Conflict management of artificial intelligence / machine learning operations in communications networks

Abstract

An Artificial Intelligence / Machine Learning (AI / ML) conflict management method for a wireless communications network, the method comprising: obtaining information on a plurality of AI / ML functionaliti

Description

BACKGROUND Field

[0001] Certain examples of the present disclosure provide techniques relating to conflict management of Artificial Intelligence (Al) and / or Machine Leaning (ML) in communications systems. For example, certain examples of the present disclosure provide methods, apparatus and systems for Al and / or ML conflict management in 3rd Generation Partnership Project (3GPP) networks such as 5th Generation (5G) and 6th Generation (6G) networks. Description of Related Art

[0002] The content of the following documents is referred to below and / or their content provides background information and context that the following disclosure should be considered in view of: [1] Moysen, Jessica, et al. "Conflict resolution in mobile networks: a selfcoordination framework based on non-dominated solutions and machine learning for data analytics [application notes]." IEEE Computational Intelligence Magazine 13.2, 52-64, May 2018. [2] 3GPP Technical Specification (TS) 28.104 Technical Specification Group Services and System Aspects; Management and orchestration; Management Data Analytics (MDA), v17.2.0, January 2023. [3] Rawal, Atul, et al. "Recent Advances in Trustworthy Explainable Artificial Intelligence: Status, Challenges, and Perspectives." IEEE Transactions on Artificial Intelligence, vol. 3, 852-866, December 2022. [4] 3GPP Technical Specification (TS) 28.628 Technical Specification Group Services and System Aspects; Telecommunication management; SelfOrganizing Networks (SON) Policy Network Resource Model (NRM) Integration Reference Point (IRP); Information Service (IS), v17.0.0, March 2022. [5] RP-213599, Study on Artificial Intelligence (Al) / Machine Learning (ML) for NR Air Interface, 3GPP TSG RAN Meeting #94e, December 2021 [6] 3GPP TS 38.413, Technical Specification Group Radio Access Network; NG-RAN; NG Application Protocol (NGAP), v17.3.0, January 2023. [7] 3GPP TS 38.413, January 2023. [8] 3GPP TS 38.331 v17.3.0, January 2023. [9] 3GPP TS 38.423, January 2023.

[10] 3GPP TS 38.300, January 2023.

[11] O-RAN. WG1.O-RAN-Architecture-Description: “O-RAN Working Group 1; O-RAN Architecture Description, v07.00, 2022 (Note: the example versions shown for each TS are non-limiting, other versions of the TS may be 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 are now widely deployed, and development of Sixth Generation (6G) Systems is in progress.

[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. New frameworks and architectures are also being developed as part of 5G networks in order to increase the range of functionality and use cases available through 5G networks.

[0006] In recent years, autonomous systems that incorporate machine learning solutions have become more prevalent in telecommunication networks, performing tasks such as prediction, planning, control, etc. These systems are different from conventional software components and present novel challenges and risks that may not be manageable through traditional software engineering practices. A reason for this difference is that the logic of these ML solutions is not defined by source code or specifications, but rather it is determined by the training process, the training data, and the input data at inference time.

[0007] A new framework developed as part of 5G networks (and beyond - B5G) is the use of artificial intelligence I machine learning (AI / ML), which may be used for the optimisation of the operation of 5G networks. In particular, AI / ML analytics and AI / ML model outputs are expected to drive many strategic decisions in B5G and 6G application functions and network orchestration entities. However, the use of AI / ML for network management may lead to conflicts between entities that are performing different optimisation activities. Al Conflict in Mobile Networks and O-RAN

[0008] The following describes the concept of Al conflict at any level of autonomous cellular network (e.g. RAN, CN, or in any other network entity(-ies) and / or functions) and / or an external entity(-ies) and / or function(s) and / or a given UE (or a set of UEs). Conflict in Mobile Networks: in current networks conflict may occur between different network entities and / or functionalities, or even in the same network entity and / or function. For example, conflict may take place between RAN node elements, components and services due to the adversarial competition for network resources that are key role for achieving a certain target(s) (e.g. maximize throughput, reuse RF bands, etc.). This conflict could pose a threat to the network management and KPIs / KPMs performance. Al Conflict in Mobile Networks: is the state of disagreement, dispute or contradiction on applying policies, following rules, configuring, reconfiguring parameters of one or more of the network elements between two or more of Al algorithms and / or human operators to allow allocating the necessary (are competed on) network resources to achieve the intent goal(s). As a result of the Al conflict, one or more of the network key performance indictors may suffer from low, medium or high degradation. Al Conflict in Open-RAN (O-RAN): The O-RAN architecture supports conflict mitigation function to resolve potentially overlapping or conflicting requests from multiple xApps in the Near-Real-Time RAN Intelligent Controller

[11] , However, even Al Conflict in O-RAN is not defined clearly, the main three types of Al conflicts are: • In “Direct Conflicts”: The conflicts can be observed directly. • For “Indirect Conflicts”: Not observed directly, but the dependences between Al applications are observed. • In “Implicit Conflicts”: Can’t be observed directly, even the dependence between Al applications are not obvious. 3GPP Background (AI / ML work, Al conflict use cases)

[0009] Background information on AI / ML work is set out in Appendix I and Appendix II.

[0010] 3GPP standards have many use-cases where the configuration and reconfiguration of parameters in the network may lead to a conflict, examples of which are set out below. However, the approaches provided by the present disclosure are not limited to addressing these example conflicts but may be applied to any form of AI / ML-related conflict in a communications system. Use-case 1: MLB-MRO conflict - the multi-objective problem formulation in [1] investigates how to solve / optimize / mitigate the mobility load balancing (MLB) problem while the same network elements are working to optimize the mobility robustness optimization (MRO) in parallel and at the same time. In this case the MLB tries to balance traffic load function of the cells by transferring the excess traffic (the UEs in congested cells) to less loaded neighbouring cells and vice versa. The parameters that govern UE cell selection, such as handover, thresholds, cell Individual offset, hysteresis margins and times to trigger a handover event can push such change to occur to achieve the goal of a higher system capacity, by distributing UE traffic across the available radio resources in the system. At the same time, the MRO is targeting how to minimize the call drops due to radio link failures which depending on how the before mentioned handover parameters are reconfigured (e.g., too-early handovers for some UEs cause a radio link failures due to high propagation losses with the new cell and for too-late handovers the a radio link failures due to losing serving base station connection before establishing a new connection with one of the neighbouring cells). This conflict loop require a conflict management to solve or mitigate [2], Use-Case 2: CCO-COC-ESM conflict - the Capacity and Coverage Optimization (CCO), Cell Outage Compensation (COC) and Energy Saving Management (ESM) has been discussed in [4] as an example of changes to the coverage and / or capacity of one or more cells during the same time period, that lead to the following issue: one of the cells (called C1) detects an outage issue within the cell, then, the COC function will try to compensate for the outage by reconfiguring the RF configuration of some compensation candidate cells, e.g., TX power, antenna tilt and antenna azimuth of a neighbour cells (called C2, C3 and / or C4). Before the outage at C1 is compensated, the CCO function may detect the degrading of coverage related KPI (e.g., success rate of RRC connection establishments, cell throughput) of C1 and its neighbour cells (C2 and C3) and make a conclusion that there is a coverage problem in this KPI degraded area. Meanwhile, ESM is operating on C2 to compensate for the coverage of its neighbouring C4 which is going into energy saving state. From the time point at which the outage of C1 is detected until C1 has been compensated by C2 and C3, during this period, if there is no coordination among COC, CCO and ESM, there will be possibly different settings for adjusting TX power, antenna tilt and antenna azimuth of C2 for COC, CCO or ESM purposes respectively. It is most likely that the adjustment from COC, ESM and optimization from CCO may conflict in the common affected outage compensation candidate cell(s) (C2 in this use-case).

[0011] Without solving the conflict between the Al applications in mobile networks, the Al applications actions may eventually result in performance degradation in the system due to the incoherence or contradictions of the Al applications' actions. Conflict management determines which Al application action takes precedence or should be compromised within the actions' set of all the Al applications that conflict with each other. The following are some of the issues that may be addressed by the approaches provided by this disclosure. The present disclosure is not limited to addressing these issues and may address only some of and / or different issues. • How to manage conflict between AI / ML models, between functionalities of different model or model’s functionalities, model’s tasks and / or model’s objectives? • How to detect conflict between AI / ML models, between functionalities of different model or model’s functionalities, model’s tasks and / or model’s objectives? • How to rank, classify and report conflict between AI / ML models, between functionalities of different model or model’s functionalities, model’s tasks and / or model’s objectives? • How to prevent and / or resolve conflict between AI / ML models, between functionalities of different model or model’s functionalities, model’s tasks and / or model’s objectives? • How to mitigate conflict if prevention and / or resolving between AI / ML models, between functionalities of different model or model’s functionalities, model’s tasks and / or model’s objectives is not possible? • What is the behaviour of the network and / or UE in the case of conflict in AI / ML operation? SUMMARY

[0012] 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.

