Member UE selection for AI / ML

Abstract

A method for selecting (UE) for artificial intelligence / machine learning (AI / ML) in a wireless communication system. The system comprises an application function (AF), a network exposure function (NEF), and a network data analytics function (NWDAF). The method comprises: transmitting, from the AF to the NEF, a first message for requesting UE selection assistance and including member UE filtering criterion and a specific parameter depending on the member UE filtering criteria. The UE filtering criteria includes end-to-end data volume transfer time. Transmitting, from the NEF to the NWDAF, a second message including the specific parameter depending on the criteria and an indication of a request for end-to-end data volume transfer time analytics. Transmitting, from the NWDAF to the NEF, a third message including analytics associated with end to-end data volume transfer time. Deriving, by the NEF, a list of one or more candidate UEs that fulfil the criteria. Transmitting, from the NEF to the AF, a fourth message including I-JES that fulfil the criteria. wherein the specific parameter includes one or more of a data volume of UL / DL data, a target number of data transmission repetitions, a target time interval between data transmissions, and a request for geographical distribution of the UEs.

Description

BACKGROUND Field Certain examples of the present disclosure provide one or more techniques for supporting member User Equipment (UE) selection, for example for supporting Artificial Intelligence (Al) / Machine Learning (ML) federated learning, for example in a 3rd Generation Partnership Project (3GPP) 5th Generation (5G) New Radio (NR) network. Description of the Related Art Herein, the following document(s) may be referenced and the contents thereof are incorporated into the present disclosure: [1] 3GPP TS 22.261 (e.g. V19.5.0) [2] 3GPP TS 23.501 (e.g. V18.4.0) [3] 3GPP TS 23.502 (e.g. V18.4.0) [4] 3GPP TS 23.503 (e.g. V18.4.0) [5] 3GPP TS 23.288 (e.g. V18.4.0) Various acronyms, abbreviations and definitions used in the present disclosure are defined at the end of this description. Overview of AI / ML In AI / ML operation, AI / ML models and / or data may be transferred across the AI / ML applications (AFs), 5G Core (5GC) and UEs (including the AI / ML server in the UE). The AI / ML operation may be divided into two main phases: model training and inference. During model training and inference, multiple rounds of interaction may be required. As described in 3GPP TS 22.261 [1], certain AI / ML operation types may be categorised into three types: model splitting, model sharing, and distributed / federated learning. The requirements, frequency and / or volume of data transmission may differ for different AI / ML processing phrases and / or operation types. In clause 6.40 of 3GPP TS 22.261 [1], AI / ML model transfer of three types of AI / ML operations to be supported in Release 18 are described as follows: - AI / ML operation splitting between AI / ML endpoints The AI / ML operatlon / model is split into multiple parts according to the current task and environment. The intention is to offload the computation-intensive, energy-intensive parts to network endpoints, whereas leave the privacysensitive and delay-sensitive parts at the end device. The device executes the operation / model up to a specific part / layer and then sends the intermediate data to the network endpoint. The network endpoint executes the remaining parts / layers and feeds the inference results back to the device. AI / ML model / data distribution and sharing over 5G system Multi-functional mobile terminals might need to switch the AI / ML model in response to task and environment variations. The condition of adaptive model selection is that the models to be selected are available for the mobile device. However, given the fact that the AI / ML models are becoming increasingly diverse, and with the limited storage resource in a UE, it can be determined to not pre-load all candidate AI / ML models on-board. Online model distribution (i.e. new model downloading) is needed, in which an AI / ML model can be distributed from a NW endpoint to the devices when they need it to adapt to the changed AI / ML tasks and environments. For this purpose, the model performance at the UE needs to be monitored constantly. Distributed / Federated Learning over 5G system The cloud server trains a global model by aggregating local models partially-trained by each end devices. Within each training iteration, a UE performs the training based on the model downloaded from the Al server using the local training data. Then the UE reports the interim training results to the cloud server via 5G UL channels. The server aggregates the interim training results from the UEs and updates the global model. The updated global model is then distributed back to the UEs and the UEs can perform the training for the next iteration. Overview of Support for AI / ML Operation This section describes certain 5GC enablers for supporting distributed / federated learning, model sharing and model operation splitting AI / ML operations in the application layer. The Application Function (AF) that aims to provide an AI / ML service to UE(s) may request 5GC assistance for federated learning, splitting AI / ML operation etc. as described in clause 5.46.2 of 3GPP TS 23.501 [2], The AF may subscribe to the Network Exposure Function (NEF) by sending a list of target UEs, filtering criteria and other corresponding requirements for member UE selection. Based on the AF subscription information, the NEF may collect information and service data of the corresponding candidate UEs and determine the UE(s) that can fulfil the filtering criteria and any other requirements. The NEF notifies the AF of the list of candidate UE(s) that fulfil the certain filtering criteria and requirements provide by the AF. The detailed procedures are described, for example, in clause 4.15.13 of 3GPP TS 23.502 [3]. The list of candidate UE(s) may become the UEs for this AI / ML operation depending on the AF internal policies and final decision. During the AI / ML operation (e.g. federated learning operation), the AF may update the filtering criteria and / or list of target UE(s) to improve or maintain the service quality. This may be done for various reasons, for example if some of the UE cannot fulfil the Quality of Service (QoS) requirement, if some UEs move out of an area of interest resulting in there not being enough UEs to participate the AI / ML operation, if the transfer time of the UEs cannot fulfil the required thresholds, etc. Based on an updated request from the AF, the NEF may update the list of candidate UE(s), and then inform the AF about the new list of candidate UE(s) that fulfil the requirements and / or inform the AF of any UE(s) that cannot fulfil the requirements. Alternatively, the AF may select a list of UE(s) for the AI / ML operation (e.g. distributed / federated learning) without NEF involvement, depending on operator policies, as described in (informative) Annex I of 3GPP TS 23.502 [3], To schedule the AI / ML traffic more efficiently and avoid 5G System (5GS) congestion, the AF that provides the AI / ML service may negotiate with 5GC on a preferred time window for the AI / ML operation (e.g. model transfer and / or inference data transfer) using the Planned Data Transfer with QoS (PDTQ) requirements described in clause 6.1.2.7 of 3GPP TS 23.503 [4], At the time the AI / ML operation starts (e.g. when, or just before, the distributed / federated learning starts), the AF discovers a suitable NEF and requests the NEF to provide the QoS for the list of UEs (for example each UE identified by its UE Internet Protocol (IP) address) that were selected, as described in clause 4.15.13 of 3GPP TS 23.502 [3]. The AF may subscribe to QoS Monitoring for those AF requests for QoS that result in a successful resource allocation. The AF may provide one or more of the following, that are derived from the performance requirements listed in clause 7.10 of 3GPP TS 22.261 [1]: QoS parameters, QoS profiles, QoS requirements, corresponding 5G QoS Identifiers (5Qls), etc. As result of the subscription to NEF to provide the list of UEs that fulfil certain filtering criteria, the AF may be notified about changes in the list of candidate UEs. As such, the AF may request a new preferred time window for the AI / ML operation using the Planned Data Transfer with QoS, or may request to provide a QoS with an updated list of UEs. Among the updated UEs, some may not require resources allocation any longer, and some newly selected UEs may require resource allocation and QoS Monitoring. The AF that aims to provide an AI / ML operation (e.g. model sharing) may request assistance from the 5GC, as described in clause 4.15.3.2.3 in 3GPP TS 23.502 [3] by subscribing to the NEF to be notified on the traffic volume shared between the UE and the AI / ML application server. This may help the AI / ML application server to determine how large the model is, and then, for example, select large models less frequently than small models. In addition, to support AI / ML operations, one or more of the following may be used: • AF requests to the 5G System in the context of 5GS assistance to AI / ML operations in the application layer shall be authorized by the 5GC using existing mechanisms (before Release 18). • User Plane Function (UPF) may perform traffic detection for AI / ML traffic, as defined in clause 5.8.2.4 of 3GPP TS 23.501 [2], and the network may support charging for AI / ML traffic. • Information exposure to an authorized third party. • External parameter provisioning. • Network data analytics support. Overview of Member UE Selection In the 3GPP Release 18 specification, the 5G System can support UE member selection assistance functionality to assist the AF in selecting members I member UE(s) that are able to participate in some services (e.g. the application AI / ML operations, including federated learning, split learning / inference, reinforcement learning, etc.). The member UE selection assistance functionality may be hosted by NEF. Member UE selection procedures are described in clause 4.15.13 of 3GPP TS 23.502 [3], The features of the member UE selection assistance functionality hosted by NEF include the following: • Receiving a request from the AF that shall include a list of target member UE(s) (which may not necessarily be a part of the subsequent updated request), one or more filtering criteria, as specified in Table 4.15.13.2-1 of 3GPP TS 23.502 [3] (e.g. UE current location, UE historical location, UE direction, UE separation distance, QoS requirements, Data Network Name (DNN), preferred access I Radio Access Technology (RAT) type, desired End-to-End (E2E) data volume transfer time performance, or Service Experience), optionally (a) time window(s), etc. • Referring to the filtering criteria provided by the AF and then interacting with 5GC Network Functions (NFs), using existing services to collect the corresponding data for the list of target member UEs from relevant 5GC NFs (e.g. Policy Control Function (PCF), Network Data Analytics Function (NWDAF), Access and Mobility management Function (AMF), Session Management Function (SMF)) to derive the list of candidate UE(s) (i.e. UE(s) among the list of target member UE(s) provided by the AF) which fulfil the filtering criteria. • Providing the AF with the Member UE selection assistance information, including one or more lists of candidate UE(s), and optionally other additional information (e.g. one or more recommended time window(s) for performing the application operation, QoS of each target UE, UE(s) location(s), Access / RAT type, or Service Experience. NEF may also provide the number of UEs per each filtering criterion that do not fulfil the corresponding filtering criterion. The procedures of member UE selection are specified in clause 4.15.13.1 of 3GPP TS 23.502 [3], Referring to Figure 1 of the present disclosure (a reproduction of Figure 4.15.13.1-1 of 3GPP TS 23.502 [3]): 1. AF subscribes the Member UE selection assistance information by sending a first Nnef_MemberUESelectionAssistance_subscribe request including a list of target member UEs, one or more member UE filtering criteria as listed in Table 4.15.13.2-1 and optionally, time window(s). Subsequently, the AF may only update the Member UE filtering criteria of the subscription as described in clause 4.15.13.0 by invoking Nnef_MemberUESelectionAssistance_subscribe and providing a Subscription Correlation ID, i.e. the AFdoes not provide the list of target member UEs again. 2a. [CONDITIONAL] If the AF request does not contain a Subscription Correlation ID, the NEF verifies the authorization of the AF Request and identifies which information needs to be collected for each UE in the list of target member UEs and executes the corresponding service operations based on the Member UE filtering criteria provided by the AF, e.g. events, analytics ID(s), notifications, etc. 2b. [CONDITIONAL] If the AF request contains a Subscription Correlation ID, the NEF correlates the Nnef_MemberUESelectionAssistance_Subscribe request to an existing subscription according to the Subscription Correlation ID. The NEF uses the target member UEs received in step 1 for the Member UE update using the updated filtering criteria. 3. NEF interacts with different 5GC network functions to collect the required information for each UE in the list of target member UEs. The set of interactions between the NEF and the 5GC NFs depend on the Member UE filtering criteria provided by the AF. See Table 4.15.13.2-1 for details. 4. Based on the collected information from other 5GC NFs, the NEF consolidates all the information to derive the list(s) of candidate UEs which fulfil the Member UE filtering criteria in the AF request. The NEF may derive recommended time window(s), considering the validity period(s) of the analytics used for Member UE selection criteria. During the recommended time window(s), the list(s) of candidate UE(s) can fulfil the Member UE filtering criteria. The recommended time window(s) are a subset of the time window(s) received from the AF. In different recommended time windows, the list of candidate UE(s) which fulfil the Member UE filtering criteria may be different. 5. NEF sends a Nnef_MemberUESelectionAssistance_Notify request to the AF including the list(s) of candidate UEs and possibly additional information. See clause 5.2.6.32.4 for details. The NEF performs the member UE selection assistance functionality based on an AF request to determine one or more lists of UEs that fulfil and / or cannot satisfy the filtering criteria and any other requirements. The AF invokes Nnef_MemberUESelectionAssistance_Subscribe service, as detailed in clause 5.2.6.32.2 of 3GPP TS 23.502 [3], by indicating the following parameters to NEF: Inputs, Required: Notification Target Address (+ Notification Correlation ID), at least one filtering criteria shown in Table 4.15.13.2-1. Inputs, Conditional Required: If no Subscription Correlation ID is provided in the subscription, Target of Member UE Selection Assistance Reporting (GPSI ora list of GPSIs) is required. Inputs, Optional: Application ID, Subscription Correlation ID (in the case of modification of the existing subscription), Expiry time, a set of Member UE filtering criteria shown in the Table 4.15.13.2-1 of TS 23.502, time window(s) for selecting the candidate UEs, specific parameters depending on the Member UE filtering criteria, Periodicity (the periodicity of member update), maximum number of UEs (indicates the maximum number of candidate UEs that need to be fed back). Outputs, Required: When the subscription is accepted: Subscription Correlation ID, Expiry time (required if the subscription can be expired based on the operator's policy). After sending the subscribe request, the AF expects certain outputs from NEF via Nnef_MemberUESelectionAssistance_Notify, in particular a list of candidate UE(s), and optionally corresponding time window(s), specific values of the parameters per candidate UE, and the number of UEs that cannot fulfil specific filtering criteria, as defined in clause 5.2.6.32.4 of 3GPP TS 23.502 [3]: Service operation name: Nnef_MemberUESelectionAssistance_Notify Description: NEF reports the Member UE selection assistance information to the consumer that has previously subscribed. Inputs, Required: Notification Correlation Information. Inputs, Conditional Required: At least one of the following inputs is required: - One or more list(s) of candidate UE(s). Inputs, Optional: Recommended time window for performing the application operation per list of candidate UE(s) as described in clause 4.15.13.1, specific value of the parameters that NEF gathered for the Member UE filtering criteria per candidate UE, a number for each filtering criterion that indicates the UEs in the initial list which do not meet the criterion (provided if there are multiple filtering criteria in the subscribe request). NOTE: This number can be an indication for AF to revise the corresponding filtering criterion. Outputs, Required: