Method for lifecycle management of artificial intelligence and machine learning models in wireless networks

Abstract

This disclosure generally relates to wireless communication networks, and more particularly to configuring and providing artificial intelligence (AI) and / or machine learning (ML) functions and models in terminal devices and network nodes within wireless communication networks. For example, such AI / ML functions and models may reside on the wireless terminal side or the wireless network side. The wireless terminal and the network can perform a collaborative process to determine, configure, activate, or deactivate a set of AI / ML functions and models for adaptive prediction and inference provided by the network, through a series of triggering messages.

Description

Technical Field

[0001] This disclosure generally relates to wireless communication networks, and more particularly to configuring and providing artificial intelligence (AI) and / or machine learning (ML) capabilities and models in terminal devices and network nodes. Background Technology

[0002] In wireless communication systems, determining adaptive network configurations of communication resources, particularly within the air communication interface, can require lengthy and resource-intensive measurement / reporting processes and / or significant computational power. These types of communication resource configurations may include, but are not limited to, beam management, channel state information (CSI) feedback compression and decompression, and wireless terminal localization / orientation. The correlation between various network conditions and these adaptive resource configurations can be learned through artificial intelligence (AI) techniques and models, and used to assist in providing wireless communication resources. Therefore, it may be desirable to provide an efficient mechanism for configuring, selecting, and providing various AI models deployed on terminal devices or wireless network nodes based on the capabilities of the terminal devices and the wireless network conditions between these terminal devices and the wireless network. Summary of the Invention

[0003] This disclosure generally relates to wireless communication networks, and more particularly to configuring and providing artificial intelligence (AI) and / or machine learning (ML) functions and models in terminal devices and network nodes. For example, such AI / ML functions and models may reside on the wireless terminal side or the wireless network side. The wireless terminal and the network may perform a collaborative process to determine, configure, activate, or deactivate a set of AI / ML functions and models for adaptive prediction and inference provided by the network, through a series of triggering messages.

[0004] In one example implementation, a method performed by a user equipment (UE) communicating with a network (NW) in a wireless communication network is disclosed. The method may include: performing a capability reporting process to transmit the UE's AIML capabilities to the NW in an Artificial Intelligence or Machine Learning (AIML) capability report; performing an applicable AIML reporting process to transmit a set of applicable AIML features, functions, or models of the UE to the NW; receiving an AIML configuration for one or more applicable AIML features, functions, or models selected by the NW; and performing a prediction based on the AIML configuration using the selected one or more applicable AIML features, functions, or models.

[0005] In the above example implementation, the capability reporting process includes: the UE sending an AIML capability report request to the NW, and sending the AIML capability report to the NW after receiving an AIML capability query from the NW.

[0006] In any of the above example implementations, the AIML capability report request sent by the UE to the NW is triggered by at least one of the following: at least one of the AIML features, functions, or models stored at the UE has been changed; the computing resource requirements of at least one of the AIML features, functions, or models stored at the UE have been changed; or at least one AIML feature, function, or model has been added to the UE or configured for the UE by the NW.

[0007] In any of the above example implementations, the AIML capability report request includes uplink RRC signaling, which includes uplink UE Assistance Information (UAI) messages.

[0008] In any of the above example implementations, the AIML capability report request includes at least one of the following: one or more indications for indicating the AIML features, functions, or models that have been changed at the UE; or one or more AIM change indications for indicating the type of change of the AIML features, functions, or models that have been changed at the UE.

[0009] In any of the above example implementations, the AIML capability report includes at least one of the following: a list of one or more AIML-based features supported by the UE; one or more AIML-based functions associated with and supported by the UE; one or more AIML-based models supported by the UE and associated with the one or more AIML-based functions or the one or more AIML-based features; an indication of the supported radio frequency bands for each AIML-based feature, function, or model; the computational resource consumption associated with one or more AIML-based features, functions, or models; the maximum computational resources that the UE can provide to support the AIML-based features, functions, or models; or one or more configurations, scenarios, contexts, or conditions associated with the training of the AIML-based features, functions, or models.

[0010] In any of the above example implementations, the set of applicable AIML features, functions, or models includes at least one of the following: AIML-based spatial beam prediction; AIML-based temporal beam prediction; AIML-based channel state information (CSI) feedback compression or decompression; or AIML-based temporal CSI prediction.

[0011] In any of the above example implementations, executing the applicable AIML report includes: receiving an applicable AIML report request from the NW and sending the applicable AIML report to the NW.

[0012] In any of the above example implementations, the applicable AIML report request includes a system information message or a special downlink RRC message constructed for an AIML report request for an applicable AIML feature, function, or model at the UE.

[0013] In any of the above example implementations, the applicable AIML report is included in a UE assistance information message, an uplink MAC CE, or a special uplink RRC message for reporting applicable AIML features, functions, or models at the UE.

