Ai / ML model related communication between UE and ran

By using a common transaction identifier for AI/ML model transfer sessions, the communication system addresses data size-related issues, enhancing reliability and efficiency in UE-RAN data transmission.

WO2026235075A1PCT designated stage Publication Date: 2026-11-12NEC CORP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/JP2026/016938
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-06
Filing Date
2026-04-28
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

The large data size of AI/ML model parameters and training data for UE-RAN communication leads to significant signaling overhead, resource utilization issues, congestion, and increased power consumption, particularly in the Uu interface, affecting end-to-end reliability and service degradation.

Method used

Implementing a common transaction identifier for initial and subsequent messages during AI/ML model transfer sessions between RAN nodes and UEs, utilizing the control plane to manage data transmission efficiently and reduce the burden on the RRC layer.

Benefits of technology

Enhances the reliability and efficiency of AI/ML model data transfer by minimizing signaling overhead and resource utilization, reducing congestion, and maintaining service quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure JP2026016938_12112026_PF_FP_ABST
    Figure JP2026016938_12112026_PF_FP_ABST
Patent Text Reader

Abstract

A wireless communication system is described that may have at least a RAN node, and at least one UE that support the training and use of two-sided artificial intelligence / machine learning (AI / ML) models. In particular, herein described are procedures for AI / ML model transfer, data collection and data transfer (e.g., model datasets and model parameters) for AI / ML models.
Need to check novelty before this filing date? Find Prior Art

Description

AI / ML MODEL RELATED COMMUNICATION BETWEEN UE AND RAN

[0001] The present disclosure relates to a communication system and to parts thereof.

[0002] The disclosure has particular but not exclusive relevance to wireless communication systems and devices thereof operating according to the 3rd Generation Partnership Project (3GPP) standards or equivalents or derivatives thereof (including Long-Term Evolution (LTE)-Advanced, Next Generation or 5G / 6G networks, future generations, and beyond). The present disclosure in particular, but not exclusively, relates to artificial intelligence (AI) / machine learning (ML) model transfer, data collection and data transfer (e.g., model datasets and / or model parameters) for AI / ML models, and / or the related signalling and communications in a network for such transfers.

[0003] Earlier developments of the 3GPP standards were referred to as the Long-Term Evolution (LTE) of Evolved Packet Core (EPC) network and Evolved Universal Mobile Telecommunications System (UMTS) Terrestrial Radio Access Network (E-UTRAN), also commonly referred as '4G'. More recently, the term '5G' and 'new radio' (NR) is used to refer to an evolving communication technology that supports a variety of applications and services. Various details of 5G networks are described in, for example, the 'NGMN 5G White Paper' V1.0 by the Next Generation Mobile Networks (NGMN) Alliance, which document is available from https: / / www.ngmn.org / 5g-white-paper.html. 3GPP intends to support 5G by way of the so-called 3GPP Next Generation (NextGen) radio access network (RAN) and the 3GPP NextGen core network.

[0004] Under the 3GPP standards, a NodeB (or an eNB in LTE, and gNB in 5G) is the radio access network (RAN) node (or simply 'access node', 'access network node' or 'base station') via which communication devices (user equipments or 'UEs') connect to a core network and communicate with other communication devices or remote servers. For simplicity, the present application may use the term access network node, RAN node (or simply RAN) or base station to refer to any such access nodes.

[0005] For simplicity, the present application will use the term mobile device, user device, or UE, to refer to any communication device that is able to connect to the core network via one or more RAN nodes. Although the present application may refer to mobile devices in the description, it will be appreciated that the technology described can be implemented on any communication devices (mobile and / or generally stationary) that can connect to a communication system for sending / receiving data, regardless of whether such communication devices are controlled by human input or software instructions stored in memory. For example, such a communication device may be operable by a human or may be a partially or fully automated (e.g., machine-type-communication (MTC) / Internet of Things (IoT)) device.

[0006] The RAN architecture may be distributed with the RAN node structure split into two or more parts. In some RAN implementations there are two parts, known as the Central Unit (CU) - sometimes referred to as a 'control unit' - and the Distributed Unit (DU), connected by an F1 interface. This enables the use of a 'split' architecture in which the typically 'higher' CU layers (for example, but not necessarily or exclusively, Packet Data Convergence Protocol (PDCP) and Radio Resource Control (RRC) layers) and the, 'lower' DU layers (for example, but not necessarily or exclusively, Radio Link Control (RLC), Media (sometimes referred to as 'Medium') Access Control (MAC), and Physical (PHY) layers) are separated between a particular CU, and one or more DUs that are connected to and controlled by that CU via the F1 interface. Thus, for example, the higher layer CU functionality for a number of RAN nodes may be implemented centrally (for example, by a single processing unit, or in a cloud-based or virtualised system), whilst retaining the lower layer DU functionality locally separately for each RAN node.

[0007] In more recently proposed RAN distributed architectures, in addition to the CU and DU, the concept of a Radio Unit (RU) - sometimes referred to as a 'remote unit' - has been introduced. In this architecture the RU is responsible for handling the digital front end (DFE), digital beamforming functionality and, typically, the functionality of the lower parts of the PHY layer, whilst the DU typically handles the higher parts of the PHY layer and the RLC and MAC layers. The CU in this architecture continues to be responsible for controlling one or more DUs (each DU corresponding to a different respective gNB) and to handle higher layer signalling (typically RRC and PDCP layers).

[0008] The actual functional split between the CU and DUs (and potentially RUs where applicable) of these distributed architectures is flexible allowing the functionality to be optimised for different use cases. Effectively, the split architecture enables a 5G network to use a different distribution of protocol stacks between CU and DUs (and potentially RUs) depending on, for example, midhaul availability and network design.

[0009] The choice of how to split functions in the architecture depends on, among other things, factors related to radio network deployment scenarios, constraints and intended supported use cases. Key considerations include: the need to support a specific quality of service for each service offered and for real / non-real time applications; support of specific user density and load demand in a given geographical area; and available transport networks with different performance levels.

[0010] In 5G, core network entities comprise logical nodes (or 'functions') including control plane functions (CPFs) and one or more user plane functions (UPFs). The CPFs include, amongst other things, one or more Access and Mobility Management Functions (AMFs), a session management function (SMF), an Authentication Server Function (AUSF), a Unified Data Management (UDM) entity for managing user specific data, a Policy Control Function (PCF), an Application Function (AF), a Security Anchor Function (SEAF), an Authentication credential Repository and Processing Function (ARPF), and / or the like. The AMF generally corresponds to the mobility management entity (MME) in 4G and performs many of the functions performed by the MME. Each UPF combines functionality of both the S-GW and P-GW - specifically user plane functionality of the S-GW (SGW-U) and user plane functionality of the P-GW (PGW-U). The SMF provides session management functionality (that formed part of MME functionality in 4G). The SMF also combines the some of the functionality provided by the S-GW and P-GW - specifically control plane functionality of the S-GW (SGW-C) and control plane functionality of the P-GW (PGW-C). The SMF also allocates IP addresses to each UE.

[0011] Some recent developments in 3GPP relate to the use of AI and ML, often abbreviated to AI / ML. Predictions or inferences generated using an AI / ML model can be used as part of various methods for improving the reliability or efficiency of communication in the network. For example, AI / ML models can be used to predict the path of a UE based on previous mobility of the UE, used for cell and / or beam management, or used in methods of encoding and transmitting information. An AI / ML model may be hosted at a RAN node, and the RAN node may perform control of communication resources or control related to the status of a UE (e.g., control of UE mobility, or control of a radio resource control (RRC) state of the UE) based on an inference (e.g., determination or prediction) generated using the AI / ML model. The RAN node may also transmit an inference generated using the model to another node in the network, for use at the other node.

[0012] Alternatively, an AI / ML model may be hosted at two nodes of the network, for example at a RAN node and at a UE. In this case, the RAN node and the UE may both make determinations or predictions using the model. For example, the UE may use the model as part of an encoding process for encoding (and / or compressing) channel state information (CSI) for transmission to the RAN node, and the RAN node may use the same model as part of a corresponding decoding (and / or decompression) process for decoding the CSI received from the UE.

[0013] It will be appreciated that the use of AI / ML models may also be extended to other procedures and methods performed in a communication system to further improve the reliability or efficiency of communication in the network. For example, AI / ML models may be extended to generate inferences about the positions of UEs in a communication system using, for example, UE-assisted information and / or location management function (LMF)-based positioning assistance, and / or RAN node-based positioning assistance to facilitate life cycle management (LCM) operations specific to positioning accuracy enhancements in the communication system.

[0014] An AI / ML data collection entity in the network may provide AI / ML training data to an AI / ML model training function. The collected data may be, for example, CSI data such as one or more model dataset comprising N samples of target CSI (e.g., CSI-RSs) and / or CSI feedback corresponding to the input and output of a nominal encoder respectively of a node and / or UE. In another example, the collected data may comprise model parameters corresponding to a nominal encoder of a node and / or UE, while in yet another example, the collected data may be model parameters corresponding to a nominal encoder of a node and / or UE, along with a model dataset comprising N samples of target CSI (e.g., CSI-RSs). Alternatively, the collected data may be any other suitable data related to the training of the AI / ML model.

[0015] In the examples given above, collected data comprising a model dataset comprising N samples of target CSI (e.g., CSI-RSs) and / or CSI feedback corresponding to the input and output of a nominal encoder respectively of a node and / or UE typically has a size of around 225 megabytes (MBs). Collected data comprising model parameters corresponding to a nominal encoder of a node and / or UE typically has a size of around 11.6 MBs. Collected data comprising model parameters corresponding to a nominal encoder of a node and / or UE, along with a model dataset comprising N samples of target CSI (e.g., CSI-RSs) typically has a size of around 11.6 MBs plus the size of the N samples.

[0016] Accordingly, the model parameters and / or training data for the AI / ML model typically has a large data size, and so the collection and transmission of such data in the network must be carefully considered to avoid congestion of the network or inefficient use of the available communication resources.

[0017] There are a number of possible options for data collection for two-sided AI / ML models. For example, in an 'non-Over-The-Air' (non-OTA) approach, a first termination entity in the network for transmission of the network-sided AI / ML data for use in the training of a UE-side model may be provided in a RAN node, the core network, or inside an operations and maintenance (OAM) domain which may receive the network-sided AI / ML data for the two-sided model from a node (e.g., RAN node) of the network that hosts and trains a network-sided AI / ML model of the two-sided AI / ML model. In these options, a control plane tunnel could be used to transmit the network-sided AI / ML data, and either the NAS layer or RRC layer could be used. That first termination entity in the network for transmission of the network-sided AI / ML data for the two-side model may then transfer the network-sided AI / ML data to a UE-side training entity for the model training of the UE-sided part of the two-sided AI / ML model. That UE-side training entity may, for example, be a server inside a mobile network operator (MNO) network, or an over-the-top (OTT) server, or another server for data collection and UE-sided AI / ML model training.

[0018] In another example, in an OTA approach, a first termination entity in the network for transmission of the network-sided AI / ML data for use in the training of a UE-side model may be provided in a RAN node, the core network, or inside an OAM domain (if needed) which may receive the network-sided AI / ML data for the two-sided model from a node (e.g., RAN node) of the network that hosts and trains a network-sided AI / ML model of the two-sided AI / ML model. In these options, a control plane tunnel could be used to transmit the network-sided AI / ML data, and either the NAS layer or RRC layer could be used. That first termination entity in the network for transmission of the network-sided AI / ML data for the two-side model may then transfer the network-sided AI / ML data to a UE that subsequently transfers the network-sided AI / ML data to a UE-side training entity for the model training of the UE-sided part of the two-sided AI / ML model. That UE-side training entity may, for example, be a server inside an MNO network, or OTT server, or another server for data collection and UE-sided AI / ML model training.

[0019] Alternatively, in an OTA approach, the network-sided AI / ML data for use in the training of a UE-side model may be transmitted directly by a node (e.g., RAN node) of the network that hosts and trains a network-sided AI / ML model of the two-sided AI / ML model to a node (e.g., RAN node) serving a UE, for transfer to the UE, and then subsequent transfer to a UE-side training entity for the model training of the UE-sided part of the two-sided AI / ML model. That UE-side training entity may, for example, be a server inside an MNO network, or an OTT server, or another server for data collection and UE-sided AI / ML model training.

[0020] In the case of the OTA scenarios described above, a RAN node of the network may send the network-sided AI / ML data for use in the training of a UE-side model and / or model parameters directly to the UE for subsequent forwarding to a UE-side training entity via the control plane (CP) between the RAN node of the network and the UE (Scenario #1). Alternatively, the core network of the network may send the network-sided AI / ML data for use in the training of a UE-side model and / or model parameters to the UE via a RAN node (Scenario #2). Alternatively, the OAM of the network may send the network-sided AI / ML data for use in the training of a UE-side model and / or model parameters to the UE via a RAN node (Scenario #3).

[0021] NPL 1: 'NGMN 5G White Paper' V1.0 by the Next Generation Mobile Networks (NGMN), available from https: / / www.ngmn.org / 5g-white-paper.html.

[0022] In all of the scenarios described above however, issues persist. For example, the large amount of data and / or model parameters to be transferred between the RAN node and the UE over the Uu interface may have a significant impact on the signalling overhead and resource utilisation associated with the Uu interface, which may in turn result in congestion and / or increased power consumption.

[0023] Furthermore, given that a UE is typically configured to support a data buffer size of around 35 kB in, for example, scenario #1 where the RAN node may send the network-sided AI / ML data and / or model parameters directly to the UE for subsequent forwarding to a UE-side training entity via the CP, issues may arise where the amount of data and / or model parameters to be sent to the UE over the CP is larger than the amount of data that can be supported by the data buffer size of the UE.

[0024] Whilst the transmission of such large amounts of data and / or model parameters may typically be robustly facilitated through functions such as data segmentation and data retransmission typically performed by the RRC layer of the CP, such overloading of the CP with large model datasets may lead to undesirable disruptions in core control message transmission resulting in service degradation, reliability issues, and the like.

[0025] Challenges to support end-to-end (E2E) reliability also need careful consideration in the context of data sharing for two-sided AI / ML models as the data and / or model parameters may need to be shared between a plurality of different RAN nodes during UE mobility and during different RRC station transitions.

[0026] There is therefore a need for enhanced procedures for control plane-based data / model parameter information sharing between RAN nodes and UEs to reduce the impact on the RRC layer of the control plane between the RAN nodes and UEs, in the context of control plane-based data / model parameter information sharing for the training and use of two-sided AI / ML models.

[0027] The present specification aims to disclose apparatus and methods that at least contribute to addressing one or more of the above needs and / or issues.

[0028] The disclosure has a method performed by a user equipment, UE, the method comprising transmitting to, or receiving from, a radio access network, RAN, node, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; and receiving, as part of the current session, from the RAN node, at least one message for transferring the information related to the first AI / ML model to the UE, wherein the initial message, and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.

[0029] The disclosure has a method performed by radio access network, RAN, node, the method comprising receiving from, or transmitting to, a user equipment, UE, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; and transmitting, as part of the current session, to the UE, at least one message for transferring the information related to the first AI / ML model to the UE, wherein the initial message and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.

[0030] The disclosure has a user equipment, UE, comprising means for transmitting to, or receiving from, a radio access network, RAN, node, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; and means for receiving, as part of the current session, from the RAN node, at least one message for transferring the information related to the first AI / ML model to the UE, wherein the initial message and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.

[0031] The disclosure has a radio access network, RAN, node, comprising means for receiving from, or transmitting to, a user equipment, UE, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; and means for transmitting, as part of the current session, to the UE, at least one message for transferring the information related to the first AI / ML model to the UE, wherein the initial message and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.