Operation execution result indication. The AF may determine the UE(s) that participate the AI / ML operation and / or the operation time window considering the outputs of NEF and the AF internal logic. Overview of Filtering Criteria of Member UE Selection The filtering criteria of member UE selection and the corresponding procedures are specified in clause 4.15.13.2 to clause 4.15.13.6 of 3GPP TS 23.502 [3], The filtering criteria may include QoS requirements, access type or RAT type of Protocol Data Unit (PDU) session, End-to-End (E2E) data volume transfer time analytics, UE current and historical location, UE separation distance, service experience analytics and DNN. If AF sends the E2E data volume transfer time analytics as one of the filtering criteria to NEF, the NEF will subscribe to, or send a request to, the NWDAF to obtain the corresponding output analytics. As specified in clause 6.18 of 3GPP TS 23.288 [5], in order to assist with the AI / ML operation, based on the NEF subscription or request, the NWDAF may derive the transfer time analytics by considering the following filter information and input parameters: The consumer of these analytics indicates in the request or subscription: - Analytics Filter Information including: - Optionally, DNN; - Optionally, S-NSSAI; - Optionally, Application ID; - Optionally, Area of Interest (AOI(s)): restricts the scope of the E2E data volume transfer time analytics to the provided area; - Optionally, a list of analytics subsets that are requested (see clause 6.18.3); - QoS requirements (e.g. 5QI, QoS Characteristics); - Optionally, either a target number of repeating data transmissions or a target time interval between data transmissions within the Analytics target period; - Data Volume UL / DL: indicates a specific data volume transmitted once from UE to AF and / or from AF to UE; - A request for geographical distribution (i.e. the Aois) of the UEs. An Analytics target period indicates the time period over which the statistics or predictions are requested. - In a subscription, the Notification Correlation Id and the Notification Target Address are included. - Optionally, preferred level of accuracy of the analytics. - Optionally, preferred level of accuracy per analytics subset (see clause 6.18.3). - Optionally, preferred order of results for the list of E2E data volume transfer time: - ordering criterion: "E2E data volume transfer time", - order: ascending or descending. - Optionally, Reporting Thresholds, which apply only for subscriptions and indicate conditions on the levels to be reached for the respective analytics subsets (see clause 6.18.3) - Optionally, maximum number of UEs. The Data Volume UL / DL and a request for geographical distribution (i.e. the Aois) of the UEs are two new parameters added into the subscription or the request of the NWDAF consumer, as agreed in 3GPP S2-2313630 in the SA2 160 meeting. The above information is presented as background information only to assist with an understanding of the present disclosure. No determination has been made, and no assertion is made, as to whether any of the above might be applicable as prior art with regard to the present invention. SUMMARY It is an aim of certain examples of the present disclosure to address, solve and / or mitigate, at least partly, at least one of the problems and / or disadvantages associated with the related art, for example at least one of the problems and / or disadvantages described herein. It is an aim of certain examples of the present disclosure to provide at least one advantage over the related art, for example at least one of the advantages described herein. The present invention is defined in the independent claims. Advantageous features are defined in the dependent claims. Embodiments or examples disclosed in the description and / or figures falling outside the scope of the claims are to be understood as examples useful for understanding the present invention. In accordance with an aspect of the present disclosure, there is provided a method for selecting one or more user equipment (UE) for participating in an artificial intelligence / machine learning (AI / ML) operation in a wireless communication system, the wireless communication system comprising an application function (AF), a network exposure function (NEF), and a network data analytics function (NWDAF), and the method comprising: transmitting, from the AF to the NEF, a first message for requesting UE selection assistance and including member UE filtering criterion and a specific parameter depending on the member UE filtering criteria, wherein the member UE filtering criteria includes an end-to-end data volume transfer time; transmitting, from the NEF to the NWDAF, a second message including the specific parameter depending on the member UE filtering criteria and an indication of a request for end-to-end data volume transfer time analytics; transmitting, from the NWDAF to the NEF, a third message including one or more analytics associated with end-to-end data volume transfer time; deriving, by the NEF, based on the analytics and the member UE filtering criteria, a list of one or more candidate UEs that fulfil the member UE filtering criteria; and transmitting, from the NEF to the AF, a fourth message including the list of one or more candidate UEs that fulfil the member UE filtering criteria, wherein the specific parameter depending on the member UE filtering criteria includes one or more of a data volume of UL / DL data, a target number of data transmission repetitions, a target time interval between data transmissions, and a request for geographical distribution of the UEs. In an example, the method further comprises: collecting, by the NWDAF, input data for end-to-end data volume transfer time analytics based on the specific parameter depending on the member UE filtering criteria. In an example, the NWDAF uses the data volume as a reference to collect the input data for the end-to-end data volume transfer time analytics. In an example, the NWDAF derives the analytics associated with end-to-end data volume transfer time based on the input data. In an example, the analytics associated with the end-to-end data volume transfer time are associated with a data volume similar to the data volume of the specific parameter depending on the member UE filtering criteria. In an example, the data volume is an expected or an observed data volume from a UE to an AF or from an AF to a UE of an AI / ML-based service. In an example, the geographical distributions of UEs includes areas of interest (Aoi) of the UEs. In an example, the first message includes an initial UE list, and the second message includes an indication of one or more UEs from the initial UE list that are a target of the analytics. In an example, the one or more UEs from the initial list are determined based on one or more of the member UE filtering criteria and the specific parameter depending on the member UE filtering criteria. In an example, the deriving includes: deriving the list of one or more candidate UEs using an average or variance of end-to-end data volume transfer time of the specific data volume of UL / DL data. In an example, the method further comprises: selecting, by the NEF, the NWDAF based on a request sent to a further network entity including an Analytics ID indicating end-to-end data volume transfer time. In an example, the request is a Nudm_UECM_Get message or a Nnrf_NFDiscovery_Request message. In an example, the method further comprises: verifying, by the NEF, an authorization of the request for UE selection assistance; and identifying information that needs to be collected and executed based on the member UE filtering criteria. In an example, the first message is a Nnef_MemberUESelectionAssistance_subscribe message. In an example, the second message is an Nnwdaf_AnalyticsSubscription_Subscribe message and the third message is a Nnwdaf_AnalyticsSubscription_Notify message. In an example, the second message is a Nnwdaf_Analyticslnfo_Request message and the third message is a Nnwdaf_Analyticslnfo_Request response message. In an example, the first, second, and / or third messages includes an Analytics ID indicating the end-to-end data volume transfer time. According to an aspect of the present disclosure there is provided a wireless communication system comprising an application function (AF), a network exposure function (NEF), and a network data analytics function (NWDAF), wherein the wireless communication system is configured to: transmit, from the AF to the NEF, a first message for requesting UE selection assistance and including member UE filtering criteria and a specific parameter depending on the member UE filtering criteria, wherein the member UE filtering criteria includes an end-to-end data volume transfer time; transmit, from the NEF to the NWDAF, a second message including the specific parameter depending on the member UE filtering criteria and an indication of a request for end-to-end data volume transfer time analytics; transmit, from the NWDAF to the NEF, a third message including one or more analytics associated with end-to-end data volume transfer time; derive, by the NEF, based on the analytics and the member UE filtering criteria, a list of one or more candidate UEs that fulfil the member UE filtering criteria; and transmit, from the NEF to the AF, a fourth message including the list of one or more candidate UEs that fulfil the member UE filtering criteria, wherein the specific parameter depending on the member UE filtering criteria includes one or more of a data volume of UL / DL data, a target number of data transmission repetitions, a target time interval between data transmissions, and a request for geographical distribution of the UEs. In an example, the method further comprises: collecting, by the NWDAF, input data for end-to-end data volume transfer time analytics based on the specific parameter depending on the member UE filtering criteria. In an example, the NWDAF uses the data volume as a reference to collect the input data for the end-to-end data volume transfer time analytics. In an example, the NWDAF derives the analytics associated with end-to-end data volume transfer time based on the input data. In an example, the analytics associated with the end-to-end data volume transfer time are associated to a data volume similar to the data volume of the specific parameter depending on the member UE filtering criteria. In an example, the data volume is an expected or an observed data volume from a UE to an AF or from an AF to a UE of an AI / ML-based service. In an example, the geographical distributions of UEs includes areas of interest (Aoi) of the UEs. In an example, the wireless communication system is configured to implement any of the abovedetailed methods. Other aspects, advantages and salient features of the invention will become apparent to those skilled in the art from the following detailed description taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a reproduction of Figure 4.15.13.1-1 of 3GPP TS 23.502 [3]: 5GC assistance to Member UE selection and update; Figure 2 is a reproduction of Figure 4.15.13.6.2-1 of 3GPP TS 23.502 [3]: Assistance to member UE selection for end-to-end data volume transfer time related filtering criteria; and Figure 3 is a block diagram of an exemplary network entity that may be used in certain examples of the present disclosure. DETAILED DESCRIPTION The following description of examples of the present disclosure, with reference to the accompanying drawings, is provided to assist in a comprehensive understanding of the present invention, as defined by the claims. The description includes various specific details to assist in that understanding but these are to be regarded as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the examples described herein can be made without departing from the scope of the invention. The same or similar components may be designated by the same or similar reference numerals, although they may be illustrated in different drawings. Detailed descriptions of techniques, structures, functions, operations or processes known in the art may be omitted for clarity and conciseness, and to avoid obscuring the subject matter of the present invention. The terms and words used herein are not limited to the bibliographical or standard meanings, but, are merely used to enable a clear and consistent understanding of the invention. Throughout the description and claims of this specification, the words “comprise”, “include” and “contain” and variations of the words, for example “comprising” and “comprises”, means “including but not limited to”, and is not intended to (and does not) exclude other features, elements, components, integers, steps, processes, operations, functions, characteristics, properties and / or groups thereof. Throughout the description and claims of this specification, the singular form, for example “a”, “an” and “the”, encompasses the plural unless the context otherwise requires. For example, reference to “an object” includes reference to one or more of such objects. Throughout the description and claims of this specification, language in the general form of “X for Y” (where Y is some action, process, operation, function, activity or step and X is some means for carrying out that action, process, operation, function, activity or step) encompasses means X adapted, configured or arranged specifically, but not necessarily exclusively, to do Y. Features, elements, components, integers, steps, processes, operations, functions, characteristics, properties and / or groups thereof described or disclosed in conjunction with a particular aspect, embodiment, example or claim are to be understood to be applicable to any other aspect, embodiment, example or claim described herein unless incompatible therewith. The skilled person will appreciate that the techniques described herein may be used in any suitable combination. Certain examples of the present disclosure provide one or more techniques for supporting UE member selection, for example for supporting Al / ML federated learning, for example in a 3GPP 5G NR network. However, the skilled person will appreciate that the present invention is not limited to these examples, and may be applied in any suitable system or standard, for example one or more existing and / or future generation wireless communication systems or standards, including any existing or future releases of the same standards specification, for example 3GPP 5G, 5G-advanced or 6th Generation (6G). The functionality of the various network entities and other features disclosed herein may be applied to corresponding or equivalent entities or features in the same or any other suitable communication systems or standards. Corresponding or equivalent entities or features may be regarded as entities or features that perform the same or similar role, function or purpose within the network. For example, the functionality of a base station or the like (e.g. eNB, gNB, NB, Radio Access Network (RAN) node, access point, wireless point, transmission / reception point, central unit, distributed unit, radio unit, remote radio head, etc.) in the examples below may be applied to any other suitable type of entity performing RAN functions, and the functionality of a UE or the like (e.g. electronic device, user device, mobile station, subscriber station, customer premises equipment, terminal, remote terminal, wireless terminal, vehicle terminal, etc.) in the examples below may be applied to any other suitable type of device. A particular network entity may be implemented as a network element on a dedicated hardware, as a software instance running on a dedicated hardware, and / or as a virtualised function instantiated on an appropriate platform, e.g. on a cloud infrastructure. The skilled person will appreciate that the present invention is not limited to the specific examples disclosed herein. For example: • The techniques disclosed herein are not limited to 3GPP 5G. • One or more entities in the examples disclosed herein may be replaced with one or more alternative entities performing equivalent or corresponding functions, processes or operations. • One or more of the messages in the examples disclosed herein may be replaced with one or more alternative messages, signals or other type of information carriers that communicate equivalent or corresponding information. • One or more further elements or entities may be added to the examples disclosed herein. • One or more non-essential elements or entities may be omitted in certain examples. • The functions, processes or operations of a particular entity in one example may be divided between two or more separate entities in an alternative example. • The functions, processes or operations of two or more separate entities in one example may be performed by a single entity in an alternative example. • Information carried by a particular message in one example may be carried by two or more separate messages in an alternative example. • Information carried by two or more separate messages in one example may be carried by a single message