[0014] In any of the above example implementations, the applicable AIML report includes a bit string indicating the applicability of AIML features, functions, or models supported within the UE capability.

[0015] In any of the above example implementations, the applicable AIML report request includes an RRC reconfiguration message, and the applicable AIML report is included in an RRC reconfiguration complete message.

[0016] In any of the above example implementations, the applicable AIML report includes a bit string indicating the applicability of AIML features, functions, or models supported within the AIML capabilities of the UE.

[0017] In any of the above example implementations, the applicable AIML report request includes at least one of the following: a list of additional conditions associated with at least one AIML feature, function, or model; one or more indicators for identifying the at least one AIML feature, function, or model associated with the additional conditions; or one or more indicators for identifying at least one AIML feature, function, or model, wherein the applicable AIML report is requested for the at least one AIML feature, function, or model.

[0018] In any of the above example implementations, the one or more indicators are provided by cell group, by frequency band, or by cell.

[0019] In any of the above example implementations, the list of additional conditions includes at least one of the following: an indication of cell range; the antenna height of the base station in the current cell; non-line-of-sight probability; indoor / outdoor conditions; a downlink transmission beamcodebook indication for AIML model training; the antenna array dimension of the base station; the downtilt angle of the base station's antenna; or at least one beam pattern.

[0020] In any of the above example implementations, the UE sending the applicable AIML report to the NW is triggered by at least one of the following: the RRC configuration associated with the applicable AIML report has been established, and no other indications of applicable AIML features, functions, or models have been previously sent to the NW; at least one AI / ML function or model for the AIML-based feature has been activated; or at least one AI / ML function or model for the configured AIML-based feature stored in the UE has changed since the UE last made an applicable AIML report.

[0021] In some other example implementations, a method performed by a network node communicating with a user equipment (UE) in a wireless communication network is disclosed. The method may include: receiving an AIML capability report from the UE, the AIML capability report indicating the AIML capabilities of the UE; obtaining an applicable AIML report from the UE, the applicable AIML report indicating a set of applicable AIML features, functions, or models for the UE; generating an AIML configuration for one or more selected applicable AIML features, functions, or models; and sending the AIML configuration to the UE so that the UE can perform predictions using the selected one or more applicable AIML features, functions, or models according to the AIML configuration.

[0022] In the above example implementation, the method may further include: receiving an AIML capability report request from the UE before receiving the AIML capability report from the UE.

[0023] In any of the above example implementations, the method may further include: sending an AIML capability query to the UE so that the UE can respond via the AIML capability report.

[0024] In any of the above example implementations, the AIML capability report request includes uplink RRC signaling, which includes uplink UE Assistance Information (UAI) messages.

[0025] In any of the above example implementations, the AIML capability report request includes at least one of the following: one or more indications for indicating the AIML features, functions, or models that have been changed at the UE; or one or more AIML change indications for indicating the type of change to the AIML features, functions, or models that have been changed at the UE.

[0026] In any of the above example implementations, the AIML capability report includes at least one of the following: a list of one or more AIML-based features supported by the UE; one or more AIML-based functions associated with and supported by the UE; one or more AIML-based models supported by the UE and associated with the one or more AIML-based functions or the one or more AIML-based features; an indication of the supported radio frequency bands for each AIML-based feature, function, or model; the computational resource consumption associated with one or more AIML-based features, functions, or models; the maximum computational resources that the UE can provide to support the AIML-based features, functions, or models; or one or more configurations, scenarios, contexts, or conditions associated with the training of the AIML-based features, functions, or models.

[0027] In any of the above example implementations, the set of applicable AIML features, functions, or models includes at least one of the following: AIML-based spatial beam prediction; AIML-based temporal beam prediction; AIML-based channel state information (CSI) feedback compression or decompression; or AIML-based temporal CSI prediction.

[0028] In any of the above example implementations, the method may further include: sending an applicable AIML report request to the UE, so that the UE sends the applicable AIML report in response.

[0029] In any of the above example implementations, the applicable AIML report request includes a system information message or a special downlink RRC message constructed for an AIML report request for an applicable AIML feature, function, or model at the UE.

[0030] In any of the above example implementations, the applicable AIML report is included in a UE assistance information message, an uplink MAC CE, or a special uplink RRC message for reporting applicable AIML features, functions, or models at the UE.

[0031] In any of the above example implementations, the applicable AIML report includes a bit string indicating the applicability of AIML features, functions, or models supported within the UE capability.

[0032] In any of the above example implementations, the applicable AIML report request includes an RRC reconfiguration message, and the applicable AIML report is included in an RRC reconfiguration complete message.

[0033] In any of the above example implementations, the applicable AIML report includes a bit string indicating the applicability of AIML features, functions, or models supported within the AIML capabilities of the UE.

