Method and apparatus for support of AI / ML-based positioning

Abstract

According to various examples of the present disclosure, there is provided a location management function (LMF) entity configured to: subscribe to or request artificial intelligence / machine learning (AI / ML) -related services from a network data analytics function (NWDAF) entity in relation to an AI / ML model for determining positioning of a user equipment (UE); and receive, from the NWDAF entity, an indication that training has been performed for the AI / ML model. According to various examples of the present disclosure, there is provided a network data analytics function (NWDAF) entity configured to: receive, from a location management function (LMF) entity, a subscription to or request for artificial intelligence / machine learning (AI / ML) -related services in relation to an AI / ML model for determining positioning of a UE; obtain data for training the AI / ML model from at least one other entity; train the AI / ML model based on the obtained data; and transmit, to the LMF entity, an indication that training has been performed for the AI / ML model; wherein the NWDAF entity includes a model training logical function (MTLF).

Description

BACKGROUND Field

[0001] Certain examples of the present disclosure relate to methods, apparatus and / or systems for training AI / ML models, performing inference with trained AI / ML models, and / or collecting data in relation to AI / ML models. In an example, there is provided a training method for AI / ML direct positioning with LMF-side models. In another example, there is provided an inference method for AI / ML direct positioning with LMF-side models. In yet another example, there is provided a data collection method for AI / ML direct positioning with LMF-side models. Various examples include combinations of the methods disclosed herein, such as use of the data collection method in the training method or the inference method, or performing the inference method based on one or more models obtained based on the training method. Description of Related Art

[0002] The content of the following documents is referred to below and / or their content provides background information that the following disclosure should be considered in the context of: [1] 3GPP TS 22.261 - Service requirements for the 5G system, SA1, Release 18 (e.g., V18.12.0), Release 19 (e.g., V19.5.0). [2] SP-231800, Study on Core Network Enhanced Support for Artificial Intelligence (Al) / Machine Learning (ML), 3GPP TSG SA Meeting #102, 11-15 December 2023, Edinburgh, UK. [3] 3GPP TS 23.273 - 5G System (5GS) Location Services (LCS); Stage 2, Release 18 (e.g. V18.4.0). [4] 3GPP TR 38.843 - Study on Artificial Intelligence (Al) / Machine Learning (ML) for NR air interface, Release 18 (e.g. V18.0.0). [5] 3GPP TR 23.700-84 - Study on Core Network Enhanced Support for Artificial Intelligence (Al) I Machine Learning (ML), Release 19 (e.g. V0.1.0). [6] 3GPP TS 23.288 - Architecture enhancements for 5G System (5GS) to support network data analytics services, Release 18 (e.g. V18.4.0). Note: indicated version numbers are provided for illustrative purposes, other (including future) versions of these documents are considered also.

[0003] Wireless or mobile (cellular) communications networks in which a mobile terminal (e.g., user equipment (UE), such as a mobile handset) communicates via a radio link with a network of base stations, or other wireless access points or nodes, have undergone rapid development through a number of generations. The 3rd Generation Partnership Project (3GPP) design, specify and standardise technologies for mobile wireless communication networks. Fourth Generation (4G) and Fifth Generation (5G) systems (5GS) are now widely deployed, while beyond 5G (B5G) and 6G systems are being considered.

[0004] 3GPP standards for 4G systems include an Evolved Packet Core (EPC) and an Enhanced-UTRAN (E-UTRAN: an Enhanced Universal Terrestrial Radio Access Network). The E-UTRAN uses Long Term Evolution (LTE) radio technology. LTE is commonly used to refer to the whole system including both the EPC and the E-UTRAN, and LTE is used in this sense in the remainder of this document. LTE should also be taken to include LTE enhancements such as LTE Advanced and LTE Pro, which offer enhanced data rates compared to LTE.

[0005] In 5G systems a new air interface has been developed, which may be referred to as 5G New Radio (5G NR) or simply NR. NR is designed to support the wide variety of services and use case scenarios envisaged for 5G networks, though builds upon established LTE technologies B5G systems, such as 6G, are currently being considered and developed, and are expected to at least partly build on 5G systems.

[0006] New frameworks and architectures are being developed as part of 5G network (and beyond, such as 6G networks) in order to increase the range of functionality and use cases available through 5G networks. One such new framework is the use of artificial intelligence I machine learning (AI / ML), which may be used for the optimisation of the operation of 5G networks.

[0007] In AI / ML operation, AI / ML models and / or data might be transferred across the AI / ML applications (e.g., application functions (AFs)), 5GC (5G core), UEs (user equipments) etc.). Without limitation, the AI / ML works could be divided into two main phases: model training and inference. During model training and inference, multiple rounds of interaction may be required.

[0008] In Section 6.40 (‘AI / ML model transfer in 5GS’) in TS 22.261 [1], three types of AI / ML operations to be supported in (at least) Release 18 and / or Release 19 are described as follows: a) AI / ML operation splitting between AI / ML endpoints The AI / ML operation / 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 privacy-sensitive 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. b) 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. c) 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.

[0009] In general, the AI / ML works can be divided into three main phases: model training, model transfer and inference. More specifically, with the introduction of federated learning, model transfer has become a crucial phase to successfully perform some AI / ML operations. Time spent for model training, inference and transmission of the AI / ML models and for output of the inference depend on computation and / or communication capabilities of participating nodes / components; hence, the time varies among different nodes / components.

[0010] 3GPP RAN Study and Work Items on AI / ML for Air Interface - AI / ML Positioning Use Case

[0011] 3GPP RAN (Radio Access Network) WGs (Working Groups) started the study and work item on AI / ML for Air Interface based on the SID (Study Item Description) approved in RP-213599. Based on the objectives of SID, RAN WGs focused on following use cases during the study of AI / ML for Air Interface: • CSI feedback enhancement. • Beam management. • Positioning accuracy enhancements for different scenarios including, e.g., those with heavy NLOS conditions.

[0012] Based on the above agreed use cases, RAN WGs evaluated the performance benefits of AI / ML based algorithms (e.g. for positioning, the statistical model(s) is / are based on that in TR 38.857 [positioning]), determined the common KPIs (Key Performance Indicators) and corresponding requirements for the AI / ML operations (e.g. the KPIs include the performance, inference latency and computational complexity of AI / ML based algorithms, and the overhead, power consumption, memory storage, etc.), evaluated the spec impacts, etc.

[0013] The outcome of the study on the AI / ML for Air Interface is documented in TR 38.843 [4],

[0014] For the Positioning accuracy enhancements objectives, the use cases were documented in clause 5 of TR 38.843 [4], The following are selected as representative sub-use cases: • Direct AI / ML positioning: AI / ML model output: UE location. • AI / ML assisted positioning: AI / ML model output: new measurement and / or enhancement of existing measurement.

[0015] More specifically, the following Cases are considered for the study: • Case 1: UE-based positioning with UE-side model, direct AI / ML or AI / ML assisted positioning. • Case 2a: UE-assisted / LMF-based positioning with UE-side model, AI / ML assisted positioning. • Case 2b: UE-assisted / LMF-based positioning with LMF (Location Management Function)-side model, direct AI / ML positioning. • Case 3a: NG-RAN node assisted positioning with gNB-side model, AI / ML assisted positioning. • Case 3b: NG-RAN node assisted positioning with LMF-side model, direct AI / ML positioning.

[0016] For positioning enhancement use cases: • For model training, training data can be generated by UE / PRU (Positioning Reference Unit) / gNB (Next generation Node B) / LMF. • For LMF-side model inference (Case 2b, Case 3b), input data can be generated by UE / gNB and terminated at LMF. • For gNB-side model inference (Case 3a), input data is internally available at gNB. • For UE-side model inference (Case 1, Case 2a), input data is internally available at UE. • For performance monitoring at the LMF side, calculated performance metrics (if needed) or data needed for performance metric calculation (if needed) can be generated by UE / gNB and terminated at LMF. • For performance monitoring at the gNB side, calculated performance metrics (if needed) or data needed for performance metric calculation (if needed) can be generated by at least gNB.