in an alternative example. • The order in which operations are performed and / or the order in which messages are transmitted may be modified, if possible, in alternative examples. Certain examples of the present disclosure may be provided in the form of an apparatus / device / network entity configured to perform one or more defined network functions and / or a method therefor. Certain examples of the present disclosure may be provided in the form of a system (e.g. network or wireless communication system) comprising one or more such apparatuses / devices / network entities, and / or a method therefor. The following problems arise in the related art. Problem 1 As described above, the Data Volume UUDL and a request for geographical distribution (i.e. the Aois) of the UEs are two new and mandatory parameters in the subscription or the request of the NWDAF consumer for E2E data volume transfer time analytics. In addition, QoS requirements are mandatory filter information of E2E data volume transfer time analytics. A target number of repeating data transmissions and / or a target time interval between data transmissions is needed by NWDAF to calculate an average value of every data volume transfer time within the Analytics target period. When deploying E2E traffic volume transfer time analytics to assist the AIML operation, the NEF and / or the AF may be the consumer of NWDAF. The AF and / or NEF should inform NWDAF of certain information, for example one or more of: expected / observed / measured Data Volume UL / DL / roundtrip, a request for geographical distribution (e.g. the Aois) of the UEs, a target number of repeating data transmissions, a target time interval between data transmissions, and QoS requirements (e.g. 5QI, QoS Characteristics). However, in the current specification, it is not clear how to provide the information mentioned above in the subscription request by AF and / or NEF to NWDAF. In particular, when the NEF is the analytics consumer, the NEF may not have any previous knowledge of the above information. It is not clear how the AF can provide the above information to the NEF. It is also not clear whether it is possible for the NEF and / or NWDAF to reuse parameters associated with other UE member filtering criterion as the filter information of E2E data volume transfer time analytics. Problem 2 In the current member UE selection assistance functionality, as described in clause 5.2.6.32.4 of 3GPPTS 23.502 [3], the AF may revise the filtering criterion based on the number of UEs that cannot fulfil the filtering criterion. However, the UE and / or network conditions could vary significantly in different time windows or at different time points. In some situations, a relatively large number of UEs may not fulfil the filtering criterion, but only within a relatively low number of time windows or time points; in other time windows or at other time points, the number of UEs not fulfilling the filtering criteria may be relatively low or zero. In this case, the AF may not need to lower the threshold and / or requirements of the filtering criterion to maintain the overall performance of the AI / ML operation, or any other types of operations / services. However, in the current specification, the AF is not aware of the corresponding time window(s) or time point(s) associated with the number of UEs that cannot fulfil a filtering criterion. Certain examples of the present disclosure provide one or more techniques to support AI / ML operations, for example federated learning. In order to support member UE selection assistance functionality, for example when using E2E data volume transfer time analytics as one of the filtering criteria for member UE selection, certain examples of the present disclosure use one or more new parameters and / or items of information. In certain examples, the parameters / information include one or more of the following: • Information relating to data transmissions. This may include, for example, a target number of repeating data transmission and / or a target time interval between data transmissions. • Information relating to QoS requirements of a service. The QoS requirements may include, for example, 5QI and / or QoS Characteristics. • Information relating to data volume. This may include, for example, an expected data volume, an observed data volume and / or a measured data volume. The expected data volume may be an expected data volume to be transferred between the UE and the AF, from the UE to the AF, and / or from the AF to the UE, and / or an expected roundtrip / UL / DL data volume. The observed / measured data volume may be an observed / measured data volume transferred between the UE and the AF, from the UE to the AF, and / or from the AF to the UE, and / or an observed / measured roundtrip / UL / DL data volume. • A request for geographical distribution of the UEs. This may include, for example, one or more Aois. The new parameters may be included in any suitable message between any suitable network entities, for example the AF request of the Member UE Selection Assistance Subscribe service operation. For example, the parameters may be included in Nnef_MemberUESelectionAssistance_Subscribe as a part of the Input, Optional: specific parameters depending on the Member UE filtering criteria, as specified in clause 5.2.6.32.2 of 3GPP TS 23.502 [3], The parameters may be considered as UE filtering information of corresponding Member UE filtering criteria (e.g., QoS, E2E data volume transfer time, UE historical location, UE current location, UE separation distance, etc.). Existing Member UE filtering criteria are specified in Table 4.15.13.2-1 of 3GPP TS 23.502 [3], By indicating the new parameters / information, for example to the NEF by the AF, the NEF is able to include the corresponding parameters / information into the request or subscription to the NWDAF. Therefore, the NWDAF is able to derive the analytics outputs to support the member UE selection assistance based on the request, for example of NEF and / or AF. Furthermore, in order to assist the AF with updating / optimising the filtering criteria of member UE selection, in certain examples of the present disclosure, information relating to one or more time windows and / or one or more time points may be provided. For example, the time windows / time points may correspond to existing specific value measurements per UE filtering criteria and / or a number of UEs that cannot fulfil the specific filtering criterion. The information may include a number of time windows / time points within which there are UEs that cannot fulfil a specific filtering criterion. The information relating to time windows / time points may be included in any suitable message between any suitable network entities. For example it may be provided by the NEF to the AF via Nnef_MemberUESelectionAssistance_Notify service operation. Based on the above information, and possibly also any other suitable information (e.g. existing / legacy information), the AF may optimise the filtering criteria, for example considering the condition of the UEs and / or the network, as well as the service requirements of the application operation (e.g. AI / ML-based service, federated learning services). Therefore, the AF is able to maintain and improve the service quality and experience. Certain examples of the present disclosure provide a method for identifying one or more UEs (e.g. performing member UE selection) for participating in an AI / ML operation (e.g. federated learning). In certain examples, the method may comprise transmitting by a first network entity (e.g. AF) that provides an AI / ML service, to a second network entity (e.g. NEF), a first message (e.g. Nnef_MemberllESelectionAssistance_Subscribe), wherein the first message comprises one or more filtering criteria (e.g. Member UE filtering criteria), wherein the first message corresponds to a request for information for identifying one or more candidate UEs (e.g. from among one or more target UEs) that fulfil the filtering criteria (and / or one or more UEs that do not fulfil the filtering criteria). In certain examples, the filtering criteria may comprise a criterion related to data transfer (e.g. an E2E data volume transfer time) between a UE and the first network entity. In certain examples, the E2E data volume transfer time may refer to a time delay for completing the transmission of a specific data volume (e.g. from UE to the first network entity and / or from the first network entity to UE). In certain examples, the data volume transfer time may be derived based on a target number of repeating data transmissions and / or a target time interval between data transmissions within an analytics target period. In certain examples, the specific data volume may be indicated by a network entity (e.g. AF and / or NEF). In certain examples, the first message may comprise information corresponding to the filtering criteria, the information based on one or more of: an expected data volume to be transferred between the UE and the first network entity (e.g. UL, DL and / or roundtrip); an observed and / or measured data volume transferred between the UE and the first network entity (e.g. UL, DL and / or roundtrip); a target number of repeating data transmissions and / or a target time interval between data transmissions; a request for geographical distribution of UEs (e.g. one or more Aois); and one or more QoS requirements (e.g. 5QI and / or QoS Characteristics). In certain examples, the filtering criterion relating to data transfer may be based on one or more of: an average and / or variance; and a specific data volume transferred between UE and the first network entity. In certain examples, the first message may comprise information indicating one or more target UEs. In certain examples, the method may comprise transmitting, from the second network entity and / or the first network entity, to a third network entity providing network analytics (e.g. NWDAF), a second message (e.g. Nnwdaf_Analyticslnfo_Request and / or Nnwdaf_AnalyticsSubscription_Subscribe), wherein the second message comprises information for obtaining analytics relating to the data transfer (e.g. E2E data volume transfer time analytics) and / or a request for geographical distribution of UEs, wherein the second message corresponds to a request for receiving the analytics. In certain examples, the information for obtaining the analytics may comprise information based on one or more of: an expected data volume to be transferred between the UE and the first network entity (e.g. UL, DL and / or roundtrip); an observed and / or measured data volume transferred between the UE and the first network entity (e.g. UL, DL and / or roundtrip); a target number of repeating data transmissions and / or a target time interval between data transmissions; a request for geographical distribution of UEs (e.g. one or more Aois); and one or more QoS requirements (e.g. 5QI and / or QoS Characteristics). In certain examples, the first message may comprise time information (e.g. indicating one or more time windows and / or time points), and the second message may comprise one or more analytics target periods based on the time information. In certain examples, the method may comprise receiving, by the first network entity or the second network entity, from the third network entity, a third message (e.g. Nnwdaf_Analyticslnfo_RequestResponse and / or Nnwdaf_AnalyticsSubscription_Notify) including the analytics. In certain examples, the method may comprise identifying, by the second network entity, based on the analytics, one or more UEs that fulfil the filtering criteria (and / or one or more UEs that do not fulfil the filtering criteria). In certain examples, the method may comprise transmitting, by the second network entity to the first network entity, a fourth message (e.g. Nnef_MemberUESelectionAssistance_Notify), wherein the fourth message comprises information identifying one or more candidate UEs that fulfil the filtering criteria (and / or one or more UEs that do not fulfil the filtering criteria). In certain examples, the fourth message may comprise information indicating the number of UEs that do not fulfil the filtering criteria. In certain examples, the fourth message may comprise information (e.g. determined based on one or more validity periods of the analytics) indicating one or more times (e.g. one or more time windows and / or time points) at which the one or more candidate UEs fulfil the filtering criteria (and / or at which one or more UEs do not fulfil the filtering criteria). In certain examples, the fourth message may comprise information indicating one or more times (e.g. one or more time windows and / or time points) associated with a number of UEs that do not fulfil the filtering criteria. In certain examples, the method may comprise modifying, by the first network entity, at least one filtering criterion based on information received from the second network entity. Certain examples of the present disclosure provide a UE, AF, NEF, NWDAF and / or other network entity (e.g. AMF, SMF) configured to perform a method according to any example, aspect, embodiment and / or claim disclosed herein. Certain examples of the present disclosure provide a network (or wireless communication system) comprising a UE, AF, NEF and / or NWDAF according to any examples, aspects, embodiments and / or claims disclosed herein. Certain examples of the present disclosure provide a computer program comprising instructions which, when the program is executed by a computer or processor, cause the computer or processor to carry out a method according to any example, aspect, embodiment and / or claim disclosed herein. Certain examples of the present disclosure provide a computer or processor-readable data carrier having stored thereon a computer program according to any example, aspect, embodiment and / or claim disclosed herein. Various examples of the present disclosure will now be described in more detail. Member UE selection assistance request by AF New Parameters As described above (Problem 1), the Data Volume UL / DLand a request for geographical distribution (e.g. Aois) of the UEs are requested as input parameters to support the NWDAF to derive the E2E data volume transfer time analytics for supporting services (e.g. AI / ML service, federated learning services, etc.). Therefore, in certain examples of the present disclosure, these parameters may be provided to the NWDAF by the service consumer (e.g. NEF), for example via Nnwdaf_AnalyticsSubscription_Subscribe service and / or Nnwdaf_Analyticslnfo_Request service. The NWDAF may consider the Data Volume UL / DL when deriving the E2E data volume transfer time analytics. The Data Volume UL / DL may include the (approximate) expected, estimated, measured and / or observed size of the (potential) AI / ML related traffic (to be) transferred between any suitable network entities, for example between the UE and the AF, between the AF and the RAN node, and / or between the RAN node and the UE. The AI / ML related traffic may be of any suitable form, for example AI / ML model, AI / ML inference data and / or signalling related to AI / ML service operating. For the UL Data Volume, the AI / ML related traffic may be transmitted from UE and / or RAN node / base station; in this case, the UL Data Volume in the consumer’s request may be the data volume estimated by the AF based on its internal logic and knowledge of the AI / ML based services, etc. As indicated in Table 6.18.3-1: E2E data volume transfer time statistics, and Table 6.18.3-2: E2E data volume transfer time predictions of 3GPP TS 23.288 [5] (reproduced below), if the consumer request the Geographical distribution of the UE(s), the NWDAF should provide the Geographical distribution of the corresponding the target UEs. Therefore, the consumer is able to determine the number of UEs in each location / position. The location / position information may be of any suitable type, for example an area of interest (Aoi), a cell, a Tracking Area (TA), etc. UEs in the same or similar location / position may experience similar network conditions, for example congestion level and / or load of the network. Accordingly, this parameter may help the consumer with determining the UEs that are suitable to participate in the AI / ML operation. Information Description > Geographical distribution of the UE(s) If requested, a list of UEs per location information. Table 1 (reproduction of Table 6.18.3-11 Table 6.18.3-2: E2E data volume transfer time statistics / predictions of 3GPP TS 23.288 [5]) The Geographical distribution of the UE(s) is only available as output analytics when the reporting target is a group of UEs or a list of UE IDs (e.g. a list of Subscription Permanent Identifiers (SUPIs) and / or GPSIs). In this case, a request for geographical distribution of UEs may be indicated using any suitable message, for example in either Nnwdaf_AnalyticsSubscription_Subscribe service or Nnwdaf_Analyticslnfo_Request service by the consumer. If the reporting target of the analytics is ‘any UE’ or a single UE, then if a request for