125, wherein the machine learning model is accessed from a vertical federated learning training client, and wherein the apparatus further comprises: means for sending a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and the analytics identifier; means for receiving the information from the network repository function; and means for selecting the vertical federated learning training client based on the information received from the network repository function.

[0300] Clause 127. The apparatus of any of clauses 121 to 126, wherein the apparatus is implemented by a second network function, wherein the apparatus further comprises means for sending a request to a network repository function to store information related to the second network function, and wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning inference client, the analyticsidentifier, or a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

[0301] Clause 128. The apparatus of any of clauses 121 to 127, wherein the apparatus is implemented by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

[0302] Clause 129. A method comprising: receiving a request from a network function of a telecommunications system to perform inferencing using a machine learning model that is one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; accessing the machine learning model associated with the vertical federated learning session identifier; performing an inferencing using the machine learning model to compute intermediate output data; and sending the intermediate data to the network function for inferencing using a federated machine learning model to compute a prediction from the intermediate output data.

[0303] Clause 130. The method of clause 129, wherein the request received from the network function comprises a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

[0304] Clause 131. The method of clause 130, wherein the machine learning model is accessed further based on the feature identifier.

[0305] Clause 132. The method of any of clauses 129 to 131, wherein accessing the machine learning model comprises sending a request to at least one of a vertical federated learning training client, a vertical federated learning training passive participant, a model training logical function, or an analytics data repository function for storage of the machine learning model.

[0306] Clause 133. The method of any of clauses 129 to 132, wherein the vertical federated learning session identifier is associated with an analytics identifier, and the request received from the network function further comprises the analytics identifier.

[0307] Clause 134. The method of any of clause 133, wherein the machine learning model is accessed from a vertical federated learning training client, and wherein the method further comprises: sending a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and the analytics identifier; receiving the information from the network repository function; and selecting thevertical federated learning training client based on the information received from the network repository function.

[0308] Clause 135. The method of any of clauses 129 to 134, wherein the method is performed by a second network function, wherein the method further comprises sending a request to a network repository function to store information related to the second network function, and wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning inference client, the analytics identifier, or a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

[0309] Clause 136. The method of any of clauses 129 to 135, wherein the method is performed by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

[0310] Clause 137. A computer-readable storage medium that is non- transitory and has instructions stored therein that, in response to execution by at least one processing circuitry, causes an apparatus to at least: receive a request from a network function of a telecommunications system to perform inferencing using a machine learning model that is one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; access the machine learning model associated with the vertical federated learning session identifier; perform an inferencing using the machine learning model to compute intermediate output data; and send the intermediate data to the network function for inferencing using a federated machine learning model to compute a prediction from the intermediate output data.

[0311] Clause 138. The computer-readable storage medium of clause 137, wherein the request received from the network function comprises a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

[0312] Clause 139. The computer-readable storage medium of clause 138, wherein the machine learning model is accessed further based on the feature identifier.

[0313] Clause 140. The computer-readable storage medium of any of clause 137 to 139, wherein the apparatus caused to access the machine learning model includes the apparatus caused to send a request to at least one of a vertical federated learning training client, a verticalfederated learning training passive participant, a model training logical function, or an analytics data repository function for storage of the machine learning model.

[0314] Clause 141. The computer-readable storage medium of any of clauses 137 to 140, wherein the vertical federated learning session identifier is associated with an analytics identifier, and the request received from the network function further comprises the analytics identifier.

[0315] Clause 142. The computer-readable storage medium of clause 141, wherein the machine learning model is accessed from a vertical federated learning training client, and wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further at least: send a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and the analytics identifier; receive the information from the network repository function; and select the vertical federated learning training client based on the information received from the network repository function.

[0316] Clause 143. The computer-readable storage medium of any of clauses 137 to 142, wherein the apparatus is implemented by a second network function, wherein the computer- readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the apparatus to further send a request to a network repository function to store information related to the second network function, and wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning inference client, the analytics identifier, or a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

[0317] Clause 144. The computer-readable storage medium of any of clauses 137 to 143, wherein the apparatus is implemented by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

[0318] Clause 145. An apparatus comprising means for performing the method of any of clauses 129 to 136.

[0319] Clause 146. A computer-readable medium comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 129 to 136.

[0320] Clause 147. A computer-readable storage medium comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 129 to 136.

[0321] Clause 148. A computer program comprising instructions that, in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 129 to 136.