[0034] In any of the above example implementations, the applicable AIML report request includes at least one of the following: a list of additional conditions associated with at least one AIML feature, function, or model; one or more indicators for identifying the at least one AIML feature, function, or model associated with the additional conditions; or one or more indicators for identifying at least one AIML feature, function, or model, wherein the applicable AIML report is requested for the at least one AIML feature, function, or model.

[0035] In any of the above example implementations, the one or more indicators are provided by cell group, by frequency band, or by cell.

[0036] In any of the above example implementations, the list of additional conditions includes at least one of the following: an indication of cell range; the antenna height of the base station in the current cell; non-line-of-sight probability; indoor / outdoor conditions; a downlink transmission beamcodebook indication for AIML model training; the antenna array dimension of the base station; the downtilt angle of the base station's antenna; or at least one beam pattern.

[0037] A UE or NW according to any of the above methods is disclosed. The UE or NW may include a processor and a memory, wherein the processor is configured to read computer code from the memory to cause the UE or NW to perform the method according to any of the above methods.

[0038] A non-transitory computer-readable program medium is also disclosed, on which computer code is stored. When executed by a processor of a UE or NW node according to any of the above methods, the computer code is configured to cause the processor to implement any of the above methods.

[0039] The above embodiments and other aspects and alternatives to their implementations will be described in more detail in the accompanying drawings, description and claims below. Attached Figure Description

[0040] Figure 1 An example wireless communication network including a wireless access network, a core network, and a data network is shown.

[0041] Figure 2 An example radio access network is shown, including multiple mobile stations / terminals or user equipment (UEs) communicating with each other via an over-the-air wireless communication interface, as well as radio access network nodes.

[0042] Figure 3An example radio access network (RAN) architecture is shown.

[0043] Figure 4 An example communication protocol stack, including various network layers, is shown in a wireless access network node or wireless terminal device.

[0044] Figure 5 An example overall process for AI / ML lifecycle management for UE-side AI models is shown.

[0045] Figure 6 An example procedure for AI / ML capability reporting triggered by a UE is shown.

[0046] Figure 7 An example uplink MAC CE for reporting available UE AI / ML functions is shown.

[0047] Figure 8 An example process for testing applicable AI / ML functions is shown.

[0048] Figures 9A to 9E An example RRC configuration structure for AI / ML functions is shown.

[0049] Figure 10 An example DL MAC CE for signaling activation / deactivation of AI / ML functions is shown.

[0050] Figure 11 An example procedure for attaching change reports to UEs is shown. Detailed Implementation

[0051] This disclosure will now be described in detail with reference to the accompanying drawings, which form part of this disclosure and illustrate specific examples of embodiments by way of illustration. However, this disclosure may be implemented in a variety of different forms, and therefore the subject matter covered or claimed is intended to be construed as not being limited to any of the embodiments set forth below.

[0052] Throughout the specification and claims, terms may have nuanced meanings beyond their explicitly stated meanings, implied or suggested by the context. Similarly, the phrases “in one embodiment” or “in some embodiments” as used herein do not necessarily refer to the same embodiment, and the phrases “in another embodiment” or “in other embodiments” as used herein do not necessarily refer to different embodiments. Likewise, the phrases “in one implementation” or “in some implementations” as used herein do not necessarily refer to the same implementation, and the phrases “in another implementation” or “in other implementations” as used herein do not necessarily refer to different implementations. For example, the claimed subject matter is intended to include a combination of exemplary embodiments or implementations, in whole or in part.

[0053] Generally, terms can be understood, at least in part, based on their use in the context. For example, terms such as “and,” “or,” or “and / or” as used herein can include a variety of meanings, which can depend at least in part on the context in which these terms are used. Generally, if “or” is used in an associative list, such as A, B, or C, it is intended to indicate A, B, and C in an inclusive sense, and A, B, or C in an exclusive sense. Furthermore, the terms “one or more” or “at least one” as used herein, depending at least in part on the context, can be used to describe any feature, structure, or characteristic in a singular sense, or can be used to describe a combination of features, structures, or characteristics in a plural sense. Similarly, terms such as “a,” “an,” or “the / that” can also be understood to convey either a singular or a plural usage, depending at least in part on the context. Moreover, the terms “based on” or “determined by” can be understood to not necessarily convey a set of exclusive factors, and may allow for the presence of other factors that are not necessarily explicitly described, also depending at least in part on the context.

[0054] This disclosure generally relates to wireless communication networks, and more particularly to configuring and providing artificial intelligence (AI) and / or machine learning (ML) functions and models in terminal devices and network nodes. For example, such AI / ML functions and models may reside on the wireless terminal side or the wireless network side. The wireless terminal and the network may perform a collaborative process to determine, configure, activate, or deactivate a set of AI / ML functions and models for adaptive prediction and inference provided by the network, through a series of triggering messages.