[0017] The conclusions of the study on AI / ML for Air Interface were documented in clause 8 of TR 38.843 [4], For positioning accuracy enhancements, the conclusions include: Direct AI / ML positioning and AI / ML assisted positioning were identified and selected as the representative sub-use cases. Evaluation results have shown that in considered evaluation scenarios (i.e., InF-DH, and other I nF scenarios), both direct AI / ML positioning and AI / ML assisted can significantly improve the positioning accuracy compared to existing RAT-dependent positioning methods. Various aspects of AI / ML for positioning accuracy enhancement were investigated and evaluated as described in clause 6.4 that provides summary of evaluation results from different sources. Based on the conducted analysis, it is recommended to proceed with normative work for AI / ML based positioning. The necessity, feasibility and potential enhancements to facilitate the support of AI / ML for positioning accuracy enhancements with NR RAT-dependent positioning methods were studied and the outcome are outlined in clause 7. It is recommended to specify necessary measurement, signalling and procedure to facilitate training, inference, monitoring and / or other LCM operations for both direct AI / ML positioning and AI / ML assisted positioning, specifically: specify necessary signalling of data collection; investigate the necessity of other information for supporting data collection, and if needed, specify during normative work investigate on the necessity and signalling details of measurement enhancements, and if needed, specify during normative work investigate on the necessity and signalling details of monitoring method(s), and if needed, specify during normative work A variety of enhancements for measurements (e.g., based on extensions to current positioning measurements or with new measurements) were also identified as potentially beneficial (e.g., trade-off positioning accuracy requirement and signalling overhead) and are recommended to be investigated further and if needed, specified during normative work.

[0018] Based on the outcome of the study work, RAN WGs agreed the Rel-19 work item on Artificial Intelligence (Al) / Machine Learning (ML) for NR Air Interface during RAN 102 meeting (Dec 2023). It was considered that the RAN Rel-19 work item may require coordination with SA / SA WGs of the ongoing study / work, as it may relate to SA / SA WGs required work.

[0019] The above Rei-18 study work on AI / ML for NR Air Interface carried out by RAN WGs may also have SA / SA2 impacts, e.g. SA2 work may require to support AI / ML based Positioning considering the above conclusions.

[0020] 3GPPSA2 Rel-19 Study in AIML (FS AIML CN)

[0021] New SID on Core Network Enhanced Support for Artificial Intelligence (Al) / Machine Learning (ML) was approved in SP-231800 [2] in TSG SA Meeting #102 (Dec 2023). In WT#1.4 of the Study Item Description (SID): WT1.4: Study whether and how to consider enhancements to LCS to support AI / ML based Positioning considering the conclusions in 3GPP TR 38.843. NOTE 3: UE data collection, model delivery and transfer to the UE and model identification / management are not within the scope of WT#1.4

[0022] Based on the conclusion of Rel-18 RAN WGs study item on AI / ML for NR Air Interface, SA2 evaluated the potential SA2 impacts of the RAN work on WT#1.4 (e.g. see S2-2401829), with some observations provided as follows. Model inference for AI / ML positioning could be at LMF side, and input data (e.g. SRS or PRS measurements, etc.) for model inference from UE and RAN needs to be collected by the LMF; Model training for AI / ML positioning should be at CN side, and data (e.g. measurements and UE location result, etc.) for model training needs to be collected by the CN node. Model performance monitoring for AI / ML positioning could be at LMF side, and data collection (e.g. ground truth data, input data for AI / ML positioning prediction, etc.) may be needed.

[0023] After discussion, in WG SA2 Meeting #160-Ad Hoc-e (Jan 2024), it was concluded that SA2 would study the problem documented in WT1.4 of SP-231800 [2] with the agreement to focus on Case 2b and Case 3b described above, since those were the cases with clear impact on the 5GC. SA2 would work to address the above issues as required during Rel-19 study and normative phases.

[0024] Based on the agreed Key Issue (KI) description of Kl#1: Enhancements to LCS to support Direct AI / ML based Positioning which is documented in clause 5.2.1 of TR 23.700-84 [5], SA2 need to solve the following questions during Rel-19 Study phase to fully support Core Network Enhanced Support for Artificial Intelligence (Al) / Machine Learning (ML): This key issue aims to provide solutions for whether and how to consider enhancements to support AI / ML based Positioning for Cases 2b, 3b as defined in TR 38.843, which will investigate the following aspects: Study whether and how an AI / ML model for Direct AI / ML positioning (i.e. case 2b / 3b) is handled: Which entity trains the model for Direct AI / ML positioning and if the entity that train the model and the consumer are different, how the Model consumer gets the trained AI / ML model; How the Model consumer uses the trained model to perform inference and / or derive UE position; Define procedures for data collection with objective to train AI / ML models for Direct AI / ML positioning; Whether and how to support Direct AI / ML positioning with additional 5GC enhancements. How to monitor model performance for ML models used for Direct AI / ML based positioning. NOTE 1: UE data collection, model delivery and transfer to the UE and model identification / management are not within the scope of this key issue. NOTE 2: What data to be collected for the model training / model inference / model performance monitoring for LMF-sided model needs to be coordinated with RAN WG. NOTE 3: Any potential impacts for case 1 / 2a / 3a in TR 38.843, are out of the scope and any potential alignment work will be based on the possible requirements defined by RAN WGs considering the conclusions in 3GPP TR 38.843.

[0025] However, based on the current 3GPP SA2 architecture and framework, there is a lack of solutions for AI / ML direct positioning support at the 5GC in existing SA2 specifications. SUMMARY

[0026] 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.

[0027] Various examples of the present disclosure are set out in the claims.

[0028] According to an aspect of the present disclosure, there is provided a location management function (LMF) entity configured to: subscribe to or request artificial intelligence / machine learning (AI / ML) -related services from a network data analytics function (NWDAF) entity in relation to an AI / ML model for determining positioning of a user equipment (UE); and receive, from the NWDAF entity, an indication that training has been performed for the AI / ML model.

[0029] According to various examples, the LMF entity is further configured to: receive the AI / ML model from the NWDAF entity.

[0030] According to various examples, the indication is included in a notification from the NWDAF entity.

[0031] According to various examples, the notification notifies that the AI / ML model has been trained according to a request of the LMF entity.

[0032] According to various examples, if the AI / ML-related services are subscribed to or requested by using a specific service operation, the notification is for a corresponding notification service operation.

[0033] According to various examples, the specific service operation is Nnwdaf_MLModelProvision subscribe service operation, and the corresponding notification service operation is Nnwdaf_MLModelProvision notify service operation.

[0034] According to various examples, the LMF entity is further configured to transmit an analytics ID indicating the AI / ML model to the NWDAF entity.

[0035] According to various examples, the LMF entity is further configured to obtain measurements from the UE and / or next generation radio access node (NG-RAN).

[0036] According to various examples, the obtained measurements from the UE include positioning reference signal (PRS), and / or the obtained measurements from the NG-RAN include sounding reference signal (SRS).

[0037] According to various examples, the LMF entity is further configured to determine a location of the UE based on the AI / ML model and at least one of the obtained measurements and data collected from at least one other entity.

[0038] According to various examples, the location of the UE is estimated based on performing inference using: the AI / ML model and at least one of the obtained measurements and collected data.

[0039] According to various examples, the LMF entity is further configured to train the AI / ML model using the obtained measurements.

[0040] According to various examples, the LMF entity is included in an apparatus comprising at least one processor, a receiver and a transmitter.

[0041] According to another aspect of the present disclosure, there is provided a network data analytics function (NWDAF) entity configured to: receive, from a location management function (LMF) entity, a subscription to or request for artificial intelligence / machine learning (AI / ML) -related services in relation to an AI / ML model for determining positioning of a UE; obtain data for training the AI / ML model from at least one other entity; train the AI / ML model based on the obtained data; and transmit, to the LMF entity, an indication that training has been performed for the AI / ML model; wherein the NWDAF entity includes a model training logical function (MTLF).

[0042] According to various examples, the NWDAF entity is further configured to transmit the trained AI / ML model to the LMF entity.

[0043] According to various examples, the indication is included in a notification transmitted to the LMF entity.

[0044] According to various examples, the notification notifies that the AI / ML model has been trained according to a request of the LMF entity.