[0013] In accordance with a first aspect of the present disclosure, there is provided an Artificial Intelligence / Machine Learning (AI / ML) conflict management method for a wireless communications network, the method comprising: obtaining information on a plurality of AI / ML functionalities of the wireless communications network; detecting a conflict between the plurality of AI / ML functionalities based on the information on the plurality of AI / ML functionalities; and if a conflict is detected, identifying a conflict management action to prevent, resolve or mitigate the conflict, and providing to at least one of the plurality of AI / ML functionalities an indication of the conflict management action.

[0014] In an example, the detecting a conflict comprises detecting a conflict based on one or more of an objective, a determined parameter, and a task of the plurality of AI / ML functionalities.

[0015] In an example, the detecting a conflict comprises creating or updating a conflict profile corresponding to the detected conflict and including information on the detected conflict in the conflict profile.

[0016] In an example, the conflict profile includes information on one or more of AI / ML functionality ID, AI / ML model ID, conflict ID, conflict type, conflict severity, conflict parameters, conflict cause, conflict detection / monitoring criteria, conflict event trigger, conflict timestamp, conflict management action, conflict reporting, and conflict outcome.

[0017] In an example, the detecting a conflict comprises detecting a conflict based on one or more of a variance of a parameter(s) determined by one or more of the plurality of AI / ML functionalities, and a network performance indicator.

[0018] In an example, the information on the plurality of AI / ML functionalities includes one or more of an ID of the respective AI / ML functionality, an objective of the respective AI / ML functionality, and a parameter determined by the respective AI / ML functionality.

[0019] In an example, the conflict management action includes one or more of updating permissions of one or more of the plurality of AI / ML functionalities, configuring collaboration between the plurality of AI / ML functionalities, determining a parameter of one or more of the plurality of AI / ML functionalities, and configuring a mediator between the plurality of AI / ML functionalities.

[0020] In an example, the method further comprises determining a type of a detected conflict between the plurality of AI / ML functionalities as a direct conflict, an indirect conflict, an implicit conflict, or an other type of conflict.

[0021] In an example, the method further comprises ranking a detected conflict based on the severity of the conflict.

[0022] In an example, the obtaining information on the plurality of AI / ML functionalities includes receiving a conflict management registration message from one or more of the plurality of AI / ML functionalities, the conflict management registration message including information on the respective AI / ML functionalities.

[0023] In an example, the method further comprises receiving a conflict management subscription message from one or more of the plurality of AI / ML functionalities for subscribing to a conflict management procedure.

[0024] In an example, the detecting a conflict includes detecting one or more of a past conflict, a current conflict, and a potential conflict between the plurality of AI / ML functionalities.

[0025] In an example, the detecting a conflict includes identifying the conflict based on the information on the plurality of AI / ML functionalities, and monitoring for an occurrence of the identified conflict.

[0026] In an example, the method further comprises providing a conflict report / notification to one or more of the plurality of AI / ML functionalities including information on the detected conflict.

[0027] In an example, each of the plurality of AI / ML functionalities is associated with a network entity.

[0028] In an example, the method is performed by one or more of the associated network entities.

[0029] In an example, the plurality of AI / ML functionalities are each associated with a user equipment (UE) entity, a radio access network (RAN) entity, a core network (CN) entity, an operation, administration and maintenance (OAM) entity, an application function (AF) entity, and an external entity.

[0030] In an example, the method is performed by one or more conflict management functions.

[0031] In an example, the plurality of AI / ML functionalities are each one of an AI / ML function, an AI / ML model, AI / ML model functionality, an AI / ML application, an AI / ML application function, an AI / ML network entity, and an AI / ML network function.

[0032] In accordance with a second aspect of the present disclosure, there is provided a conflict management entity for a communications network, wherein the conflict management entity is configured to: obtain information on a plurality of AI / ML functionalities of the wireless communications network; detect a conflict between the plurality of AI / ML functionalities based on the information on the plurality of AI / ML functionalities; and if a conflict is detected, identify a conflict management action to prevent, resolve or mitigate the conflict, and provide to at least one of the plurality of AI / ML functionalities an indication of the conflict management action.

[0033] In an example, the conflict management entity is a logical entity of the wireless communications network.

[0034] In an example, the conflict management entity is included in one or more of a radio access network (RAN) entity, a core network (CN) entity, a user equipment (UE) entity, an operation, administration and maintenance (OAM) entity, and a external entity.

[0035] In accordance with a third aspect of the present disclosure, there is provided a method of a wireless communication network for Artificial Intelligence / Machine Learning (AI / ML) conflict management, the wireless communication network comprising a plurality of AI / ML functionalities, the method comprising: obtaining information on the plurality of AI / ML functionalities; detecting a conflict between the plurality of AI / ML functionalities based on the information on the plurality of AI / ML functionalities; and if a conflict is detected, identifying a conflict management action to prevent, resolve or mitigate the conflict, providing to at least one of the plurality of AI / ML functionalities an indication of the conflict management action, and performing the conflict management action by the at least one of the plurality of AI / ML functionalities.

[0036] In an example, the wireless communications network is a 3GPP wireless communications network.

[0037] 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

[0038] Embodiments / examples of the present disclosure are further described hereinafter with reference to the accompanying drawings, in which: Figure 1 provides an example of an AI / ML conflict management procedure involving two NW Entities X, Y (or NW Functions X, Y, or NW Entity X and NW Function Y) in two Actions A and B; Figure 2 provides an example of an AI / ML conflict registration procedure involving a Client and CMC, in AI / ML CONFLICT REGISTRATION REQUEST and AI / ML CONFLICT REGISTRATION RESPONSE actions; Figure 3 provides an example of an AI / ML conflict registration procedure involving a UE (or a group of UEs) and NG-RAN (gNB); Figure 4 provides an example of an AI / ML conflict management procedure between two NG-RAN (gNBs) using Xn interface; Figure 5 provides an example of a AI / ML CONFLICT EVENT REPORT / INDICATION procedure (Class 2); Figure 6 provides an example of an interactions between conflict management stages; and Figure 7 provides a block diagram of an exemplary network entity / function that may be used in certain examples of the present disclosure. DETAILED DESCRIPTION

[0039] 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 invention. 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.

[0040] The same or similar components may be designated by the same or similar reference numerals, although they may be illustrated in different drawings.

[0041] 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.

[0042] 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.

[0043] 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.

[0044] 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.

[0045] Throughout the description, the expression “at least one of A, B and / or C” (or the like) 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.

[0046] 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.

[0047] 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.

[0048] The following examples are applicable to, and use terminology associated with, 3GPP 4G (e.g., LTE) and / or 5G (e.g., NR). However, the skilled person will appreciate that the techniques disclosed herein are not limited to these examples or to 3GPP 4G (e.g., LTE) and / or 5G (e.g., NR), 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 (e.g., B5G, 5G-Advanced, 6G etc.). The skilled person will appreciate that the techniques disclosed herein may be applied in any existing or future releases of 3GPP 4G (e.g., LTE) and / or 5G (e.g., NR) and / or 5G Advanced and / or 6G, and / or (3GPP Release 17, 18, 19, 20, etc.) 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.

[0049] Furthermore, the following also applies to the present disclosure: • The terms functionality / use-case / configuration / scenario / site may be used interchangeably. • The terms model and model functionality may be used interchangeably. • This disclosure also apply to non-3GPP entities. • The concepts, proposals, solutions, methods, embodiments, figures, and / or examples, presented in this disclosure, would apply to various type of communication systems, such as 4G, 4G-Advanced, 5G, 5G-Advanced, and 6G. Moreover, the above may also apply (in full or part or modified) to systems of Non-Terrestrial Networks (NR-NTN and / or loT-NTN and / or UAV, etc.), in addition to Terrestrial Networks (TN).

[0050] A particular network entity may be implemented as a network element on 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.

[0051] 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 4G or 5G or 5G-Advanced and also apply to B5G and 6G systems. • 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.

[0052] 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. AI / ML Conflict Management

[0053] The following is a summary of the main ideas introduced by this disclosure:

[0054] A new framework including a new network entity (or entities) and / or network function (or functions) for handling conflict management procedures and actions related to conflict management procedures of AI / ML operations in communications networks. Although a new network entity (or entities) and / or network function (or functions) are referred to, existing entities / functions may be repurposed and / or new functionality introduced into an existing entity to provide the required functionality.