[0322] Many modifications and other implementations of the disclosure set forth herein will come to mind to one skilled in the art to which the disclosure pertains having the benefit of the teachings presented in the foregoing description and the associated figures. Therefore, it is to be understood that the disclosure is not to be limited to the specific implementations disclosed and that modifications and other implementations are intended to be included within the scope of the appended claims. Moreover, although the foregoing description and the associated figures describe example implementations in the context of certain example combinations of elements and / or functions, it should be appreciated that different combinations of elements and / or functions may be provided by alternative implementations without departing from the scope of the appended claims. In this regard, for example, different combinations of elements and / or functions than those explicitly described above are also contemplated as may be set forth in some of the appended claims. Although specific terms are employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.

Claims

CLAIMS1. An apparatus comprising: means for assigning a vertical federated learning session identifier to a vertical federated learning model training session; means for selecting network functions of a telecommunications system to perform model training of machine learning models in a common sample space but with different feature spaces; and means for sending a request to each of the network functions to perform model training of a respective machine learning model of the machine learning models in the common sample space with a respective feature space of the different feature spaces, wherein the request comprises the vertical federated learning session identifier for storage of the machine learning model in association with the vertical federated learning session identifier.

2. The apparatus of claim 1, wherein the apparatus further comprises: means for receiving intermediate output data from the network functions based on the model training; means for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and means for storing the federated machine learning model in association with the vertical federated learning session identifier.

3. The apparatus of claim 2, wherein the different feature spaces are associated with respective feature identifiers, and wherein the federated machine learning model is stored in further association with the respective feature identifiers associated with the different feature spaces.

4. The apparatus of claim 2 or claim 3, wherein the means for storing the federated machine learning model comprises means for sending a request to an analytics data repositoryfunction for storage of the federated machine learning model at the analytics data repository function.

5. The apparatus of any of claims 2 to 4, wherein the apparatus further comprises: means for receiving a request for the federated machine learning model; and means for sending at least one of the federated machine learning model, information for retrieval of the federated machine learning model, the vertical federated learning session identifier, an analytics identifier, or respective feature identifiers with which the different feature spaces are associated.

6. The apparatus of any of claims 1 to 5, wherein the different feature spaces are associated with respective feature identifiers, and wherein the request sent to each of the network functions further comprises a feature identifier of the respective feature identifiers.

7. The apparatus of any of claims 1 to 6, wherein the federated learning session identifier is associated with an analytics identifier, and wherein the request sent to each of the network functions further comprises the analytics identifier.

8. The apparatus of any of claims 1 to 7, wherein the apparatus further comprises: means for sending a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and an analytics identifier; and means for receiving the information from the network repository function, wherein a network function of the network functions is selected based on the information received from the network repository function.

9. The apparatus of any of claims 1 to 8, wherein the apparatus is implemented by a network function, wherein the apparatus further comprises means for sending a request to a network repository function to store information related to the network function, and wherein the request comprises at least one of an indication that the network function is operable as a vertical federated learning training server, an indication that the network function is operable as a verticalfederated learning training active participant, an analytics identifier, or an identifier of a vendor of the network function.

10. The apparatus of any of claims 1 to 9, wherein the apparatus further comprises: means for selecting a network function to operate as an active participant for the model training of the machine learning models; and means for sending a request to the network function to carry out the model training with the network functions as active participants, the request comprising the vertical federated learning session identifier and an analytics identifier.

11. The apparatus of any of claims 1 to 10, wherein the apparatus is implemented by a network data analytics function, a model training logical function, or an application function of the telecommunications system.

12. An apparatus comprising: means for receiving a request from a network function of a telecommunications system to perform model training of a machine learning model that is to be one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier; means for training the machine learning model in the common sample space with a respective feature space of the different feature spaces, the machine learning model trained to determine intermediate output data; means for sending the intermediate data to the network function for training, based on the intermediate output data, a federated machine learning model to compute a prediction; and means for storing the machine learning model in association with the vertical federated learning session identifier.

13. The apparatus of claim 12, wherein the request received from the network function comprises a feature identifier associated with the respective feature space with which the machine learning model is trained.

14. The apparatus of claim 13, wherein the federated machine learning model is stored in further association with the feature identifier.

15. The apparatus of any of claims 12 to 14, wherein the means for storing the machine learning model comprises means for sending a request to an analytics data repository function for storage of the machine learning model at the analytics data repository function.

16. The apparatus of any of claims 12 to 15, wherein the federated learning session identifier is associated with an analytics identifier, and wherein the request further comprises the analytics identifier.

17. The apparatus of any of claims 12 to 16, wherein the apparatus further comprising: means for receiving a request for the machine learning model comprising the vertical federated learning session identifier and a feature identifier with which the respective feature space is associated; and means for providing a response comprising at least one of the machine learning model, or information for retrieval of the machine learning model.