[0032] Examples of apparatus and methods will now be described, by way of example, with reference to the accompanying drawings in which:

[0033] Fig. 1 schematically illustrates a mobile ('cellular' or 'wireless') communication system;Fig. 2 illustrates a functional framework for AI / ML models, and how various entities of the framework may interact with one another, that may be implemented in the communication system of Fig. 1;Fig. 3 schematically illustrates a method of training an AI / ML model, and of monitoring the performance of the AI / ML model, which may be implemented in the communication system of Fig. 1;Fig. 4 illustrates an example of a protocol stack architecture for AI / ML model, model dataset, and / or model parameter transfer, which may be implemented in the communication system of Fig. 1;Fig. 5 illustrates another example of a protocol stack architecture for AI / ML model, model dataset, and / or model parameter transfer, which may be implemented in the communication system of Fig. 1;Fig. 6 illustrates an example processing of AI / ML adaption layer (AAL) information bits into an AAL Data protocol data unit (PDU);Fig. 7 illustrates an example sequence diagram of a general AAL transmission procedure (RRC Connection-based) that may be implemented in the communication system 1;Fig. 8 illustrates an example sequence diagram of an adapted version of the transmission procedure of Fig. 7 that supports message retransmissions;Fig. 9 illustrates an example broadcasted message structure that may be used for AAL transmission procedures with RRC connectionless-based transmission;Fig. 10 illustrates another example sequence diagram of a general AAL transmission procedure (RRC Connectionless) that may be implemented in the communication system 1;Fig. 11 is a schematic block diagram illustrating the main components of a UE 3 as shown in Fig. 1; andFig. 12 is a schematic block diagram illustrating the main components of a RAN node 5 for the communication system 1 shown in Fig. 1.

[0034] <Overview>  An exemplary communication system will now be described in general terms, by way of example only, with reference to Fig. 1.

[0035] Fig. 1 schematically illustrates a mobile ('cellular' or 'wireless') communication system 1 to which the examples described herein are applicable.

[0036] In the communication system 1, user equipments (UEs) 3 (3-1, 3-2, 3-3) (e.g., mobile telephones and / or other mobile or stationary devices) can communicate with each other via a (radio) access network ((R)AN) node 5 that operates according to one or more compatible radio access technologies (RATs). In the illustrated example, the RAN node 5 comprises a base station 5 or 'gNB' 5 operating one or more associated cells. Communication via the RAN node 5 is typically routed through a core network 7 (e.g., a 5G / 6G core network, evolved packet core network (EPC), or any other generations core network).

[0037] As those skilled in the art will appreciate, whilst three UEs 3 and one RAN node 5 are shown in Fig. 2 for illustration purposes, the system, when implemented, will typically include other RAN nodes 5 and UEs 3.

[0038] Each RAN node 5 controls one or more associated cells either directly, or indirectly via one or more other nodes (such as home base stations, relays, remote radio heads, distributed units, and / or the like). It will be appreciated that the RAN nodes 5 may be configured to support 4G, 5G, 6G and / or later generations, and / or any other 3GPP or non-3GPP communication protocols.

[0039] The UEs 3 and their serving RAN node 5 are connected via an appropriate air interface (for example the so-called 'Uu' interface and / or the like). Neighbouring RAN nodes 5 may be connected to each other via an appropriate RAN node to RAN node interface (such as the so-called 'X2' interface, 'Xn' interface, and / or the like).

[0040] The core network 7 includes a number of logical nodes (or 'functions') for supporting communication in the communication system 1. In this example, the core network 7 comprises control plane functions (CPFs) 10 and one or more network node entities for the communication of user data (e.g., user plane functions (UPFs) 11). The CPFs 10 include one or more network node entities for the communication of control signalling (e.g., Access and Mobility Management Functions (AMFs) 10-1), one or more network node entities for session management (e.g., Session Management Functions (SMFs) 10-2) and a number of other functions 10-n (such as, for example an Authentication Server Function (AUSF) which facilitates security processes).

[0041] The communication system 1 also includes an Operations, Administration and Maintenance (OAM) 14 comprising one or more OAM functions for provisioning and managing network or elements within the wider communication system 1. The OAM 14 may be responsible for the storage and analysis of some radio-related measurements and may perform some data analytics functions including some RAN analytics. The OAM 14 may, for example, communicate with one or more of the CPFs 10.

[0042] The RAN node 5 is connected to the core network nodes via appropriate interfaces (or 'reference points') such as an N2 reference point between the RAN node 5 and the AMF 10-1 for the communication of control signalling, and an appropriate reference point (e.g., reference point N3) between the RAN node 5 and each UPF 11 for the communication of user data. The UEs 3 are each connected to the AMF 10-1 via a non-access stratum (NAS) connection over an appropriate reference point (e.g. N1 reference point (analogous to the S1 reference point in LTE)). It will be appreciated that N1 communication are routed transparently via the RAN node 5.

[0043] Each UPF 11 is connected to an external data network 20 (e.g., an IP network such as the internet) via an appropriate reference point (e.g., N6 reference point) for communication of the user data.

[0044] The AMF 10-1 performs mobility management related functions, maintains the NAS connection with each UE 3 and manages UE registration. The AMF 10-1 is also responsible for managing paging. The AMF 10-1 receives user information sent through the network and forwards the information to the SMF 10-2.

[0045] The SMF 10-2 is connected to the AMF 10-1 via an appropriate reference point (e.g., N11 reference point). The SMF 10-2 provides session management functionality (that formed part of MME functionality in LTE) and additionally combines some control plane functions (provided by the serving gateway and packet data network gateway in LTE). The SMF 10-2 uses user information provided via the AMF 10-1 to determine what session manager would be best assigned to the user. The SMF 10-2 may be considered effectively to be a gateway from the user plane to the control plane of the network. The SMF 10-2 also allocates IP addresses to each UE 3.

[0046] The RAN node 5 of the communication system 1 is configured to operate at least one cell on an associated time-division duplex (TDD) carrier that operates in unpaired spectrum and / or at least one cell on an associated frequency-division duplex (FDD) carrier that operates in paired spectrum.

[0047] The RAN node 5 is also configured for transmission of, and the UEs 3 are configured for the reception of, control information and user data via a number of downlink (DL) physical channels and for transmission of a number of physical signals. The DL physical channels correspond to resource elements (REs) carrying information originated from a higher layer, and the DL physical signals are used in the physical layer and correspond to REs which do not carry information originated from a higher layer.

[0048] The DL physical channels may include, for example, a physical downlink shared channel (PDSCH), a physical broadcast channel (PBCH), and a physical downlink control channel (PDCCH). The PDSCH carries data sharing the PDSCH's capacity on a time and frequency basis. The PDSCH can carry a variety of items of data including, for example, user data, UE-specific higher layer control messages mapped down from higher channels, system information blocks (SIBs), and paging. The PDCCH carries downlink control information (DCI) for supporting a number of functions including, for example, scheduling the downlink transmissions on the PDSCH and also the uplink data transmissions on a physical uplink shared channel (PUSCH). The PBCH provides UEs 3 with the Master Information Block (MIB). It also, in conjunction with the PDCCH, supports the synchronisation of time and frequency, which aids cell acquisition, selection and re-selection.

[0049] The RAN node 5 also transmits DL physical signals that do not carry any data, such as, for example, reference signals (RSs) and synchronisation signals (SSs). A reference signal (sometimes known as a pilot signal) is a signal with a predefined special waveform known to both the UE 3 and the RAN node 5. The reference signals may include, for example, cell specific reference signals, UE-specific reference signal (UE-RS), downlink demodulation signals (DMRSs), and channel state information reference signal (CSI-RS).

[0050] Similarly, the UEs 3 are configured for transmission of, and the RAN node 5 is configured for the reception of, control information and user data via a number of uplink (UL) physical channels corresponding to REs carrying information originated from a higher layer, and UL physical signals which are used in the physical layer and correspond to REs which do not carry information originated from a higher layer. The physical channels may include, for example, the PUSCH, a physical uplink control channel (PUCCH), and / or a physical random-access channel (which may be referred to either as a 'PRACH' or a 'RACH' - the term (P)RACH will be used herein). The UL physical signals may include, for example, DMRSs for an UL control / data signal, and / or sounding reference signals (SRSs) used for UL channel measurement.

[0051] The RAN node 5 is also configured to transmit synchronisation signal (SS) / physical broadcast channel (PBCH) blocks (SSBs) periodically in the cell or cells that it operates. The SSB includes both synchronisation signals (e.g., a primary synchronisation signal (PSS) and a secondary synchronisation signal (SSS)) and the PBCH carrying a master information block (MIB) that provides at least part of the minimum system information for accessing the corresponding cell (e.g., parameters required for acquiring system information block 1 (SIB1) which carries other minimum system information).

[0052] Each UE 3 may receive an SSB, and the UE 3 may assume that reception occasions of a PBCH, PSS and SSS are in consecutive symbols and form the SSB (also referred to as a SS / PBCH block). The PSS is a part of the SSB that aids in initial cell detection and synchronisation. It provides coarse timing and frequency synchronisation for UEs 3. The PSS consists of a predefined sequence of complex-valued symbols transmitted over a specific frequency range. The SSS is another component of the SSB that provides additional information for fine-grained synchronisation and cell identification. It carries the cell identity group and provides the necessary information to determine the exact physical cell ID (PCI) of the serving cell (e.g., a unique identifier assigned to each cell within the network that helps UEs 3 differentiate between neighbouring cells and synchronize with the correct cell.

[0053] The RAN node 5 may transmit several SSBs corresponding to different DL beams. The total number of SSBs may be confined, for example, within a 5 ms duration as an SS burst. The periodicity of the SSB transmissions may be indicated to the UE 3 using any suitable signalling (e.g., per serving cell using ssb-periodicityServingCell). The periodicity value for the SSB may be, for example, greater than or equal to 20 ms. For initial cell selection, the UE 3 may be configured to assume that an SS burst occurs with a periodicity of 2 frames. The UE 3 may also be provided with an indication of which SSBs within a 5 ms duration are transmitted (e.g., using ssb-PositionsInBurst). The UE 3 may also be provided with an indication of an absolute transmit power value of the SSS (e.g., using ss-PBCH-BlockPower) which may range from -60 to 50 dBm. Furthermore, the UE 3 may also be provided with an indication of the subcarrier spacing (SCS) used for the SSB (e.g., using ssbSubcarrierSpacing). It will be appreciated that all indications to the UE 3 about the SSB may be indicated to the UE 3 via any appropriate information element (IE) (e.g., ServingCellConfigCommon provided in a SIB1 message, or any other appropriate dedicated message).

[0054] Each UE 3 is configured to search for SSBs when scanning for a cell to camp on and to decode the associated PBCH before proceeding to decode other system information transmitted on the PDSCH. Each UE 3 is also configured to perform measurements on specific resources configured for the SSBs, for example reference signal received power (RSRP), reference signal received quality (RSRQ), and / or signal to interference and noise ratio (SINR) measurements or the like. Each UE 3 may also be configured to perform radio resource management (RRM) measurements when triggered to do so by a RAN node 5.

[0055] <AI / ML>  The communication system 1 supports the use of artificial intelligence (AI) and machine learning (ML), often abbreviated to AI / ML in accordance with recent developments in cellular communication technology (e.g., as part of the work of the 3GPP) that those skilled in the art will be familiar with. These AI / ML features make use of trained AI / ML models to make one or more predictions or inferences, from a set of one or more input vectors, which can be used in the network (e.g., for improving the reliability or efficiency of communication in the network).

[0056] In respect of the communication system 1, for example, AI / ML models could potentially be trained and used for predicting the path of a UE 3 based on previous mobility of the UE 3, used for beam management, or used in methods of encoding and transmitting information. An AI / ML model may be hosted at a RAN node 5 (or any other suitable network node), and the RAN node 5 may perform control of communication resources for UEs 3 it serves, and / or perform control related to the status of a UE 3 (e.g. control of UE mobility, or control of a radio resource control, RRC, state of the UE 3) based on an inference (e.g. determination or prediction) generated using the AI / ML model. The RAN node 5 may also transmit an inference generated using the model to another node in the network, for use at the other node. An AI / ML model may also be hosted the UE 3, or at a plurality of locations within the network, for example at both the RAN node 5 and at the UE 3. For example, the RAN node 5 and the UE 3 may both make determinations and / or predictions using the same model or different models.

[0057] The support for such AI / ML features may involve different levels of collaboration between the network (RAN node 5 and / or core network 7) and a UE 3 served by the network when deploying and using such AI / ML features. For example, three possible 'network-UE collaboration levels' that may be supported are:  Level x: Involving no collaboration between the network and the UE 3. Specifically, level x is an implementation-based AI / ML operation without any dedicated AI / ML-specific enhancement.  Level y: Signalling-based collaboration without AI / ML model transfer. For example, this level is applicable when model training is performed offline, and models are registered to both a RAN node 5 and the UE 3. Here, the RAN node 5 and the UE 3 are aware of available models (before operation), and the RAN node 5 is only required to activate / deactivate the models residing at the UE 3 when needed.  Level z: Signalling-based collaboration with AI / ML model transfer (e.g., where an AI / ML model is transferred to the UE 3 when needed.

[0058] The AI / ML model types that are supported in the communication system 1 may include, for example:  Single-sided model: A single-sided AI / ML model is an AI / ML model that is deployed (hosted) only at the UE side or at the network side. For example, an AI / ML model may be hosted (stored, for generating inferences) at a UE 3, RAN node 5 , or a central entity of the communication system 1, or an operations, administration, and maintenance (OAM) / over-the-top (OTT) server. When the AI / ML model is used at a UE 3, a RAN node 5, a central entity of the communication system 1, or an OAM / OTT server only, the AI / ML model may be referred to as a 'single-sided' model. An example of this type of 'single-sided' model is an AI / ML model for beam prediction in time, which can be deployed at the UE side.  However, even when the model is a single-sided model, it will be appreciated that the model need not necessarily be trained at the node at which it is deployed (e.g., a UE 3 or a RAN node 5). For example, the model could be trained at the RAN node 5 (or at another node in the network such as a core network node / function - e.g., a central entity of the communication system 1, or an OAM / OTT server - and then transferred to the UE 3 for use at the UE 3.  Two-sided model: A 'two-sided' model is an AI / ML model (or model pair) that has one AI / ML model hosted at one node (e.g., the UE 3), and a corresponding AI / ML model hosted at another node (e.g., a RAN node 5) - it will be appreciated that any pair of network nodes may be used. Such a two-sided model may also be referred to as a 'paired' AI / ML model. An inference using a two-sided model is performed jointly across the nodes at which the AI / ML models of the two-sided model are deployed. The joint inference may comprise, for example, a first part of the inference being performed at one node (e.g., the UE 3 or RAN node 5), and then the remaining part may be performed by the other (e.g., the RAN node 5 or UE 3). It will be appreciated that whilst the AI / ML model hosted at the different nodes may be the same AI / ML model, they need not necessarily be the same model. One example of this type of model is, for example only, channel state information (CSI) compression, where the UE performs CSI compression and network performs CSI decompression. As with the single-sided model case, the two-sided model (or models) may be trained at any suitable network node, and then transmitted to the UE 3 and the RAN node 5 (or other respective node or nodes).

[0059] A general discussion of how AI / ML may be implemented in the communication system 1 will now be provided, by way of example only, with reference to Figs. 2 and 3.

[0060] Fig. 2 illustrates a functional framework for AI / ML models, and how various entities of the framework may interact with one another, that may be implemented in the communication system 1.

[0061] The entities include a data collection entity 241, a model training function 243, a model inference function 245, an actor 247, a management function 249, and a model storage entity 251.

[0062] The model storage entity 251 may be a reference point for protocol terminations for model transfer and delivery. The AI / ML models could be stored at any suitable node in the network.

[0063] The data collection entity 241 provides training data to the model training function 243, inference data to the model inference function 245, and monitoring data to the management function 249. The collected data may be, for example, data regarding mobility (e.g., handover of a UE 3, or a location of the UE 3). The data may be obtained, for example, by a UE 3 or a RAN node 5 (e.g., by receiving a measurement report from a UE 3, or by receiving data from another RAN node 5 or a core network node / function) and transmitted to another RAN node 5 or core network node that generates the AI / ML model inference output (or alternatively, the same RAN node 5 that obtains the data may generate the AI / ML model output).