[0055] The new network entity (or entities) and / or network function (or functions) is / (are) involved in one (or more) conflict management procedure(s) in / with at least one AI / ML model (or model functionality). Examples of the conflict management procedures, include one (or more, or combination of) the following: • Conflict Registration, Conflict Detection, Conflict Monitoring, Conflict Ranking, Conflict Mitigation, Conflict Resolution, Conflict Avoidance, Conflict Update / Modification, Conflict Handling, and / or any other procedures related to conflict management.

[0056] The new network entity (or entities) and / or network function (or functions) is / (are) involved in one (or more) conflict management action(s) in / with at least one AI / ML model (or model functionality). Examples of the conflict management actions, include one (or more, or combination of) the following: • Conflict Registration Request / Response / Acknowledge / Notification / Report / Failure; • Conflict Detection Request / Response / Acknowledge / Notification / Report / Failure; • Conflict Monitoring Request / Response / Acknowledge / Notification / Report / Failure; • Conflict Ranking Request / Response / Acknowledge / Notification / Report / Failure; • Conflict Mitigation Request / Response / Acknowledge / Notification / Report / Failure; • Conflict Update / Modification Request / Response / Acknowledge / Notification / Report / Failure; • Other actions related to Conflict management procedures.

[0057] The naming of the actions and entities / network functions are not limited to those set out above and below and may take alternative names even though the underlying function etc. may remain the same. Consequently, it is the content of the functionality / content of the actions and entities / network functions that it of primary relevance.

[0058] Figure 1 shows an example of an AI / ML conflict management procedure with two actions A S110 and B S112, between two network (NW) entities X102 ,Y104 (or NW Functions X, Y, or NW Entity X and Network Function Y), that can be any of the above mentioned examples of procedures and actions. Entities / Network Functions X and Y may for example also be referred to as first and second Entities / Network Functions, receiving and transmitting Entities / Network Functions, and Source and Destination / Target Entities / Network Functions. Likewise, Actions A and B may be referred to as first and second actions / messages, steps etc. for example. The information transmitted by Actions A and B may include information such as conflict management related information and AI / ML model (or model functionality) information, but are not limited to these and may carry any suitable information related to the procedure being implemented, such as the examples of Figures 2 to 6 for example. The variation in naming and message content also applies to the procedures of Figures 2 to 6.

[0059] In one example, one or both of the entities of Figure 1 may be new network entities and / or functions and may be included in (or part of, or co-located with) RAN, CN, a dedicated internal / external entity (or function), a server, database, cloud, OAM, Application Function (AF), and / or a UE (or a set of UEs), etc.

[0060] The new network entity (or entities) and / or network function (or functions) is (are) involved in at least one conflict management procedure and / or action with a network entity (and / or a network function) termed Conflict Management Controller (CMC) although the entity is not limited to being term the Conflict Management Controller. • In a related example, the CMC is a logical entity (or a function) that is located (or co-located) in (with) at least one new (or existing) network entity and / or function, and / or UE (or a group of UEs) and / or database, edge, cloud, server, application function. • In one example, the CMC could be part of RAN, CN (e.g. AMF, SMF, UPF, UDM, and / or any other internal or external network entity / function), and server, OAM, etc.

[0061] The actions A and B may take the form of the any of the interactions described below, such as those described with reference to Figures 2 to 6. For example, the new network entity (or entities) and / or network function (or functions) is / (are) involved in one (or more) conflict management action(s) in / with at least one AI / ML model (or model functionality). In other words, the overall / general interaction set out in Figure 1 may be used for any Conflict management procedure / conflict management stage and / or action, where the naming of the messages and their content may be appropriately specified. The conflict management procedure or stages thereof may also be a class 1 or class 2 procedures. Examples of the conflict management actions that may be performed using the general procedure of Figure 1, include one (or more, or combination of) the following • Conflict Registration Request / Response / Acknowledge / Notification / Report / Failure; • Conflict Detection Request / Response / Acknowledge / Notification / Report / Failure; • Conflict Monitoring Request / Response / Acknowledge / Notification / Report / Failure; • Conflict Ranking Request / Response / Acknowledge / Notification / Report / Failure; • Conflict Mitigation Request / Response / Acknowledge / Notification / Report / Failure; • Conflict Update / Modification Request / Response / Acknowledge / Notification / Report / Failure; • Other actions related to Conflict management procedures.

[0062] The new network function and / or network entity that is involved in at least one conflict management procedure, for at least one AI / ML model (e.g. multiple tasks, objectives, functionality, use cases, scenarios, configurations, etc.), may perform, for example, one (or more) of the following actions: trigger, initiate, indicate, request, subscribe / notify, modify, provide, inform, report, negotiate, recommend, collect, update, response, inform, process, acknowledge, reject, wait, postpone, do nothing, or any other suitable action related to conflict management procedures.

[0063] Figure 2 shows an example of an AI / ML conflict registration procedure with request S210 (e.g. AI / ML CONFLICT REGISTRATION REQUEST) and response S212 (e.g. AI / ML CONFLICT REGISTRATION RESPONSE) actions, between a Client 202 and CMC 204. The term “Client” is used to refer to any network entity (or function), UE (or a group of UEs), server, Al Application(s), or Application Function (AF). As set out above in relation to Figure 1, the actions / messages S210, S212 may carry any suitable / required information such as conflict related information and AI / ML model (or model functionality) information but are not limited to these. Furthermore, the response message may include a request approval (or Acknowledge), a request refusal (or reject), or a request modification or any other request-related information for example. Upon receipt of the request at the CMC, the CMC may process it and perform actions required for conflict registration, such as those set out below with respect to Figure 6.

[0064] Figure 3 shows an example of AI / ML conflict (event) registration procedure with request S310 and response S312 actions, between a UE 302 (or a group of UEs) and NG-RAN(s) or AMF 304. • In one example, the conflict registration procedure uses existing RRC signalling / message and / or NAS signalling / messages. • In an alternative example, the conflict registration procedure uses new procedure may use newly defined RRC signalling / messages and / or NAS signalling / messages. For example, AI / ML CONFLICT (EVENT) REGISTRATION REQUEST message and AI / ML CONFLICT (EVENT) REGISTRATION RESPONSE message.

[0065] The solutions, embodiments, examples, figures, other description in this disclosure, proposed with new network entity (or entities) and / or function (or functions) can also be used with / apply to existing network entities (e.g. gNB, eNB, NG-RAN, MME, UPF, SMF, AMF, other) and signalling and / or messages (e.g. RRC, NAS), and / or system information broadcast (e.g. periodically or on-demand) and / or messages and / or lEs, and / or interfaces X2, Xn, S1, NG, F1, E1, in addition to possible interaction with NWDAF (for analytics and predictions purposes).

[0066] Figure 4 shows an example of AI / ML conflict management procedure (with request S410 and response S412 actions), between a first NG-RAN(s) 402 and a second NG-RAN(s) or AMF 404. • In one example, the conflict management procedure (or any other suitable procedure naming) is carried over existing interfaces X2, Xn, S1, NG, F1, E1. Additionally, the procedure may involve other network entities, such as gNB, eNB, MME, UPF, SMF, etc. and / or any related signalling and / or messages / Information Elements (lEs). • In an alternative example, the conflict management procedure may use a newly defined interfaces, signalling, messages and / or lEs. • An example, of possible naming of the conflict management procedure and / or messages / actions, but not limited to this naming, is AI / ML CONFLICT MANAGEMENT REQUEST message and AI / ML CONFLICT MANAGEMENT RESPONSE message.

[0067] In an embodiment, the conflict detection procedure and conflict monitoring procedure can use a new class 1 procedure (e.g. AI / ML Conflict Detection Request and AI / ML Conflict Detection Response, and AI / ML Conflict Monitoring Request and AI / ML Conflict Monitoring Response, or other type of related actions), and / or new messages and / or new lEs, and / or using existing messages and / or lEs. In another example of class 1 procedures / messages, AI / ML Conflict Detection Report and AI / ML Conflict Detection Acknowledge, and / or AI / ML Conflict Monitoring Report and AI / ML Conflict Monitoring Acknowledge. The acknowledgement message / action is sent by target entity (or another NW entity and / or function).

[0068] In an embodiment, the conflict (event) detection or conflict monitoring outcome is reported and / or indicated using a new class 2 procedure (and / or a new message and / or new lEs), and / or using an existing message and / or lEs. An example is shown in Figure 5.