[0045] According to various examples, if the AI / ML-related services are subscribed to or requested by using a specific service operation, the notification is for a corresponding notification service operation.

[0046] According to various examples, the specific service operation is Nnwdaf_MLModel Provision subscribe service operation, and the corresponding notification service operation is Nnwdaf_MLModelProvision notify service operation.

[0047] According to various examples, the NWDAF entity is further configured to receive an analytics ID indicating the AI / ML model from the LMF entity.

[0048] According to various examples, the data is obtained from at least one of the LMF entity, application management function (AMF), NG-RAN, or gateway mobile location centre (GMLC).

[0049] According to various examples, the obtained data includes location services (LCS) quality of service (QoS).

[0050] According to various examples, the LMF entity is co-located with the NWDAF entity.

[0051] According to another aspect of the present disclosure, there is provided a method fora location management function (LMF) entity, the method comprising: subscribing to or requesting artificial intelligence / machine learning (AI / ML) -related services from a network data analytics function (NWDAF) entity in relation to an AI / ML model for determining positioning of a user equipment (UE); and receiving, from the NWDAF entity, an indication that training has been performed for the AI / ML model.

[0052] According to various examples, the method is modified to be in accordance with any one or more of the examples relating to the LMF entity given above.

[0053] According to another aspect of the present disclosure, there is provided a method for a network data analytics function (NWDAF) entity, the method comprising: receiving, from a location management function (LMF) entity, a subscription to or request for artificial intelligence / machine learning (AI / ML) -related services in relation to an AI / ML model for determining positioning of a UE; obtaining data for training the AI / ML model from at least one other entity; training the AI / ML model based on the obtained data; and transmitting, to the LMF entity, an indication that training has been performed for the AI / ML model; wherein the NWDAF entity includes a model training logical function (MTLF).

[0054] According to various examples, the method is modified to be in accordance with any one or more of the examples relating to the NWDAF entity given above.

[0055] According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium configured to store instructions which, when executed by at least one processor or a computer, cause the at least one processor or the computer to perform (or assist in performing) a method according to any one or more of the aspects and / or examples given above.

[0056] Other aspects, advantages, and salient features of the disclosure 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

[0057] Embodiments / examples of the present disclosure are further described hereinafter with reference to the accompanying drawings, in which: Figure 1 is a call flow illustrating a training procedure (i.e. a method) for AI / ML direct positioning with LMF-side models according to various examples of the present disclosure. Figure 2 is a call flow illustrating an interference procedure (i.e. a method) for AI / ML direct positioning with LMF-side models according to various examples of the present disclosure. Figure 3 is a call flow illustrating a data collection procedure (i.e. a method) for AI / ML direct positioning with LMF-side models according to various examples of the present disclosure. Figure 4 is a block diagram illustrating an example structure of a network entity in accordance with certain examples of the present disclosure. Figure 5 is Figure 6.1.2-1 from TS 23.273 [3] - “5GC-MT-LR Procedure for the commercial location services”. Figure 6 is a method flow diagram illustrating a method according to various examples of the present disclosure. Figure 7 is a method flow diagram illustrating a method according to various examples of the present disclosure. DETAILED DESCRIPTION

[0058] The following description of examples of the present disclosure, with reference to the accompanying drawings, is provided to assist in a comprehensive understanding of certain examples of the present disclosure. 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 or disclosure.

[0059] The same or similar components may be designated by the same or similar reference numerals, although they may be illustrated in different drawings.

[0060] Detailed descriptions of techniques, structures, constructions, functions or processes known in the art may be omitted for clarity and conciseness, and to avoid obscuring the subject matter of the present disclosure.

[0061] 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 disclosure.

[0062] Throughout the description 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.

[0063] Throughout the description 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.

[0064] Throughout the description, the expression “at least one of A, B and / or C” (or the like), the expression “and / or”, and the expression “one or more of A, B and / or C” (or the like) should be seen to separately include all possible combinations, for example: A, B, C, A and B, A and C, A and B and C.

[0065] Throughout the description 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.

[0066] Features, elements, components, integers, steps, processes, operations, functions, characteristics, properties and / or groups thereof described or disclosed in conjunction with a particular aspect, embodiment or example are to be understood to be applicable to any other aspect, embodiment or example described herein unless incompatible therewith.

[0067] Certain examples of the present disclosure relate to methods, apparatus and / or systems for training AI / ML models, performing inference with trained AI / ML models, and / or collecting data in relation to AI / ML models. In an example, there is provided a training method for AI / ML direct positioning with LMF-side models. In another example, there is provided an inference method for AI / ML direct positioning with LMF-side models. In yet another example, there is provided a data collection method for AI / ML direct positioning with LMF-side models. Various examples include combinations of the methods disclosed herein, such as use of the data collection method in the training method or the inference method, or performing the inference method based on one or more models obtained based on the training method.

[0068] The following examples are applicable to, and use terminology associated with, 3GPP 5G. However, the skilled person will appreciate that the techniques disclosed herein are not limited to these examples or to 3GPP 5G, 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. The skilled person will appreciate that the techniques disclosed herein may be applied in any existing or future releases of 3GPP 5G NR or any other relevant standard. For example, the functionality of the various network entities and other features disclosed herein may be applied to corresponding or equivalent entities or features in other 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, operation or purpose within the network. In particular, the following disclosure should be considered at least in relation to 6G also, which is expected to use at least part of the 5G architecture, or equivalent, and to which the present disclosure also relates.

[0069] 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.

[0070] The skilled person will appreciate that the present disclosure is not limited to the specific examples disclosed herein. For example: • The techniques disclosed herein are not limited to 3GPP 5G, B5G or 6G. • 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, entities and / or messages may be added to the examples disclosed herein. • One or more non-essential elements, entities and / or messages 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 may be modified, if possible, in alternative examples. • The transmission of information between network entities is not limited to the specific form, type and / or order of messages described in relation to the examples disclosed herein.

[0071] 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. Such an apparatus / device / network entity 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). Certain examples of the present disclosure may be provided in the form of a system (e.g., a network) comprising one or more such apparatuses / devices / network entities, and / or a method therefor.

[0072] It will be appreciated that examples of the present disclosure may be realized in the form of hardware, software or a combination of hardware and software. Certain examples of the present disclosure may provide a computer program comprising instructions or code which, when executed, implement a method, system and / or apparatus in accordance with any aspect, example and / or embodiment disclosed herein. Certain embodiments of the present disclosure provide a machine-readable storage storing such a program.

[0073] A network according to one or more of the examples disclosed herein may include one or more of a Network Data Analytics Function (NWDAF) entity, an Access and Mobility Management Function (AMF) entity, a Session Management Function (SMF) entity, a Network Slice Selection Function (NSSF) entity, a Network Repository Function (NRF) entity, Application Function (AF) entity, and an Operation and Maintenance (OAM) entity. The network may include one or more Service Consumers (including one or more of the entities mentioned above and / or one or more other entities) that receive analytics from NWDAF. The skilled person will appreciate that a network may omit one or more of the entities mentioned above and / or may comprise one or more additional entities

[0074] As described above, based on the current 3GPP SA2 architecture and framework, there is a lack of solutions for AI / ML direct positioning support at the 5GC in existing SA2 specifications. Therefore, new solutions are required to solve, or at least mitigate, the above issues during SA2 Rel-19 study and normative phase to support Rel-19 Core Network Enhanced Support for Artificial Intelligence (AI)ZMachine Learning (ML), e.g. AI / ML direct positioning support at the 5GC. Accordingly, the present disclosure provides various examples which are intended to solve, address or mitigate one or more of said issues.

[0075] To this end, various examples of the present disclosure include procedures (i.e. methods) to support training, inference and data collection aspects of AI / ML Direct Positioning with LMF-side models as defined above.

[0076] It should be noted that, herein, the LMF may be a standalone network function (NF) or the LMF may be co-located or combined with another 5GC NF. For example, the LMF is co-located with a Network Data Analytics Function (NWDAF) containing Analytical Logical Function (AnLF).

[0077] Various examples of the present disclosure are now described in combination with Figure 1.

[0078] Figure 1 illustrates (via a call flow diagram) a training procedure for AI / ML direct positioning with LMF-side models. More generally, Figure 1 may be regarded as a method for training an AI / ML direct positioning model, where the method is performed by one or more entities in a network.