[0055] Wireless Network Overview like Figure 1As shown in Figure 100, the example wireless communication network may include wireless terminal equipment or user equipment (UE) 110, UE 111, and UE 112, carrier network 102, various service applications 140, and other data networks 150. The wireless terminal equipment or UE may alternatively be referred to as a wireless terminal. For example, carrier network 102 may include access network nodes 120 and 121, and a core network 130. Carrier network 110 may be configured to transmit voice, data, and other information (collectively referred to as data traffic) between the UE and service application 140 or between the UE and other data networks 150. Access network nodes 120 and 121 may be configured as various wireless access network nodes (WANNs, alternatively referred to as wireless base stations) to interact with the UE at one end of the communication session and with the core network 130 at the other end. The term "access network" may be used more broadly to refer to the combination of wireless terminal devices 110, 111, and 112 with access network nodes 120 and 121. The radio access network may also be referred to as the Radio Access Network (RAN). The core network 130 may include various network nodes configured to control communication sessions and perform network access management and traffic routing. Service applications 140 may be hosted by various application servers deployed outside of but connected to the core network 130. Similarly, other data networks 150 may also be connected to the core network 130.

[0056] exist Figure 1 In the example wireless communication network 100, each UE can communicate with each other through a wireless access network. For example, UE 110 and UE 112 can be connected to the same access network node 120 and communicate through the same access network node 120. Each UE can communicate with each other through both the access network and the core network. For example, UE 110 can be connected to access network node 120, and UE 111 can be connected to access network node 121, and thus, UE 110 and UE 111 can communicate with each other through access network nodes 120 and 121, as well as the core network 130. UEs can also communicate with service applications 140 and data networks 150 through the core network 130. In addition, each UE can communicate directly with each other through sidelink communication (as shown in 113).

[0057] Figure 2An example system diagram of a radio access network 120 is also shown, which includes a WANN 202 serving UE 110 and UE 112 via an air interface 204. The radio transmission resources of the air interface 204 include a combination of frequency resources, time resources, and / or spatial resources. Each of UE 110 and UE 112 can be a mobile or fixed terminal device equipped with a mobile access unit such as a SIM / USIM (Subscriber Identity Module / Universal Subscriber Identity Module) module to access the wireless communication network 100. Both UE 110 and UE 112 can be implemented as terminal devices including, but not limited to, mobile phones, smartphones, tablets, laptops, in-vehicle communication devices, roadside communication devices, sensor devices, smart home appliances (such as televisions, refrigerators, and ovens), or other devices capable of wireless communication over a network. Figure 2 As shown, each of these UEs (such as UE 112) may include transceiver circuitry 206 coupled to one or more antennas 208 for enabling wireless communication with WANN 120 or another UE (such as UE 110). Transceiver circuitry 206 may also be coupled to processor 210, which may also be coupled to memory 212 or other storage devices. Memory 212 may be transient or non-transient and may store computer instructions or code that, when read and executed by processor 210, cause processor 210 to implement the methods described herein.

[0058] Similarly, WANN 120 may include a wireless base station or other wireless network access point capable of wirelessly communicating with one or more UEs via air interface 204 and with core network 130. For example, WANN 120 may not be implemented in the form of a 2G base station, 3G nodeB, LTE eNB, 4G LTE base station, 5G NR base station of 5G gNB, 5G centralized unit base station, 5G distributed unit base station, or 6G base station. Each type of WANN may be configured to perform a corresponding set of wireless network functions. WANN 202 may include transceiver circuitry 214 coupled to one or more antennas 216, which may include various forms of antenna towers 218, to enable wireless communication with UEs 110 and 112. Transceiver circuitry 214 may be coupled to one or more processors 220, which may be further coupled to memory 222 or other storage devices. The memory 222 may be transient or non-transient, and may store instructions or code that, when read and executed by one or more processors 220, cause one or more processors 220 to perform the various functions of the WANN 120 described herein.

[0059] like Figure 2 In the example shown, data packets in a radio access network can be transmitted as Protocol Data Units (PDUs). The data included can be encapsulated as PDUs at various network layers, with nested and / or layered protocol headers. When a connection is established between the transmitting and receiving ends (e.g., a radio link control (RRC) connection), communication between the transmitting device or receiving end (these two terms can be used interchangeably) and the receiving device or receiving end (these two terms can also be used interchangeably) can occur between the PDUs. Either the transmitting or receiving device can be a wireless terminal device (such as… Figure 2 Devices 110 and 120), or wireless access network nodes (such as...) Figure 2 (Node 202). Each device can be both a sending device and a receiving device for bidirectional communication.

[0060] Figure 1 The core network 130 may include various geographically distributed and interconnected network nodes to provide network coverage over the service area of ​​the carrier network 102. These network nodes may be implemented as dedicated hardware network nodes. Alternatively, these network nodes may be virtualized and implemented as virtual machines or software entities. Each of these network nodes may be configured with one or more types of network functions that collectively provide provisioning and routing functions for the core network 130.