[0069] In Figure 5, a source entity 502 initiates the procedure, by sending, e.g. AI / ML CONFLICT EVENT REPORT / INDICATION message, or AI / ML CONFLICT DETECTION (MONITORING) REPORT / INDICATION message S510 (or any other suitable message naming) to the target entity 504 (and / or conflict management controller). Upon reception of the AI / ML CONFLICT EVENT REPORT I INDICATION message, or AI / ML CONFLICT DETECTION (MONITORING) REPORT I INDICATION message (or any other suitable message naming), the target entity (and / or conflict management controller) shall, if supported, based on the conflict reason indicated (e.g. by the source entity using a new cause value IE, e.g. Functionality Conflict, or any other suitable cause value), take appropriate action.

[0070] The new network function and / or network entity that is involved in at least one conflict management procedure, for at least one AI / ML model (e.g. multiple tasks, objectives, functionality, use cases, scenarios, configurations, etc.), may perform, for example, one (or more) of the following actions: trigger, initiate, indicate, request, subscribe / notify, modify, provide, inform, report, negotiate, recommend, collect, update, response, inform, process, acknowledge, reject, wait, postpone, do nothing, or any other suitable action related to conflict management procedures.

[0071] New signalling procedures, messages, and / or lEs to support the conflict management procedures related to at least one AI / ML model (e.g. multiple tasks, objectives, functionality, use cases, scenarios, configurations, etc.) may be introduced.

[0072] This disclosure may extend to a mix of new and existing network entities (and / or functions) and / or interfaces and / or signalling and / or messages, and / or lEs to implement the disclosed AI / ML conflict management procedures.

[0073] The new network entity (and / or function) may assign a Conflict Profile (CP) for example to a given conflict event. This CP is used during all conflict management exchanges in at least one network entity (and / or function). For example, the Conflict Profile may include one or more of the following information: • source entity ID (or source group of entity ID(s)) • target entity ID (or source group of entity ID(s)) • conflict event ID, or conflict session ID, or list of conflict sessions IDs. • conflict parameters (e.g. RSRP, energy, data rate, delay, other) • conflict cause (e.g. conflict between multiple objectives, functionalities, model parameters and / or configurations, other) • conflict type (e.g. direct, in-direct, implicit conflict, other) • conflict detection / monitoring criteria (e.g. objects, actions, tasks, outcomes, functionality, error performance, accuracy performance, other) or conflict threshold (e.g. %, probability, other) policy (e.g. preconfigured by the network, OAM, orAP) • conflict event trigger (e.g. condition, threshold, other) • conflict severity (Ranking) ID • conflict outcome decisions set ID

[0074] Figure 6 provides an example interaction between the conflict management stages, such as stages including Conflict Registration, Conflict Detection, Conflict Monitoring, Conflict Ranking, Conflict Mitigation, Conflict Resolution, Conflict Avoidance, Conflict Update / Modification, Conflict Handling, and / or any other procedures related to conflict management. It should be noted that the interactions of Figure 6 are merely example interactions and the ordering thereof are not limited to that illustrated in Figure 6. For example, one or more of the interactions in Figure 6 may be skipped and / or removed. Messages sent as part of the interactions may also be sent directly between entities / network functions even though an intermediate entity / stage is shown in Figure 6. Additional messages may also be introduced in one or more of the stages. In some examples, the messages / interactions of Figure 6 may be combined, and / or class 1 or class 2 implementations used for any of the interactions. The naming of the messages / interactions and of the stages may also differ even though their functionality may be unchanged. The stages of Figure 6 may be implemented by the CMC but they may also be performed by any appropriate part of a communications system and also may be distributed between different parts of a communications system, for example, one or more entities implementing the stages may be located in the UE, RAN, the core network or outside of the core network or any combination of these. Existing or new network entities / functions may also be used to implement individual or combined parts of the functionality set out in Figure 6. The stages and interactions of Figure 6 may also be combined. However, in the example of Figure 6, the interactions are between the Network Entity and a CMC, which in combination implement the stages.

[0075] The interaction of Figure 6 starts with a Conflict Management stage 604 in which a Conflict Management Subscription Request is sent at S650 from the NW Entity (or NE Function) 602, and a Conflict Management Subscription Response is sent back at S652.

[0076] The NW entity (or Function) 602 requests registration at the Conflict Registration stage 612 by sending a Conflict Management Registration Request at S654. The Conflict Registration stage 612 then requests to add the conflict Al ID to the CMC database at S656 by sending a Conflict Management and Registration Request to the CMC database 614. A Conflict Management Registration Response is sent to the NW Entity 602 at S658

[0077] Once the registration is complete, the CMC database 614 then requests the Conflict Detection / Monitoring stage 610 to start analysing the data (e.g. the new network entity (or entities) and / or network function (or functions) information that has been saved in the database for a network data analysis function) for detecting conflict(s) between the NW Entity (or NE Function) 602 and the other registered NW Entity(s) (or NW Function(s)) by sending a Conflict Management Detection / Monitoring Request at S660, where the other registered NW Entity(s) (or NW Function(s)) may have previously registered using an equivalent procedure to NW Entity (or Function) 602. Then the existing Conflict Profiles (CP(s)) under monitoring will be updated and new conflicts will have new CP(s). The Conflict Detection / Monitoring stage 610 will send all the CP(s) send back to the CMC database 614 as a Conflict Management Detection / Monitoring Response at S662. In addition, the Conflict Detection / Monitoring stage 610 will send requests to both the Conflict Classification stage 608 and Conflict Ranking 606 stages using a Conflict Management Classification and a Conflict Management Ranking Request (or a combined request) at S664. Also, the Conflict Detection / Monitoring stage 610 may send CPs notification to the NW entity (or Function) 602 at S666 by sending a Conflict Management CP Notification Request.

[0078] Then the NW entity (or Function) 602 requests classification from the Conflict Classification stage 608 by sending a Conflict Management Classification Request at S668. Then the Conflict Classification stage 608 responds back to the NW entity (or Function) 602 though a Conflict Management Classification Response at S670. Similarly, the NW entity (or Function) 602 requests the rankings from the Conflict Ranking stage 606 by sending a Conflict Management Ranking Request at S672, then the Conflict Ranking stage 606 responds back to the NW entity (or Function) 602 with a Conflict Management Ranking Response at S674.

[0079] At S676 a Conflict Management Actions Response is sent to the NW entity, where this message includes actions related to a detected conflict, such as an acknowledgement, and / or info about solving, preventing or mitigating a conflict. Such a message is illustrated as being sent as part of a conflict ranking stage but may be sent as part of other stages or a separate stage and may be sent by any entity that is suitable for generating such information, such as a CMC or other entity. There may also be default actions that may be taken in response to a conflict or potential conflict, in order to solve or mitigate the conflict. Such default responses may be stored at the CMC (in the database for example) and provided to AI / ML functions as required.

[0080] Subsequently, the NW entity (or Function) 602 can request the conflict report from the Conflict Reporting stage 618 by sending a Conflict Management Report Request at S678. Once the request is received, the Conflict Reporting stage 618 responds back with a Conflict Management Report Response at step S680. Then the procedure may then repeat the foregoing interactions where S682 and S684 correspond to S650 and S652.

[0081] A detailed description of the stages of Figure 6 is below. AI / ML Conflict Management Stages

[0082] In the following we describe example conflict management stages (such as those of Figure 6) as part of an interaction between a NW entity and CMC, the example equally applies to interaction between two NW entities (or functions): Conflict Registration

[0083] In this stage the Al application(s) register(s) in the CMC database to provide the related information:

[0084] Step 1: the Al application (or NW entity) sends a register request to the CMC. This message may include information related Al application ID, model (or model ID) and / or model functionality (or functionality ID) and conflict (conflict session ID, if available). Additionally / optionally, other parameters as described in this disclosure may be included.

[0085] Step 2: the CMC may register the conflict event and assign a conflict session ID (if not previously assigned) and may store together with Al application ID and optionally, other parameters described in this disclosure, in the CMC (e.g. database)

[0086] Step 3a: the CMC analyses the requested Al application ID and / or other information related to the conflict (and involved model, or models, or model functionality (-ies)).

[0087] Step 3b: the CMC responds that the conflict event is registered successfully with an assigned session ID. Additional / optionally, the CMC sends session ID, other information related to the conflict event, and / or a set of permissions (e.g., this permissions set may allow the Al application to change / control a set of parameters values / range or update policies). Conflict Detection / Monitoring

[0088] In this stage the CMC detects and monitors any conflict(s) between the Al applications, or AI / ML models (or models functionalities, models use cases, models objectives, or models tasks, etc.) by using different methods (e.g., monitoring frequent changings in the network element(s), KPI(s), etc.):

[0089] Stepl : the CMC fetches all related information of the registered Al applications (or models functionalities, models use cases, models objectives, or models tasks, etc.) that are in conflict from the CMC database (and / or other network entity or function, or server, or cloud, or database). Additionally, optionally current and historical KPI(s), permissions, relation, other information.