[0079] For the purposes of this description, the entities in Figure 1 may be labelled as follows: • 100 - NG-RAN (Next Generation Radio Access Network). • 200 - AMF (Access and Mobility Management Function). • 300 - LMF (Location Management Function). • 400 - NWDAF containing MTLF (Model Training Logical Function). • 500 - GMLC (Gateway Mobile Location Centre). • 600 - OAM (Operations and Management). • 700 - AF (Application Function).

[0080] However, it will be appreciated that, in more general examples, reference could instead be made to a first entity 100, a second entity 200, a third entity 300, a fourth entity 400, a fifth entity 500, a sixth entity 600 and a seventh entity 700; where each entity is arranged to perform the associated operation(s) indicated below. Furthermore, any one or more of entities in Figure 1 may be combined or co-located. For example, as described above, the LMF and the NWDAF can be colocated, in which case an operation which indicates interaction between these two entities is performed internally (for example, not necessitating use of a transmitter / recei ver).

[0081] It will also be understood that various examples relate to individual entities shown in Figure 1. For instance; various examples are directed to the LMF, in which case such examples can focus on operations in which the LMF is involved and omit any other operations; while some other examples are directed to the NWDAF, in which case such examples can focus on operations in which the NWDAF is involved and omit any other operations. Additionally, yet further examples relate to any combination of the individual entities, and so may include the operations performed by these entities while omitting operations performed by other entities not included in the combination.

[0082] In operation S110, LMF 300 subscribes to training services from NWDAF 400 containing MTLF for an AI / ML model to be used for direct positioning. This subscription may be carried out with a new service operation or by leveraging existing services, such as Nnwdaf_MLModelProvision, Nnwdaf_MLModellnfo or Nnwdaf_MLModelTraining, using the Subscribe or Request service operation. A new analytics ID may be used to identify the model for AI / ML direct positioning. A number of optional parameters as defined in clauses 7.5.2, 7.6.2 and 7.10.2 of TS 23.288 [6] may be provided as service operation inputs. In an example, LMF 300 transmits a request relating to training to NWDAF 400, such as a subscription request.

[0083] Although not shown in Figure 1, the training subscription may be triggered by LMF 300 receiving a previous request from another network entity. That is, operation S110 may be performed in response to a trigger or based on a condition such as the LMF 300 receiving a request from another entity (not shown in Figure 1), where this request causes the LMF 300 to subscribe or request training services from the NWDAF 400. The trigger (e.g. the request from the other entity) itself is not depicted in Figure 1 may not be considered part of the training procedure illustrated in Fig. 1; that is, an operation of receiving or detecting the trigger, or the trigger occurring, is optional.

[0084] In operation S120, data collection is performed by NWDAF 400 containing MTLF to train the model(s) for direct positioning. More details on data collection subprocedure can be found in Figure 3, described below.

[0085] For example, the NWDAF 400 collects data (such as metrics) from one or more of the other entities shown in Figure 1 or one or more other entities, such as other NFs or parts of the network. The NWDAF 400 may identify entities from which suitable or relevant data may be obtained and / or may identify the type of data required for the training, such as data which relates to positioning services. Based on this, the NWDAF 400 may communicate with the other entity / entities to acquire the data.

[0086] In operation S130, NWDAF 400 containing MTLF trains the AI / ML model(s).

[0087] In operation S140, NWDAF 400 containing MTLF notifies LMF 300 that the requested AI / ML model for direct positioning has been trained. This notifying may be performed using the Notification service operation of the same service used in step 1; for example, if the LMF 300 used NnwdaLMLModelProvision, NnwdaLMLModellnfo or NnwdaCMLModelTraining using the Subscribe or Request service operation, the NWDAF 400 uses a corresponding Notification service operation.

[0088] As mentioned above, any entity in Figure 1 may be co-located, or have its functionality combined, with one or more of the other entities in Figure 1. Accordingly, in a case where LMF 300 and NWDAF 400 are co-located (or combined), operations S110, S130 and S140 are essentially performed by the same entity. In other words, in such a case it is not necessary to describe communications between different entities for facilitating operations S110 and S140.

[0089] In an example, the operations of Figure 1 can be labelled as follows: • S110 - 1. Training subscription. • S120 - 2. Data collection sub-procedures for training of AI / ML direct positioning model(s). • S130 - 3. Training of AI / ML model. • S140 - 4. Training notification

[0090] Various examples of the present disclosure are now described in combination with Figure 2.

[0091] Figure 2 illustrates (via a call flow diagram) an inference procedure for AI / ML direct positioning with LMF-side models. More generally, Figure 2 may be regarded as a method for performing inference with an AI / ML direct positioning model, where the method is performed by one or more entities in a network.

[0092] For the purposes of this description, the entities in Figure 2 may be labelled as follows: • 100 - NG-RAN (Next Generation Radio Access Network). • 200 - AMF (Access and Mobility Management Function). • 300 - LMF (Location Management Function). • 400 - NWDAF containing MTLF (Model Training Logical Function). • 500 - GMLC (Gateway Mobile Location Centre). • 600 - OAM (Operations and Management). • 700 -AF (Application Function). • 800 - UE (User Equipment). • 900 - UDM (Unified Data Manager).

[0093] However, it will be appreciated that, in more general examples, reference could instead be made to a first entity 100, a second entity 200, a third entity 300, a fourth entity 400, a fifth entity 500, a sixth entity 600, a seventh entity 700, an eighth entity 800 and a ninth entity 900; where each entity is arranged to perform the associated operation(s) indicated below. Furthermore, any one or more of entities in Figure 2 may be combined or co-located. For example, as described above, the LMF and the NWDAF can be co-located, in which case an operation which indicates interaction between these two entities is performed internally (for example, not necessitating use of a transmitter / receiver). It will be appreciated that an entity shown in Figure 2 may correspond to the same entity shown in Figure 1 (e.g. the LMF 300 of Figure 2 may be the same as the LMF 300 in Figure 1, as appropriate).

[0094] It will also be understood that various examples relate to individual entities shown in Figure 2. For instance; various examples are directed to the LMF, in which case such examples can focus on operations in which the LMF is involved and omit any other operations; while some other examples are directed to the NWDAF, in which case such examples can focus on operations in which the NWDAF is involved and omit any other operations. Additionally, yet further examples relate to any combination of the individual entities, and so may include the operations performed by these entities while omitting operations performed by other entities not included in the combination.

[0095] The example of Figure 2 assumes that a trained AI / ML model (or models) is available for direct positioning. Such a model / models may be provided through performing a method according to Figure 1, such as described above.

[0096] In operation S210, steps 1 to 11 in clause 6.1.2 of TS 23.273 [3] are followed. In various examples, these steps, or the detail of these steps, is omitted from the procedure of Figure 2. In other words, while it may be useful to understand that these steps of §6.1.2 of TS 23.273 [3] have been performed as an initial operation, it is not necessary to include these steps 1 to 11 (or a subset thereof) in a method according to the present disclosure. For reference, steps 1 to 11 of TS 23.273 [3] are provided in an Annex below, and corresponding Figure 6.1.2-1 of TS 23.273 [3] is included herein as Figure 5.

[0097] It should be noted that, in examples according to the method of Figure 2, it is assumed that the location services consumer does not need to be aware that AI / ML direct positioning is used by the network to estimate the UE location.

[0098] In operation S220, AI / ML Direct Positioning preparations are performed to support Case 2b and Case 3b as defined in TR 38.843 [4], The detail of this step may be understood with reference to TR 38.384 [4] and / or to related studies for such preparations. It is noted that the AI / ML direct positioning preparations need not be performed by the entities shown in Figure 2 (i.e., UE 800, NG-RAN 100, AMF 200 and LMF 300), but this is merely exemplary based on one envisaged way of making the preparations. In general, operation S220 can be omitted from a method according to Figure 2 because the detail of this operation is not necessary for implementing the procedure.

[0099] In optional operation S230, which relates to Case 2b, the UE 800 provides measurements to the LMF 300 (e.g. positioning reference signal, PRS). For example, performing operation S230 (i.e. including it in the method) is conditional on the method relating to case 2b.