[0061] Return to the wireless radio access network (RAN). Figure 3 An example RAN 340 communicating with core network 310 and radio terminals UE1 through UE7 is shown. RAN 340 may include one or more types of radio base stations or WANN 320 and WANN 321, which may include, but are not limited to, gNB, eNodeB, NodeB, or other types of base stations. RAN 340 may perform backhaul with core network 310. WANN 320 may further include multiple independent access network nodes, for example, in the form of a Central Unit (CU) 322 and one or more Distributed Units (DUs) 324 and 326. CU 322 is connected to DU1 324 and DU2 326 via various interfaces, such as F1 interfaces. The F1 interface may also include, for example, an F1-C interface and an F1-U interface, which can be used to carry control plane information and user plane data, respectively. In some embodiments, CU may be a gNB Central Unit (gNB-CU), and DU may be a gNB Distributed Unit (gNB-DU). While the various implementations described below are provided in the context of 5G cellular wireless networks, the basic principles described herein are applicable to other types of wireless access networks, including but not limited to other current or future generations of cellular networks such as 4G LTE and 6G networks, as well as Wi-Fi (Wireless Fidelity), Bluetooth, ZigBee, and WiMax (Worldwide Interoperability for Microwave Access) networks.

[0062] Each UE can be connected to the network via the air interface through the WANN 320. Each UE can be served by at least one cell. Each cell is associated with a coverage area. These cells are alternatively referred to as serving cells. The coverage areas between cells may partially overlap. Each UE may actively communicate with at least one cell when it is potentially connectable to or able to connect to more than one cell. Figure 1 In the example, UE1, UE2, and UE3 can be served by cell 1 330 of DU1, while UE4 and UE5 can be served by cell 2 332 of DU1, and UE6 and UE7 can be served by cell 3 associated with DU2. In some implementations, a UE can be served by two or more cells simultaneously. Each UE can be mobile, and the signal strength and quality at the UE from the various cells can depend on the UE's location and mobility.

[0063] In some example implementations, Figure 3The cells shown can be alternatively referred to as serving cells. Serving cells can be grouped into serving cell groups (CGs). A serving cell group can be a master CG (MCG) or a secondary CG (SCG). Each type of cell group may contain one master cell and one or more secondary cells. For example, a master cell in an MSG may be referred to as a PCell, while a master cell in an SCG may be referred to as a PScell. Secondary cells in either an MCG or an SCG may be referred to as SCells. Master cells including both PCells and PScells may be collectively referred to as spCells (special cells). All of these cells may be referred to as serving cells or cells. Unless otherwise specified, the terms "cell" and "serving cell" are generally used interchangeably. The term "serving cell" may refer to a cell that is currently serving, will be serving, or may be serving a UE. In other words, a "serving cell" may not currently be serving a UE. Although the various embodiments described below may sometimes refer to one of the types of serving cells described above, the basic principles apply to all types of serving cells in both types of serving cell groups.

[0064] 4 further illustrates in Figures 1 to 3 A simplified diagram of the various network layers involved in transmitting user plane PDUs from transmitting device 402 to receiving device 404 in an example wireless access network. Figure 4 It is not intended to include all the necessary equipment components or network layers for handling the transmission of PDUs. Figure 4 This illustrates that the data encapsulated in the upper network layer 420 of the transmitting device 402 can pass through the packet data convergence protocol layer (PDCP) of the transmitting device. Figure 4 The PDUs (not shown in the diagram) and the radio link control (RLC) layer 422, the physical (PHY) layer and radio interface (as shown in 406) of the transmitting and receiving devices, and the media access control (MAC) layer 434 and RLC layer 432 of the receiving device are transmitted to the corresponding upper layer 430 of the receiving device 304 (such as the radio resource control or RRC layer). Various network entities in each of these layers can be configured to handle the transmission and retransmission of PDUs.

[0065] exist Figure 4 In the middle, the upper layer 420 can be referred to as layer 3 or L3, while the intermediate layers (such as RLC layers and / or MAC layers and / or PDCP layers) are... Figure 4(Not shown in the diagram) can be collectively referred to as layer 2 or L2, and the term layer 1 is used to refer to layers such as the physical layer and the radio interface association layer. In some instances, the term "lower layer" can be used to refer to the set of L1 and L2, while the term "higher layer" can be used to refer to layer 3. In some cases, the term "lower layer" can be used to refer to layers among L1, L2, and L3 that are lower than the current reference layer. Control signaling can be initiated and triggered within each of the layers from L1 to L3 and within the various network layers therein. These signaling messages can be encapsulated and concatenated into lower-layer packets and transmitted through allocated control or data air radio resources and interfaces. The term "layer" typically includes its various corresponding entities. For example, the MAC layer covers the corresponding MAC entities that can be created. Layer 1, for example, covers the PHY entity. Layer 2, for example, covers the MAC layer / entity, RLC layer / entity, Service Data Adaptation Protocol (SDAP) layer, and / or PDCP layer / entity.