[0090] Step 2: the CMC detects the conflict(s) between all registered Al applications (or models functionalities, models use cases, models objectives, or models tasks, etc.) in the CMC database (and / or other network entity or function, or server, or cloud, or database).

[0091] Step 3: the CMC creates a Conflict Profile (CP) for each detected conflict, and additionally, the CMC may (optionally) performs the following : • saves all CPs in the CMC database. • triggers Conflict Classification and update conflict class for each CP(s) in the CMC database. • triggers Conflict Ranking and update conflict rank for each CP(s) in the CMC database.

[0092] Step 4: CMC sends a conflict notification(s) or / and report(s) or / and CP(s) to all related Al applications in the CMC database, and additionally, CMC may send (optionally): • the conflict report including the related CP(s) • the conflict notification(s) / report(s) to the Al application (or NW entity) Conflict Classification

[0093] In this stage the CMC classifies the conflict(s) between the Al applications (or models functionalities, models use cases, models objectives, or models tasks, etc.):

[0094] Step 1: the Client (e.g., Al application, NW entity or CMC stages) requests a conflict classification by sending a classification message with related information (e.g. CP(s) / report(s)).

[0095] Step 2: the CMC classifies (e.g., direct, indirect, implicit, or another conflict class) all received conflict(s) from the Client.

[0096] Step 3: the CMC responds by sending conflict(s) classes to the Client Conflict Ranking

[0097] In this stage the CMC ranks the conflict(s) between the Al applications (or models functionalities, models use cases, models objectives, or models tasks, etc.):

[0098] Step 1: the Client (e.g., Al application, NW entity or CMC stages) requests a conflict ranking by sending a ranking message with related information (e.g. CP(s) / report(s)).

[0099] Step 2: the CMC ranks (e.g., severity score starting withl for low severity until 5 for highest conflict severity) all received conflict(s) from the Client.

[00100] Step 3: the CMC responds by sending conflict(s) ranks to the Client Conflict Handling

[00101] In this stage the CMC handles the conflict(s) between the Al applications (or models functionalities, models use cases, models objectives, or models tasks, etc.) using different methods, where the handling may be performed across a number of stages / components associated with the CMC:

[00102] Stepl: the CMC triggers Conflict Detection procedure.

[00103] Step 2: the CMC receives all the detected conflict(s) (e.g., CP(s)) from the Conflict Detection

[00104] Step 3: the CMC handles the conflict(s). the following are possible handling examples: • The CMC may handle the conflict by preventing, solving or mitigating. • The CMC updates the set of permissions for each related Al application (or NW entity) • The CMC creates a handling report with handling details (e.g., using collaboration, mediation, investigation or others) to the related Al application (or NW entity)

[00105] Step4: go back to “Stepl” Conflict Reporting

[00106] In this stage the CMC sends a conflict report to the Al application(s) (or models functionalities, models use cases, models objectives, or models tasks, etc.) (Or NW entity) on demand or periodically:

[00107] Step 1: the Client (e.g., Al application, NW entity or CMC stages) reguests a conflict report(s) by sending a reporting message with related information (e.g. report frequency = every T second, report type= subscription level).

[00108] Step 2: the CMC sends a conflict report(s) or / and notification(s) or / and CP(s) to the Client and / or all related conflict Al applications) (or models functionalities, models use cases, models objectives, or models tasks, etc.):

[00109] Various example implementations of the foregoing proposals and also variations thereof are set out in the following paragraphs, where these may be combined in any suitable manner and / or implemented alone.

[00110] In an example, the AI / ML conflict detection at a given entity (or a group of entities) is based on a pre-configured conflict criteria, conflict policy, and / or other method for determining a conflict event.

[00111] In an example, the AI / ML conflict monitoring at a given entity (or a group of entities) is based on a pre-configured conflict criteria, conflict policy, and / or other method for determining a conflict event.

[00112] In an example, the entity (or group of entities) that detects the AI / ML conflict case, will send an indication and / or report of a conflict detection event (or conflict event) to another entity (or group of entities).

[00113] In an example, the entity (or group of entities) that monitors the AI / ML conflict case (event) will send an indication and / or report of outcome of conflict monitoring of a given conflict event to at least one other entity (or function).

[00114] In an example, the entity reporting conflict detection event is termed source entity (or source group of entities, or any other suitable naming) and the entity receiving the indication and / or report of conflict event is termed a target entity (or target group of entities).

[00115] In an example, the entity reporting conflict monitoring event is termed source entity (or source group of entities, or any other suitable naming) and the entity receiving the indication and / or report of conflict event is termed a target entity (or target group of entities).

[00116] In an example, the source entity indicate and / or report a conflict detection event (or conflict event) to the conflict management controller. In another example, the conflict management controller is the target entity.

[00117] In an example, the source entity indicating and / or reporting the conflict event assigns a conflict ID (conflict event ID) and include this ID in the indication and / or report of conflict event that is sent to the target entity.

[00118] In an example, the conflict management controller detects the conflict event and inform all entities involved in the conflict (and / or other entities).

[00119] In an example, the conflict management controller detects the conflict event and request information, e.g. related to the conflict and / or AI / ML model operation, from all entities involved in the conflict (and / or other entities).

[00120] In an example, entities (or group entities) report (and / or indicate) the conflict event to conflict management controller entity that assigns conflict ID and send this ID to all entities involved in the reported or indicated conflict event.

[00121] In another example, the source entity (or source group of entities) includes, as part of the conflict indication and / or report, information related to AI / ML model(s) and / or Model(s) functionality (-ies). For example, the report includes information on AI / ML model ID, functionality ID, other information related to model operation and life-cycle management. In another example, the report includes information related to the conflict as listed below: • All CP(S) involved in the conflict • Each Conflict detecting (starting) and updating ( Monitoring ) / Handling (ending time) • Each Conflict Class • Each conflict Rank • Each Conflict Root-Cause Parameters • Each Conflict Severity • Each Conflict Handling Method Applied Conflict Management Example

[00122] This is an example of a conflict between the two Al applications (MLB and MRO) where each Al application is optimizing a different objective by tuning independent parameters and / or dependent parameters (e.g., CIO, TTT, HYS). The stages below corresponds to some of those described with reference to Figure 6; however, it should be noted that procedure below is merely illustrative and therefore the stages / functionality of Figure 6 may be combined in any suitable manner to detect and mitigate conflicts between AI / ML models / functions. (Stage 1) Conflict Registration:

[00123] Step 1: both MLB and MRO are Al application(s) (or NW entity(S)) which send a register request with information related to the Al application (e.g. Application ID, etc.) and / or model (e.g. model ID, model functionality, functionality ID, etc.) to the CMC.

[00124] Step 2: CMC registers the Al application (e.g. Application ID, etc.) and / or model (e.g. model ID, model functionality, functionality ID, etc.) in the CMC database for both the MLB and MRO respectively (e.g., the MLB registers the following A3 even tuneable parameters: Cell Individual Offset (CIO), hysteresis margin (HYS) or time to trigger (TTT). At the same time the MLB registers the following parameters: Cell Individual Offset (CIO) and time to trigger (TTT) only).

[00125] Step 3a: CMC analyses the requested Al application (e.g. based on Application ID, other information) and / or model (e.g. based on model ID, model functionality, functionality ID, and / or other information) with the CMC database (e.g., analyse the effect of CIO on both the MLB and MRO, similarly for other registered parameters, the CMC complete the analysis).

[00126] Step 3b: CMC respond with a set of permissions (e.g., this permissions set allows the MRL to change only the CIO and HYS only and for MLB to change both the CIO and the TTT). (Stage 2) Conflict Detection:

[00127] The CMC detects the conflict(s) between the MLB and MRO by using different methods (e.g., monitoring frequent changings in changing the HYS and high variance in the CIO values within short period of time.) following the steps:

[00128] Step 1: CMC fetches all related information (e.g., CIO, HYS,TTT cells status, MLB KPI-System Capacity and MRO KPI- Radio Link Failure) of the registered MLB and MRO in the CMC database.

[00129] Step 2: CMC detects two conflicts between the MLB and MRO in the CMC database

[00130] Step 3: CMC creates a CP for each detected conflict (e.g. for the MRO-MLB-CP: High frequent change in CIO, TTT and high rate of Radio Link Failure I for the MLB-MRO-CP: High variance in HYS changed values and System Capacity reduction ). Additionally, the following are examples of CMC handling : • Save all CPs in the CMC database. • Triggers Conflict Classification and update conflict class for each CP(s) in the CMC database. • Triggers Conflict Ranking and update conflict rank for each CP(s) in the CMC database.