[00100] In optional operation S240, which relates to case 3b, the NG-RAN 100 node provides measurements to the LMF 300 (e.g. sounding reference signal, SRS). For example, performing operation S240 (i.e. including it in the method) is conditional on the method relating to case 3b.

[00101] In operation S250, data collection is performed by a number of the entities in Figure 2 (in the illustrated example, these are one or more of NG-RAN 100, AMF 200, LMF 300, NWDAF 400, GMLC 500, UDM 900, OAM 600, AF 700). Data collection is required to perform inference for AI / ML Direct Positioning (i.e. using the model(s)) and to estimate the UE location. For example, LMF 300 performs data collection to obtain data (e.g. data suitable for performing inference or for estimating UE location) from, or in cooperation with, one or more of the other entities shown in Figure 2 (or from another entity / entities in the network). Details on the data collection procedure (sub-procedure) can be found in Figure 3, described below.

[00102] In operation S260, AI / ML Direct Positioning inference is performed at LMF 300. For example, LMF 300 acquires the collected data, or a result or portion thereof, following operation S250, and in operation S260 uses this for performing inference. In various examples, as part of or based on performing inference, the location of the UE is estimated by the LMF 300.

[00103] In operation S270, Steps 13 to 24 in cl. 6.1.2 of TS 23.273 [3] are followed to deliver the estimated UE location to the consumer. In various examples, these steps, or the detail of these steps, are omitted from the procedure of Figure 2. In other words, while it may be useful to understand that these steps of §6.1.2 of TS 23.273 [3] have been performed as an initial operation, it is not necessary to include these steps 13 to 24 (or a subset thereof) in a method according to the present disclosure. For reference, steps 13 to 24 of TS 23.273 [3] are provided in an Annex below, and corresponding Figure 6.1.2-1 of TS 23.273 [3] is included herein as Figure 5 as mentioned above.

[00104] In an example, the operations of Figure 2 can be labelled as follows: • S210 - 1. 5GC-MT-LR procedure for location services as per cl. 6.1.2 of TS 23.273 [3], steps 1-11. • S220 - 2. AI / ML Direct Positioning preparations. • S230 - 3a. Case 2b measurements. • S240 - 3b. Case 3b measurements. • S250 - 4. Data collection sub-procedures for inference of AI / ML direct positioning model(s). • S260 - 5. AI / ML Direct Positioning inference. • S270 - 6. 5GC-MT-LR procedure for location service as per cl. 6.1.2 of TS 23.273 [3], steps 13-24.

[00105] Various examples of the present disclosure are now described in combination with Figure 3.

[00106] Figure 3 illustrates (via a call flow diagram) a data collection procedure (or sub-procedure(s)) for AI / ML direct positioning with LMF-side models. More generally, Figure 1 may be regarded as a method for collecting data for use in performing inference for AI / ML direct positioning models.

[00107] For the purposes of this description, the entities in Figure 3 may be labelled as follows: • 100 - NG-RAN (Next Generation Radio Access Network). • 200 - AMF (Access and Mobility Management Function). • 300 - LMF (Location Management Function). • 400 - NWDAF containing MTLF (Model Training Logical Function). • 500 - GMLC (Gateway Mobile Location Centre). • 600 - OAM (Operations and Management). • 700 -AF (Application Function).

[00108] However, it will be appreciated that, in more general examples, reference could instead be made to a first entity 100, a second entity 200, a third entity 300, a fourth entity 400, a fifth entity 500, a sixth entity 600, and a seventh entity 700; where each entity is arranged to perform the associated operation(s) indicated below. Furthermore, any one or more of entities in Figure 3 may be combined or co-located. For example, as described above, the LMF and the NWDAF can be colocated, in which case an operation which indicates interaction between these two entities is performed internally (for example, not necessitating use of a transmitter / receiver). It will be appreciated that an entity shown in Figure 3 may correspond to the same entity shown in Figure 1 and / or Figure 2 (e.g. the LMF 300 of Figure 3 may be the same as the LMF 300 in Figure 1 or the LMF 300 in Figure 2, as appropriate).

[00109] It will also be understood that various examples relate to individual entities shown in Figure 3. For instance; various examples are directed to the LMF, in which case such examples can focus on operations in which the LMF is involved and omit any other operations; while some other examples are directed to the NWDAF, in which case such examples can focus on operations in which the NWDAF is involved and omit any other operations. Additionally, yet further examples relate to any combination of the individual entities, and so may include the operations performed by these entities while omitting operations performed by other entities not included in the combination.

[00110] The method (or sub-procedure) illustrated in Figure 3 may be part of the training and inference procedures, examples of which are illustrate in Figures 1 and 2 respectively. In other words, noting the reference to Figure 3 in the descriptions of Figure 1 and Figure 2, it will be appreciated that various examples of the present disclosure include combinations of Figures 1 and 3 and Figures 2 and 3, with the method of Figure 3 performed at the indicated point in the method of Figures 1 or 2.

[00111] In operation S310, NWDAF 400 collects data from AMF 200 (e.g. UE ID; that is, the NWDAF 400 transmits UE ID to AMF 200 such that AMF 200 collects data based on the UE ID). The NWDAF 400 may obtain (e.g. request) the data from AMF 200 using the Namf_EventExposure service or the like.

[00112] In operation S320, NWDAF 400 collects data (e.g. first data) from GMLC 500 (e.g. at least one of past UE location(s), positioning method(s), or LCS (location services) QoS). For example, the NWDAF 400 collects the data from the GMLC 500 using Ngmlcjocation service or the like.

[00113] In operation S330, NWDAF 400 collects data (e.g. second data) from OAM 600 (e.g. at least one of UE speed, or UE orientation). This may be performed according to the data collection principles from OAM described in clause 6.2.3 of TS 23.288 [6],

[00114] In operation S340, NWDAF 400 collects data (e.g. third data) from AF 700 (e.g. UE trajectory). The NWDAF 400 may collect the data from the AF 700 using the Naf_EventExposure service or the like.

[00115] In operation S350, if LMF 300 is a standalone 5GC NF, LMF 300 may collect data directly from other 5GC NF and, optionally, NG-RAN 100 (e.g. for Case 3b) when performing inference for AI / ML Direct Positioning. For example, operation S350 is conditional upon LMF 300 being standalone (e.g. not co-located with NWDAF 400). If operation S350 is performed or included in the method, then LMF 300 collects data from one or more of the entities (e.g. NFs) in the network (or from one or more other entities).

[00116] In an example, the operations of Figure 3 can be labelled as follows: • S310 - 1. Namf_EventExposure. • S320 - 2. Ngmfc_Location. • S330 - 3. Data collection for OAM. • S340 - 4. Naf_EventExposure. • S350 - 5. LMF-based data collection.

[00117] In terms of impacts on services, entities and interfaces, various examples herein disclose the following: i) at NWDAF, support for model sharing to LMF as well as enhancing the support for model training services to enable AI / ML positioning inference at LMF, and ii) support for inference for AI / ML Direct Positioning and data collection.

[00118] Figure 4 is a block diagram of an exemplary apparatus, or network entity, that may be used in examples of the present disclosure. The skilled person will appreciate said 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.

[00119] The entity 1000 comprises a processor (or controller) 1001, a transmitter 1003 and a receiver 1005. The receiver 1005 is configured for receiving one or more messages from one or more other network entities, for example as described above. The transmitter 1003 is configured for transmitting one or more messages to one or more other network entities, for example as described above. The processor 1001 is configured for performing one or more operations, for example according to the operations as described above.

[00120] Figure 6 illustrates a method according to an example of the present disclosure. The method is performed by an LMF entity.

[00121] In operation 610, the LMF entity subscribes to, or requests, AI / ML-related services from a NWDAF entity in relation to an AI / ML model for determining positioning of a UE.

[00122] In operation 620, the LMF entity receives from the NWDAF entity, an indication that training has been performed for the AI / ML model.

[00123] It will be appreciated that the present disclosure includes further examples whereby the method of Figure 6 is modified according to any one or more of the other examples describes herein or according to any one or more of the claims.

[00124] Figure 7 illustrates a method according to another example of the present disclosure. The method is performed by a NWDAF entity including a model training logical function (MTLF).