[0064] The model training function 243 performs the AI / ML model training, validation, and testing, and may generate model performance metrics as part of a model testing procedure. The model training function 243 may output a trained AI / ML model to the model storage entity 251 (though it will be appreciated that the output model may be stored at locations other than model storage entity 251).

[0065] The model inference function 245 provides AI / ML model inference output (e.g., predictions or decisions), and the actor 247 is a function or node that receives the output from the model inference function 245 and triggers or performs corresponding actions (e.g., a RAN node 5 that increases / reduces its transmit power or initiates a handover procedure for a UE 3). The AI / ML model inference output may be, for example, a prediction of mobility (e.g., expected path, route or trajectory, inter-cell, or inter-beam mobility, or expected handover) of the UE 3, or one or more model parameters for use in encoding or decoding transmissions between the RAN node 5 and the UE 3. The model inference function 245 may receive an AI / ML model from the model storage entity 251, and inference data from the data collection entity 241 for use with the AI / ML model. The model inference function 245 may also output monitoring data for use at the management function 249 and receive information indicating an AI / ML to activate or deactivate from the management function 249.

[0066] The management function 249 receives monitoring data from the data collection entity 241 and may also receive monitoring data from the model inference function 245. The management function 249 may transmit to the model storage entity 251, an indication of an AI / ML model to be transmitted for use at the model inference function 245. The management function 249 may also transmit to the model training function 243, performance feedback or a retraining request for the AI / ML model.

[0067] The functions and entities illustrated in Fig. 2 may be co-located at a single node of the communication system 1 (e.g., at the RAN node 5 or core network node / function) or may be distributed amongst a plurality of network nodes (e.g., a plurality of the RAN nodes 5).

[0068] By way of example only, terms referred to by 3GPP in the context of this framework include:  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. Model training can be performed offline or online or combination of both.  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, which 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.  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.  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.  Model monitoring: A procedure that monitors the inference performance of the AI / ML model.  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.  Supervised learning: A process of training a model from input and its corresponding labels.  Unsupervised leaning: 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 (also referred to as 'state') and a feedback signal (also referred to as 'reward') resulting from the model's output (also referred to as 'action') in an environment the model is interacting with.

[0069] The data collection by the data collection entity 241 may be performed at various nodes of the communication system 1 (e.g., at one or more RAN nodes 5 or UEs 3). Particularly advantageous methods of obtaining, at the UE 3, data for an AI / ML model, and transmitting the AI / ML data from the UE 3 to the RAN node 5, will be described in more detail later.

[0070] Fig. 3 schematically illustrates of a method of training an AI / ML model, and of monitoring the performance of the AI / ML model. As illustrated in Fig. 3, stored data / features may first be extracted in a data extraction step. In the data validation step, a determination of whether to proceed with training or retraining the AI / ML model is made (e.g., based on the extracted data). In the data preparation stage, the data is prepared for use in training the AI / ML model. For example, the data may be cleaned (e.g., filtered), subject to a transformation, or modified in any other suitable manner. The data may also be divided in training data, validation data and test data sets in the data preparation stage.

[0071] In the model training step, the AI / ML model is trained (or retrained) using training data prepared in the data preparation step. It will be appreciated that any suitable training method can be used to train the AI / ML model (e.g., a method that comprises supervised learning or unsupervised learning). In the model evaluation step, the AI / ML model is evaluated (e.g., a prediction accuracy of the AI / ML model is evaluated) using a test data set (which may be generated in the data preparation step). In the model validation step, a determination of whether the AI / ML model is suitable for deployment in the communication system 1 is made (e.g., based on the results of the model evaluation step).

[0072] In the model serving step, the AI / ML model is deployed for use in the communication system 1. AI / ML model deployment may comprise compiling a trained AI / ML model, packaging the model into an executable format, and delivering the AI / ML model to a target device. For example, the AI / ML model may be transmitted to the RAN node 5 and / or the UE 3, for use at the RAN node 5 and / or the UE 3 to generate predictions or determinations using the AI / ML model as part of a prediction service. In the performance monitoring step, the performance of the deployed AI / ML model is monitored. The predictive performance of the AI / ML model may be monitored by comparing predictions generated using the model with one or more measurements. For example, when the AI / ML model is used to predict a location of a UE 3, the prediction accuracy of the AI / ML model may be assessed using a measurement of an actual location of the UE 3. If the AI / ML model is used for predicting future measurement results (e.g., the measured Reference Signal Received Power (RSRP) of reference signals) at some point in time, the prediction accuracy of the AI / ML model may be assessed using actual measurement results acquired by the UE 3 when that point in time is reached. If the AI / ML model is used for determining model parameters for use in encoding and decoding data transmitted between a RAN node 5 and a UE 3, the model may be assessed based on the performance of the encoding and / or decoding processes. In the retraining trigger step, retraining of the AI / ML model is triggered (e.g., because the prediction accuracy of the AI / ML model has fallen below an acceptable threshold accuracy, or because a performance of a method that uses inferences from the AI / ML model has fallen below an acceptable threshold performance), and the method returns to the data extraction step.

[0073] Each step of the method of Fig. 3 may be executed at a single node of the communication system 1 (including at the RAN node 5), or alternatively steps of the method may be distributed between a plurality of different nodes (or indeed one or more of these steps may be performed online or offline).

[0074] As discussed above, information collected by nodes / functions in the communication system 1 (e.g., at a UE 3 and / or RAN nodes 5) can be used as training data for an AI / ML model and used as inference data for use in generating one or more model inferences using the AI / ML model. The information used as training data, monitoring data, and / or to generate the one or more model inferences may be referred to as 'AI / ML information' or 'AI / ML data.'

[0075] <Configuration information for AI / ML>  Configuration information for an AI / ML model (which may be referred to as "AI / ML configuration information") may be exchanged between nodes in the communication system. For example, a core network node may transmit AI / ML configuration information to a RAN node 5 (or any other entity of the network that supports an AI / ML-based prediction functionality) that hosts an AI / ML model. The AI / ML configuration information may include a list of supported use cases for the AI / ML model (the AI / ML model need not necessarily be for predicting UE mobility). The supported use cases may be, for example: energy saving; traffic steering; anomaly detection; quality of experience (QoE) optimisation; mobility robustness optimisation (MRO); RAN slice service level agreement (SLA) assurance; massive multiple-input multiple-output (MIMO) beamforming optimisation; network slice subnet instance (NSSI) resource allocation; optimisation coverage and capacity optimisation (CCO); mobility load balancing (MLB); RACH optimisation; or UE transmission power optimisation. The AI / ML configuration information may include an indication of a particular AI / ML model to use for a particular use case. The AI / ML configuration information may also include an indication of whether feedback is required (e.g., from another network node). The feedback may include, for example, communication performance feedback (e.g., indicating a communication performance for communication between a UE 3 and a RAN node 5).

[0076] <Data / Parameter Sharing for Two-Sided Models>  As described above, an AI / ML model may be implemented as a two-sided AI / ML model (or a model pair) that has one AI / ML model hosted at one node (e.g., the UE 3), and a corresponding AI / ML model hosted at another node (e.g., a RAN node 5). An inference using a two-sided AI / ML model is performed jointly across the nodes at which the AI / ML models of the two-sided AI / ML model are deployed. With such two-sided AI / ML models, as with the single-sided AI / ML model case, the two-sided AI / ML model (or models) may be trained at any suitable network node, and then the trained AI / ML model (e.g., the model parameters of the trained AI / ML model) transmitted to the UE 3 and the RAN node 5 (or other respective node or nodes).

[0077] Nevertheless, it will be appreciated that, the two-sided AI / ML model (or models) may be deployed at the UE 3 and RAN node 5 at the outset, and each AI / ML model of the two-sided AI / ML model may be trained at the UE 3 and RAN node 5 respectively, based on model data shared between the UE 3 and RAN node 5. For example, a first AI / ML model of the two-sided AI / ML model (e.g., at the RAN node 5) may be trained, and then appropriate data / model parameters may be sent by the RAN node 5 to the UE 3 (e.g., directly, or indirectly via an appropriate other node) for training a second AI / ML model of the two-sided model (e.g., at the UE 3).

[0078] Moreover, each AI / ML model of the two-sided AI / ML model may be deployed / trained at the RAN node 5 (or the UE 3) initially, and may then be distributed to the UE 3 (or the RAN node 5) upon request for that two-sided AI / ML model from the UE 3 (or the RAN node 5). Similarly, each AI / ML model of the two-sided AI / ML model may be deployed / trained at the RAN node 5 (and / or the UE 3) initially and may then be distributed to another RAN node 5 (and / or another UE 3) upon request for that two-sided AI / ML model.

[0079] It will be appreciated that, each AI / ML model of the two-sided AI / ML model may be (further) trained locally at the UE 3 and / or RAN node 5 to which it was distributed. For example, a given AI / ML model of the two-sided AI / ML model may be provided to the UE 3 and later appropriate training data may be sent by the to the UE 3 (e.g., directly or via an appropriate other node) for (further) training the model of the two-sided model at the UE 3.

[0080] Beneficially, each UE 3 and RAN node 5 of the communication system 1 are mutually configured to support one or more enhancements for supporting AI / ML model, (e.g., AI / ML reference model), model dataset, and / or model parameter sharing between the RAN node 5 and the UE 3, e.g., in the context of the training and use of two-sided AI / ML models in the communication system 1.

[0081] Beneficially, for example, as described in more detail later, each UE 3 and RAN node 5 of the communication system 1 may be mutually configured to support one or more protocol stack architectures for control plane (CP)-based AI / ML reference model, model dataset, and / or model parameter sharing between RAN nodes 5 and UEs 3 in the context of the training and use of two-sided AI / ML models in the communication system 1.

[0082] Beneficially, for example, as described in more detail later, each UE 3 and RAN node 5 of the communication system 1 may be mutually configured to support one or more procedures for supporting control plane-based AI / ML model, model dataset, and / or model parameter transmission between RAN nodes 5 and UEs 3 in the context of the training and use of two-sided AI / ML models in the communication system 1.

[0083] Beneficially, for example, as described in more detail later, each UE 3 and RAN node 5 of the communication system 1 may be mutually configured to support one or more procedures for supporting control plane-based AI / ML model, model dataset, and / or model parameter broadcasting by RAN nodes 5 to UEs 3 in the context of the training and use of two-sided AI / ML models in the communication system 1.

[0084] <AI / ML Adaptation Layer (AAL) Transmission>  As mentioned above, each UE 3 and RAN node 5 of the communication system 1 may be mutually configured to support one or more protocol stack architectures for CP-based AI / ML model, model dataset, and / or model parameter sharing between RAN nodes 5 and UEs 3 in the context of the training and use of two-sided AI / ML models in the communication system 1.

[0085] A number of possible protocol stack architectures for supporting such transfer will now be described by way of example only.

[0086] <Protocol Stack Architectures for AAL Transmission>  Fig. 4 illustrates an example of a first protocol stack architecture for AI / ML model, model dataset, and / or model parameter transfer that may be implemented in the communication system 1.

[0087] As shown in Fig. 4, the protocol stack architecture comprises a UE protocol stack comprising an AI / ML Model (Dataset, Parameters) layer 4-2; an AI / ML adaption layer (AAL) 4-4; a radio resource control (RRC) layer 4-6; a packet data convergence protocol (PDCP) layer 4-8; a radio link control (RLC) layer 4-10, a medium access control (MAC) layer 4-12; and a physical (PHY) layer 4-14.

[0088] It will be appreciated that the UE 3 has at least one functional entity respectively corresponding to each layer for performing the functionality (data processing, data transfer / transmission etc.) of that layer.

[0089] The protocol stack architecture also comprises a RAN node protocol stack. The RAN node protocol stack includes a number of UE-side layers for communication with corresponding layers of the UE protocol stack over one or more associated interfaces. The RAN node protocol stack also includes (not shown) a number of core network-side layers for communication with the core network 7 over one or more associated interfaces.

[0090] The RAN node protocol stack comprises an AI / ML Model (Dataset, Parameters) layer 4-16; an AAL 4-18; an RRC layer 4-20; a PDCP layer 4-22; a RLC layer 4-24, a MAC layer 4-26; and a PHY layer 4-28.

[0091] It will be appreciated that the RAN node 5 has at least one functional entity respectively corresponding to each layer for performing the functionality (data processing, data transfer / transmission etc.) of that layer.

[0092] The purpose of the layers mentioned above will be understood by the skilled person and so are only briefly introduced below for the purposes of completeness:

[0093] AI / ML Model (Dataset, Parameters) layer 4-2, 4-16: Each AI / ML Model (Dataset, Parameters) layer entity of the AI / ML Model (Dataset, Parameters) layer 4-2, 4-16 of the UE 3 and the RAN node 5 may perform operations and functions associated with training, executing, and storing AI / ML models, including model datasets and model parameters used for such training and execution.

[0094] AI / ML Adaption layers 4-4, 4-18: Each AAL entity of the AALs 4-4, 4-18 of the UE 3 and the RAN node 5, perform operations and functions associated with the AAL such as the handling of the transmission of AI / ML model between the UE 3 and the RAN node 5, including the transmission therebetween of a reference AI / ML model, one or more model datasets (e.g., for training), one or more model parameters and the like). Each AAL entity of the AALs 4-4, 4-18 of the UE 3 and the RAN node 5, may also, where appropriate, perform operations and functions of data segmentation, data retransmission, and the like. For example, each AAL entity of the AALs 4-4, 4-18 is configured for receiving appropriate AAL data packets (e.g., AAL service data units (SDUs) from the layer above and AAL PDUs from the layer below), for processing those AAL data packets (AAL SDUs / AAL PDUs) appropriately as they pass through the AAL to form processed data packets (AAL PDUs / AAL SDUs) for subsequent transfer to the next layer as appropriate (e.g., to the layer below in the case of an AAL SDU or to the layer above in the case of an AAL PDU).

[0095] RRC layers 4-6, 4-20: Each RRC layer entity of the RRC layers 4-6, 4-20 of UE 3 and the RAN node 5, perform operations and functions associated with the RRC layer (i.e., part of the network layer) such as managing the establishment, modification, and release of connections between the UE 3 and the RAN node 5. This may include both initial connection setup and subsequent handovers. The RRC layer entities also control the establishment, modification, and release of radio bearers, the handling of the configuration and release RAN protocols, such as measurement reporting, mobility procedures, and beam management, as well as mobility management, radio resource management (RRM), and security control. The RRC layer entities also support traffic and signalling messages between the UE 3 and the RAN node 5 via a full signaling message-based data transmission path between the UE 3 and the RAN node 5.

[0096] PDCP layers 4-8, 4-22: Each PDCP layer entity of the PDCP layers 4-8, 4-22 of the UE 3 and RAN node 5 perform operations and functions associated with the PDCP layer (i.e., part of the data link layer) such as the transfer of user plane and control plane data, header compression, and ciphering and integrity protection procedures.

[0097] RLC layers 4-10, 4-24: Each RLC layer entity of the RLC layers 4-10, 4-24 of the UE 3 and RAN node 5, perform operations and functions associated with the RLC layers 4-10, 4-24 (i.e., part of the data link layer) such as transfer of upper layer PDUs, error corrections through automatic repeat request (ARQ), duplication detection, protocol error detection and recovery, and the like.

[0098] MAC layers 4-12, 4-26: Each MAC layer entity of the MAC layers 4-12, 4-26 of the UE 3 and RAN node 5, perform operations and functions associated with the MAC layers 4-12, 4-26 (i.e., part of the data link layer) such as ensuring reliable and efficient communication between two or more devices, and providing unique identification of each device.

[0099] PHY layers 4-14, 4-28: Each PHY layer entity of the PHY layers 4-14, 4-28 of the UE 3 and RAN node 5, perform operations and functions associated with the PHY layers 4-14, 4-28 such as establishing, maintaining, and deactivating the physical connection between the UE 3 and the RAN node 5, and the transmission of individual bits between the UE 3 and the RAN node 5.