[00131] Step 4: CMC sends a conflict notification(s) or / and report(s) or / and CP(s) to both the MLB and MRO registered Al applications in the CMC database. Additionally, the following are examples of CMC handling : • the conflict report including the related CP(s) • send the conflict notification(s) / report(s) to the Al application ( or NW entity) (Stage 3) Conflict Classification:

[00132] The CMC classifies the conflicts between the MLB and MRO following the steps below:

[00133] Step 1: The Clients (e.g. the MLB and the MRO) request a conflicts classification by sending a classification message with related information (e.g. CP(s) / report(s)).

[00134] Step 2: CMC classifies the conflicts of setting up the cell CIO, TTT and HYS to two conflicts (e.g., in this example, the same parameters are in conflict are CIO and TTT, then this is a direct class one. While the HYS is indirect as it depends on the CIO and / or TTT) for both the MLB and the MRO.

[00135] Step 3: CMC response by sending conflicts classes to the Client (e.g., sending the first Client-MLB: that the 1st conflict is with MRO and it is a direct one and related to the CIO and TTT values and the 2nd is indirect and related to HYS. Similarly sending the 2nd Client-MRO: that the conflict is with MLB and it is a direct one and related to the CIO and TTT values. While the 2nd is indirect and related to HYS). (Stage 4) Conflict Ranking:

[00136] The CMC ranks the conflicts ( Direct and the indirect) between the MLB and the MRO:

[00137] Step 1: Client (e.g., MLB and MRO classified conflicts from Stage3) requests a conflict ranking by sending a ranking message with related information (e.g. the MLB and MRO CP(s) / report(s)).

[00138] Step 2: CMC ranks (e.g., MLB-MRO direct 1st conflict high severity score 5 and MLB-MRO indirect 2nd conflict with low severity 2) all received conflicts from both Clients (The MLB and MRO).

[00139] Step3: CMC response by sending conflict(s) ranks to the Client. (Stage 5) Conflict Handling:

[00140] Step 1: CMC triggers Conflict Detection / Monitoring.

[00141] Step 2: CMC receives all the detected conflicts (e.g., 1st MLB-MRO and 2nd MLB-MRO conflicts) from the Conflict Detection / Monitoring

[00142] Step 3: CMC handles the conflict (e.g., giving the permission for both the MRO and MLB to collaboratively choose the reasonable CIO and HYS values for the 1st conflict and add a mediator to allow the MRO to find the best TTT value that mitigate the effect on the MLB KPI)

[00143] Step 4: go back to “Step 1” (Stage 6) Conflict Reporting:

[00144] In this stage the CMC sends a conflict report to the MLB and MRO periodically:

[00145] Step 1: The MLB and MRO requests a conflict report(s) by sending a reporting message with report frequency = every 600 seconds, report type= subscription level 1 and 2, respectively.

[00146] Step 2: CMC sends a conflict report(s) or / and notification(s) or / and CP(s) to the MLB and MRO# Example 3GPP Specification Modification

[00147] In view of the examples set out above, the following provides example AI / ML Conflict Management Parameters. In particular, the IE below defines the conflict related parameters that may be exchanged during conflict management procedures and actions and may be introduced into the 3GPP specification. lE / Group Name Presence Range IE type and reference Semantics description Criticality Assigned Criticality NW Entity ID (or NW Function ID), or source group of Entity ID(s) M - Conflict Event / Session ID M - Conflict Type O e.g. direct, in-direct, implicit conflict, other YES ignore Conflict Severity 0 (e.g. RSRP, energy, data rate, delay, other YES ignore Conflict Parameters 0 YES ignore Conflict Cause o e.g. conflict between multiple objectives, functionalities, model parameters and / or configurations, other YES ignore Conflict Detection / Monitoring Criteria o YES ignore Conflict Timestamp 0 YES ignore Conflict Outcome 0 YES ignore Conflict Handling Proposal 0 e.g. proposal how to handle the conflict YES ignore Model ID (or list of model IDs) M - Functionality ID (or list of functionalities IDs) M - Model identification type O ENUMERATED (model-ID-based, functions I ity-based-ID, both, other) Model identification is model-ID-based, functionality-based, both, or other YES ignore Model information 0 YES ignore Functionality information 0 YES ignore Conflict Registration O ENUMERATED (UE-side, NW-side, two-sided) Indicates to register a model for conflict event(s) YES ignore Conflict Reporting Frequency o INTEGER (1.. xx, ...) Indicates the reporting frequency Conflict event(s) in Units: second, minutes, other YES ignore Conflict management Information 0 XXX Indicates other parameters and information relate to the Conflict management. YES ignore

[00148] 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.

[00149] Figure 7 is a block diagram of an exemplary network entity / function that may be used in examples of the present disclosure, such as the techniques disclosed in relation to any of the preceding figures. For example, any of the network entities, network function etc. may be provided in the form of the network entity illustrated in Figure 7. The skilled person will appreciate that a network entity / function 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.

[00150] The entity 700 comprises a processor (or controller) 701, a transmitter 703 and a receiver 705. The receiver 705 is configured for receiving one or more messages from one or more other network entities, for example as described above. The transmitter 703 is configured for transmitting one or more messages to one or more other network entities, for example as described above. The processor 701 is configured for performing one or more operations, for example according to the operations as described above.

[00151] Further examples in accordance with the present disclosure are set out below, where the examples may be combined in any appropriate form and also combined with any of the approaches set out above.

[00152] An Artificial Intelligence / Machine Learning (AI / ML) conflict management procedure for a communications network, the procedure comprising: transmitting from a first network entity / function to a second network entity / function a first AI / ML conflict management message; processing the first AI / ML conflict management message at the second network entity / function; and transmitting from the second network entity / function to the first network entity / function, a second AI / ML conflict management message in response to the first AI / ML conflict management message and based on a result of the processing of the first AI / ML conflict management message.

[00153] In an example, the procedure further includes transmitting from a third network entity / function to the second network entity / function a third AI / ML conflict management message; processing the third AI / ML conflict management message at the second network entity / function; and transmitting from the second network entity / function to the third network entity / function, a fourth AI / ML conflict management message in response to the third AI / ML conflict management message and based on a result of the processing of the third AI / ML conflict management message.

[00154] In an example, the second AI / ML conflict management message and the fourth Al ML conflict management message are transmitted based on the processing of the first and second Al / ML conflict management messages.

[00155] In an example, the first, second and third network entities / functions are any one of a User Equipment (UE), an Access and Mobility Management Function (AMF), a Next Generation Radio Access Network (NG-RAN), a Conflict Management Controller, or any network entity / function including AI / ML functionality.

[00156] In an example, the second network entity is an AI / ML Conflict Management Controller (CMC) entity.

[00157] In an example, the processing includes one or more of detecting, registering, ranking, monitoring, mitigating, preventing, classifying, reporting, and recording conflicts between AI / ML network functions.

[00158] In an example, the processing includes detecting and / or monitoring conflicts between AI / ML functions at the second and third network entities / functions.

[00159] In an example, the CMC includes a conflict database storing one or more conflict profiles, each conflict profile indicting one or more potential, detected and / or monitored conflicts between AI / ML network functions.

[00160] In an example, the first and second AI / ML conflict management messages form one of a plurality of conflict management stages.

[00161] In an example, the plurality of conflict management stages include any of conflict registration, conflict detection, conflict monitoring, conflict ranking, conflict mitigation, conflict update / modification, conflict subscription, conflict classification, conflict management actions, conflict event reporting, and conflict management reporting.

[00162] In an example, each of the plurality of stages is performed using a class 1 or a class 2 procedure.

[00163] An AI / ML Conflict Management Controller (CMC) entity for a communications network, the CMC configured to: receive from a first network entity / function a first AI / ML conflict management message; process the first AI / ML conflict management message; and transmit to the first network entity / function, a second AI / ML conflict management message in response to the first AI / ML conflict management message and based on a result of the processing.

[00164] In an example, the processing comprises one or more of conflict registration, conflict detection, conflict monitoring, conflict ranking, conflict mitigation, conflict update / modification, conflict subscription, conflict classification, conflict management actions, conflict event reporting, and conflict management reporting.

[00165] For all of the examples / aspects / embodiments etc. described above / herein, it should be considered that the corresponding features / operations apply in any order or combination, and that furthermore there exists the possibility to omit one or more features / operations.

[00166] Moreover, for all of the examples, embodiments, aspects etc. above, these apply to at least LTE, NR, NR NTN or loT NTN (note this list is merely to give some examples and should not be seen as limiting), including any related signalling / messages on any of the inferences X2, Xn, NG, S1, F1, etc (again, this list is merely to give some examples and should not be seen as limiting).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.