[00125] In operation 710, the NWDAF entity receives, from a LMF entity, a subscription to or request for AI / ML-related services in relation to an AI / ML model for determining positioning of a UE.

[00126] In operation 720, the NWDAF entity obtains data fortraining the AI / ML model from at least one other entity.

[00127] In operation 730, the NWDAF entity trains the AI / ML model based on the obtained data.

[00128] In operation 740, the NWDAF entity transmits, to the LMF entity, an indication that training has been performed for the AI / ML model.

[00129] It will be appreciated that the present disclosure includes further examples whereby the method of Figure 7 is modified according to any one or more of the other examples describes herein or according to any one or more of the claims

[00130] It will be appreciated that, in each example / embodiment / aspect etc. described above, one or more features or operations may be omitted, modified or moved (e.g., to change the order of the features or the operations), if desired and appropriate. Additionally, one or more features or operations from any example / embodiment may be combined with features or operations from any other example / embodiment. In particular, regardless of whether or not a pointer towards a combination of features / examples is found herein, the present disclosure should be considered to include all combinations of two or more of the embodiments, examples etc. disclosed herein, and all combinations of two or more of the features disclosed herein.

[00131] 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 or example 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.

[00132] 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.

[00133] 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 and / or aspect 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.

[00134] While the present disclosure has been shown, illustrated 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 disclosure.

[00135] The reader's attention is directed to all papers and documents which are filed concurrently with or previous to this specification in connection with this application and which are open to public inspection with this specification, and the contents of all such papers and documents are incorporated herein by reference.

[00136] Annex-Steps 1 to 24 of Clause 6.1.2 of TS 23.273 [31, see Figure 5 also 1. The LCS Client, the NF or the AF (via NEF) sends a request to the (H)GMLC fora location and optionally a velocity for the target UE which may be identified by an GPSI or an SU PI. The request may include the required QoS, supported GAD shapes and other attributes. (H)GMLC (for 1a, 1c) or NEF (for 1b) authorizes the LCS Client, the NF or the AF for the usage of the LCS service. If the authorization fails, step 2-23 are skipped and (H)GMLC (for 1a, 1c) or NEF (for 1b) responds to the LCS Client, the NF or the AF the failure of the service authorization in step 24. In some cases, the (H)GMLC derives the GPSI or SUPI of the target UE and possibly the QoS from either subscription data or other data supplied by the LCS Client, the NF or the AF. The LCS request may carry also the Service Identity (see TS 22.071 [2]) and the Codeword and the service coverage information. The (H)GMLC may verify that the Service Identity received in the LCS request matches one of the service identities allowed for the LCS client orAF. If the service identity does not match one of the service identities for the LCS client orAF, the (H)GMLC shall reject the LCS request. Otherwise, the (H)GMLC can map the received service identity in a corresponding service type. The LCS service request may include a scheduled location time if a current location of the UE is required at a specific time in the future. The LCS service request may include integrity requirements including Time-to-Alert (TTA), Target Integrity Risk (TIR) and Alert Limit(AL). Definitions of these parameters are specified in TS 38.305 [9], NOTE 1: In this release of specification, integrity requirements are for GNSS integrity and RAT-dependent integrity. If the LCS service request contains the pseudonym of the target UE and the (H)GMLC cannot resolve the PMD address from the pseudonym, the (H)GMLC itself determines the verinym (GPSI or SUPI) of the target UE. If the (H)GMLC can resolve the address of PMD from the pseudonym, the HGMLC requests the verinym from its associated PMD. If (H)GMLC is not able to obtain the verinym of the target UE, the (H)GMLC shall cancel the location request. If a scheduled location time is not included and the requested type of location is "current or last known location" and the requested maximum age of location information is available, the (H)GMLC verifies whether it stores the previously obtained location estimate of the target UE. If the HGMLC stores the location estimate and timestamp of the location estimate (if available) and the location estimate satisfies the requested accuracy and the requested maximum age of location, the (H)GMLC checks the result of the privacy check at step 2. If the result of the privacy check for call / session unrelated class is "Location allowed without notification" then steps 3-23 may be skipped. 1b-1 AF sends the Nnef_EventExposure_Subscribe to the NEF. 1b-2 The NEF identifies based on the QoS attribute received from the location request that higher than cell-ID level location accuracy is required and invokes the Ngmlc_Location_ProvideLocation_Request service operation to the (H)GMLC, which contains the attributes received from the AF request. The NEF may also invoke the Ngmlc_Location_ProvideLocation_Request service operation to the (H)GMLC for lower than cell-ID location accuracy as an implementation option or if a scheduled location time is included. 1c. The NF (e.g. NWDAF) invokes the Ngmlc_Location_ProvideLocation service operation to the (H)GMLC. If location is required formore than one UE, the steps following below may be repeated and in that case the NEF or HGMLC receiving location request, shall verify whether the number of Target UEs in the Nnef_EventExposure_Subscribe, Ngmlc_Location_ProvideLocation orLCS request is equal to or less than the Maximum Target UE Number of the LCS client. If Maximum Target UE Number is exceeded, the NEF or HGMLC shall reject the Nnef_EventExposure_Subscribe, Ngmlc_Location_ProviceLocation or LCS request, the steps 2-23 are skipped, and then GMLC respond to the client with proper error cause in the step 24. NOTE 2: If cell-ID level or lower than cell-ID level location accuracy is required in the location request, the NEF may invoke an Namf_EventExposure_Subscribe service operation to subscribe location event reporting from the AMF for the target UE as further described in clause 6.5. 2. The (H)GMLC invokes a Nudm_SDM_Get service operation towards the UDM of the target UE to get the privacy settings of the UE identified by its GPSI or SUPI. The UDM returns the target UE Privacy setting of the UE. The (H)GMLC checks the UE LCS privacy profile. If the target UE is not allowed to be located, steps 3-23 are skipped. The UDM may also reply (H)GMLC with an LPHAP indication, if stored in the UE LCS subscriber data. 3. The (H)GMLC invokes a Nudm_UECM_Get service operation towards the UDM of the target UE with GPSI or SUPI of this UE. The UDM returns the network addresses of the current serving AMF and additionally the address of a VGMLC (for roaming case). If the location request is an immediate location request, the (H)GMLC checks the country codes of the serving node addresses. If the (H)GMLC finds the current AMF is out of the service coverage of the (H)GMLC, the (H)GMLC returns an appropriate error message to the LCS client, the NF or the AF (via NEF). GMLC may determine the LMF ID based on the LCS data for an LCS Client / AF. In case a group ID is provided or derived from the location request, GMLC determines the LMF ID based on the provisioned Group ID. GMLC may be configured with an LMF ID, irrelevant to any LCS client / AF. When the GMLC receives a MT location request from LCS client / AF, GMLC determines the LMF ID for all LCS client / AF. NOTE 3: The UDM is aware of the serving AMF address at UE registration on an AMF as defined in clause 4.2.2.2.2 of TS 23.502 [ 19]. The UDM is aware of a serving VGMLC address at UE registration on an AMF as defined in clause 4.2.2.2.2 of TS 23.502