[0100] As seen in Fig. 4, in this example, each AI / ML AAL 4-4, 4-18 may be a signalling layer located between the corresponding AI / ML model (Dataset, Parameters) layer 4-2, 4-16 and the corresponding RRC layer 4-6, 4-20. Accordingly, AAL PDUs are transferred to the RRC layer 4-6, 4-20 when formed from AAL SDUs (corresponding to PDUs transferred from the corresponding AI / ML model (Dataset, Parameters) layer 4-2, 4-16). Similarly, AAL PDUs (corresponding to RRC SDUs) are received from the RRC layer 4-6, 4-20 and processed to form AAL SDUs (for transfer to the corresponding AAL 4-4, 4-18).

[0101] It will be appreciated that the description above of the various layers and associated entities are by way of example only, and that those layers and associated entities may also be able to support performance of any other appropriate operations associated with their respective layers, as would be known by the skilled person.

[0102] Fig. 5 illustrates another example of a protocol stack architecture for signaling-based AI / ML model, model dataset, and / or model parameter transfer.

[0103] As shown in Fig. 5, the protocol stack architecture comprises a UE protocol stack comprising an AI / ML Model (Dataset, Parameters) layer 4-2; an AAL 4-4; a packet data PDCP layer 4-8; an RLC layer 4-10, a MAC layer 4-12; and an PHY layer 4-14.

[0104] It will be appreciated that the UE 3 has at least one functional entity respectively corresponding to each layer for performing the functionality (data processing, data transfer / transmission etc.) of that layer.

[0105] The protocol stack architecture also comprises an RAN node protocol stack. The RAN node protocol stack includes a number of UE-side layers for communication with corresponding layers of the UE protocol stack over one or more associated interfaces. The RAN node protocol stack also includes (not shown) a number of core network-side layers for communication with the core network 7 over one or more associated interfaces.

[0106] The RAN node protocol stack comprises an AI / ML Model (Dataset, Parameters) layer 4-16; an AAL 4-18; a PDCP layer 4-22; an RLC layer 4-24, a MAC layer 4-26; and a PHY layer 4-28.

[0107] It will be appreciated that the RAN node 5 has at least one functional entity respectively corresponding to each layer for performing the functionality (data processing, data transfer / transmission etc.) of that layer.

[0108] The purpose of the layers mentioned above, and functions of the associated entities is essentially as described above with reference to Fig. 4, as will be understood by the skilled person, and will not be repeated in its entirety.

[0109] For example, as described above, each AAL entity of the AALs 4-4, 4-18 of the UE 3 and the RAN node 5, may also, where appropriate, perform operations and functions of data segmentation, data retransmission, and the like. For example, each AAL entity of the AALs 4-4, 4-18 is configured for receiving appropriate AAL data packets (e.g., AAL service data units (SDUs) from the layer above and AAL PDUs from the layer below), for processing those AAL data packets (AAL SDUs / AAL PDUs) appropriately as they pass through the AAL to form processed data packets (AAL PDUs / AAL SDUs) for subsequent transfer to the next layer as appropriate (e.g., to the layer below in the case of an AAL SDU or to the layer above in the case of an AAL PDU).

[0110] However, in the example of Fig. 5, each AAL 4-4, 4-18 is a general protocol layer located between the corresponding AI / ML model (Dataset, Parameters) layer 4-2, 4-16 and the corresponding PDCP layer 4-8, 4-22. Accordingly, AAL PDUs are transferred to the PDCP layer 4-8, 4-22 when formed from AAL SDUs (corresponding to PDUs transferred from the corresponding AI / ML model (Dataset, Parameters) layer 4-2, 4-16). Similarly, AAL PDUs (corresponding to RRC SDUs) are received from the PDCP layer 4-8, 4-22 and processed to form AAL SDUs (for transfer to the corresponding AAL 4-4, 4-18).

[0111] There now follows a more detailed description of the AALs 4-4, 4-18 of the protocol stacks of Figs. 5 and 6, including the use of those AALs 4-4, 4-18 for AI / ML adaption layer-based signalling and transmissions between the UE 3 and the RAN node 5.

[0112] <AI / ML Adaption Layer (AAL)>  The sharing and transmission of AI / ML models (e.g., AI / ML reference models), model datasets for training those AI / ML models, and / or model parameters for configuring those AI / ML models between the UE 3 and the RAN node 5 may take place in accordance with either of the protocol stack architectures of Figs. 4 and 5.

[0113] Possible operations of the AAL entities of the AALs 4-4, 4-18 will now be described in more detail with reference specifically to operations at the AAL 4-18 of the RAN node 5. It will nevertheless be appreciated that the corresponding description applies in respect of the AAL entity of the AAL 4-4 of the UE 3 shown in the protocol stacks of Figs. 4 and 5.

[0114] For example, with reference to the protocol stack of Figs. 4 and 5, the AAL 4-18 of the RAN node 5 may receive AI / ML models (e.g., AI / ML reference models), model datasets for training those AI / ML models, and / or model parameters for configuring those AI / ML models from the AI / ML Model (Dataset, Parameters) layer 4-16 of the RAN node 5 in the form of an appropriate SDU (e.g., an AAL SDU, or the like) from the higher AI / ML model (dataset, parameters) layer 4-16 of the RAN node 5.

[0115] Fig. 6 illustrates an example processing of AAL information bits into an AAL Data PDU.

[0116] As shown in Fig. 6, upon receiving the AAL SDU, the AAL 4-18 of the RAN node 5 may segment the AAL SDU into a plurality of AAL SDU segments (e.g., AAL SDU segment #1…#N). For example, the AAL 4-18 of the RAN node 5 may receive the AAL SDU, and may then segment that AAL SDU into AAL SDU segments of a fixed size. One or more of the AAL SDU segments may be associated with a specific transaction that is used to carry a set of transmissions for AI / ML model transfer, model dataset sharing, AI / ML model parameter configurations, and the like. Accordingly, each AAL SDU segment may be marked by an independent segment number or sequence number associated with a specific transaction.

[0117] Each AAL SDU segment (e.g., AAL SDU segment #1) may then subsequently be respectively (re)assembled / packaged by the AAL 4-18 of the RAN node 5 into an appropriate AAL data PDU which may comprise an appropriate header (e.g., an AAL header) and a data payload (e.g., AAL data payload) that contains data of that AAL SDU segment (e.g., AAL SDU segment #1).

[0118] Each AAL data PDU (including the header and data payload) may be packaged by the AAL 4-18 of the RAN node 5 with a fixed size. For example, each AAL data PDU may have an appropriate (relatively large) fixed size (e.g., 45 k-bytes) and the AAL data PDU may be further segmented at lower layers. Alternatively the AAL data PDU may have a lower fixed size (e.g., 9 k-bytes) with no further segmentation being performed at lower layers. In this case, it will be appreciated that the last AAL data PDU packaged by the AAL 4-18 of the RAN node 5 may be packaged with padding where appropriate if the data in the last AAL SDU is less than the fixed size of the AAL data PDU.

[0119] Once packaged into AAL data PDUs, the AAL 4-18 of the RAN node 5 may then deliver each AAL data PDU to a lower layer (e.g., RRC layer 4-20 or PDCP layer 4-22 of the RAN node 5). The AAL 4-18 of the RAN node 5 may not support data PDU reordering to save the buffer at the RAN node 5 and / or the UE 3, in which case, the AAL 4-18 of the RAN node 5 may deliver each AAL data PDU to a lower layer (e.g., RRC layer 4-20 or PDCP layer 4-22 of the RAN node 5) sequentially in the order in which the AAL data PDUs are packaged up by the AAL 4-18 of the RAN node 5. It will be appreciated that in this case, the AAL 4-18 of the RAN node 5 may assume that reliable, in-sequence delivery can be supported by the lower layers (e.g. RRC layer 4-20 or PDCP layer 4-22) of the RAN node 5.

[0120] Alternatively, rather than the AAL 4-18 of the RAN node 5 segmenting the AAL SDU into a plurality of AAL SDU segments (e.g., AAL SDU segment #1…#N) as described above, a layer above the AAL 4-18 of the RAN node 5 may perform any segmentation that is required, and in this case one AAL SDU may be equal to one AAL data PDU (without including the header of the AAL PDU).

[0121] As mentioned above, the AAL 4-18 of the RAN node 5 may be a signalling layer be placed over the RRC layer 4-20 of the RAN node 5, or alternatively, the AAL 4-18 of the RAN node 5 may be a general protocol layer placed above PDCP layer 4-22 of the RAN node 5.

[0122] Whilst AAL data PDUs are described above, it will be appreciated that the AAL 4-18 of the RAN node 5 may support both the packaging and transmission of control PDUs and / or data PDUs. For example, when packaging a PDU, the AAL 4-18 of the RAN node 5 may include an appropriate header to the data payload of the AAL PDU that includes a field that indicates whether the AAL PDU is a control PDU or a data PDU.

[0123] AAL data PDUs packaged by the AAL 4-18 of the RAN node 5 may be used to carry (a portion of) the transmission for AI / ML reference model, model datasets for training that AI / ML model, model parameter configurations, and the like. The AAL data PDUs may include both a header and a message body (e.g., a data payload portion).

[0124] On the other hand AAL control PDUs packaged by the AAL 4-18 of the RAN node 5 may be used to help control transmissions (e.g., of data PDUs) e.g. by providing acknowledgements (or negative acknowledgements) for data PDU transmissions, or the like. Moreover, an AAL control PDU may be used to send appropriate messages (e.g., an abort message, or the like), from the RAN node 5 to the UE 3. An AAL control PDU may include only a header, and no message body (e.g., no data payload portion).

[0125] Alternatively, AAL control PDU packaging and transmission may not be supported by the AAL 4-18 of the RAN node 5. For example, rather than supporting specific AAL control PDUs, AAL control PDUs may be implemented as a special AAL data PDU with only a header, and no data payload.

[0126] When the AAL 4-18 of the RAN node 5 is placed over the RRC layer 4-20 of the RAN node 5 (as shown in Fig. 4), the AAL data PDUs (or AAL control PDUs) coming from the AAL 4-18 of the RAN node 5 may be carried by a signalling Radio bearer (SRB) within the RRC layer 4-20 of the RAN node 5, and the RRC layer 4-20 of the RAN node 5 may treat those AAL data PDUs (or AAL control PDUs) as a special type of signalling (e.g., a NAS PDU). Such SRB may, for example, be a new type of SRB, or alternatively it may be an SRB mapped to an existing SRB (e.g. SRB4) of the RRC layer 4-20 of the RAN node 5. The priority of that SRB may be (pre)defined or configured from the RAN node 5 to the UE 3. For example, the priority of that SRB may be (pre)defined or configured to be the lowest priority among all SRBs.

[0127] Alternatively, when the AAL 4-18 of the RAN node 5 is placed over the PDCP layer 4-22 of the RAN node 5 (as shown in Fig. 5), the AAL data PDU (or AAL control PDU) coming from the AAL 4-18 of the RAN node 5 may be mapped to a data radio bearer (DRB) carried by the PDCP entity of the PDCP layer 4-22 of the RAN node 5. The PDCP layer 4-22 of the RAN node 5 may, for example, treat the AAL data PDU (or AAL control PDU) as a special data packet generated by the RAN node 5. The priority of that DRB may be (pre)defined or configured from the RAN node 5 to the UE 3.

[0128] <RRC Connection-based Transmission>  The AAL entity of each AAL 4-4, 4-18 may be only established between the RAN node 5 and the UE 3 when the UE 3 is in an RRC connected state, and the AAL entity of each AAL 4-4, 4-18 may be released by the UE 3 and the RAN node 5 when the UE 3 leaves an RRC connected state. Accordingly, AAL transmission between the RAN node 5 and the UE 3 may be point to point transmission. In this scenario, transmissions between the RAN node 5 and the UE 3 are 'reliable' transmissions that can support acknowledgement, retransmission, and the like.

[0129] It will be appreciated that when the AAL entity of each AAL 4-4, 4-18 is released by the UE 3 and the RAN node 5 when the UE 3 leaves an RRC connected state, the UE 3 may flush, from its buffer, incomplete reception of the AI / ML reference model, model parameters, model datasets, and the like. However, the UE 3 may keep received portions of the AI / ML reference model, model parameters, model datasets, and the like, when the UE 3 is only temporarily out of coverage (e.g. RLF, Handover Failure), which may be controlled by a specific timer. For example, the UE 3 may only flush its buffer at the expiry of a timer that is triggered when the UE 3 leaves an RRC connected state. However, if the UE 3 enters an RRC connected state prior to that timer expiring the UE 3 may not flush its buffer.

[0130] <RRC Connectionless-based Transmission>  Conversely, the AAL entity of each AAL 4-4, 4-18 may be established between the RAN node 5 and the UE 3 irrespective of whether or not the UE 3 is in an RRC connected state. In this case, the AAL entity of each AAL 4-4, 4-18 is not released when the UE 3 leaves an RRC connected state.

[0131] There now follows a detailed description of the format and possible content of the AAL PDUs that may be supported for transmission.

[0132] <AAL PDU Header>  As described above with reference to Fig. 6, when generating AAL PDUs, the AAL 4-4, 4-18 may package an AAL SDU segment (e.g., AAL SDU segment #1) into an AAL PDU comprising an AAL PDU header and an AAL data payload. That AAL header of the AAL PDU may, for example, include one or more fields that provide specific header-type information to the lower layers that received that AAL PDU.

[0133] For example, in the case where the AAL PDU is a data PDU, the AAL PDU header may include: an AI / ML model identifier (ID) which indicates the AI / ML model that the current AAL PDU transmission is associated. That AI / ML model ID may be a global model ID, a local Model ID, or a cell ID plus a local model ID.

[0134] In the case where the AAL PDU is a data PDU, the AAL PDU header may include a transaction ID which indicates a purpose of current AAL PDU transmission (e.g., AI / ML model transfer, model dataset sharing, model parameter configuration, or the like).

[0135] In the case where the AAL PDU is a data PDU, the AAL PDU header may include a sequence number or segmentation number which indicates a particular sequence number or segmentation number associated with the AAL PDU transmission, which may be used to detect data PDU loss or duplicates within one transaction (i.e., within one set of AAL PDU transmissions).

[0136] In the case where the AAL PDU is a data PDU, the AAL PDU header may include a last transmission indication to indicate whether the current AAL PDU transmission is a last AAL PDU transmission within the transaction. For example, the last transmission indication may indicate when a transaction has ended (i.e., the last AAL PDU transmission of a transaction has ended). Additionally (or alternatively), the AAL PDU header may include the length of the padding bits or information bits used for the last AAL PDU transmission.

[0137] In the case where the AAL PDU is a data PDU, the AAL PDU header may include an acknowledgement flag to indicate whether or not the transmitter of the AAL PDU (e.g., RAN node 5) wishes the receiver of the AAL PDU (e.g., UE 3) to provide an acknowledgement of successful reception of the AAL PDU.

[0138] In the case where the AAL PDU is a control PDU, the AAL PDU header may include transaction ID which indicates a purpose of current AAL PDU transmission (e.g. feedback, abort transmissions, or the like).

[0139] In the case where the AAL PDU is a control PDU, the AAL PDU header may include a sequence number or segmentation number which Indicates a particular sequence number or segmentation number associated with an AAL data PDU transmission for which the AAL control PDU is providing feedback.

[0140] In the case where the AAL PDU is a control PDU, the AAL PDU header may include an acknowledgement bit to indicate acknowledgement or non-acknowledgement (or negative acknowledgement) of the reception of a data PDU. For example, an acknowledgement bit may be used by an AAL data PDU receiver by receiver (e.g., UE 3) to feedback its reception status of a data PDU from a sender (e.g., RAN node 5).

[0141] <AAL PDU Type and Content>  As mentioned above, when generating AAL PDUs, the AAL entity of the AAL 4-4, 4-18 may package an AAL SDU segment (e.g., AAL SDU segment #1) into an AAL PDU comprising an AAL PDU header and an AAL data payload. The contents of that AAL data payload of the AAL PDU may vary depending on: i) whether the AAL PDU is a control PDU or a data PDU, and ii) the purpose of the AAL PDU.