[0066] AI / ML-assisted wireless network provisioning and configuration AI and ML (often referred to as AL) can facilitate more efficient configuration and provisioning in wireless networks. At the heart of a general AI framework are various AI models. AI models typically contain a large number of model parameters, which are determined through a training process where correlations in the training dataset are learned and embedded into the trained model parameters. Therefore, the trained model parameters can be used to generate inferences or predictions from a set of input datasets. AI models are particularly well-suited for situations where there is little traceable determinism or analytical derivation path between the input data and the output, but correlations in the data can be identified from historical data and embedded into the AI ​​model through the training process.

[0067] In wireless communication systems such as those described above, determining adaptive network configurations may rely on empirical characteristics and may require lengthy measurement processes and / or substantial computational power. These types of configurations may include, but are not limited to, air interface beam management, channel state information (CSI) feedback compression and decompression, and wireless terminal location. The correlation between various network conditions and these adaptive configurations can be learned using AI techniques. Therefore, using AI models to assist network configuration can help reduce measurement and computational requirements, resulting in more agile and efficient network configurations. Thus, it may be desirable to provide a mechanism for providing AI models based on the AI ​​capabilities of the terminal device and network conditions to assist in the adaptive determination of these network configurations.

[0068] For example, AI technology can be applied to beam management in air communication interfaces. In current implementations, beam management typically relies on exhaustive search beam scans and measurements. In other words, the network (NW) performs a comprehensive beam scan by sending a sufficient number of reference signals. The UE can be configured to monitor and measure each reference signal and then report the measurements to the NW so that the NW can determine the optimal beam the UE should switch to. However, this process is resource- and power-intensive and time-consuming. By training an AI model that embeds learned correlations between various network condition parameters, the optimal beam can be inferred with reasonable accuracy using fewer measurements (or fewer reference signals), thus achieving so-called AI / ML-based spatial beam management. In some implementations, the AI ​​model can help identify inferences about the best candidate beams for future time instances using other network conditions, and then only scan and measure the candidate beams to select the beam to use in the future time instance, thus achieving so-called AI / ML-based temporal beam management. Furthermore, since beam configuration is closely related to the UE's location, AI technology can also be further used to infer or predict the UE's trajectory or location, thereby indirectly helping to select the optimal beam.

[0069] For example, AI technology can be applied to Channel State Information (CSI) feedback. Traditionally, CSI feedback can be achieved using codebooks known to both the UE and NW. The UE can measure the CSI and obtain the measurement result, then map the measurement result to the nearest vector in the codebook and send the index of that vector to the NW to save air interface resources. However, since the codebook is not infinite and cannot change dynamically over time, there is always a certain degree of mismatch, leading to uncontrolled CSI feedback errors as the radio environment changes. Therefore, AI can be applied, for example, to the compression-decompression of CSI feedback. Specifically, the CSI report can be compressed by a UE-side AI model and decompressed by a corresponding NW-side AI model. Such an AI model can be initially trained and continuously evolved over time and with the accumulation of network conditions. In some example implementations, the AI ​​model can help identify inferences about future CSI feedback values ​​using other network conditions, and then only measure the current reference signaling used for CSI, thus achieving so-called AI / ML-based CSI prediction.

[0070] For example, AI technology can be applied to UE positioning. Traditional UE positioning methods rely on Positioning Reference Signal (PRS) or Sounding Reference Signal (SRS). Examples include the DL positioning reference signal and the uplink sounding reference signal. Regardless of the alternative method used, the line-of-sight (LOS) beam is the key beam to be identified in order to generate the most accurate position estimate through trilateration on the NW side. However, in most cases, it is difficult to identify the LOS beam from other non-line-of-sight (NLOS) beams, resulting in inaccurate UE positioning. On the other hand, trained AI models can identify various patterns and correlations in the PRS and SRS to extract LOS information and provide more accurate UE positioning.

[0071] In some example implementations, the AI ​​model can reside on the wireless terminal side. In other implementations, the AI ​​model can reside on the network (NW) side. The AI ​​model can be provided as a service. The AI ​​model can be retrained and updated as needed. The UE or NW can determine which models to retrieve or retain, and how to update and configure these models to assist wireless communication, including but not limited to beam management, CSI feedback, UE positioning, etc. The selection, retrieval, storage, and execution of the AI ​​model can depend on the capabilities of the terminal device and / or NW. The terminal and NW can be configured to transmit these capabilities and network conditions to provide the selection, configuration, and use of AI models over time; this is often referred to as AI / ML Lifecycle Management (LCM).