[19] . NOTE 4: The HGMLC can also query the HSS of the target UE fora serving MME address as described in clause 9.1.1 of TS 23.271 [4], The EPC-MT-LR procedure described in clause 9.1.15 of TS 23.271 [4], excluding the UE availability event, may then be performed instead of steps 4-23, e.g. if the HSS returns an MME address but the UDM does not return an AMF address. 4. Fora non-roaming case, this step is skipped. In the case of roaming, the HGMLC may receive an address of a VGMLC (together with the network address of the current serving AMF) from the UDM in step 3, otherwise, the HGMLC may use the NRF service in the HPLMN to select an available VGMLC in the VPLMN, based on the VPLMN identification contained in the AMF address received in step 3. The HGMLC then sends the location request to the VGMLC by invoking the Ngmlc_Location_ProvideLocation service operation towards the VGMLC. In the cases when the HGMLC did not receive the address of the VGMLC, or when the VGMLC address is the same as the HGMLC address, or when both PLMN operators agree, the HGMLC sends the location service request message to the serving AMF. In this case, step 4 is skipped. If the result of privacy check indicates that the verification based on current location is needed, the HGMLC shall send a location request to the VGMLC (in the case of roaming) or to the AMF (in the case of nonroaming) indicating "positioning allowed without notification" and VGMLC shall invoke an Namf_Location_ProvidePositioninglnfo Request service operation towards the AMF at step 5. H-GMLC also provides the LCS client type ofAF, if received in step 41b 2, or LCS client type of LCS client and other attributes to be sent to AMF in step 5. 5. In the case of roaming, the VGMLC first authorizes that the location request is allowed from this HGMLC, PLMN or from this country. If not, an error response is returned. The (H)GMLC or VGMLC invokes the Namf_Location_ProvidePositioninglnfo service operation towards the AMF to request the current location of the UE. The service operation includes the SU PI, the client type and may include the required LCS QoS, supported GAD shapes, scheduled location time, service type and other attributes as received or determined in step 1. If received in step 2, the (H)GMLC or VGMLC provides the LPHAP indication to the AMF. To support location service in PNI-NPN with signalling optimisation, the H-GMLC also includes a contact address (Notification Target Address, e.g. a URI) and a Notification Correlation ID, which is used by LMF to provide location determination to H-GMLC directly. NOTE 5: The location request forwarded at step 4 and step 5 may also carry the result of the privacy check in step 2 which may include a codeword provided by the LCS Client or AF and an indication of a privacy related action as described in clause 5.4. 6. If the UE is in CM IDLE state, the AMF initiates a network triggered Service Request procedure as defined in clause 4.2.3.3 of TS 23.502 [ 19] to establish a signalling connection with the UE. If signalling connection establishment fails, step 7-13 are skipped and the AMF answers to the GMLC in step 14 with the last known location of the UE (i.e. Cell ID) together with the age of this location. 7. If the indicator of privacy check related action indicates that the UE must either be notified or notified with privacy verification and if the UE supports LCS notification (according to the UE capability information), a notification invoke message is sent to the target UE, indicating the identity of the LCS client and the, service type (if that is both supported and available) and whether privacy verification is required. 8. The target UE notifies the UE user of the location request and, if privacy verification was requested, waits for the user to grant or withhold permission. The UE then returns a notification result to the AMF indicating, if privacy verification was requested, whether permission is granted or denied for the current LCS request. If the UE user does not respond after a predetermined time period, the AMF shall infer a "no response" condition. The AMF shall return an error response in step 14 and if roaming VGMLC in step 15 to the HGMLC if privacy verification was requested and either the UE user denies permission or there is no response with the indication received from the (H)GMLC indicating barring of the location request and steps 10-13 are skipped. The notification result may also indicate the Location Privacy Indication setting for subsequent LCS requests; i.e whether subsequent LCS requests, if generated, will be allowed or disallowed by the UE. The Location Privacy Indication may also indicate a time for disallowing the subsequent LCS requests. 9. The AMF invokes the Nudm_ParameterProvision_Update (LCS privacy) service operation to store in the UDM the Location Privacy Indication information received from the UE. The UDM may then store the updated UE privacy setting information into the UDR as the "LCS privacy" Data Subset of the Subscription Data. 10-13. Step 10-13 are the same as steps 6-9 defined in clause 6.1.1 with the addition that LMF can also perform 6.11.4 as positioning procedure and service type may be indicated towards the LMF and the exception that the LMF may determine the UE location in local coordinates or geographical co-ordinates or both. If the supported GAD shapes is not received in step 11 or Local Co-ordinates is not included in the supported GAD shapes, the LMF shall determine a geographical location. If a scheduled location time is provided at step 5, steps 11 and 12 include the following additional differences. 11. The AMF includes the scheduled location time in the Nlmf_Location_DetermineLocation service operation sent towards the LMF. If received in step 5, the AMF provides the LPHAP indication to the LMF. If H-GMLC contact address is received in step 5, AMF also includes the H-GMLC contact address and the Notification Correlation ID in the Nlmf_Location_DetermineLocation service operation and sends towards the LMF. 12. If received in step 11 of the LPHAP indication, the LMF may determine an appropriate positioning method by taking into account the LPHAP indication. When sending a location request to the UE, the LMF may include the scheduled location time. NOTE 6: If integrity requirements are received in step 11, LMF may determine to use GNSS positioning method and RAT-dependent positioning method. NOTE 7: LMF does not deliver the scheduled location time to NG-RAN as part of step 12. NOTE 8: The LMF may send a location request to the UE at step 12 containing the scheduled location time sometime before the scheduled location time to allow the UE to enter CM Connected state shortly before the scheduled location time. If H-GMLC contact address and the Notification Correlation ID is received in step 11, the LMF responds to AMF in the Nlmf_location_determineLocation Response to indicate that the location determination will be sent directly to GMLC. In this case, the LMF determines to use local AMF for obtaining Non-UE Associated Network Assistance Data, as described in clause 6.11.3. When LMF determines the UE location, it executes the step 28 as described in clause 6.3.1. 14. The AMF returns the Namf_Location_ProvidePositioninglnfo Response towards the (V)GMLC (or HGMLC for roaming when the NL3 reference point is not supported) to return the current location of the UE. The service operation includes the location estimate, its age and accuracy and may include information about the positioning method and the timestamp of the location estimate. If indicated in step 13 that the location determination will be sent directly to GMLC, AMF responds to GMLC with Namfjocation_providePositioninglnfo Response to indicate that GMLC will receive the location from LMF directly. 15. In the case of roaming, the VGMLC forwards the location estimation of the target UE, its age, its accuracy and optionally the information about the positioning method received at step 14 to the HGMLC. For non-roaming scenario, this step is skipped. 16. If the privacy check in step 2 indicates that further privacy checks are needed, the (H)GMLC shall perform an additional privacy check in order to decide whether the (H)GMLC can forward the location information to the LCS client orAF or send a notification if the result of the privacy check requires the notification and verification based on current location. One example when this additional privacy check is needed is when the target UE user has defined different privacy settings for different geographical locations. When an additional privacy check is not needed, the (H)GMLC skips steps 17-23. 17. If the result of privacy checks in step 16 indicates that the notification (and verification) based on current location is needed, and in the case of roaming, the (H)GMLC shall send a location request to the VGMLC with location type indicating "notification only". 18. The (H) GMLC or VGMLC invokes the Namf_Location_ProvidePositioninglnfo service operation towards the AMP to request notification (and verification) based on current location. 19. If the UE is in CM IDLE state, the AMF initiates a network triggered Service Request procedure as defined in clause 4.2.3.3 of TS 23.502 [ 19] to establish a signalling connection with the UE. 20. If the indicator of privacy check related action indicates that the UE must either be notified or notified with privacy verification and if the UE supports LCS notification, the AMF sends a notification invoke message to the target UE, indicating the identity of the LCS client and the service type (if that is both supported and available) and whether privacy verification is required. 21. Step 21 is the same as step 8. 22. The AMF returns the Namf_Location_ProvidePositioninglnfo Response towards the (V)GMLC (or HGMLC for roaming when the NL3 reference point is not supported) with an indication of the result of notification and verification procedure performed in steps 20-21. 23. In the case of roaming, the VGMLC forwards an indication of the result of notification and verification procedure to the HGMLC. For non-roaming scenario, this step is skipped. 24. The (H)GMLC sends the location service response to the LCS Client, the NF or the AF (via the NEF) if the target UE is allowed to be located by the LCS Client, the NF or the AF. Accordingly, NEF invokes Nnef_EventExposure_Notify or sends Nnef_EventExposure_Subscribe Response to the AF. If the location request from the LCS Client contained the pseudonym and the (H)GMLC resolved the verinym from the pseudonym in step 1, the (H)GMLC shall use the pseudonym of the target UE in the location response to the external LCS client. If the external LCS client orAF requires it, the (H)GMLC may first transform the universal location coordinates provided by the AMF into some local geographic reference system. The (H)GMLC may record charging information both for the LCS Client orAF and internetwork revenue charges from the AMF's network. The location service response from the (H)GMLC to the LCS Client, the NF or the AF may contain the information about the positioning method used and the indication whether the obtained location estimate satisfies the requested accuracy or not. If in step 2, step 15, step 16 or step 23 the (H)GMLC identifies that the target UE is not allowed to be located by the LCS Client, the NF or the AF, it rejects the LCS service request, and optionally indicate in the response the reason of the rejection, i.e. the target UE is not allowed to be located. If the LCS QoS Class is Assured and (H)GMLC detects that requested accuracy is not achieved, the (H)GMLC sends error response including failure cause. 5 Acronyms and Definitions (as may be used herein) 3GPP 3rd Generation Partnership Project NW Network 5G 5th Generation NWDAF Network Data Analytics Function 5GC 5G Core QAM Operations and Management 5QI 5G QoS Identifier OS Operating System 5GS 5G System PCF Policy Control Function 5GSM 5G System Session Management PCC Policy and Charging Control 5GMM 5G System Mobility Management PCO Protocol Configuration Options AF Application Function PDR Packet Detection Rule Al Artificial Intelligence PDU Protocol Data Unit AIML Artificial Intelligence / Machine PMF Performance Measurement Function Learning PRS Positioning Reference Signal AM Acknowledged Mode PRU Positioning Reference Unit AMF Access and Mobility Management PSA PDU session anchor Function QFI QoS Flow Identifier (ID) AS Application Server QoE Quality of Experience ASP Application Service Provider QoS Quality of Service ATSSS Access Traffic Steering Switching RACH Random Access Channel & Splitting RAN Radio Access Network AUSF Authentication Server Function RAT Radio Access Technology CDRX Connected Mode Discontinuous RLC-AM Radio Link Control Acknowledge Reception Mode CSI Channel Status Information RLC-UM Radio Link Control Unacknowledge DCAF Data Collection Application Mode Function RSD Route Selection Descriptor DNAI Data Network Access Identifier SA Standalone DNN Data Network Name SBA Service-Based Architecture DNS Domain Name Server SBI Service-Based Interface DRB Data Radio Bearer SCEF Service Capability Exposure Function eNB Evolved Node B SCP Service-Based Communication Proxy EPS Evolved Packet System SCTP Stream Control Transmission FQDN Fully Qualified Domain Name Protocol GBR Guaranteed Bit Rate SDAP Service Data Adaptation Protocol GMLC Gateway Mobile Location Centre SDU Service Data Unit gNB Next generation Node B SIM Subscriber Identity Module GPSI Generic Public Subscription SLA Service Level Agreement Identifier SM Session Management IAB Integrated Access and Backhaul SMF Session Management Function ID Identity / ldentifier SN Secondary Node HoT Industrial Internet of Things S-NSSAI Single Network Slice Selection IMEI International Mobile Equipment Assistance Information Identities SRS Sounding Reference Signal IP Internet Protocol SSC Session and Service Continuity l-SMF Intermediate SMF SUPI Subscription Permanent Identifier LMF Location Management Function TAI Tracking Area Identity MA-PDU Multiple Access PDU TE Terminal Equipment ML Machine Learning TM Transparent Mode MME Mobility Management Entity TS Technical Specification MN Master Node UDM Unified Data Manager MNO Mobile Network Operator UDR Unified Data Repository MPTCP MultiPath TCP UE User Equipment MT Mobile Termination UL Uplink NAS Non-Access Stratum UM Unacknowledged Mode NEF Network Exposure Function UP User Plane NRF Network Repository Function UPF User Plane Function NG-RAN Next Generation Radio Access URLLC Ultra-Reliable and Low-Latency Network Communication NG-eNB Next Generation eNB URSP UE Route Selection Policy NSA Non-Standalone XRM Extended Reality and Media NSSF Network Slice Selection Function