Geographical distribution is indicated, for example in either Nnwdaf_AnalyticsSubscription_Subscribe service or Nnwdaf_Analyticslnfo_Request service by the consumer, then the NWDAF is not able to provide the Geographical distribution of the UE(s). As described in clause 6.18.1 of 3GPP TS 23.288 [5], if a target number of repeating data transmissions or a target time interval between data transmissions is given, the E2E data volume transfer time can be provided as an average value of every data volume transfer time within the Analytics target period. The QoS requirements (e.g. 5QI, QoS Characteristics, etc.) may be mandatory filter information of the E2E data volume transfer time analytics, which should be provided by the consumer via the request or subscription, for example by the NEF if the NEF is the consumer of the NWDAF analytics. However, when the NEF is the consumer of the E2E data volume transfer time analytics, the NEF may not have prior knowledge of the target number of repeating data transmission or target time interval or the QoS requirements (e.g. 5QI, QoS Characteristics). Therefore, in the current specification, the NEF is not able to include the target number of repeating data transmission or target time interval or the QoS requirements in the subscription or the request to the NWDAF. In this case, the NWDAF cannot use the QoS requirements as analytics filter information, i.e. for analytics outputs reporting. The consumer of the E2E data volume transfer time analytics may be AF or NEF, as specified in 3GPP TS 23.288 [5], In certain examples of the present disclosure, the consumer of the analytics (e.g. NEF or AF) may indicate certain information to the NWDAF, for example one or more of: Data Volume UL / DL / roundtrip (e.g. expected, observed and / or measured), a request for geographical distribution (e.g. Aois) of the UEs, QoS requirements, target number of repeating data transmission, and target time interval between data transmissions. The information may be indicated, for example, via either Nnwdaf_AnalyticsSubscription_Subscribe service or Nnwdaf_Analyticslnfo_Request service. In certain examples, the request for geographical distribution of the UEs may be indicated by the consumer only if the target of reporting is a group of UEs and / or a list of UE IDs (e.g. a list of SUPIs and / or GPSIs) and / or if the output analytics are associated with and / or applies to a group of UEs and / or a list of UE IDs. The consumer of the E2E data volume transfer time analytics may be NEF. Based on the existing procedures in the current specification, the NEF may not know the Data Volume UL / DL / roundtrip (expected, observed and / or measured), QoS requirements, the target number of repeating data transmissions, the target time interval between data transmissions, and / or that a request for geographical distribution (e.g. Aois) of the UEs is made. Therefore, the NEF may not be able to provide the above information, for example in either Nnwdaf_AnalyticsSubscription_Subscribe service or Nnwdaf_Analyticslnfo_Request service operations. The NEF collects data formember UE selection assistance functionality based on AF request. The NEF may not have any prior knowledge of the service (e.g. AI / ML or federated learning service), for example if the member UE selection is the initial selection for a service, not an update of the member UE selection. Therefore, in certain examples of the present disclosure, the AF may inform the NEF of certain information, for example one or more of: Data Volume UL / DL / roundtrip (e.g. expected, observed and / or measured), a request for geographical distribution (e.g. Aois) of the UEs, QoS requirements, target number of repeating data transmission, and target time interval between data transmissions. For example, the information may be included in the Nnef_MemberUESelectionAssistance_Subscribe service operation from AF to NEF, for example along with or as part of the E2E data volume transfer time filtering criterion. In summary, to support certain analytics (e.g. E2E data volume transfer time analytics) for supporting certain operations (e.g. AI / ML services or a federated learning operation), certain examples of the present disclosure provide on or more of the following items of information: • Information relating to data transmissions. This may include, for example, a target number of repeating data transmission and / or a target time interval between data transmissions. • Information relating to QoS requirements of a service. The QoS requirements may include, for example, 5QI and / or QoS Characteristics. • Information relating to data volume. This may include, for example, an expected data volume, an observed data volume and / or a measured data volume. The expected data volume may be an expected data volume to be transferred between the UE and the AF, from the UE to the AF, and / or from the AF to the UE, and / or an expected roundtrip / UL / DL data volume. The observed / measured data volume may be an observed / measured data volume transferred between the UE and the AF, from the UE to the AF, and / or from the AF to the UE, and / or an observed / measured roundtrip / UL / DL data volume. • A request for geographical distribution of the UEs. This may include, for example, one or more Aois. Some or all of the above information may be indicated in any suitable messages between any suitable network entities. For example, the information may be provided by an analytics (e.g. E2E data volume transfer time analytics) consumer (e.g. NEF and / or AF) to NWDAF in the Nnwdaf_AnalyticsSubscription_Subscribe and / or Nnwdaf_Analyticslnfo_Request messages. In the case that NEF is the analytics consumer, the information may be provided by AF to NEF in the Nnef_MemberllESelectionAssistance_Subscribe request message, for example as part of the UE filtering information / specific parameters associated with Member UE filtering criteria. The Member UE filtering criteria may include one or more of the UE filtering criteria defined in Table 4.15.13.2-1 of TS 23.502 [3], for example QoS, E2E data volume transfer time, UE historical location, UE current location, UE separation distance, etc. The request for geographical distribution of UEs may be indicated by the AF to the NEF, for example via Nnef_MemberUESelectionAssistance_Subscribe request service. If the NEF receives the request, for example in the subscription request service, the NEF may include the request to the NWDAF. In certain examples, the NEF may include the request for UEs geographical distribution to the NWDAF only if the target of analytics reporting is a group of UEs and / or a list of UE IDs (e.g. SUPIs and / or GPSIs) and / or if the output analytics applies to a group of UEs and / or a list of UE IDs (e.g. SUPIs and / or GPSIs). The request for geographical distribution of UEs and (expected) Data Volume / size of expected data I specific data volume may be associated with E2E data volume transfer time filtering criteria. In certain examples, if the AF include the E2E data volume transfer time filtering criteria in the Nnef_MemberUESelectionAssistance_Subscribe service operation, the AF may include the request for geographical distribution of UEs and (expected) Data Volume / size of expected data / specific data volume, the target number of repeating data transmission, the target time interval, and / or the QoS requirements (e.g. 5QI, QoS Characteristics) into Nnef_MemberUESelectionAssistance_Subscribe service operation. Enhancements to Member UE filtering criteria In order to clarify the how to indicate the new parameter(s) from AF and NEF, and how the NEF will use the new parameter(s) associated with different analytics, and inform the NWDAF of the corresponding parameter(s), one or more E2E data volume transfer time Member UE filtering criteria (for example, one or more existing criteria) may be extended (and / or replaced) with one or more of the new parameters, for example as indicted in Table 2 below. The E2E data volume transfer time is described in the current specification as follows: Indicate the target end-to-end data volume transfer time that refers to a time for completing the transmission of a specific data volume between UE and AF, e.g. the average and variance of End-to-end data volume transfer time. In certain examples, the specific data volume between and AF may be: the expected data volume to be transferred between the UE and the AF, from the UE to the AF, and / or from the AF to the UE, and / or the roundtrip / UL / DL data volume; and / or the observed / measured data volume transferred between the UE and the AF, from the UE to the AF, and / or from the AF to the UE, and / or the roundtrip / UL / DL data volume. In certain examples, the AF may indicate the request for geographical distribution (e.g. the Aois) of the UEs, and the NEF may include the request for geographical distribution (e.g. the Aois) of the UEs to the NWDAF to derive the E2E data volume transfer time analytics. Therefore, the target E2E data volume transfer time (e.g. the average and / or variance of E2E data volume transfer time) may also be associates with the geographical distribution (i.e. the Aois) of the UEs. In certain examples, the data volume transfer time may be derived by considering the target number of repeating data transmissions and / or the target time interval between data transmissions within the analytics target period. In certain examples, the target repetition number of data transmissions or a target time interval between data transmissions may be given within the E2E data volume transfer time as part of the parameters associated with the E2E data volume transfer time member UE filtering criterion. The NEF may inform the above parameter to the NWDAF. If it is given, the E2E data volume transfer time may be provided, for example as an average value of the data volume transfer time considering the target repetition number of data transmissions or a target time interval between data transmissions within the analytics target period by the NWDAF. In certain examples, the NEF may also inform the NWDAF of using QoS requirements (e.g. 5QI, QoS Characteristics) as filter information of E2E data volume transfer time analytics, when the NEF deploys the E2E data volume transfer time as one of the Member UE filtering criteria. In certain examples, when deploying the E2E data volume transfer time member UE filtering criterion, the UE filtering information may include the E2E data volume transfer time UL and DL. Therefore, the NEF may use the target E2E data volume transfer time UL and DL to select the UEs that can fulfil this filtering criterion. Member UE filtering criteria Description of filtering criteria for member UEs selected by NEF UE filtering information Detailed description clause End-to-end data volume transfer time Indicate the target end-to-end data volume transfer time that refers to a time for completing the transmission of a specific data volume between UE and AF, e.g. the average and variance of End-to-End data volume transfer time Indicate the target end-to-end data volume transfer time that refers to a time for NF service: Nnwdaf_AnalyticsSubscription / Nnwdaf_Analyticslnfo Filter: Target = GPSI(s) or SUPI(s), expected / observed / measured data volume, QoS reguirements, the target number of repeating data transmission or target time interval, target number of data__________transmission repetitions or target time interval Analytics ID: E2E data volume transfer time 4.15.13.6 completing the transmission of the expected / observed / measured data volume between UE and AF that might be indicated by the AF, e.g. the average and variance of End-to-End data volume transfer time. The transfer time is associated to Geographical distribution of the UE(s), If the reguest for geographical distribution (i.e. the Aois) of the UEs is included in the AF reguest. And the Geographical distribution of the UEfs) provided by the E2E transfer time analytics will be used to by the NEF to select candidate UEs. The data volume transfer time might be derived by considering the target number of repeating data transmission or target time interval within the Analytics target period. Table 2 (reproduction of Table 4.15.13.2-1 of 3GPP TS 23.502 [3]: Description of Member UE filtering criteria extended with new parameters indicated with bold and underline) Enhancements to the procedure for E2E data volume transfer time related member UE filtering criteria In certain examples of the present disclosure, specific procedures for E2E data volume transfer time related member UE filtering criteria may operate according to 3GPP TS 23.502 [3], clause 4.15.13.6, modified as indicated below by bold and underline. In certain examples: the E2E data volume transfer time filtering criteria may be expressed as an average and / or variance; the E2E data volume transfer time may be for a specific data volume between UE and AF; the data volume indicated by the AF may be associated with the average and / or the variance of the end-to-end data volume transfer time in this filtering criterion. 4.15.13.6.1 General An AF may invoke Nnef_MemberUESelectionAssistance_Subscribe service operation with end-to-end data volume transfer time related filtering criteria for receiving a list of UEs that fulfil the filtering criteria. In addition to the mandatory parameters, the AF may also include in the request: - End-to-end data volume transfer time filtering criteria: this may include the average end-to-end data volume transfer time for a specific data volume between UE and AF and / or the variance of the end-to-end data volume transfer time. An Area of Interest: location area of the candidate UEs. - Time windows for selecting the candidate UEs: start time and stop time. - expected data volume to be transferred between the UE and the AF or from the UE to the AF or from the AF to the UE, the roundtrip / UL / DL data volume; or the observed / measured data volume transferred between the UE and the AF or from the UE to the AF or from the AF to the UE, the roundtrip / UL / DL data volume. The Data Volume may be associated to the average and / or the variance of the end-to-end data volume transfer time in the End-to-end data volume transfer time filtering criteria. The expected, observed or measured data volume from UE to AF and / or from AF to UE of which the end-to-end data volume transfer time to be derived. - the target number of repeating data transmission or target time interval / the target number of data transmission repetitions or target time interval. - A reguest for geographical distribution of UEs. - QoS reguirements (e.g. 5QI, QoS Characteristics). 4.15.13.6.2 Member UE Selection assistance with end-to-end data volume transfer time related filtering criteria 1. AF subscribes to the member UE selection assistance functionality by invoking Nnef_MemberUESelectionAssistance_subscribe request including the Application ID, DNN / S NSSAI, Aoi, and the end-to-end data volume transfer time related filtering criteria including the average end-to-end data volume transfer time and / or the variance of the transfer time, the target number of repeating data transmission or target time interval, expected / observed / measured Data Volume UL / DL / Roundtrip, reguest for geographical distribution (i.e. the Aois) of the UEs, QoS reguirements (e.g. 5QI, QoS Characteristics), the expected, observed or measured Data Volume UL / DL between AF and UE, the target number of data transmission repetitions or target time interval, reouest for geographical distribution (i.e. the Aois) of the UEs. 2. NEF verifies the authorization of the AF Request and identifies which information needs to be collected and executed based on the end-to-end data volume transfer time related filtering criteria provided by the AF. 3. S-NSSAI / DNN are not included in the request from the AF, the NEF derives the S-NSSAI and DNN which this Application has access to. NEF discovers and selects the NWDAF(s) by invoking Nudm_UECM_Get or Nnrf_NFDiscovery_Request including Analytics ID = E2E data volume transfer time, Aoi, S-NSSAI, etc. 4. NEF sends an Analytics request / subscribe to NWDAF by invoking Nnwdaf_AnalyticsSubscription_Subscribe / Nnwdaf_Analyticslnfo_Request including Analytics ID = E2E data volume transfer time, Application ID, DNN, S-NSSAI, Aoi, and target UEs based on the initial UE list obtained from the AF, and the parameters associated to End-to-end data volume transfer time filtering criterion or E2E data volume transfer time analytics, i.e. the target number of repeating data transmission or target time interval, expected / observed / measured Data Volume UL / DL / Roundtrip, reguest for geographical distribution (i.e. the Aois) of the UEs, QoS reguirements (e.g. 5QI, QoS Characteristics), the target number of data transmission repetitions or target time interval, the expected, observed or measured Data Volume UL / DL, reguest for geographical distribution (i.e. the Aois) of the UEs, etc. 5. The NWDAF collects data from multiple sources for end-to-end data volume transfer time analytics as specified in clause 6.18.2 of TS 23.288