[0072] AI / ML lifecycle management (LCM) for AI features / functions / models on the UE side AI / ML LCM can be performed at various levels or granularities. In one example implementation, AI / ML LCM can be performed at the level of AI / ML-based feature groups (e.g., AI / ML-based beam management groups, AI / ML-based CSI feedback groups), where the AI / ML-based beam management group (e.g., feature group) includes AI / ML-based spatial beam management aspects (e.g., features) and AI / ML-based temporal beam management aspects (e.g., features), and the AI / ML-based CSI feedback group includes AI / ML-based CSI feedback enhancement aspects (e.g., features) and AI / ML-based CSI prediction aspects (e.g., features). In another example implementation, AI / ML LCM can be performed at the AI / ML-based feature level. In another example implementation, AI / ML LCM can be performed at the AI / ML function level. In another example implementation, AI / ML LCM can be performed at the AI / ML model level. The AI / ML terms feature group, feature, function, and model are used only to denote the various levels at which AI / ML can be configured and applied. They can be hierarchically related. They may overlap in some cases. These levels can be divided in any suitable way to facilitate the configuration and management of AI / ML usage. Each of these levels can be hierarchical in itself. For example, an AI / ML model may include lower-level AI / ML models as internal components. AI / ML features may include lower-level sub-features, and similarly, AI / ML functions may include lower-level sub-functions.

[0073] For example, an AI / ML model can refer to a specifically trained algorithm that processes one or more inputs to generate a prediction as output. AI / MI models may include components such as various types of neural networks, regression algorithms, support vector machine algorithms, K-nearest neighbor algorithms, random forests, K-means clustering, principal component analysis, Bayesian networks, etc.

[0074] For example, AI / ML models can be trained to support specific AI features or AI functions. AI / ML features or functions can be implemented using different AI models, which may differ in their internal architecture, hyperparameters, training parameters, inputs, computational resource requirements, complexity, and prediction accuracy. For instance, an AI / ML function can comprise one or a set of AI / ML models.

[0075] For example, AI / MI-based features can include one or more AI / ML functions. In some example implementations, AI / ML features can be synonymous with AI / MI functions (e.g., an AI / ML function can represent a corresponding AI / ML-based feature). As another example, an AI / ML-based feature group can include one or more AI / ML-based features to form a feature class.

[0076] As an example only, AI / ML features can refer to the following categories: AI / ML-based spatial beam prediction, AI / ML-based temporal beam prediction, AI / ML-based CSI feedback compression and decompression, AI / ML-based CSI prediction, AI / ML-based temporal / spatial cell measurement result prediction for mobility, AI / ML-based temporal / spatial beam prediction for mobility, etc.

[0077] In some example implementations, AI / ML LCM may include at least one of the following non-restrictive aspects: • AI / ML features (feature groups) / functions / model recognition.

[0078] • Applicable AI / ML features / functions / model reports between wireless terminals and NW.

[0079] • AI / ML feature / function / model control (including selection, activation, deactivation, and rollback).

[0080] • AI / ML feature / function / model performance monitoring.

[0081] AI / ML features / functions / models can be located on the UE side or the NW side. The LCM of AI / MI features / functions / models on the UE side can depend on the UE capabilities.

[0082] UE-side AI / ML LCM - Overview Figure 5 The document illustrates an example implementation of the aforementioned LCM aspects of the AI / ML model for the radio terminal side (UE side) involving UE 502 and NW 504, which includes the following general steps: • Step 1: The UE capability reporting process can be performed between UE 502 and NW 504 for one or more AI / ML-based features / functions / models.

[0083] Step 2: NW 504 and UE 502 can additionally perform the applicable AI / ML feature / function / model reporting process.

[0084] Step 3: NW 504 and UE 502 can perform a preparation phase to test AI / ML features / functions / models.

[0085] • Step 4: NW 504 can configure the RRC configuration of AI / ML features / functions / models to UE 502 based on UE capabilities and / or applicable features / functions / model reports.

[0086] Step 5: NW 504 can send a message to UE 502 to activate / deactivate AI / ML features / functions / models.

[0087] Step 6: UE 502 can activate or deactivate AI / ML features / functions / models accordingly and perform predictions and inferences.

[0088] Step 7: UE 502 can then perform a prediction / inference or actual measurement and send the predicted and inferred or actual measured values ​​to NW 504 via one or more inference reports and / or measurement reports.

[0089] Therefore, in the above general example process, the UE can first report a set of supported AI / ML-based features / functions / models to the NW based on the UE's capabilities. Then, the UE and the NW can collaborate to determine: (1) among the supported AI / ML-based features / functions / models, the applicable or suitable features / functions / models for the UE to activate, based on capabilities, network conditions and various other factors, and / or (2) among the supported features / functions / models, the features / functions / models that are no longer applicable or suitable for the UE to activate.