Claims

1. A location management function (LMF) entity configured to:subscribe to or request artificial intelligence / machine learning (AI / ML) -related services from a network data analytics function (NWDAF) entity in relation to an AI / ML model for determining positioning of a user equipment (UE); andreceive, from the NWDAF entity, an indication that training has been performed for the AI / ML model.

2. The LMF entity of claim 1, further configured to:receive the AI / ML model from the NWDAF entity.

3. The LMF entity of any one of the previous claims, wherein the indication is included in a notification from the NWDAF entity.

4. The LMF entity of claim 3, wherein the notification notifies that the AI / ML model has been trained according to a request of the LMF entity.

5. The LMF entity of claim 3 or claim 4, wherein if the AI / ML-related services are subscribed to or requested by using a specific service operation, the notification is for a corresponding notification service operation.

6. The LMF entity of claim 5, wherein the specific service operation is Nnwdaf_MLModelProvision subscribe service operation, and the corresponding notification service operation is Nnwdaf_MLModelProvision notify service operation.

7. The LMF entity of any one of the previous claims, further configured to transmit an analytics ID indicating the AI / ML model to the NWDAF entity.

8. The LMF entity of any one of the previous claims, further configured to:obtain measurements from the UE and / or next generation radio access node (NG-RAN).

9. The LMF entity of claim 8, wherein the obtained measurements from the UE include positioning reference signal (PRS), and / or the obtained measurements from the NG-RAN include sounding reference signal (SRS).

10. The LMF entity of claim 8 or claim 9, further configured to: determine a location of the UE based on the AI / ML model and at least one of the obtained measurements and data collected from at least one other entity.

11. The LMF entity of claim 10, wherein the location of the UE is estimated based on performing inference using: the AI / ML model and at least one of the obtained measurements and collected data.

12. The LMF entity of any one of claims 8 to 11, further configured to: train the AI / ML model using the obtained measurements.

13. The LMF entity of any one of the previous claims, wherein the LMF entity is included in an apparatus comprising at least one processor, a receiver and a transmitter.

14. A network data analytics function (NWDAF) entity configured to:receive, from a location management function (LMF) entity, a subscription to or request for artificial intelligence / machine learning (AI / ML) -related services in relation to an AI / ML model for determining positioning of a user equipment (UE);obtain data for training the AI / ML model from at least one other entity;train the AI / ML model based on the obtained data; andtransmit, to the LMF entity, an indication that training has been performed for the AI / ML model;wherein the NWDAF entity includes a model training logical function (MTLF).

15. The NWDAF entity of claim 14, further configured to: transmit the trained AI / ML model to the LMF entity.

16. The NWDAF entity of claim 14 or claim 15, wherein the indication is included in a notification transmitted to the LMF entity.

17. The NWDAF entity of claim 16, wherein the notification notifies that the AI / ML model has been trained according to a request of the LMF entity.

18. The NWDAF entity of claim 16 or claim 17, wherein if the AI / ML-related services are subscribed to or requested by using a specific service operation, the notification is for a corresponding notification service operation.

19. The NWDAF entity of claim 18, wherein the specific service operation is Nnwdaf_MLModelProvision subscribe service operation, and the corresponding notification service operation is Nnwdaf_MLModelProvision notify service operation.

20. The NWDAF entity of any one of claims 14 to 19, further configured to: receive an analytics ID indicating the AI / ML model from the LMF entity.

21. The NWDAF entity of any one of claims 14 to 20, wherein the data is obtained from at least one of the LMF entity, application management function (AMF), NG-RAN, or gateway mobile location centre (GMLC).

22. The NWDAF entity of claim 21, wherein the obtained data includes location services (LCS) quality of service (QoS).

23. The LMF entity or the NWDAF entity of any one of the previous claims, wherein the LMF entity is co-located with the NWDAF entity.

24. A method for a location management function (LMF) entity, the method comprising: subscribing to or requesting artificial intelligence / machine learning (AI / ML) -related services from a network data analytics function (NWDAF) entity in relation to an AI / ML model for determining positioning of a user equipment (UE); andreceiving, from the NWDAF entity, an indication that training has been performed for the AI / ML model.

25. A method for a network data analytics function (NWDAF) entity, the method comprising:receiving, from a location management function (LMF) entity, a subscription to or request for artificial intelligence / machine learning (AI / ML) -related services in relation to an AI / ML model for determining positioning of a user equipment (UE);obtaining data for training the AI / ML model from at least one other entity;training the AI / ML model based on the obtained data; andtransmitting, to the LMF entity, an indication that training has been performed for the AI / ML model;wherein the NWDAF entity includes a model training logical function (MTLF).39

Images