[0142] For example, where the AAL PDU is a data PDU, the AAL PDU may include a request for an AI / ML model. For example, where the AAL PDU is generated by the UE 3 to be sent to the RAN node 5, the AAL PDU may include a request for the RAN node 5 to provide an AI / ML model to the UE 3.

[0143] In another example, where the AAL PDU is a data PDU, the AAL PDU may include an AI / ML model. For example, where the AAL PDU is generated by the RAN node 5 to be sent to the UE 3, the AAL PDU may include an AI / ML model to be provided by the RAN node 5 to the UE 3.

[0144] In another example, where the AAL PDU is a data PDU, the AAL PDU may include a request for one or more model datasets for use with an AI / ML model (e.g., for training the AI / ML model). For example, where the AAL PDU is generated by the UE 3 to be sent to the RAN node 5, the AAL PDU may include a request for the RAN node 5 to provide one or more model datasets for use with an AI / ML model at the UE 3.

[0145] In another example, where the AAL PDU is a data PDU, the AAL PDU may include one or more model datasets for use with an AI / ML model (e.g., for training the AI / ML model). For example, where the AAL PDU is generated by the RAN node 5 to be sent to the UE 3, the AAL PDU may include one or more model datasets for use with an AI / ML model at the UE 3.

[0146] In another example, where the AAL PDU is a data PDU, the AAL PDU may include a request for one or more model parameters for AI / ML model configuration. For example, where the AAL PDU is generated by the UE 3 to be sent to the RAN node 5, the AAL PDU may include a request for the RAN node 5 to provide one or more model parameters for use by the UE 3 to configure an AI / ML model at the UE 3.

[0147] In another example, where the AAL PDU is a data PDU, the AAL PDU may include one or more model parameters for use in configuring an AI / ML model. For example, where the AAL PDU is generated by the RAN node 5 to be sent to the UE 3, the AAL PDU may include one or more model parameters for use by the UE 3 to configure an AI / ML model at the UE 3.

[0148] In yet another example, where the AAL PDU is a control PDU, the AAL PDU may include an appropriate indication of feedback to indicate whether or not a data PDU has been correctly received or not. For example, where the AAL PDU is generated by the UE 3 to be sent to the RAN node 5, the AAL PDU may include an indication of whether a previous data PDU sent by the RAN node 5 to the UE 3 was successfully received or not.

[0149] In yet another example, where the AAL PDU is a control PDU, the AAL PDU may include an appropriate abort indication to indicate that a transaction between the RAN node 5 and UE 3 is to be aborted. For example, where the AAL PDU is generated by the RAN node 5 to be sent to the UE 3, the AAL PDU may include an indication that a current transaction between the RAN node 5 and UE 3 is to be aborted.

[0150] <AAL Procedures - RRC Connection-based>  As mentioned above, each UE 3 and RAN node 5 of the communication system 1 may be mutually configured to support one or more procedures for supporting control plane-based AI / ML reference model, model dataset, and / or model parameter transmission between RAN nodes 5 and UEs 3 in the context of the training and use of two-sided AI / ML models in the communication system 1.

[0151] Such a procedure may, beneficially, be based on the principles of another procedure used for point-to-point communication between one entity (e.g., a server) and a target device (e.g., a UE 3). For example, the procedures may be based on existing techniques that use an appropriate positioning protocol (e.g., the so-called LTE Positioning Protocol (LPP) or the like). In a positioning protocol such as LPP, for example, point-to-point transmission is provided between a location server (e.g., an enhanced-serving mobile location centre (E-SMLC) or the like, location management function (LMF) or service location protocol (SLP)) and a target device (e.g., the UE 3) in order to position the target device using position-related measurements obtained by one or more reference sources.

[0152] During such a positioning protocol transmission, a positioning protocol session is established which is used between a location server and the target device in order to obtain location related measurements or a location estimate, or to transfer assistance data between nodes in the network.

[0153] Typically, a single positioning protocol session is used to support a single location request (e.g., for a single mobile terminal location request (MT-LR), a mobile originating location request (MO-LR), or a network induced location request (NI-LR)), and multiple positioning protocol sessions can be used between the same endpoints to support multiple different location requests.

[0154] Each positioning protocol session may comprise one or more positioning protocol transactions, with each positioning protocol transaction performing a single operation (capability exchange, assistance data transfer, location information transfer, or the like). Such positioning protocol transactions are typically realised as positioning protocol procedures in a RAN node 5 with an appropriate positioning protocol layer in the protocol stack of the nodes of the network. Such positioning protocol transactions are thus typically indicated at the positioning protocol level with a transaction ID in order to associate messages with one another (e.g., request and response). Furthermore, messages within a transaction may be linked by a common transaction identifier, or the like.

[0155] In more detail, typically, each positioning protocol transaction may involve the exchange of one or more positioning protocol messages between the location server and the target device. The general format of a positioning protocol message consists of a set of common fields followed by a body. The body (which may be empty) contains information specific to a particular message type, and each message type contains information specific to one or more positioning methods and / or information common to all positioning methods.

[0156] There now follows a description of an AAL procedure, for AAL transactions, that makes beneficial use of similar principles to those of positioning protocol transactions such as those described above.

[0157] Fig. 7 illustrates an example sequence diagram of a general AAL transmission procedure (RRC Connection-based) that may be implemented in the communication system 1.

[0158] A session for AAL transmissions may be initiated by either the RAN node 5 or the UE 3 by sending an AAL message for an AI / ML transaction X (e.g. to request the transmission of a reference AI / ML model). As shown in Fig. 7, by way of example only, at step S702 a session for AAL signalling is initiated by the UE 3. Upon initiating the session for AAL signalling, the UE 3 sends an initial message to the RAN node 5 for a transaction X.

[0159] Subsequently, at step S704, the RAN node 5 and / or UE 3 may exchange further messages with one another associated with the transaction X. For example, in the case where the initial message sent at step S702 requests that the RAN node 5 provide a reference AI / ML model to the UE 3, one of the further messages sent at step S704 may include an appropriate message sent by the RAN node 5 to the UE 3 to provide the UE 3 with the requested reference model.

[0160] In another subsequent message, the UE 3 may then request that the RAN node 5 provide one or more model datasets for use with the provided AI / ML model (e.g., for training purposes), and then in a next subsequent message, the RAN node 5 may provide the UE 3 with the one or more requested model datasets, and so on.

[0161] At step S706, the RAN node 5 (or UE 3 - not shown), may send a last message to the UE 3 (or RAN node 5 - not shown) for the transaction and then terminate the whole transaction between the RAN node 5 and the UE 3.

[0162] At step S708, following termination of the previous transaction between the UE 3 and the RAN node 5, the UE 3 or the RAN node 5 may initiate another transaction between the RAN node 5 and the UE 3 (e.g. as part of the same session). As shown in Fig. 7, by way of example only, following termination of the previous transaction (transaction X), the UE 3 initiates a new transaction with the RAN node 5 by sending an initial message to the RAN node 5 for a transaction Y.

[0163] It will be appreciated that in the procedure described above with reference to Fig. 7, within each transaction (e.g., transaction X or Y), all follow-up messages following the initial message sent at step S702 (or step S708) may contain the same transaction identifier, and the last message of the transaction may include an appropriate information element (IE) to indicate that the message is the last message of that specific transaction (e.g., a 'last transmission' indication set to 'TRUE', or '1', or the like).

[0164] It will be appreciated that although only one transaction at a time is shown in Fig. 7, multiple transaction may in fact be performed in parallel and that in this case, each individual transaction will use its own unique transaction IDs. It will also be appreciated that upon termination of a transaction, the unique transaction ID associated with that transaction may then be re-used for a subsequently initiated transaction.

[0165] <Duplication Detection Support>  The above procedure described with reference to Fig. 7 may be adapted to support duplication detection when the RAN node 5 and / or the UE 3 sends a specific message within a transaction more than once unnecessarily.

[0166] For example, the RAN node 5, or UE 3, as the message sender, may, include a sequence number or segmentation number in all the AAL PDUs that they transmit (i.e., in all of the messages that they send in the procedure of Fig. 7) during a particular transaction (e.g., transaction X).

[0167] That sequence number or segmentation number may also be distinct for different data PDUs sent in the same direction (e.g., RAN node 5 to UE 3; or UE 3 to RAN node 5) for same AI / ML transaction e.g., transaction X. For example, the sequence or segmentation number included in a message sent in a specific direction (e.g., RAN node 5 to UE 3) for a same AI / ML transaction (e.g., transaction X) may start with a value of zero in the initial message and increase monotonically in each succeeding message. It will be appreciated that in this case, sequence / segmentation numbers used in the uplink and downlink may be independent (e.g., can be the same).

[0168] The UE 3 or RAN node 5, as the message receiver may then record the most recent received sequence / segmentation number for each transaction (e.g., transaction X). If a message is received carrying the same sequence number or segmentation number as that last received for the associated transaction, the UE 3 or RAN node 5 receiving that message may discard the message. Otherwise (i.e., if the sequence / segmentation number is different, or if no sequence / segmentation number was previously received, or if no sequence / segmentation number is included), the UE 3 or RAN node 5 receiving that message may process that message.

[0169] The RAN node 5 or UE 3, as the message sender, may delete all sending and receiving sequence / segmentation numbers when the associated transaction (e.g., transaction X) is terminated.

[0170] Acknowledgement Support  The above procedure described with reference to Fig. 7 may be adapted to support the provision of an acknowledgement (ACK) or negative acknowledgement (NACK) upon successful or unsuccessful reception of a message by the RAN node 5 and / or the UE 3. For example, each AAL data PDU may carry an appropriate acknowledgement flag to indicate that the RAN node 5 (or UE 3), as the message sender, wishes the message receiver (e.g., the UE 3 (or the RAN node 5)) to acknowledge the reception of the message sent in the current transaction. By way of example only, each AAL data PDU may include an appropriate IE (e.g., acknowledgementFlag, or the like) that is set to 'TRUE,' or '1', or the like, to indicate that the message sender wishes the message receiver to acknowledge reception of the message.

[0171] Additionally, each AAL data PDU that includes such an appropriate IE may also include a sequence number or segmentation number to indicate the order / position of that AAL data PDU in the set of AAL data PDUs sent during a specific transaction (e.g., transaction X).

[0172] Upon reception of the AAL message (i.e., data AAL PDU), which includes the IE acknowledgementFlag set to TRUE, a receiver may return a control AAL PDU with an acknowledgement response / indication (i.e., a special AAL message with an acknowledgement response). For example, the control AAL PDU may include an ACK indication and the same sequence number or segmentation number of the message being acknowledged. Alternatively, an empty data PDU may be used to carry such an ACK indication and the sequence number or segmentation number of the message being acknowledged.

[0173] It will be appreciated that an acknowledgement may be returned for each received message sent if requested irrespective of whether the received message is deemed to be a duplicated message or not.

[0174] Once a sender receives an acknowledgement for the AAL message, and provided that the included sequence / segmentation number matches the sequence / segmentation number in the previous message to be acknowledged, the sender may send the next message.

[0175] It will be appreciated that no message reordering is needed at the receiver since the 'stop-and-wait' method of sending messages described above ensures that messages normally arrive in the correct order.

[0176] Retransmission Support  The above procedure described with reference to Fig. 7 may be adapted to support retransmission of messages during a specific transaction (e.g., transaction X). For example, the AAL may support retransmissions of AAL data PDUs between the UE 3 and the RAN node 5 in the procedure of Fig. 7.

[0177] It will be appreciated that the support for such retransmission may be in conjunction with, and based on, support of acknowledgement and / or duplicate detection capabilities at the AAL. For example, when an AAL message (e.g. a data AAL PDU) which requires acknowledgement is sent by the sender (e.g., the UE 3), the sender may only send the subsequent message (e.g. next data AAL PDU) following a positive acknowledgement from the receiver (e.g., RAN node 5) for the previous message sent.

[0178] On the other hand, when an AAL message (e.g. a data AAL PDU) which requires acknowledgement is sent and an acknowledgement is not received, that AAL message may be resent by the sender (e.g., the UE 3) following a (pre)configured timeout period - i.e., in the absence of an acknowledgement from the RAN node 5 of successful reception of the message before the configured timeout period ends. Alternatively, the timeout period may be determined by sender (e.g., the UE 3) implementation.

[0179] Similarly, an AAL message may be resent by the sender (e.g., the UE 3) if a negative acknowledgement is received in respect of that AAL message (e.g., sent by the intended recipient (e.g. the RAN node 5) when a message carrying an expected sequence / segmentation number is missing).

[0180] Alternatively, blind retransmission by the sender (e.g., the UE 3) may be allowed with a (pre)configured number of transmissions, e.g., M blind retransmissions. If after those M retransmissions, the transmission is still unacknowledged, the sender (e.g., the UE 3) may abort all activity for the associated transaction (e.g., transaction X).

[0181] Fig. 8 illustrates an example sequence diagram of an adapted version of the procedure of Fig. 7 that supports message retransmissions.

[0182] The procedure shown in Fig. 8 relates specifically to downlink (DL) transmissions from the RAN node 5 to the UE 3. Nevertheless, it will be appreciated that a similar procedure may be implemented for uplink (UL) transmissions from the UE 3 to the RAN node 5.

[0183] A session for AAL transmissions may be initiated by either the RAN node 5 or the UE 3 by sending an AAL message for an AI / ML transaction X (e.g. to request the transmission of a reference AI / ML model). As shown in Fig. 8, by way of example only, at step S802 a session for AAL transmissions is initiated by the RAN node 5.

[0184] As shown in Fig. 8, at step S802, the RAN node 5 may send a message N (e.g., a first segment of model parameter configuration) to the UE 3 for transaction X. That message N may also include a request for acknowledgement along with a sequence number or segmentation number.

[0185] At step S804, by way of example only, if the UE 3 does not receive the message N correctly and / or the UE 3 is unable to decode the message N, the UE 3 may return a negative acknowledgement (NACK) message to the RAN node 5 for message N for the transaction X. That NACK message may also include a sequence number or segmentation number indicating that the NACK message corresponds to message N.

[0186] At step S806, in response to receiving the NACK message for message N at step S804 for transaction X, or if the RAN node 5 does not receive a positive acknowledgement (ACK) or NACK message from the UE 3 within a (pre)configured timeout period, the RAN node 5 may retransmit the message N. That retransmitted message N may also include the same sequence / segmentation number that was included in the message N sent at step S804.

[0187] At step S808, by way of example only, if the message N is received by the UE 3 and the UE 3 is able to decode the whole message (including the sequence number or segmentation number), the UE 3 may return an ACK message for message N to RAN Node 5.

[0188] It will be appreciated that the retransmission at step S806 may be repeated one or more times if the response message sent at step S808 is a NACK message for the message N, or if an ACK message is still not received (e.g., step S808 does not occur) before the expiry of a (pre)configured timeout period (which may be timed relative to the transmission of the original message N or relative to the most recent retransmission). Optionally, however, if a (pre)configured maximum number of retransmissions M is reached, then the RAN node 5 may abort all AAL transmissions and activity associated with a current transaction X. In the case where the RAN node aborts all AAL procedures and activity associated with a current transaction X, the RAN node 5 may send an appropriate AAL control PDU to the UE 3 to trigger the termination of all AAL procedures and activity.

[0189] Once an ACK message is received by the RAN node 5 (e.g., at either step S804 or S808), the RAN node 5 may send a next message N+1 (e.g. the second segment of model parameter configuration) for the current transaction X to the UE 3 when this message is available (e.g. when there are more model parameters that needs to be configured to the UE 3).