18. The apparatus of any of claims 12 to 17, wherein the apparatus is implemented by a second network function, wherein the apparatus further comprises means for sending a request to a network repository function to store information related to the second network function, and wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning training client, an indication that the second network function is operable as a vertical federated learning training passive participant, an analytics identifier, or a feature identifier associated with the respective feature space with which the machine learning model is trained.

19. The apparatus of any of claims 12 to 18, wherein the apparatus is implemented by a network data analytics function, or a model training logical function, or an application function of the telecommunications system.

20. An apparatus comprising: means for receiving a request to provide at least one of analytics or one or more predictions for a data analytics procedure associated with an analytics identifier; means for selecting a network function to provide information about a vertical federated learning session associated with the analytics identifier; means for sending a request comprising the analytics identifier to the network function, the request sent to the network function for a federated machine learning model; means for receiving a response from the network function comprising a vertical federated learning session identifier and information about the federated machine learning model; means for accessing the federated machine learning model associated with the vertical federated learning session identifier; means for sending requests to network functions of a telecommunications system to perform inferencing using the machine learning models trained in a common sample space but with different feature spaces, the requests comprising the vertical federated learning session identifier and the analytics identifier for retrieval of the machine learning models based on the vertical federated learning session identifier and the analytics identifier; means for receiving intermediate output data from the network functions based on the inferencing; and means for performing an inferencing, based on the intermediate output data and using the federated machine learning model, to compute a prediction.

21. The apparatus of claim 20, wherein the means for accessing the federated machine learning model comprises means for accessing the federated machine learning model and respective feature identifiers associated with the different feature spaces, and wherein the apparatus further comprises means for selecting the network functions based on the respective feature identifiers.

22. The apparatus of claim 21, wherein each of the requests sent to a respective network function of the network functions further comprises a feature identifier of the respective feature identifiers.

23. The apparatus of any of claims 20 to 22, wherein the means for accessing the federated machine learning model comprises means for sending a request to at least one of a vertical federated learning training server, a vertical federated learning training active participant, a model training logical function, or an analytics data repository function for retrieval of the federated machine learning model.

24. The apparatus of any of claims 20 to 23, wherein the apparatus further comprises: means for sending a request to a network repository function for information related to at least one network function to provide the information about the vertical federated learning session, the request comprising at least one of the analytics identifier or a vendor identifier of the at least one network function; and means for receiving the information from the network repository function, wherein the network function to provide the information about the vertical federated learning session is selected based on the information received from the network repository function.

25. The apparatus of any of claims 20 to 24, wherein the apparatus is implemented by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

26. An apparatus comprising: means for receiving a request from a network function of a telecommunications system to perform inferencing using a machine learning model that is one of machine learning models trained in a common sample space but with different feature spaces, the request comprising a vertical federated learning session identifier;means for accessing the machine learning model associated with the vertical federated learning session identifier; means for performing an inferencing using the machine learning model to compute intermediate output data; and means for sending the intermediate data to the network function for inferencing using a federated machine learning model to compute a prediction from the intermediate output data.

27. The apparatus of claim 26, wherein the request received from the network function comprises a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

28. The apparatus of claim 27, wherein the machine learning model is accessed further based on the feature identifier.

29. The apparatus of any of claims 26 to 28, wherein the means for accessing the machine learning model comprises means for sending a request to at least one of a vertical federated learning training client, a vertical federated learning training passive participant, a model training logical function, or an analytics data repository function for storage of the machine learning model.

30. The apparatus of any of claims 26 to 29, wherein the vertical federated learning session identifier is associated with an analytics identifier, and the request received from the network function further comprises the analytics identifier.

31. The apparatus of claim 30, wherein the machine learning model is accessed from a vertical federated learning training client, and wherein the apparatus further comprises: means for sending a request to a network repository function for information related to at least one network function that supports model training for a feature identifier and the analytics identifier; means for receiving the information from the network repository function; andmeans for selecting the vertical federated learning training client based on the information received from the network repository function.

32. The apparatus of any of claims 26 to 31, wherein the apparatus is implemented by a second network function, wherein the apparatus further comprises means for sending a request to a network repository function to store information related to the second network function, and wherein the request comprises at least one of an indication that the second network function is operable as a vertical federated learning inference client, the analytics identifier, or a feature identifier associated with the one of the different feature spaces with which the machine learning model is trained.

33. The apparatus of any of claims 26 to 32, wherein the apparatus is implemented by a network data analytics function, an analytics logical function, or an application function of the telecommunications system.

Images