[50] , 6. The NWDAF provides the required output analytics to the consumer NF as specified in clause 6.18.4 of TS 23.288

[50] by means of either Nnwdaf_Analyticslnfo_Request response or Nnwdaf_AnalyticsSubscription_Notify, depending on the service used in step 4. 7. Based on the analytics received from the NWDAF, the NEF consolidates results and derives the list(s) of candidate UE(s) that fulfil the filtering criteria requested by the AF. The NEF may use the average end-to-end data volume transfer time of the specific / expected / measured / observed volumes of UL / DL data and / or the variance to derive the list(s) of candidate UEs and / or the Geographical distribution of the UEfSL that meet the requirements from the AF. 8. NEF sends a Nnef_MemberUESelectionAssistance_Notify request to the AF including the list(s) of candidate UE(s) and additional information. Corresponding explanations of the new parameters in the AF member UE selection assistance request may be defined based on the following modification of clause 6.18 of 3GPP TS 23.288 [5] to indicate to the NWDAF how to deploy / use these parameters for analytics outputs generations or reporting. 6.18 End-to-end data volume transfer time analytics 6.18.1 General Clause 6.18 describes how NWDAF can provide E2E data volume transfer time analytics, in the form of statistics or predictions or both, to a service consumer. NWDAF collects E2E data volume transfer time related input data from 5GC NFs, OAM and AF. The consumer can either subscribe to analytics notifications (i.e. a Subscribe-Notify model) or request a single notification (i.e. a Request-Response model). The E2E data volume transfer time refers to a time delay for completing the transmission of a specific data volume from UE to AF, or from AF to UE. One or more specific data volume might be indicated by the consumer, i.e. the NEF. The data volume might be the expected data volume to be transferred between the UE and the AF or from the UE to the AF or from the AF to the UE, the roundtrip / UL / DL data volume; or the observed / measured data volume transferred between the UE and the AF or from the UE to the AF or from the AF to the UE, the roundtrip / UL / DL data volume. The NWDAF may use the corresponding data volume as reference to collect the relevant input data, I.e. for the input parameters in Table 6.18.2-1 / -2 / -3, the service data might be associated to the indicated data volume or a piece of data / service / traffic has the similar data volume. For example, in Table 6.18.2-3 Transmitted UL / DL data volume are the same / similar / close to / approximate to the data volume indicated by the consumer, and the UL / DL transmission time duration are associated to the Transmitted UL / DL data volume. And the NWDAF may derive the corresponding E2E data volume transfer time for the data volume provided by the NEF. The NWDAF can derive the outputs that associated to the data volume that is the same / similar / close to! approximate to the data volume indicated for better support of AIML-based services. The data volume might be the expected data volume to be transferred between the UE and AF of an AI / ML-based service, or the observed or measured data volume transferred between the UE and AF based on consumer’s preliminary knowledge. The NWDAF uses the data volume indicated by the consumer as a reference to collect the input data for deriving the statistics and prediction of the end-to-end data volume transfer time. If a target number of repeating data transmissions or a target time interval between data transmissions is given, the E2E data volume transfer time can be provided as an average value of every data volume transfer time within the Analytics target period. The E2E data volume transfer time analytics may be used to assist an AF hosting AI / ML-based services, e.g. for member selection of federated learning. The E2E data volume transfer time analytics maybe provided as defined in clause 6.18.3 for a UE individually or a list of UEs. Reuse the information associated to other filtering criteria or procedures / combine the information of different filtering criteria Alternatively, for the data volume, QoS requirements, the request for geographical distribution (i.e. the Aois) of the UEs, the NEF may store and apply the data volume and QoS requirements of a service it receives in other procedures for the same service, or based on the information within other filtering criteria for member UE selection provided by the AF. For example, the NEF may store the data volume and / or the QoS requirements of a service receives during Setting up / update an AF session with required QoS procedure, Setting up / update Multimember AF session with required QoS, Negotiations for planned data transfer with QoS requirements etc. The NEF may link the previously received data volume and / or the QoS requirements of a service (i.e. AIML service) to the newly received Nnef_MemberUESelectionAssistance_Subscribe request. Then the NEF will include the data volume into the request or the subscription to the NWDAF to obtain the analytics, i.e. the End-to-end data volume transfer time analytics. The NEF may link the previously received data volume to the Nnef_MemberUESelectionAssistance_Subscribe request by matching the IDs associated to / indicated by the different procedure, i.e. by matching the application IDs. If the ID that links to the previously received data volume and / or QoS requirements is the same as that in the Nnef_MemberUESelectionAssistance_Subscribe, the NEF can determine that the previously received data volume and / or QoS requirement can be applied / linked to the received Nnef_MemberllESelectionAssistance_Subscribe request. And therefore, the NEF indicate the data volume and / or the QoS requirements to the NWDAF. The NWDAF may apply this data volume as a reference to collect input data and derive the output analytics, i.e. the NWDAF may collect the service data from multiple NFs that are associated to the data volume that is the same / similar to / approximate to / within an range / not exceed a threshold of the data volume received from the consumer. The NWDAF will use the QoS requirements as a one of the filters for output analytics reporting. In another example, the NEF may consider the QoS requirements of the services (AIML or Federated learning services) in QoS Member UE filtering criteria as filter information of the NWDAF analytics formember UE selection. Then the NEF informs the NWDAF of the corresponding QoS requirements to use as filter information. The NEF may use the request for geographical distribution (i.e. the Aois) of the UEs within other filtering criteria (i.e. UE current location, UE historical location, UE direction, UE separation distance, etc.) to support the E2E data volume transfer time filtering criterion. The NEF may link / combine the filtering criteria in the same AF request for member UE selection (i.e. Nnef_MemberUESelectionAssistance_subscribe), or the NEF identifies whether the different member UE selection filtering criteria link to the same service (AIML services) based on Application ID, DNN or other information in the filter information. Optimisation of the time windows for member UE selection During the member UE selection procedures, the AF may indicate the time window(s) for selecting the candidate UEs in the Nnef_MemberUESelectionAssistance_Subscribe request. The time window(s) for selecting the candidate UE(s) is used by the NEF when subscribing / requesting to NWDAF. The NEF maps the time window(s) for selecting the candidate UE(s) to the Analytics target period, which should be included in the Nnwdaf_AnalyticsSubscription_Subscribe or Nnwdaf_Analyticslnfo_Request service operations. Based on the AF request, the NEF collects and consolidate the required data to derive the UEs that fulfil the member UE selection filtering criteria. The NEF may also derive recommended time window(s) for the service operation, considering the validity period(s) of the analytics used for Member UE selection criteria. Within the recommended time window(s), the list(s) of candidate UE(s) can fulfil the Member UE filtering criteria. The recommended time window(s) are a subset of the time window(s) received from the AF. In different recommended time windows, the list of candidate UE(s) which fulfil the Member UE filtering criteria may be different. The NEF may also indicate the specific value of the parameters that NEF gathered for the Member UE filtering criteria per candidate UE and / or the number of UEs that cannot fulfil per the filtering criterion (if there are multiple filtering criteria in the AF subscribe request). Those information could be used by the NEF to revise the corresponding filtering criterion and therefore achieve the optimised service operation performance by considering the trade-off between the service quality, UE and network condition, the number of the UEs that are eligible to participate the service (i.e. AIML service), etc. The UE and / or network conditions could vary from time to time. The UE and / or network conditions might be significantly different in different time windows or at different time points, for example due to load and traffic variation of the network, the services the UE and the network are operating, the variation of the control policies etc. There are some possibilities that only a large number of UEs cannot fulfil certain filtering criterion within a small number of time windows or very a few time points. In this case, the AF may not need to lower the threshold / the requirements of the filtering criterion to make more UEs can satisfy the requirements of the service operation and lower the service quality. Instead, the AF may determine to abandon some candidate time windows for operating the service. Therefore, it will be beneficial to inform the AF of the time windows or time points within / at which there are one or more UE cannot fulfil a specific filtering criterion, or the number of the time window within which there are one or more UE cannot fulfil a specific filtering criterion. However, in the current spec, the AF is not aware of the corresponding time window(s) or time points associated to the number of UEs that cannot fulfil this filtering criterion. If the filtering criteria is related to analytics outputs, the time window might be applied. If the filtering criteria is related to some real-time or historical data, i.e. QoS measurement, UE current location, UE historical location etc., the time points might be applied. The NEF may indicate time window / time point or the number of time windows to the AF together with the existing specific value of the parameters and / or the number of the UE, i.e. via Nnef_MemberUESelectionAssistance_Notify or Nnef_MemberUESelectionAssistance_Notify service operation. Certain examples of the present disclosure may operate according to the modified versions of 3GPP TS 23.288 and / or 3GPP TS 23.502 according to the attached annexes to this description. Figure 3 is a block diagram of an exemplary network entity that may be used in examples of the present disclosure. For example, a UE / network entity (e.g. NEF, AF, NF, NWDAF) I base station (e.g. eNB, gNB) in the examples of Figures 1 and 2 may comprise an entity of Figure 3. The skilled person will appreciate that a network entity may be implemented, for example, as a network element on a dedicated hardware, as a software instance running on a dedicated hardware, and / or as a virtualised function instantiated on an appropriate platform, e.g. on a cloud infrastructure. The entity 300 comprises a processor (or controller) 301, a transmitter 303 and a receiver 305. The receiver 305 is configured for receiving one or more messages from one or more other network entities, for example as described above. The transmitter 303 is configured for transmitting one or more messages to one or more other network entities, for example as described above. The processor 301 is configured for performing one or more operations, for example according to the operations as described above. The techniques described herein may be implemented using any suitably configured apparatus and / or system. Such an apparatus and / or system may be configured to perform a method according to any aspect, embodiment, example or claim disclosed herein. Such an apparatus may comprise one or more elements, for example one or more of receivers, transmitters, transceivers, processors, controllers, modules, units, and the like, each element configured to perform one or more corresponding processes, operations and / or method steps for implementing the techniques described herein. For example, an operation / function of X may be performed by a module configured to perform X (or an X-module). The one or more elements may be implemented in the form of hardware, software, or any combination of hardware and software. It will be appreciated that examples of the present disclosure may be implemented in the form of hardware, software or any combination of hardware and software. Any such software may be stored in the form of volatile or non-volatile storage, for example a storage device like a ROM, whether erasable or rewritable or not, or in the form of memory such as, for example, RAM, memory chips, device or integrated circuits or on an optically or magnetically readable medium such as, for example, a CD, DVD, magnetic disk or magnetic tape or the like. It will be appreciated that the storage devices and