[0190] <UE Mobility Considerations>  In the procedures of Figs. 7 and 8 described above, it is assumed that multiple RAN nodes (or cells) may share the same reference AI / ML model, model datasets, and / or model parameters with the UE 3 in a designated area. Hence, when the UE 3 moves from the coverage of a first RAN node 5 (or cell) to the coverage a second RAN node 5 (or cell) in the same designated area as the first RAN node 5, the UE 3 may keep any received portions of the AI / ML reference model, model datasets, and / or model parameters that it received from the first RAN node 5 previously, and furthermore the UE 3 may also continue reception, from the second RAN node 5, of the AI / ML reference model, the model datasets, and / or the model parameters, associated with the AI / ML reference model previously received from the first RAN node 5.

[0191] It will be appreciated that multiple RAN nodes 5 (or cells) within a same designated area may store the AI / ML reference model, model datasets, and / or model parameters associated with the same AI / ML reference model.

[0192] By way of example only, during a mobility (e.g., handover) preparation stage, the first (e.g., source) RAN node 5 may interact with the second (e.g., target) RAN node to determine whether or not the second (target) RAN node 5 supports transmission of the same AI / ML reference model, model dataset, and / or model parameters as the first (source) RAN node 5. The first (source) RAN node 5 may then forward a status indication of the transmission for the AI / ML reference model, model datasets, and / or model parameters for the UE 3. For example, a model ID, a transaction ID, and / or a sequence number or segmentation number of the transaction for UE 3 may be forwarded to the second (target) RAN node 5.

[0193] Following a successful handover, an AAL entity may be established at the UE 3 and the second (target) RAN node 5 to continue the further transmission of a remaining part of the AI / ML reference model, model datasets, and / or model parameters. The second (target) RAN node 5 may then start transmission to the UE 3 with the next sequence number or segmentation number following the sequence number or segmentation number indicated to the second (target) RAN node 5 in the transmission status received from first (source) RAN node 5. Alternatively, the second (target) RAN node 5 may generate the transmission from the very beginning of the sequences or segments of the AI / ML information bits - in other words the whole transaction (e.g., transaction X) may be restarted.

[0194] Additionally, the UE 3 may indicate its reception status of part (or all) of the AI / ML reference model, model dataset, and / or model parameters to the second (target) RAN node 5 during handover, RRC reestablishment, RRC resume, or the like. The second (target) RAN node 5 may then start transmission to the UE 3 with the next sequence number or segmentation number following the sequence number or segmentation number indicated to the second (target) RAN node 5 in the transmission status received from first (source) RAN node 5.

[0195] It will be appreciated that the principle described above may be used between the UE 3 and the RAN node 5 to transmit other types of data. For example, the UE 3 may use the AAL to report its collected data for monitoring the performance of a certain AI / ML model operated in the UE 3 to the RAN node 5.

[0196] <AAL Procedures with RRC Connectionless-based Transmission>  In the case where AAL communication between the RAN node 5 and UE 3 is RRC connectionless-based, all or part of an AI / ML reference model, one or more model datasets, and / or model parameters may be broadcast or multicast to the UE 3 via system information blocks (SIB), or an appropriate Multicast Common Channel-like (MCCH-like) mechanism.

[0197] In this scenario, all or part of the AI / ML reference model, the one or more model datasets, and / or the model parameters may be periodically broadcast to the UE 3 (in e.g., segments) via a data sharing channel (e.g., PDSCH), and the UE 3 may receive a portion / segment of all or part of the AI / ML reference model, the one or more model datasets, and / or the model parameters with each broadcast. The periodicity of the broadcast to the UE 3 may follow a configuration provided to the UE 3 within a MCCH-like common channel, and the MCCH-like common channel may be used to announce the model ID and the transaction ID of a current broadcast.

[0198] It will be appreciated that as all or part of the AI / ML reference model, the one or more model datasets, and / or the model parameters are broadcast, reception of the AI / ML reference model, the one or more model datasets, and / or the model parameters by the UE 3 may occur irrespective of the RRC state of the UE 3; for example, the UE 3 may receive the AI / ML reference model, the one or more model datasets, and / or the model parameters irrespective of whether it is in an RRC connected state, an RRC inactive, or an RRC idle state.

[0199] It will also be appreciated that as all or part of the AI / ML reference model, the one or more model datasets, and / or the model parameters are broadcast, the UE 3 may only receive and process the AI / ML reference model, the one or more model datasets, and / or the model parameters in the order of the broadcasts. For example, the UE 3 may only receive and process a segment N+1 if the segment that the UE 3 previously received was segment N. Once the segment N+1 has been processed, the UE 3 may discard segment N, and await reception of segment N+2. Furthermore, to avoid the UE 3 incorrectly receiving a wrong segment during a broadcast over the data sharing channel (e.g., PDSCH), the segment number of an upcoming segment may be announced in the MCCH-like common channel first.

[0200] It will also be appreciated that reception of all or part of the AI / ML reference model, the one or more model datasets, and / or the model parameters by the UE 3 may continue when the UE 3 moves from one cell to another cell as long as both cells can provide the same AI / ML reference model, one or more model datasets, and / or model parameters the model ID and the transaction ID are the same. For example, the UE 3 may continue to receive the AI / ML reference model, the one or more model datasets, and / or the model parameters associated with a specific transaction (e.g., transaction X) as it moves between cells providing those cells are within a specific area scope (which may, for example, be configured via a list of cells being provided to the UE 3 that are within a specific area scopes).

[0201] Additionally, the UE 3 may be configured to keep a received AI / ML reference model (or received part thereof), a received one or more model datasets (or received part thereof), and / or received model parameters (or received part thereof) for a particular AI / ML model even if the UE 3 experiences radio link failure (RLF) or handover failure. Fig. 9 illustrates an example broadcasted message structure that may be used for AAL procedures with RRC connectionless-based transmission as described above.

[0202] As shown in Fig. 9, the (segmented) AI / ML reference model, the (segmented) one or more model datasets, and / or the (segmented) model parameters may be broadcast by the RAN node 5 to the UE 3 via an appropriate system information block (e.g., SIB1).

[0203] In the SIB1 broadcasts to the UE 3 by the RAN node 5, an appropriate indication (e.g., IE) may be provided to indicate to the UE 3 that the RAN node 5 supports AI / ML model information broadcasting, including the broadcasting of resource information necessary for receiving and reading an AI / ML common channel (e.g., an MCCH-like channel).

[0204] The UE 3 may then use the information provided in the SIB1 to receive and read the AI / ML common channel (e.g., an MCCH-like channel) over which the RAN node 5 may transmit appropriate messages to the UE 3 to indicate to the UE 3 one or more configurations of scheduling resources (including the periodicity of those resources) that may be used by the RAN node 5 for the transmission of AI / ML model related transmissions, including, by way of example only, a model ID, transaction ID, and the like (e.g., a AI / ML reference model, one or more model datasets, and / or model parameters).

[0205] Periodic broadcasts of the AI / ML reference model, the one or more model datasets, and / or the model parameters may then be sent by the RAN node 5 in AAL data PDU segments via the scheduling resources configured over the AI / ML common channel (e.g., an MCCH-like channel). For example, as shown in Fig. 9, the RAN node 5 may broadcast a plurality of AAL data PDUs (segment #1 to #4) on a periodic basis, which may then be received by the UE 3.

[0206] There now follows a detailed description of the format and possible content of the AAL PDUs that may be supported for broadcast.

[0207] <AAL PDU Header>  As described above with reference to Fig. 6, when generating AAL PDUs, the AAL 4-18 may package an AAL SDU segment (e.g., AAL SDU segment #1) into an AAL PDU comprising an AAL PDU header and an AAL data payload. That AAL header of the AAL PDU may, for example, include one or more fields that provide specific header-type information to the lower layers that received that AAL PDU.

[0208] For example, in the case where the AAL PDU is a data PDU, the AAL PDU header may include: an AI / ML model identifier (ID) which indicates the AI / ML model that the current AAL PDU transmission is associated. That AI / ML model ID may be a global model ID, a local Model ID, or a cell ID plus a local model ID.

[0209] In the case where the AAL PDU is a data PDU, the AAL PDU header may include a transaction ID which indicates a purpose of current AAL PDU transmission (e.g., AI / ML model transfer, model dataset sharing, model parameter configuration, or the like).

[0210] In the case where the AAL PDU is a data PDU, the AAL PDU header may include a sequence number or segmentation number which Indicates a particular sequence number or segmentation number associated with the AAL PDU transmission, which may be used to detect data PDU loss or duplicates within one transaction (i.e., within one set of AAL PDU transmissions).

[0211] In the case where the AAL PDU is a data PDU, the AAL PDU header may include a last transmission indication to indicate whether the current AAL PDU transmission is a last AAL PDU transmission within the transaction. For example, the last transmission indication may indicate when a transaction has ended (i.e., the last AAL PDU transmission of a transaction has ended). Additionally (or alternatively), the AAL PDU header may include the length of the padding bits or information bits used for the last AAL PDU transmission.

[0212] In the case where the AAL PDU is a data PDU, the AAL PDU header may include an indication of the length of the last transmission when the AAL PDU header includes an indication that the current transmission is the last transmission. That length of the last transmission may, for example, indicate the length of the padding bits or information bits used for the last transmission for a specific transaction (e.g., transaction X).

[0213] AAL PDU Type and Content  As mentioned above, when generating AAL PDUs, the AAL entity of the AAL 4-18 may package an AAL SDU segment (e.g., AAL SDU segment #1) into an AAL PDU comprising an AAL PDU header and an AAL data payload. The contents of that AAL data payload of the AAL PDU may vary depending on: i) whether the AAL PDU is a control PDU or a data PDU, and ii) the purpose of the AAL PDU.

[0214] For example, where the AAL PDU is a data PDU, the AAL PDU may include an AI / ML model. For example, where the AAL PDU is generated by the RAN node 5 to be sent to the UE 3, the AAL PDU may include an AI / ML model to be provided by the RAN node 5 to the UE 3.

[0215] In another example, where the AAL PDU is a data PDU, the AAL PDU may include one or more model datasets for use with an AI / ML model (e.g., for training the AI / ML model). For example, where the AAL PDU is generated by the RAN node 5 to be sent to the UE 3, the AAL PDU may include one or more model datasets for use with an AI / ML model at the UE 3.

[0216] In another example, where the AAL PDU is a data PDU, the AAL PDU may include one or more model parameters for use in configuring an AI / ML model. For example, where the AAL PDU is generated by the RAN node 5 to be sent to the UE 3, the AAL PDU may include one or more model parameters for use by the UE 3 to configure an AI / ML model at the UE 3.

[0217] <AAL Transmission Procedures - RRC Connectionless-based>  As mentioned above, each UE 3 and RAN node 5 of the communication system 1 may be mutually configured to support one or more procedures for supporting control plane-based AI / ML reference model, model dataset, and / or model parameter broadcasting by RAN nodes 5 to UEs 3 in the context of the training and use of two-sided AI / ML models in the communication system 1.

[0218] Fig. 10 illustrates another example sequence diagram of a general AAL transmission procedure (RRC Connectionless) that may be implemented in the communication system 1.

[0219] A session for AAL transmissions may be initiated by the RAN node 5 by sending an AAL message for an AI / ML transaction X (e.g. for providing an AI / ML model) to the UE 3. As shown in Fig. 10, by way of example only, at step S1002 a session for AAL signalling is initiated by the RAN node 5. Upon initiating the session for AAL signalling, the RAN node 5 sends an initial message to the UE 3 for a transaction X.

[0220] Subsequently, at step S1004, the RAN node 5 may broadcast further messages to the UE 3 associated with the transaction X. For example, in the case where the initial message sent at step S1002 provided a reference AI / ML model to the UE 3, one of the further messages sent at step S1004 may include an appropriate message broadcast by the RAN node 5 to the UE 3 to provide the UE 3 with AI / ML model datasets, or the like.

[0221] At step S1006, the RAN node 5 may broadcast a last message to the UE 3 for the transaction and then terminate the whole transaction between the RAN node 5 and the UE 3.

[0222] At step S1008, following termination of the previous transaction between the UE 3 and the RAN node 5, the RAN node 5 may initiate another transaction between the RAN node 5 and the UE 3 (e.g. as part of the same session). As shown in Fig. 10, by way of example only, following termination of the previous transaction (transaction X), the RAN node 5 initiates a new transaction with the UE 3 by broadcasting an initial message to the UE 3 for a transaction Y.

[0223] It will be appreciated that in the procedure described above with reference to Fig. 10, within each transaction (e.g., transaction X or Y), all follow-up messages following the initial message broadcast at step S1002 (or step S1008) may contain the same transaction identifier, and the last message of the transaction may include an appropriate information element (IE) to indicate that the message is the last message of that specific transaction (e.g., a 'last transmission' indication set to 'TRUE', or '1', or the like).

[0224] It will be appreciated that although only one transaction at a time is shown in Fig. 10, multiple transaction may in fact be performed in parallel and that in this case, each individual transaction will use its own unique transaction IDs. It will also be appreciated that upon termination of a transaction, the unique transaction ID associated with that transaction may then be re-used for a subsequently initiated transaction.

[0225] <Duplication Detection Support>  The above procedure described with reference to Fig. 10 may be adapted to support duplication detection when the RAN node 5 and / or the UE 3 sends a specific message within a transaction more than once unnecessarily.

[0226] For example, the RAN node 5, as the message sender, may, include a sequence number or segmentation number in all the AAL data PDUs that they transmit (i.e., in all of the messages that they send in the procedure of Fig. 10) during a particular transaction (e.g., transaction X).

[0227] That sequence number or segmentation number may also be distinct for different data PDUs sent in the same direction (e.g., RAN node 5 to UE 3) for same AI / ML transaction e.g., transaction X. For example, the sequence or segmentation number included in a message sent in a specific direction (e.g., RAN node 5 to UE 3) for a same AI / ML transaction (e.g., transaction X) may start with a value of zero in the initial message and increase monotonically in each succeeding message.

[0228] The UE 3, as the message receiver may then record the most recent received sequence / segmentation number for each transaction (e.g., transaction X). If a message is received carrying the same sequence number or segmentation number as that last received for the associated transaction, the UE 3 receiving that message may discarded the message. Otherwise (i.e., if the sequence / segmentation number is different, or if no sequence / segmentation number was previously received, or if no sequence / segmentation number is included), the UE 3 receiving that message may process that message.

[0229] <UE Mobility Considerations>  In the procedure of Fig. 10 described above, it is assumed that multiple RAN node 5 (or cell) may share the same reference AI / ML model, model datasets, and / or model parameters with the UE 3 in a designated area. Hence, when the UE 3 moves from the coverage of a first RAN node 5 (or cell) to the coverage a second RAN node 5 (or cell) in the same designated area as the first RAN node 5, the UE 3 may keep any received segments / portions of the AI / ML reference model, model datasets, and / or model parameters that it received from the first RAN node 5 previously, and furthermore the UE 3 may also continue reception, from the second RAN node 5, of an AI / ML reference model, model datasets, and / or model parameters, associated with the AI / ML reference model previously received from the first RAN node 5. It is assumed that RAN nodes (or cells) in a designated area may broadcasting the AI / ML reference model, one or more model dataset, and / or model parameters.

[0230] During UE mobility in an RRC idle mode or RRC inactive mode, the UE 3 may read the SIB1 of a new camping cell (hosted by e.g., the RAN node 5) to decide if the new camping cell supports the broadcast of AI / ML related information bits. At the same time, the UE 3 may further read the AI / ML specific common channel (e.g., the MCCH-like channel) indicated by the SIB1 of the new camping cell to determine if the new camping cell broadcasts the same AI / ML reference model, model dataset, and / or model parameters as it recently received from its previous camping cell.

[0231] If the new camping cell broadcasts the same AI / ML reference model, model dataset, and / or model parameters as the UE 3 recently received from its previous camping cell, the UE 3 may continue to receive remaining segments of the AI / ML reference model, model dataset, and / or model parameters from the new camping cell.

[0232] It will be appreciated that upon receiving those remaining segments, the UE 3 may drop a newly received segment if the segment number of that segment is less than the segment number of the last segment the UE 3 received from its previous camping cell (i.e., duplicates are avoided). The UE 3 may only process a segment whose segment number is the next segment number following the last segment it received from the previous camping cell.