[00167] Additionally, where the figures illustrating example method flows include text in relation to a specific step / operation, it will be appreciated that this text is simply an example of the corresponding step / operation, where a more general definition (such as may be found in the description of the corresponding step) may apply for the step / operation.

[00168] Additionally, regarding all of the above, one or more features or operations etc. from any example / embodiment may be combined with features or operations from any other example / embodiment. That is, the present disclosure should be considered to include all combinations of examples / embodiments disclosed herein, as appropriate, as well as combinations of individual features within and between each example / embodiment, as appropriate.

[00169] 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.

[00170] 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.

[00171] 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.

[00172] 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.

[00173] 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. Acronyms and Definitions 3GPP 3rd Generation Partnership Project 5G 5th Generation 5GC 5G Core 5QI 5G QoS Identifier 5GS 5G System 5GSM 5G System Session Management 5GMM 5G System Mobility Management AF Application Function Al Artificial Intelligence AM Acknowledged Mode AMF Access and Mobility Management Function AS Application Server ASP Application Service Provider AUSF Authentication Server Function CDN Content Delivery Network DCAF Data Collection Application Function DNAI Data Network Access Identifier DNN Data Network Name DNS Domain Name Server DRB Data Radio Bearer eNB Evolved Node B EPC Evolved Packet Core FEC Forward Error Correction FQDN Fully Qualified Domain Name GBR Guaranteed Bit Rate gNB Next generation Node B GPSI Generic Public Subscription Identifier HSS Home Subscriber Service IAB Integrated Access and Backhaul ID Identity / ldentifier HoT Industrial Internet of Things IMEI International Mobile Equipment Identities IP Internet Protocol l-SMF Intermediate SMF LADN Local Area Data Network LL SSM Lower Layer SSM MBMS Multimedia Broadcast / Multicast Service MBS Multicast / Broadcast Service MBSF Multicast / Broadcast Service Function MBSTF Multicast / Broadcast Service Transport Function MB-SMF Multicast / Broadcast Session Management Function MB-UPF Multicast / Broadcast User Plane Function ML Machine Learning MME Mobility Management Entity MN Master Node MNF Monitoring Network Function MNO Mobile Network Operator MT Mobile Termination NAS Non-Access Stratum NEF Network Exposure Function NRF Network Repository Function NG-RAN Next Generation Radio Access Network NG-eNB Next Generation eNB NSA Non-Standalone NSSF Network Slice Selection Function NTN Non-Terrestrial Networks NW Network NWDAF Network Data Analytics Function OS Operating System OSAPP OS Application PCF Policy Control Function PCO Protocol Configuration Options PDR Packet Detection Rule PDU Protocol Data Unit PTM Point To Multipoint PTP Point to Point QFI QoS Flow Identifier (ID) QoS Quality of Service RACH Random Access Channel RAN Radio Access Network RRC Radio Resource Control RSD Route Selection Descriptor SA Standalone SDAP Service Data Adaptation Protocol SDU Service Data Unit SGW Serving Gateway SIM Subscriber Identity Module SLA Service Level Agreement SM Session Management SMF Session Management Function SN Secondary Node S-NSSAI Single Network Slice Selection Assistance Information SSB Synchronization Signal Block SSM Source Specific IP Multicast address SSC Session and Service Continuity SRB Signaling Radio Bearer SUPI Subscription Permanent Identifier TA Tracking Area TAI Tracking Area Identity TE Terminal Equipment TM Transparent Mode TMGI Temporary Mobile Group Identity TS Technical Specification UAV Unmanned Aerial Vehicle UDM Unified Data Manager UDR Unified Data Repository UE User Equipment UL Uplink UM Unacknowledged Mode UP User Plane UPF User Plane Function URLLC Ultra-Reliable and Low-Latency Communication URSP UE Route Selection Policy Appendix I (RANI - Study Item on AI / ML for NR Air Interface) RANI,#109-e Agreements end working assumotions: Terminology Description Data collection A process of collecting data by the network nodes, management entity, or UE for the purpose of AI / ML model training, data analytics and inference AI / ML Model A data driven algorithm that applies AI / ML techniques to generate a set of outputs based on a set of inputs. AI / ML model training A process to train an AI / ML Model [by learning the input / output relationship] in a data driven manner and obtain the trained AI / ML Model for inference AI / ML model Inference A process of using a trained AI / ML model to produce a set of outputs based on a set of inputs AI / ML model validation A subprocess of training, to evaluate the quality of an AI / ML model using a dataset different from one used for model training, that helps selecting model parameters that generalize beyond the dataset used for model training. AI / ML model testing A subprocess of training, to evaluate the performance of a final AI / ML model using a dataset different from one used for model training and validation. Differently from AI / ML model validation, testing does not assume subsequent tuning of the model. UE-side (AI / ML) model An AI / ML Model whose inference is performed entirely at the UE Network-side (AI / ML) model An AI / ML Model whose inference is performed entirely at the network One-sided (AI / ML) model A LF-side (AI / ML) model or a Network-side (AI / ML) model Two-sided (AI / ML) model A paired AI / ML Model(s) over which joint inference is performed, where joint inference comprises AI / ML Inference whose inference is performed jointly across the UE and the network, i.e, the first part of inference is firstly performed by UE and then the remaining part is performed by gNB, or vice versa. AI / ML model transfer Delivery of an AI / ML model over the air interface, either parameters of a model structure known at the receiving end or a new model with parameters. Delivery may contain a full model or a partial model. Model download Model transfer from the network to UE Model upload Model transfer from UE to the network Federated learning / federated training A machine learning technique that trains an AI / ML model across multiple decentralized edge nodes (e.g., UEs, gNBs) each performing local model training using local data samples. The technique requires multiple interactions of the model, but no exchange of local data samples. Offline field data The data collected from field and used for offline training of the AI / ML model Online field data The data collected from field and used for online training of the AI / ML model Model monitoring A procedure that monitors the inference performance of tire AI / ML model Supervised learning A process of training a model from input and its corresponding labels. Unsupervised learning A process of training a model without labelled data. Semi-supervised learning A process of training a model with a mix of labelled data and unlabelled data Reinforcement Learning (RL) A process of training an AI / ML model from input (a.k.a. state) and a feedback signal (a.k.a. reward) resulting from the model’s output (a.k.a. action) in an environment the model is interacting with. Model activation enable an AI / ML model for a specific function Model deactivation disable an AI / ML model for a specific function Model switching Deactivating a currently active AI / ML model and activating a different AI / ML model for a specific function RANl#110bis-e Agreements Study AI / ML model monitoring for at least the following purposes: model activation, deactivation, selection, switching, fallback, and update (including re-training). Study at least the following metrics / methods for AI / ML model monitoring in lifecycle management per use case: Monitoring based on inference accuracy, including metrics related to intermediate KPIs i. Monitoring based on system performance, including metrics related to system performance KPIs ii. Other monitoring solutions, at least following 2 options. • Monitoring based on data distribution a) Input-based: e.g., Monitoring the validity of the AI / ML input, e.g., out-of-distribution detection, drift detection of input data, or something simple like checking SNR, delay spread, etc. b) Output-based: e.g.. drift detection of output data • Monitoring based on applicable condition Note: Model monitoring metric calculation may be done at NW or UE Study performance monitoring approaches, considering the following model monitoring KPIs as general guidance iii. Accuracy and relevance (i.e., how well does the given monitoring metric / methods reflect the model and system performance) iv. Overhead (e.g.. signaling overhead associated with model monitoring) v. Complexity (e.g., computation and mcmon cost for model monitoring) vi. Latency (i.e., timeliness of monitoring result, from model failure to action, given the purpose of model monitoring) vii. FFS: Power consumption viii. Other KPIs are not precluded. Note: Relevant KPIs may vary across different model monitoring approaches. FFS: Discussion of KPIs for other LCM procedures RAN1#111 Agreements: • Regarding AI / ML model monitoring for AI / ML based positioning, to study and provide inputs on feasibility’, potential benefits (if any) and potential specification impact at least for the following aspects • At least the following are identified for further study as potential data for calculating monitoring metric o If monitoring based on model output ■ E.g., estimated UE location corresponding to model output for direct AI / ML positioning, estimated intermediate parameters) corresponding to model output for AI / ML assisted positioning, ground truth label corresponding to model inference output for both direct and AI / ML assisted positioning o If monitoring based on model input ■ E.g., measurement corresponding to model inference input o Note 1: other type of potential data for model monitoring is not precluded o Note2: combination of one or more type of potential data for monitoring is not precluded • If a given type of data is necessary’ for calculating monitoring metric, study whether and if so o How an entity can be used to provide the given ty pe of data for calculating monitoring metric ■ Companies are requested to report their assumption of the entity (or entities) used to provide the given ty pe of data for calculating monitoring metric for each case o Potential signalling for provisioning of the given type of data for calculating associated monitoring metric o Potential assistance signaling and procedure to facilitate an entity providing data for calculating monitoring metric o Potential UE-network interaction ■ E.g., model monitoring decision indication between UE and network Appendix II (RAN3 - Work Item on AI / ML for NG-RAN) RAN3#117bis-e agreements: Procedures used for Al / ML support in the NG-RAN shall be use case agnostic. Legacy information that are used to support AI / ML are transferred via existing legacy procedures (no need to signal them via other procedures) RAN3 will focus on non-split architecture use cases and procedures first and discuss split architecture use cases and procedures when completion for the non-split architecture use cases and procedures is achieved. Signalling describing the capability to support specific information predictions used for Al / ML is not pursued in this release Signalling describing the capability to supports specific Al / ML use cases is not pursued in this release AI / ML capability exchange in NG-RAN can be achieved by means of procedures for AI / ML information request, Al / ML information response and AI / ML Information Request Failure WA: Solutions for AI / ML information exchange over the NG interface are not considered as part of Rell8. Xn interface: Introduce a new Class 1 procedure for initiating the reporting of Al / ML Related Information and a Class 2 procedure for Data Reporting of Al / ML Related Information; Reporting options for the new procedure used for AI / ML Related Information to be evaluated on a case-by-case basis. Possible reporting options are one-time and periodic reporting. The new procedure is non-UE associated procedure. If needed, the procedure can be used to capture UE-associated information. The response message of the new procedure for Al / ML Related Information indicates if the requested information can be provided. Support the following UE performance information to be sent for feedback purposes: Average Packet Delay, Average UE Throughput DL, Average UE Throughput UL, Average Packet Error Rate. How to indicate validity time (e.g., implicitly with a new prediction when the previous prediction becomes invalid, explicitly with every prediction in the Al / ML output or by the request to the prediction) shall be discussed on a case by case basis. RAN3#118 agreements WA: Procedures used for AI / ML support in the NG-RAN shall be "data type agnostic". Xn interface: The request in the new Class 1 procedure for initiating the reporting of AI / ML Related Information can include an ID assigned by the requesting NG-RAN node to request for reporting, which includes the reporting parameters list of cells to report reporting periodicity The response in the new Class 1 procedure for initiating the reporting of AI / ML Related Information can include an ID assigned by the responding NG-RAN node which includes the confirmation on the reporting parameters requested. The message in the Class 2 procedure for Data Reporting of AI / ML Related Information can include the corresponding IDs assigned by the NG-RAN nodes, reports result. The "Energy Efficiency" metric should be measurable, produced and interpretable by the RAN. Start with per node granularity EE and Per cell granularity EE could be considered if it is feasible. WA: Take the EE defined in SA5 as the baseline for the energy efficiency of a gNB. What to be transfered between NG-RAN nodes is FES. UE Trajectory Prediction is transferred to the target gNB via the Handover Request. 29 04 25