storage media are embodiments of machine-readable storage that are suitable for storing a program or programs comprising instructions that, when executed, implement certain examples of the present disclosure. Accordingly, certain examples provide a program comprising code for implementing a method, apparatus or system according to any example, embodiment, aspect and / or claim disclosed herein, and / or a machine-readable storage storing such a program. Still further, such programs may be conveyed electronically via any medium, for example a communication signal carried over a wired or wireless connection. While the invention has been shown and described with reference to certain examples, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the scope of the invention, as defined by the appended claims. Abbreviations / Definitions In the present disclosure, the following acronyms / definitions may be used. 3GPP 3rd Generation Partnership Project 5G 5th Generation 5GC 5G Core 5GS 5G System 5QI 5G QoS Identifier 6G 6th Generation AF Application Function Al Artificial Intelligence Aoi Area of Interest AMF Access and Mobility Management Function DL Downlink DNN Data Network Name E2E End-to-End eNB Base Station gNB 5G Base Station GPSI General Public Subscription Identifier ID Identity / Identifier IP Internet Protocol ML Machine Learning NEF Network Exposure Function NF Network Function NR New Radio NW Network NWDAF Network Data Analytics Function OAM Operations, Administration and Maintenance PCF Policy Control Function PDTQ Planned Data Transfer with QoS PDU Protocol Data Unit QoS Quality of Service RAN Radio Access Network RAT Radio Access Technology SMF Session management Function S-NSSAI Single Network Slice Selection Assistance Information SUPI Subscription Permanent Identifier TA Tracking Area TS Technical Specification UE User Equipment UL Uplink UPF User Plane Function SA WG2 Meeting #S2-160AHE 22 - 29 January, 2024, Electronic S2-2400387 revision of S2-24xx CR-Form-v12.1 23.288 CR CHANGE REQUEST 1027 rev Current version: 18.4.0 For HELP on using this form: comprehensive instructions can be found at Proposed change affects: UICC apps| | ME| | Radio Access Network| | Core Network|~X~| Title: Clarification on E2E data volume transfer time anab / tics Source to WG: Samsung Source to TSG: SA2 Work item code: AIMLsys Date: 2024-01-12 Category: F Release: Re I-18 Use one of the following categories: Use one of the following releases: F (correction) Rel-8 (Release 8) A (mirror corresponding to a change in an earlier Rel-9 (Release 9) Rel-10 (Release 10) Rel-11 (Release 11) release) B (addition of feature), Rel-15 (Release 15) C (functional modification of feature) Rel-16 (Release 16) D (editorial modification) Rel-17 (Release 17) Detailed explanations of the above categories can Rel-18 (Release 18) be found in 3GPP TR 21.900. Reason for change: As it has been agreed in S2-2313630 in SA2 160 meeting, ‘data volume UL / DL’ and ‘A request for geographical distribution (i.e. the Aois) of the UEs' are newly introduced mandatory parameters in the request or subscription of NWDAF consumer for E2E data volume transfer time analytics. However, in the current spec, it is not clear how the NWDAF will use the data volume UL / DL to derive the statistics and prediction of the end-to-end data volume transfer time. This CR clarifies that the NWDAF derives the end-to-end data volume transfer time for the data volume UL / DL indicated by the consumer by using the data volume UL / DL as reference to collected the relevant input data. Summary of change: Clarify that the NWDAF will derive the end-to-end data volume transfer time for the data volume UL / DL indicated by the analytics consumer. Consequences if not approved: It is not clear how the NWDAF will use the data volume UL / DL to derive the outputs of end-to-end data volume transfer time analytics. Clauses affected: 6.18.1 Other specs affected: (show related CRs) Y N Other core specifications TS / TR ... CR ... Test specifications TS / TR ... CR ... O&M Specifications TS / TR ... CR ... X X X Other comments: | This CR's revision history: * * *Start of Changes * * * 6.18.1 General Clause 6.18 describes how NWDAF can provide E2E data volume transfer time analytics, in the form of statistics or predictions or both, to a service consumer. NWDAF collects E2E data volume transfer time related input data from 5GC NFs. OAM and AF. The consumer can either subscribe to analytics notifications (i.e. a Subscribe-Notify model) or request a single notification (i.e. a Request-Response model). The E2E data volume transfer time refers to a time delay for completing the transmission of a specific data volume from UE to AF, or from AF to UE. Tire data volume might be the expected data volume to be transferred between the UE and the input. data for denying the statistics and prediction of Ilie end-to-end data vohmse transfer time. If a target repetition number of data transmissions or a target time interval between data transmissions is given, the E2E data volume transfer time can be provided as an average value of the data volume transfer times within the Analytics target period. The E2E data volume transfer time analytics may be used to assist an AF or NEF with Al / ML-based services, e.g. for member UE selection of federated learning. The E2E data volume transfer time analytics may be provided as defined in clause 6.18.3 for a single UE or a list of UEs. More than one E2E data volume transfer time classes might be assigned by operator or AF to a list of UEs. The UEs might be classified into high-, medium- and low-transfer time classes with respect to the threshold(s) of the corresponding class. The service consumer may be an NF (e.g. AF, or NEF). The consumer of these analytics indicates in the request or subscription: - Analytics ID = "E2E data volume transfer time ", - Target of Analytics Reporting: a single UE (SUPI / GPSI) or a group of UEs (a list of SUPIs / GPSIs). - Analytics Filter Information, including: - Optionally, DNN; - Optionally, S-NSSA1: - Optionally. Application ID; - Optionally, Area of Interest (AOI(s)): restricts the scope of the E2E data volume transfer time analytics to the provided area; - Optionally, a list of analytics subsets that are requested (see clause 6.18.3); - QoS requirements (e.g. 5QI, QoS Characteristics); - Optionally, either a target number of repeating data transmissions or a target time interval between data transmissions within the Analytics target period; - Data Volume UL / DL: indicates a specific data volume transmitted ora;e--from UE to AF and / or from AF to UE; - A request for geographical distribution (i.e. the Aois) of the UEs. An Analytics target period indicates the time period over which the statistics or predictions are requested. - In a subscription, the Notification Correlation Id and the Notification Target Address are included. - Optionally, preferred level of accuracy of the analytics. - Optionally, preferred level of accuracy per analytics subset (see clause 6.18.3). - Optionally, preferred order of results for the list of E2E data volume transfer time: - ordering criterion: "E2E data volume transfer time", - order: ascending or descending. - Optionally, Reporting Thresholds, which apply only for subscriptions and indicate conditions on the levels to be reached for the respective analytics subsets (see clause 6.18.3) - Optionally, maximum number of UEs. * * *End of Changes SA WG2 Meeting #S2-160AHE 22 - 29 January, 2024, Electronic rews / on of S2-2400386 CR-Form-v12.1 23.502 CR CHANGE REQUEST 4682 rev Current version: 18.4.0 For HELP on using this form: comprehensive instructions can be found at http: / / www.3Qpp.oro / Change-Requests. Proposed change affects: UICC apps| | ME| | Radio Access Network! I Core Network |~x] Title: Clarification on filtering criteria of member UE selection Source to WG: Source to TSG: Samsung SA2 Work item code: AIMLsys Date: 2024-01-12 Category: F Use one of the following categories: F (correction) A (mirror corresponding to a change in an earlier release) B (addition of feature), C (functional modification of feature) D (editorial modification) Detailed explanations of the above categories can be found in 3GPP TR 21.900. Release: Re 1-18 Use one of the following releases: Rel-8 (Release 8) Rel-9 (Release 9) Rel-10 (Release 10) Rel-11 (Release 11) Rel-15 (Release 15) Rel-16 (Release 16) Rel-17 (Release 17) Rel-18 (Release 18) Q Reason for change: As it has been agreed in S2-2313630 in SA2 160 meeting, ‘data volume UL / DL’ and ‘A request for geographical distribution (i.e. the Aois) of the UEs' were newly introduced as mandatory parameters into the request or subscription ofNWDAF consumer for E2E data volume transfertime analytics in TS 23.288. However, in the current spec, the NEF is not aware of such parameters; and therefore, the NEF is not able indicate the mandatory parameters to NWDAF, as required. In order to support the mandatory parameters as it has been agreed, the AF should indicate the parameters to the AF as part of the Member UE filtering criteria information when the AF triggers the member UE selection functionality towards the NEF. The data Volume UL / DL indicated by the AF could be the expected data volume of the up-coming AIML service or the observed or measured data volume based AF knowledge, which will be used by the NWDAF to derive the data volume transfer time for the required UEs. Further alignments between TS 23.288 and TS 23.502 regarding the information in the consumer request are clarified. Summary of change: Introduce 'data volume UL / DL' and ‘request for geographical distribution (i.e. the Aois) of the UEs' into End-to-end data volume transfer time Member UE filtering criteria. Consequences if not - ‘Data Volume UL / DL’ and ‘A request for geographical distribution (i.e. approved: the Aois) of the UEs' cannot be supported as mandatory parameters indicated by the consumers to NWADF for E2E data volume transfer time analytics. Unclear spec text Clauses affected: 4.15.13.2, 4.15.13.6.1,4.15.13.6.2 Other specs affected: (show related CRs) Y N Other core specifications TS / TR ... CR ... Test specifications TS / TR ... CR ... O&M Specifications TS / TR ... CR ... X X X Other comments: | This CR's revision history: * * *Start of Changes 4.15.13.2 Member UE Filtering Criteria for 5GS assistance to Member UE selection Table 4.15.13.2-1 provides a summary7 of the Member UE filtering criteria that tire AF may request. Table 4.15.13.2-1: Description of Member UE filtering criteria Member UE filtering criteria Description of filtering criteria for member UEs selected by NEF UE filtering information Detailed description clause QoS The Quality of Service of the member UEs match or exceed the QoS of the filtering criteria NF service: Nsmf_EventExposure or Nudm_Even Exposure, Filter: target=SUPI, traffic descriptor (e.g. Application ID), DNN / S-NSSAI Event ID: QoS Monitoring 4.15.13.3 Access Type and / or RAT Type of the PDU Session Indicate the Access Type and / or RAT Type of the member UEs for the PDU Session used by the application (e.g. 3GPP / NR, Non-3GPP / WLAN, additional Access Type and RAT Type for MA PDU session) NF service: Nsmf_EventExposure Filter: a list ofGPSI(s) or SUPI(s), DNN / S-NSSAI, Event ID: Change of Access Type and / or Change of RAT Type 5.2.8.3 End-to-end data volume transfer time Indicate the target end-to-end data volume transfer time that refers to a time for completing the transmission of a specific data volume between UE and AF, e.g. the average and variance of End-to-end data volume transfer time. NF service: Nn wdaf_Analytics Sub scription / Nnwdaf_Analyticslnfo Filter: Target = GPSI(s) orSUPI(sk 4.15.13.6 Analytics ID: E2E data volume transfer time UE current location Indicate the certain area that the member UEs are currently located in. NF service: Namf_EventExposure Filter: a list ofGPSI(s) or SUPI(s) Event ID: Location Report 4.15.13.4 UE historical location Indicate the certain area that the member UEs appeared in a historical period of time NF service: Nnwdaf_AnalyticsSubscription / Nnwdaf_Analyticslnfo Filter: Visited Aoi = Target AOI, target period = historical nomadic period Analytics ID = UE mobility 4.15.13.4 UE direction Indicate the member UEs should include different moving directions NF service: Nnwdaf_AnalyticsSubscription / Nnwdaf_Analyticslnfo Filter: UE Direction Analytics ID= UE Mobility 4.15.13.4 UE separation distance (NOTE 1) Indicate the member UEs should comply with a minimum separation distance between each other NF service: Nn wdaf_Analytics S ubscription / Nnwdaf_Analyticslnfo Filter: Proximity Attributes Analytics ID: Relative Proximity 4.15.13.4 Service Experience Indicates member UEs fulfilling certain Service Experience criteria e.g., MOS value NF service: Nn wdaf_Analytics Sub scription / Nnwdaf_Analyticslnfo Filter: S-NSSAI, DNN, Application ID, DNAI, Aoi, Service Experience Contribution weight, reporting threshold (NOTE 2), Service Experience Type (NOTE 3) Analytics ID=Service Experience 4.15.13.5 DNN Indicate the DNN of the member UEs for the PDU Session used by the application NF service: Nsmf_EventExposure Filter: a list ofGPSI(s) or SUPI(s), Event ID: QFI allocation 5.2.8.3 NOTE 1: This criterion should only be applied when the number of UEs is in the range of 10's or less. NOTE 2: The Service Experience Contribution Weights signal the relative importance of each UE's Service Experience value (i.e. MOS), as defined in TS 23.288