[0233] Furthermore, while being in RRC idle mode or RRC inactive mode, the UE 3 may keep received portions of the AI / ML model, model parameters, and / or model datasets, for a (pre)configured period of time before discarding them.

[0234] <Devices of the Communication System><User Equipment>  Fig. 11 is a schematic block diagram illustrating the main components of a UE 3 as shown in Fig. 1.

[0235] As shown, the UE 3 has a transceiver circuit 31 that is operable to transmit signals to and to receive signals from a RAN node 5 via one or more antennas 33 (e.g., comprising one or more antenna elements). The UE 3 has a controller 37 to control the operation of the UE 3. The controller 37 is associated with a memory 39 and is coupled to the transceiver circuit 31. Although not necessarily required for its operation, the UE 3 might, of course, have all the usual functionality of a conventional UE (e.g., a user interface 35, such as a touch screen / keypad / microphone / speaker and / or the like for, allowing direct control by and interaction with a user) and this may be provided by any one or any combination of hardware, software, and firmware, as appropriate. Software may be pre-installed in the memory 39 and / or may be downloaded via the communication system 1 or from a removable data storage device (RMD), for example.

[0236] The controller 37 is configured to control overall operation of the UE 3 by, in this example, program instructions or software instructions stored within memory 39. As shown, these software instructions include, among other things, an operating system 41, and a communications control module 43.

[0237] The communication control module 43 is operable to control the communication between the UE 3 and its serving RAN node 5 (and other communication devices connected to the RAN node 5, such as further UEs 3 and / or core network nodes). The communication control module 43 is configured for the overall handling of uplink communications via associated uplink channels (e.g., via a physical uplink control channel (PUCCH), (physical) random access channel ((P)RACH), and / or a physical uplink shared channel (PUSCH)) including both dynamic and semi-static signalling (e.g., SRS). The communication control module 43 is also configured for the overall handling of receipt of downlink communications via associated downlink channels (e.g., of DCI via a physical downlink control channel (PDCCH) and / or a physical downlink shared channel (PDSCH)) including both dynamic and semi-persistent scheduling (e.g., SPS). The communication control module 43 is responsible, for example: for determining where to monitor for downlink control information; for determining the resources to be used by the UE 3 for transmission / reception of UL / DL communications (including interleaved resources and resources subject to frequency hopping); for managing frequency hopping at the UE side; for determining how slots / symbols are configured (e.g., for UL, DL or full duplex communication, or the like); for determining which bandwidth parts are configured for the UE 3; for determining how uplink transmissions should be encoded and the like.

[0238] It will be appreciated that the communication control module 43 may include a number of sub-modules ('layers' or 'entities') to support specific functionalities. For example, the communication control module 43 may include a PHY sub-module, a MAC sub-module, an RLC sub-module, a PDCP sub-module, an RRC sub-module, an AI / ML adaptation sub-module, AI / ML model (datasets and parameters) module, etc.

[0239] The communication control module 43 is configured, in particular, to control the UE's communications, where applicable, in accordance with any of the methods described herein.

[0240] RAN node  Fig. 12 is a schematic block diagram illustrating the main components of a RAN node 5 for the communication system 1 shown in Fig. 1.

[0241] As shown, the RAN node 5 has a transceiver circuit 51 for transmitting signals to and for receiving signals from the communication devices (such as UEs 3) via one or more antennas 53 (e.g., a single or multi-panel antenna array / massive antenna), and a core network interface 55 (e.g., comprising the N2, N3 and other reference points / interfaces) for transmitting signals to and for receiving signals from network nodes in the core network 7. Although not shown, the RAN node 5 may also be coupled to other RAN nodes via an appropriate interface (e.g., the so-called 'Xn' interface in NR). The RAN node 5 has a controller 57 to control the operation of the RAN node 5. The controller 57 is associated with a memory 59. Software may be pre-installed in the memory 59 and / or may be downloaded via the communication system 1 or from a removable data storage device (RMD), for example. The controller 57 is configured to control the overall operation of the RAN node 5 by, in this example, program instructions or software instructions stored within memory 59.

[0242] As shown, these software instructions include, among other things, an operating system 61, and a communications control module 63.

[0243] The communications control module 63 is operable to control the communication between the RAN node 5 and UEs 3 and other network entities that are connected to the RAN node 5. The communications control module 63 is configured for the overall control of the reception and decoding of uplink communications, via associated uplink channels (e.g., via a physical uplink control channel (PUCCH), a (physical) random-access channel ((P)RACH), and / or a physical uplink shared channel (PUSCH)) including both dynamic and semi-static signalling (e.g., SRS). The communications control module 63 is also configured for the overall handling the transmission of downlink communications via associated downlink channels (e.g., via a physical downlink control channel (PDCCH) and / or a physical downlink shared channel (PDSCH)) including both dynamic and semi-static signalling (e.g., CSI-RS, SSBs etc.). The communications control module 63 is also responsible, for example, for determining and scheduling the resources to be used by the UE 3 for receiving in DL / transmitting in UL, for configuring slots / symbols appropriately (e.g., for UL, DL, flexible, full duplex communication, or the like), for configuring one or more bandwidth parts for the UE 3, and for providing related configuration signalling to the UE 3.

[0244] It will be appreciated that the communications control module 63 may include a number of sub-modules (or 'layers') to support specific functionalities. For example, the communications control module 63 may include a PHY sub-module, a MAC sub-module, an RLC sub-module, a PDCP sub-module, an SDAP sub-module, an IP sub-module, an RRC sub-module, an AI / ML adaptation sub-module, AI / ML model (datasets and parameters) module, etc.

[0245] The communication control module 63 is configured, in particular, to control the RAN node's communications, where applicable, in accordance with any of the methods described herein.

[0246] <Modifications and Alternatives>  Detailed examples been described above. As those skilled in the art will appreciate, a number of modifications and alternatives can be made to the above examples whilst still benefiting from the concepts embodied therein.

[0247] It will be appreciated that description of features of and actions performed by a RAN node (base station), apply equally to distributed type base stations as to non-distributed type base stations.

[0248] It will also be appreciated that whilst information elements having specific names have been described differently named information elements but having a similar purpose may be used.

[0249] In the above description the UE and the base station are described for ease of understanding as having a number of discrete functional components or modules. Whilst these modules may be provided in this way for certain applications, for example where an existing system has been modified to implement the disclosed enhancements, in other applications, for example in systems designed with the inventive features in mind from the outset, these modules may be built into the overall operating system or code and so these modules may not be discernible as discrete entities.

[0250] In the above examples, a number of software modules were described. As those skilled in the art will appreciate, the software modules may be provided in compiled or un-compiled form and may be supplied to the UE or base station as a signal over a computer network, or on a recording medium. Further, the functionality performed by part, or all, of this software may be performed using one or more dedicated hardware circuits. However, the use of software modules is preferred as it facilitates the updating of the UE or the base station in order to update their functionalities.

[0251] Each controller may comprise any suitable form of processing circuitry including (but not limited to), for example: one or more hardware implemented computer processors; microprocessors; central processing units (CPUs); arithmetic logic units (ALUs); input / output (IO) circuits; internal memories / caches (program and / or data); processing registers; communication buses (e.g. control, data and / or address buses); direct memory access (DMA) functions; hardware or software implemented counters, pointers and / or timers; and / or the like. Various other modifications will be apparent to those skilled in the art and will not be described in further detail here.

[0252] The User Equipment (or "UE," "mobile station," "mobile device" or "wireless device") in the present disclosure is an entity connected to a network via a wireless interface.

[0253] It should be noted that the present disclosure is not limited to a dedicated communication device and can be applied to any device having a communication function as explained in the following paragraphs.

[0254] The terms "User Equipment" or "UE" (as the term is used by 3GPP), "mobile station", "mobile device", and "wireless device" are generally intended to be synonymous with one another, and include standalone mobile stations, such as terminals, cell phones, smart phones, tablets, cellular IoT devices, IoT devices, and machinery. It will be appreciated that the terms "mobile station" and "mobile device" also encompass devices that remain stationary for an extended period of time.

[0255] A UE may, for example, be an item of equipment for production or manufacture and / or an item of energy related machinery (for example equipment or machinery such as: boilers; engines; turbines; solar panels; wind turbines; hydroelectric generators; thermal power generators; nuclear electricity generators; batteries; nuclear systems and / or associated equipment; heavy electrical machinery; pumps including vacuum pumps; compressors; fans; blowers; oil hydraulic equipment; pneumatic equipment; metal working machinery; manipulators; robots and / or their application systems; tools; moulds or dies; rolls; conveying equipment; elevating equipment; materials handling equipment; textile machinery; sewing machines; printing and / or related machinery; paper converting machinery; chemical machinery; mining and / or construction machinery and / or related equipment; machinery and / or implements for agriculture, forestry and / or fisheries; safety and / or environment preservation equipment; tractors; precision bearings; chains; gears; power transmission equipment; lubricating equipment; valves; pipe fittings; and / or application systems for any of the previously mentioned equipment or machinery etc.).

[0256] A UE may, for example, be an item of transport equipment (for example transport equipment such as: rolling stocks; motor vehicles; motorcycles; bicycles; trains; buses; carts; rickshaws; ships and other watercraft; aircraft; rockets; satellites; drones; balloons etc.).

[0257] A UE may, for example, be an item of information and communication equipment (for example information and communication equipment such as: electronic computer and related equipment; communication and related equipment; electronic components etc.).

[0258] A UE may, for example, be a refrigerating machine, a refrigerating machine applied product, an item of trade and / or service industry equipment, a vending machine, an automatic service machine, an office machine or equipment, a consumer electronic and electronic appliance (for example a consumer electronic appliance such as: audio equipment; video equipment; a loud speaker; a radio; a television; a microwave oven; a rice cooker; a coffee machine; a dishwasher; a washing machine; a dryer; an electronic fan or related appliance; a cleaner etc.).

[0259] A UE may, for example, be an electrical application system or equipment (for example an electrical application system or equipment such as: an x-ray system; a particle accelerator; radio isotope equipment; sonic equipment; electromagnetic application equipment; electronic power application equipment etc.).

[0260] A UE may, for example, be an electronic lamp, a luminaire, a measuring instrument, an analyser, a tester, or a surveying or sensing instrument (for example a surveying or sensing instrument such as: a smoke alarm; a human alarm sensor; a motion sensor; a wireless tag etc.), a watch or clock, a laboratory instrument, optical apparatus, medical equipment and / or system, a weapon, an item of cutlery, a hand tool, or the like.

[0261] A UE may, for example, be a wireless-equipped personal digital assistant or related equipment (such as a wireless card or module designed for attachment to or for insertion into another electronic device (for example a personal computer, electrical measuring machine)).

[0262] A UE may be a device or a part of a system that provides applications, services, and solutions described below, as to "internet of things (IoT)," using a variety of wired and / or wireless communication technologies.

[0263] Internet of Things devices (or "things") may be equipped with appropriate electronics, software, sensors, network connectivity, and / or the like, which enable these devices to collect and exchange data with each other and with other communication devices. IoT devices may comprise automated equipment that follow software instructions stored in an internal memory. IoT devices may operate without requiring human supervision or interaction. IoT devices might also remain stationary and / or inactive for an extended period of time. IoT devices may be implemented as a part of a (generally) stationary apparatus. IoT devices may also be embedded in non-stationary apparatus (e.g., vehicles) or attached to animals or persons to be monitored / tracked.

[0264] It will be appreciated that IoT technology can be implemented on any communication devices that can connect to a communication system for sending / receiving data, regardless of whether such communication devices are controlled by human input or software instructions stored in memory.

[0265] It will be appreciated that IoT devices are sometimes also referred to as Machine-Type Communication (MTC) devices or Machine-to-Machine (M2M) communication devices. It will be appreciated that a UE may support one or more IoT or MTC applications. Some examples of MTC applications are listed in the following table. This list is not exhaustive and is intended to be indicative of some examples of machine type communication applications.

[0266] Further, the above-described UE categories are merely examples of applications of the technical ideas and exemplary examples described in the present document. Needless to say, these technical ideas and examples are not limited to the above-described UE and various modifications can be made thereto.

[0267] Various other modifications will be apparent to those skilled in the art and will not be described in further detail here.