Claims

1. An Artificial Intelligence / Machine Learning (AI / ML) conflict management method for a wireless communications network, the method comprising:obtaining information on a plurality of AI / ML functionalities of the wireless communications network;detecting a conflict between the plurality of AI / ML functionalities based on the information on the plurality of AI / ML functionalities; andif a conflict is detected,creating or updating a conflict profile corresponding to the detected conflict and including information on the detected conflict in the conflict profile,identifying a conflict management action to prevent, resolve or mitigate the conflict, and providing to at least one of the plurality of AI / ML functionalities an indication of the conflict management action.

2. The method of claim 1, wherein the detecting a conflict comprises detecting a conflict based on one or more of an objective, a determined parameter, and a task of the plurality of AI / ML functionalities.

3. The method of claims 1 or 2, wherein the conflict profile includes information on one or more of AI / ML functionality ID, AI / ML model ID, conflict ID, conflict type, conflict severity, conflict parameters, conflict cause, conflict detection / monitoring criteria, conflict event trigger, conflict timestamp, conflict management action, conflict reporting, and conflict outcome.

4. The method of any preceding claim, wherein the detecting a conflict comprises detecting a conflict based on one or more of a variance of a parameter(s) determined by one or more of the plurality of AI / ML functionalities, and a network performance indicator.

5. The method of any preceding claim, wherein the information on the plurality of AI / ML functionalities includes one or more of an ID of the respective AI / ML functionality, an objective of the respective AI / ML functionality, and a parameter determined by the respective AI / ML functionality.

6. The method of any preceding claim, wherein the conflict management action includes one or more of updating permissions of one or more of the plurality of AI / ML functionalities, configuring collaboration between the plurality of AI / ML functionalities, determining a parameter of one or29 04 25more of the plurality of AI / ML functionalities, and configuring a mediator between the plurality of AI / ML functionalities.

7. The method of any preceding claim, wherein the method further comprises determining a type of a detected conflict between the plurality of AI / ML functionalities as a direct conflict, an indirect conflict, an implicit conflict, or an other type of conflict.

8. The method of any preceding claim, wherein the method further comprises ranking a detected conflict based on the severity of the conflict.

9. The method of any preceding claim, wherein the obtaining information on the plurality of AI / ML functionalities includes receiving a conflict management registration message from one or more of the plurality of AI / ML functionalities, the conflict management registration message including information on the respective AI / ML functionalities.

10. The method of any preceding claim, wherein the method further comprises receiving a conflict management subscription message from one or more of the plurality of AI / ML functionalities for subscribing to a conflict management procedure.

11. The method of any preceding claim, wherein the detecting a conflict includes detecting one or more of a past conflict, a current conflict, and a potential conflict between the plurality of AI / ML functionalities.

12. The method of any preceding claim, wherein the detecting a conflict includes identifying the conflict based on the information on the plurality of AI / ML functionalities, and monitoring for an occurrence of the identified conflict.

13. The method of any preceding, wherein the method further comprises providing a conflict report / notification to one or more of the plurality of AI / ML functionalities including information on the detected conflict.

14. The method of any preceding claim, wherein each of the plurality of AI / ML functionalities is associated with a network entity.

15. The method of claim 14, wherein the method is performed by one or more of the associated network entities.29 04 2516. The method of any preceding claim, wherein the plurality of AI / ML functionalities are each associated with a user equipment (UE) entity, a radio access network (RAN) entity, a core network (CN) entity, an operation, administration and maintenance (OAM) entity, an application function (AF) entity, and an external entity.

17. The method of any preceding claim, wherein the method is performed by one or more conflict management functions.

18. The method of any preceding claim, wherein the plurality of AI / ML functionalities are each one of an AI / ML function, an AI / ML model, AI / ML model functionality, an AI / ML application, an AI / ML application function, an AI / ML network entity, and an AI / ML network function.

19. A conflict management entity for a communications network, wherein the conflict management entity is configured to:obtain information on a plurality of AI / ML functionalities of the wireless communications network;detect a conflict between the plurality of AI / ML functionalities based on the information on the plurality of AI / ML functionalities; andif a conflict is detected,create or update a stored conflict profile corresponding to the detected conflict and include information on the detected conflict in the conflict profile,identify a conflict management action to prevent, resolve or mitigate the conflict, and provide to at least one of the plurality of AI / ML functionalities an indication of the conflict management action.

20. The conflict management entity of claim 19, wherein the conflict management entity is a logical entity of the wireless communications network.

21. The conflict management entity of claims 19 or 20, wherein the conflict management entity is included in one or more of a radio access network (RAN) entity, a core network (CN) entity, a user equipment (UE) entity, an operation, administration and maintenance (OAM) entity, and a external entity.29 04 2522. A method of a wireless communication network for Artificial Intelligence / Machine Learning (AI / ML) conflict management, the wireless communication network comprising a plurality of AI / ML functionalities, the method comprising:obtaining information on the plurality of AI / ML functionalities;detecting a conflict between the plurality of AI / ML functionalities based on the information on the plurality of AI / ML functionalities; andif a conflict is detected,creating or updating a conflict profile corresponding to the detected conflict and including information on the detected conflict in the conflict profile,identifying a conflict management action to prevent, resolve or mitigate the conflict, providing to at least one of the plurality of AI / ML functionalities an indication of the conflict management action, andperforming the conflict management action by the at least one of the plurality of AI / ML functionalities.

23. The method and / or conflict management entity of any preceding claim, wherein the wireless communications network is a 3GPP (RTM) wireless communications network.

Images