[50] , For example, it might be that the service experience of a UE in relation to other UEs may not be as important e.g. because the data provided by such UE is not as critical to the service. NOTE 3: Indicates the type of service experience analytics, e.g. AI / ML traffic where a customized MoS apply. * * *NextChange * * * 4.15.13.6.1 General An AF may invoke NnefMemberUESelectionAssistanceSubscribe service operation with end-to-end data volume transfer time related filtering criteria for receiving a list of UEs that fulfil the filtering criteria. In addition to the mandatory parameters, the AF may also include in the request: - End-to-end data volume transfer time filtering criteria: this may include the average end-to-end data volume transfer time for a specific data volume between UE and AF and / or the variance of the end-to-end data volume transfer time. - An Area of Interest: location area of the candidate UEs. - Time windows for selecting the candidate UEs: start time and stop time. 4.15.13.6.2 Member UE Selection assistance with end-to-end data volume transfer time related filtering criteria Figure 4.15.13.6.2-1: Assistance to member UE selection for end-to-end data volume transfer time related filtering criteria 1. AF subscribes to the member UE selection assistance functionality by invoking Nnef_MemberUESelectionAssistance_subscribe request including the Application ID, DNN / S NSSAI. Aoi, and the end-to-end data volume transfer time related filtering criteria including the average end-to-end data volume transfer time and / or the variance of the transfer time, she expected, observed or 2. NEF verifies the authorization of the AF Request and identifies which information needs to be collected and executed based on the end-to-end data volume transfer time related filtering criteria provided by the AF. 3. S-NSSAI / DNN are not included in the request from the AF, the NEF derives the S-NSSAI and DNN which this Application has access to. NEF discovers and selects the NWDAF(s) by invoking NudmJJECM^Get or NnrfJsTDiscovery ^Request including Analytics ID = E2E data volume transfer time. Aoi. S-NSSAI, etc. 4. NEF sends an Analytics request / subscribe to NWDAF by invoking NnwdafAnalyticsSubscriptionSubscribe / NnwdafAnalyticsInfoRequest including Analytics ID = E2E data volume transfer time, Application ID, DNN, S-NSSAI, Aoi, and target UEs based on the initial UE list obtained from the AF, the target lumber of data ttaasmission repetitions or target tune interval. 5. The NWDAF collects data from multiple sources for end-to-end data volume transfer time analytics as specified in clause 6.18.2 of TS 23.288