[0268] For example, the whole or part of the exemplary embodiments disclosed above can be described as, but not limited to, the following supplementary notes.  (Supplementary note 1)      A method performed by a user equipment, UE, the method comprising:    transmitting to, or receiving from, a radio access network, RAN, node, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; and    receiving, as part of the current session, from the RAN node, at least one message for transferring the information related to the first AI / ML model to the UE, wherein    the initial message, and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.  (Supplementary note 2)      The method according to supplementary note 1, wherein the information related to the first AI / ML model comprises at least one of: at least part of the first AI / ML model; at least part of an AI / ML model dataset for the first AI / ML model; and / or at least part of a set of parameters for use in configuring the first AI / ML model.  (Supplementary note 3)      The method according to supplementary note 1 or 2, wherein the initial message is transmitted from the UE to the RAN node.  (Supplementary note 4)      The method according to supplementary note 3, wherein the initial message is configured for requesting provision the first AI / ML model, and the at least one message for transferring the information related to the first AI / ML model is configured for providing at least part of the first AI / ML model.  (Supplementary note 5)      The method according to supplementary note 3, wherein the initial message is configured for requesting provision an AI / ML model dataset for the first AI / ML model, and the at least one message for transferring the information related to the first AI / ML model is configured for providing at least part of the AI / ML model dataset.  (Supplementary note 6)      The method according to supplementary note 3, wherein the initial message is configured for requesting a set of parameters for use in configuring the first AI / ML model, and the at least one message for transferring the information related to the first AI / ML model is configured for providing at least part of the parameters for use in configuring the first AI / ML model.  (Supplementary note 7)      The method according to supplementary note 1 or 2, wherein the initial message is transmitted from the RAN node to the UE.  (Supplementary note 8)      The method according to supplementary note 7, wherein the initial message, and the at least one message for transferring the information related to the first AI / ML model, are each respectively configured for providing at least part of the first AI / ML model.  (Supplementary note 9)      The method according to supplementary note 7, wherein the initial message, and the at least one message for transferring the information related to the first AI / ML model, are each respectively configured for providing at least part of an AI / ML model dataset for the first AI / ML model.  (Supplementary note 10)      The method according to supplementary note 7, wherein the initial message, and the at least one message for transferring the information related to the first AI / ML model, are each respectively configured for providing at least part of a set of parameters for use in configuring the first AI / ML model.  (Supplementary note 11)      The method according to any one of supplementary notes 7 to 10, wherein the information related to the first AI / ML model is received via a broadcast or multicast transmission from the RAN node.  (Supplementary note 12)      The method according to supplementary note 11, wherein the information related to the first AI / ML model is received via one or more system information blocks and / or a multicast common channel, MCCH, based mechanism.  (Supplementary note 13)      The method according to supplementary note 11 or 12, wherein the information related to the first AI / ML model is received periodically, optionally via a data sharing channel.  (Supplementary note 14)      The method according to any one of supplementary notes 7 to 13, wherein the information related to the first AI / ML model is received regardless of whether the UE is in a radio resource control, RRC, connected state, an RRC inactive state, or an RRC idle state.  (Supplementary note 15)      The method according to any one of supplementary notes 1 to 14, wherein the method further comprises transmitting to, or receiving from the RAN node, at least one further initial message for at least one further session for transferring information related to the first AI / ML model, or related to at least one further AI / ML model, to the UE, and receiving, from the RAN node, at least one further message for transferring the information related to the first AI / ML model, or related to at least one further AI / ML model, to the UE, wherein the at least one further initial message, and each message for transferring the information related to the first AI / ML model, or related to at least one further AI / ML model, carries at least one further common transaction identifier associated with the at least one further session.  (Supplementary note 16)      The method according to supplementary note 15, wherein at least one further session for transferring the information related to the first AI / ML model, or related to at least one further AI / ML model, to the UE occurs subsequent to the current session for transferring the information related to the first AI / ML model to the UE.  (Supplementary note 17)      The method according to supplementary note 15 or 16, wherein at least one further session for transferring the information related to the first AI / ML model, or related to at least one further AI / ML model, to the UE occurs at least partially in parallel with the current session for transferring the information related to the at least one further AI / ML model to the UE.  (Supplementary note 18)      The method according to any one of supplementary notes 1 to 17, wherein the, or each, transaction identifier is configured to indicate a purpose of a message that carries that transaction identifier.  (Supplementary note 19)      The method according to any one of supplementary notes 1 to 18, wherein the at least one message for transferring the information related to the first AI / ML model includes, a final message of the current session, wherein the final message of the current session carries at least one of: information indicating that the final message is a last message of the current session; and / or information indicating a length of at least part of the final message.  (Supplementary note 20)      The method according to supplementary note 19, wherein the information indicating a length of at least part of the final message indicates a length of padding bits used in the message or a length of information bits used in the message.  (Supplementary note 21)      The method according to any one of supplementary notes 1 to 20, wherein the at least one message for transferring the information related to the first AI / ML model and / or the initial message carries an identifier of the AI / ML model to which that message relates.  (Supplementary note 22)      The method according to supplementary note 21, wherein the identifier of the AI / ML model to which that message relates comprises a global identifier, a local identifier, or a combination of a cell identifier and a local identifier.  (Supplementary note 23)      The method according to any one of supplementary notes 1 to 22, wherein the at least one message for transferring the information related to the first AI / ML model and / or the initial message carries a sequence number of a data unit carried by that message, and / or a segmentation number of a segment of a data unit carried by that message, optionally wherein the sequence number of the data unit carried by that message, and / or the segmentation number of the segment of the data unit carried by that message, is unique within the current session for data units transmitted in the same communication direction.  (Supplementary note 24)      The method according to supplementary note 23, wherein in a case where the UE receives, during the current session, a later message for transferring the information related to the first AI / ML model that includes a sequence number and / or segmentation number that was also included in an earlier message for transferring the information related to the first AI / ML model, the method further comprises discarding the later message.  (Supplementary note 25)      The method according to any one of supplementary notes 1 to 24, wherein the method further comprises transmitting, to the RAN node, at least one feedback message for positively acknowledging, or for negatively acknowledging, reception of a data unit depending on whether that data unit has been successfully received or not.  (Supplementary note 26)      The method according to supplementary note 25, wherein the at least one message for transferring the information related to the first AI / ML model and / or the initial message carries information indicating whether an acknowledgement is requested for a data unit carried by that message.  (Supplementary note 27)      The method according to supplementary note 26, wherein the at least one feedback message is transmitted in respect of a data unit, in a case where an acknowledgement is requested for that data unit.  (Supplementary note 28)      The method according to any one of supplementary notes 25 to 27, wherein the at least one feedback message carries at least one of: a transaction identifier indicating a purpose of the message is for providing feedback; a segmentation number of the data unit to which the at least one feedback message relates; and / or information indicating whether the data unit to which the at least one feedback message relates is acknowledged or negatively acknowledged.  (Supplementary note 29)      The method according to any one of supplementary notes 1 to 28, wherein the method further comprises receiving, from the RAN node, a message for aborting the current session.  (Supplementary note 30)      The method according to supplementary note 29, wherein the message for aborting the current session carries a transaction identifier indicating a purpose of the message is for aborting the current session.  (Supplementary note 31)      A method performed by radio access network, RAN, node, the method comprising:    receiving from, or transmitting to, a user equipment, UE, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; and    transmitting, as part of the current session, to the UE, at least one message for transferring the information related to the first AI / ML model to the UE, wherein    the initial message and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.  (Supplementary note 32)      The method according to supplementary note 31, wherein at least one message for transferring the information related to the first AI / ML model and / or the initial message carries information indicating whether an acknowledgement is requested for a data unit carried by that message, and, in a case where acknowledgement is requested for a previously transmitted data unit and a positive acknowledgement is not subsequently received in respect of the previously transmitted data unit, the method further comprises retransmitting, to the UE, the previously transmitted data unit in at least one further message for transferring the information related to the first AI / ML model.  (Supplementary note 33)      The method according to supplementary note 32, wherein the retransmitting of the previously transmitted data unit occurs based on a configured or predefined timeout period, relative to a time at which the previously transmitted data unit was previously transmitted.  (Supplementary note 34)      The method according to supplementary note 32 or 33, wherein the retransmitting is repeated until a positive acknowledgement is received in respect of the previously transmitted data unit, or until a configured maximum number of transmissions, or retransmissions, of the previously transmitted data unit is reached.  (Supplementary note 35)      The method according to supplementary note 34, wherein, in a case where the configured maximum number of transmissions, or retransmissions, is reached, the method further comprises transmitting, to the UE, a message for aborting the current session.  (Supplementary note 36)      The method according to any one of supplementary notes 32 to 35, wherein, transmission of a subsequent message for transferring the information related to the first AI / ML model that carries a subsequent data unit is deferred until a positive acknowledgement is received in respect of the previously transmitted data unit.  (Supplementary note 37)      The method according to any one of supplementary notes 31 to 36, wherein, in a case where the UE is to be handed over to a further RAN node during the current session, the method further comprises providing, to the further RAN node, a transmission status of the information related to the first AI / ML model being transferred to the UE as part of the current session.  (Supplementary note 38)      The method according to any one of supplementary notes 31 to 37, wherein, in a case where the UE is to be handed over to a further RAN node during the current session, method further comprises providing, to the further RAN node at least one of: the common transaction identifier; an identifier of an AI / ML model to which the current session relates; a current sequence number for data units being transmitted to the UE; and / or a current segmentation number for data units being transmitted to the UE.  (Supplementary note 39)      A user equipment, UE, comprising:    means for transmitting to, or receiving from, a radio access network, RAN, node, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; and    means for receiving, as part of the current session, from the RAN node, at least one message for transferring the information related to the first AI / ML model to the UE, wherein    the initial message and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.  (Supplementary note 40)      A radio access network, RAN, node, comprising:    means for receiving from, or transmitting to, a user equipment, UE, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; and    means for transmitting, as part of the current session, to the UE, at least one message for transferring the information related to the first AI / ML model to the UE, wherein    the initial message and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.

[0269] This application is based upon and claims the benefit of priority from Great Britain Patent Application No. 2506896.6, filed on May 6, 2025, the disclosure of which is incorporated herein in its entirety by reference.

[0270] 1 COMMUNICATION SYSTEM3 USER EQUIPMENT5 RAN NODE7 CORE NETWORK9 CELL10 CONTROL PLANE FUNCTIONS11 USER PLANE FUNCTIONS20 EXTERNAL DATA NETWORK241 DATA COLLECTION243 MODEL TRAINING245 INFERENCE247 ACTOR249 MANAGEMENT251 MODEL STORAGE31 TRANSCEIVER CIRCUIT33 ANTENNA35 USER INTERFACE37 CONTROLLER39 MEMORY41 OPERATING SYSTEM43 COMMUNICATIONS CONTROL MODULE51 TRANSCEIVER CIRCUIT53 ANTENNA55 USER INTERFACE57 CONTROLLER59 MEMORY61 OPERATING SYSTEM63 COMMUNICATIONS CONTROL MODULE

Claims

1. A method performed by a user equipment, UE, the method comprising:transmitting to, or receiving from, a radio access network, RAN, node, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; andreceiving, as part of the current session, from the RAN node, at least one message for transferring the information related to the first AI / ML model to the UE, whereinthe initial message, and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.

2. The method according to claim 1, wherein the information related to the first AI / ML model comprises at least one of: at least part of the first AI / ML model; at least part of an AI / ML model dataset for the first AI / ML model; and / or at least part of a set of parameters for use in configuring the first AI / ML model.

3. The method according to claim 1 or 2, wherein the initial message is transmitted from the UE to the RAN node.

4. The method according to claim 3, wherein the initial message is configured for requesting provision the first AI / ML model, and the at least one message for transferring the information related to the first AI / ML model is configured for providing at least part of the first AI / ML model.

5. The method according to claim 3, wherein the initial message is configured for requesting provision an AI / ML model dataset for the first AI / ML model, and the at least one message for transferring the information related to the first AI / ML model is configured for providing at least part of the AI / ML model dataset.

6. The method according to claim 3, wherein the initial message is configured for requesting a set of parameters for use in configuring the first AI / ML model, and the at least one message for transferring the information related to the first AI / ML model is configured for providing at least part of the parameters for use in configuring the first AI / ML model.

7. The method according to claim 1 or 2, wherein the initial message is transmitted from the RAN node to the UE.

8. The method according to claim 7, wherein the initial message, and the at least one message for transferring the information related to the first AI / ML model, are each respectively configured for providing at least part of the first AI / ML model.

9. The method according to claim 7, wherein the initial message, and the at least one message for transferring the information related to the first AI / ML model, are each respectively configured for providing at least part of an AI / ML model dataset for the first AI / ML model.

10. The method according to claim 7, wherein the initial message, and the at least one message for transferring the information related to the first AI / ML model, are each respectively configured for providing at least part of a set of parameters for use in configuring the first AI / ML model.

11. The method according to any one of claims 7 to 10, wherein the information related to the first AI / ML model is received via a broadcast or multicast transmission from the RAN node.

12. The method according to claim 11, wherein the information related to the first AI / ML model is received via one or more system information blocks and / or a multicast common channel, MCCH, based mechanism.

13. The method according to claim 11 or 12, wherein the information related to the first AI / ML model is received periodically, optionally via a data sharing channel.

14. The method according to any one of claims 7 to 13, wherein the information related to the first AI / ML model is received regardless of whether the UE is in a radio resource control, RRC, connected state, an RRC inactive state, or an RRC idle state.

15. The method according to any one of claims 1 to 14, wherein the method further comprises transmitting to, or receiving from the RAN node, at least one further initial message for at least one further session for transferring information related to the first AI / ML model, or related to at least one further AI / ML model, to the UE, and receiving, from the RAN node, at least one further message for transferring the information related to the first AI / ML model, or related to at least one further AI / ML model, to the UE, wherein the at least one further initial message, and each message for transferring the information related to the first AI / ML model, or related to at least one further AI / ML model, carries at least one further common transaction identifier associated with the at least one further session.

16. The method according to claim 15, wherein at least one further session for transferring the information related to the first AI / ML model, or related to at least one further AI / ML model, to the UE occurs subsequent to the current session for transferring the information related to the first AI / ML model to the UE.

17. The method according to claim 15 or 16, wherein at least one further session for transferring the information related to the first AI / ML model, or related to at least one further AI / ML model, to the UE occurs at least partially in parallel with the current session for transferring the information related to the at least one further AI / ML model to the UE.

18. The method according to any one of claims 1 to 17, wherein the, or each, transaction identifier is configured to indicate a purpose of a message that carries that transaction identifier.

19. The method according to any one of claims 1 to 18, wherein the at least one message for transferring the information related to the first AI / ML model includes, a final message of the current session, wherein the final message of the current session carries at least one of: information indicating that the final message is a last message of the current session; and / or information indicating a length of at least part of the final message.

20. The method according to claim 19, wherein the information indicating a length of at least part of the final message indicates a length of padding bits used in the message or a length of information bits used in the message.

21. The method according to any one of claims 1 to 20, wherein the at least one message for transferring the information related to the first AI / ML model and / or the initial message carries an identifier of the AI / ML model to which that message relates.

22. The method according to claim 21, wherein the identifier of the AI / ML model to which that message relates comprises a global identifier, a local identifier, or a combination of a cell identifier and a local identifier.

23. The method according to any one of claims 1 to 22, wherein the at least one message for transferring the information related to the first AI / ML model and / or the initial message carries a sequence number of a data unit carried by that message, and / or a segmentation number of a segment of a data unit carried by that message, optionally wherein the sequence number of the data unit carried by that message, and / or the segmentation number of the segment of the data unit carried by that message, is unique within the current session for data units transmitted in the same communication direction.

24. The method according to claim 23, wherein in a case where the UE receives, during the current session, a later message for transferring the information related to the first AI / ML model that includes a sequence number and / or segmentation number that was also included in an earlier message for transferring the information related to the first AI / ML model, the method further comprises discarding the later message.

25. The method according to any one of claims 1 to 24, wherein the method further comprises transmitting, to the RAN node, at least one feedback message for positively acknowledging, or for negatively acknowledging, reception of a data unit depending on whether that data unit has been successfully received or not.

26. The method according to claim 25, wherein the at least one message for transferring the information related to the first AI / ML model and / or the initial message carries information indicating whether an acknowledgement is requested for a data unit carried by that message.

27. The method according to claim 26, wherein the at least one feedback message is transmitted in respect of a data unit, in a case where an acknowledgement is requested for that data unit.

28. The method according to any one of claims 25 to 27, wherein the at least one feedback message carries at least one of: a transaction identifier indicating a purpose of the message is for providing feedback; a segmentation number of the data unit to which the at least one feedback message relates; and / or information indicating whether the data unit to which the at least one feedback message relates is acknowledged or negatively acknowledged.

29. The method according to any one of claims 1 to 28, wherein the method further comprises receiving, from the RAN node, a message for aborting the current session.

30. The method according to claim 29, wherein the message for aborting the current session carries a transaction identifier indicating a purpose of the message is for aborting the current session.

31. A method performed by radio access network, RAN, node, the method comprising:receiving from, or transmitting to, a user equipment, UE, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; andtransmitting, as part of the current session, to the UE, at least one message for transferring the information related to the first AI / ML model to the UE, whereinthe initial message and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.

32. The method according to claim 31, wherein at least one message for transferring the information related to the first AI / ML model and / or the initial message carries information indicating whether an acknowledgement is requested for a data unit carried by that message, and, in a case where acknowledgement is requested for a previously transmitted data unit and a positive acknowledgement is not subsequently received in respect of the previously transmitted data unit, the method further comprises retransmitting, to the UE, the previously transmitted data unit in at least one further message for transferring the information related to the first AI / ML model.

33. The method according to claim 32, wherein the retransmitting of the previously transmitted data unit occurs based on a configured or predefined timeout period, relative to a time at which the previously transmitted data unit was previously transmitted.

34. The method according to claim 32 or 33, wherein the retransmitting is repeated until a positive acknowledgement is received in respect of the previously transmitted data unit, or until a configured maximum number of transmissions, or retransmissions, of the previously transmitted data unit is reached.

35. The method according to claim 34, wherein, in a case where the configured maximum number of transmissions, or retransmissions, is reached, the method further comprises transmitting, to the UE, a message for aborting the current session.

36. The method according to any one of claims 32 to 35, wherein, transmission of a subsequent message for transferring the information related to the first AI / ML model that carries a subsequent data unit is deferred until a positive acknowledgement is received in respect of the previously transmitted data unit.

37. The method according to any one of claims 31 to 36, wherein, in a case where the UE is to be handed over to a further RAN node during the current session, the method further comprises providing, to the further RAN node, a transmission status of the information related to the first AI / ML model being transferred to the UE as part of the current session.

38. The method according to any one of claims 31 to 37, wherein, in a case where the UE is to be handed over to a further RAN node during the current session, method further comprises providing, to the further RAN node at least one of: the common transaction identifier; an identifier of an AI / ML model to which the current session relates; a current sequence number for data units being transmitted to the UE; and / or a current segmentation number for data units being transmitted to the UE.

39. A user equipment, UE, comprising:means for transmitting to, or receiving from, a radio access network, RAN, node, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; andmeans for receiving, as part of the current session, from the RAN node, at least one message for transferring the information related to the first AI / ML model to the UE, whereinthe initial message and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.

40. A radio access network, RAN, node, comprising:means for receiving from, or transmitting to, a user equipment, UE, an initial message for a current session for transferring information related to a first artificial intelligence (AI) / machine learning (ML) model to the UE; andmeans for transmitting, as part of the current session, to the UE, at least one message for transferring the information related to the first AI / ML model to the UE, whereinthe initial message and each message for transferring the information related to the first AI / ML model, carries a common transaction identifier associated with the current session.