[50] , 6. The NWDAF provides the required output analytics to the consumer NF as specified in clause 6.18.4 of TS 23.288

[50] by means of either Nnwdaf AnalyticsInfo Request response or Nnwdaf AnalyticsSubscription Notify, depending on tire service used in step 4. 7. Based on the analytics received from the NWDAF, the NEF consolidates results and derives the list(s) of candidate UE(s) that fulfil the filtering criteria requested by the AF. The NEF may use the average and / or the vadaxice of end-to-end data volume transfer time of specific volumes of UL / DL data radicated by the AF derive the list(s) of candidate UEs that meet the requirements from the AF. 8. NEF sends a NnefJMcmbcrUESclcctionAssistance ^Notify request to the AF including the list(s) of candidate UE(s) and additional information. * * *End of Changes

Claims

1. A method for selecting one or more user equipment (UE) for participating in an artificial intelligence / machine learning (AI / ML) operation in a wireless communication system, the wireless communication system comprising an application function (AF), a network exposure function (NEF), and a network data analytics function (NWDAF), and the method comprising:transmitting, from the AF to the NEF, a first message for requesting UE selection assistance and including member UE filtering criterion and a specific parameter depending on the member UE filtering criteria, wherein the member UE filtering criteria includes an end-to-end data volume transfer time;transmitting, from the NEF to the NWDAF, a second message including the specific parameter depending on the member UE filtering criteria and an indication of a request for end-to-end data volume transfer time analytics;transmitting, from the NWDAF to the NEF, a third message including one or more analytics associated with end-to-end data volume transfer time;deriving, by the NEF, based on the analytics and the member UE filtering criteria, a list of one or more candidate UEs that fulfil the member UE filtering criteria; andtransmitting, from the NEF to the AF, a fourth message including the list of one or more candidate UEs that fulfil the member UE filtering criteria,wherein the specific parameter depending on the member UE filtering criteria includes one or more of a data volume of UL / DL data, a target number of data transmission repetitions, a target time interval between data transmissions, and a request for geographical distribution of the UEs.

2. The method of claim 1, wherein the method further comprises:collecting, by the NWDAF, input data for end-to-end data volume transfer time analytics based on the specific parameter depending on the member UE filtering criteria.

3. The method of claim 2, wherein the NWDAF uses the data volume as a reference to collect the input data for the end-to-end data volume transfer time analytics.

4. The method of claims 2 or 3, wherein the NWDAF derives the analytics associated with end-to-end data volume transfer time based on the input data.

5. The method of claim 4, wherein the analytics associated with the end-to-end data volume transfer time are associated with a data volume similar to the data volume of the specific parameter depending on the member UE filtering criteria.

6. The method of any preceding claim, wherein the data volume is an expected or an observed data volume from a UE to an AF or from an AF to a UE of an AI / ML-based service.

7. The method of any preceding claim, wherein the geographical distributions of UEs includes areas of interest (Aoi) of the UEs.

8. The method of any preceding claim, wherein the first message includes an initial UE list, and the second message includes an indication of one or more UEs from the initial UE list that are a target of the analytics.

9. The method of claim 8, wherein the one or more UEs from the initial list are determined based on one or more of the member UE filtering criteria and the specific parameter depending on the member UE filtering criteria.

10. The method of any preceding claim, wherein the deriving includes: deriving the list of one or more candidate UEs using an average or variance of end-to-end data volume transfer time of the specific data volume of UL / DL data.

11. The method of any preceding claim, wherein the method further comprises: selecting, by the NEF, the NWDAF based on a request sent to a further network entity including an Analytics ID indicating end-to-end data volume transfer time.

12. The method of claim 11, wherein the request is a Nudm_UECM_Get message or a Nnrf_NFDiscovery_Request message.

13. The method of any preceding claim, wherein the method further comprises: verifying, by the NEF, an authorization of the request for UE selection assistance;andidentifying information that needs to be collected and executed based on the member UE filtering criteria.

14. The method of any preceding claim, wherein the first message is a Nnef_MemberUESelectionAssistance_subscribe message.

15. The method of any preceding claim, wherein the second message is an Nnwdaf_AnalyticsSubscription_Subscribe message and the third message is a Nnwdaf_AnalyticsSubscription_Notify message.

16. The method of any of claims 1 to 14, wherein the second message is a Nnwdaf_Analyticslnfo_Request message and the third message is a Nnwdaf_Analyticslnfo_Request response message.

17. The method of any preceding claim, wherein the first, second, and / or third messages includes an Analytics ID indicating the end-to-end data volume transfer time.

18. A wireless communication system comprising an application function (AF), a network exposure function (NEF), and a network data analytics function (NWDAF), wherein the wireless communication system is configured to:transmit, from the AF to the NEF, a first message for requesting UE selection assistance and including member UE filtering criteria and a specific parameter depending on the member UE filtering criteria, wherein the member UE filtering criteria includes an end-to-end data volume transfer time;transmit, from the NEF to the NWDAF, a second message including the specific parameter depending on the member UE filtering criteria and an indication of a request for end-to-end data volume transfer time analytics;transmit, from the NWDAF to the NEF, a third message including one or more analytics associated with end-to-end data volume transfer time;derive, by the NEF, based on the analytics and the member UE filtering criteria, a list of one or more candidate UEs that fulfil the member UE filtering criteria; andtransmit, from the NEF to the AF, a fourth message including the list of one or more candidate UEs that fulfil the member UE filtering criteria,wherein the specific parameter depending on the member UE filtering criteria includes one or more of a data volume of UUDL data, a target number of data transmission repetitions, a target time interval between data transmissions, and a request for geographical distribution of the UEs.

19. The method of claim 18, wherein the method further comprises:collecting, by the NWDAF, input data for end-to-end data volume transfer time analytics based on the specific parameter depending on the member UE filtering criteria.

20. The method of claim 19, wherein the NWDAF uses the data volume as a reference to collect the input data for the end-to-end data volume transfer time analytics.

21. The method of claims 19 or 20, wherein the NWDAF derives the analytics associated with end-to-end data volume transfer time based on the input data.

22. The method of claim 21, wherein the analytics associated with the end-to-end data volume transfer time are associated to a data volume similar to the data volume of the specific parameter depending on the member UE filtering criteria.

23. The method of any of claims 19 to 22, wherein the data volume is an expected or an observed data volume from a UE to an AF or from an AF to a UE of an AI / ML-based service.

24. The method of any of claims 19 to 23, wherein the geographical distributions of UEs includes areas of interest (Aoi) of the UEs.

25. The wireless communication system of claim 18, wherein the wireless communication system is configured to implement the method of any of claims 8 to 17.

Images