[0090] UE-side AI / ML LCM – UE AI / ML Capability Report The above text Figure 5 In step 1, the capability reporting process for one or more AI / ML-based features / functions / models can be triggered by the UE or by the NW, such as... Figure 6 As shown in the example. Step 608 illustrates the AI / ML capability report triggered by the NW, while step 606, which precedes 608, is for the AI / ML capability report triggered by the UE.

[0091] Specifically, in one example implementation, at step 606, the UE 602 can proactively trigger an AI / ML capability report by sending a trigger message to the NW (e.g., a base station or core network node). The NW can then accept the request and initiate the UE AI / ML capability report process, as shown in step 608. In some other example implementations, as shown in step 608, the UE AI / ML capability report can be triggered by the NW and then executed between the UE and the NW without the UE triggering step 606.

[0092] In some example implementations, the UE may support a superset of AI / ML features / functions / models (supported features / functions / models). In some cases, the UE may only download / store a subset of that superset model (available features / functions / models) locally, based on its location and other factors. While the features / functions / models supported by the superset may remain stable over time, the superset itself may still change. Even if the superset of a feature / function / model is stable, the available features / functions / models may be more volatile and may change more frequently over time. Therefore, in 606, the UE may determine to trigger a UE AI / ML capability report under at least one of the following conditions: • The AI / ML features / functions / models supported by the superset have changed at the UE, which may include, but are not limited to, the following: Since the most recent UE capability report, at least one of the supported AI / ML features / functions / models has been updated.

[0093] Since the most recent UE capability report, at least one new AI / ML feature / function / model has become supported.

[0094] Since the most recent UE capability report, at least one AI / ML feature / function / model has been removed from the superset.

[0095] Since the most recent UE capability report, the overall computing resources for hyper-centralized AI / ML features / functions / models have changed.

[0096] • The AI / ML features / functions / models actually stored at the UE have changed, which may include, but are not limited to, the following: Since the most recent UE capability report, at least one of the AI / ML features / functions / models stored at the UE has been updated.

[0097] Since the most recent UE capability report, the UE has acquired at least one new AI / ML feature / function / model.

[0098] Since the most recent UE capability report, at least one stored feature / function / model has been removed from the UE.

[0099] Since the most recent UE capability report, the overall computing resources available for AI / ML features / functions / models have changed.

[0100] • At least one of the AI / ML features / functions / models has been configured for the UE by NW.

[0101] In some example implementations, the UE trigger message in 606 can be sent using one of the following formats: uplink (UL) MAC control element (MAC CE); or UL RRC signaling (e.g., UE Assistance Information (UAI) message); or protocol signaling terminated between the AI / ML logical layer and the NW logical entity / unit.

[0102] In some example implementations, such a trigger message in UE 606 may contain or indicate at least one of the following information or information items: • One or more AI / ML feature / function / model indicators: used to indicate changed AI / ML features / functions / models.

[0103] • Change type indication, where the change type may include, for example: 1) adding AI / ML features / functions / models; 2) updating stored AI / ML features / functions / models; 3) releasing AI / ML features / functions / models from supported AI / ML features / functions / models.

[0104] NW can use the information in UE trigger message 606 to determine whether and when... Figure 6 The actual request for UE AI / ML capability report in 608.

[0105] exist Figure 6 In step 608, the NW can send a query message to the UE (either as a trigger message for the NW to initiate a UE AI / ML capability report, or as a response to the UE's trigger message in step 606 when the NW determines that a report is requested) to obtain the UE AI / ML capability report. This query message can be called a UE capability query. UECapabityEnquiring ),like Figure 6 As shown.

[0106] In response, the UE can then send a UE AI / ML capability report to the NW. The message containing this report can be called a UE capability report. UECapability ),like Figure 6 As shown. In some example implementations, UECapability A message may include or indicate at least one of the following information or information items: • Supported AI / ML-based features or groups of features (list): A list indicating the AI / ML-based features supported by the UE (these features may be supported, but may be available or unavailable to the UE).

[0107] • One or more lists of supported AI / ML-based functions: A list of AI / ML-based functions supported by the UE for AI / ML-based features (these functions may be supported, but may be available or unavailable to the UE).

[0108] • A list of one or more supported AI / ML models: This indicates a list of AI / ML models supported by the UE for AI / ML-based features or functions.

[0109] • Frequency band or frequency band list indicator: Used to indicate the frequency bands supported for each AI / ML-based feature / function / model.

[0110] • Indicator of computational resource consumption for each AI / ML feature / function / model: A quantified value indicating the computational resources that an AI / ML-based feature / function / model (if activated) will consume.

[0111] • AI / ML Overall Computing Resource Indicator: Used to indicate the maximum quantified value of computing resources that the UE can support for AI / ML.

[0112] • Indicator of preparation time for each AI / ML feature / function / model: used to indicate the maximum or minimum preparation time from when the UE receives the activation signaling of the AI / ML feature / function / model to when the AI / ML feature / function / model is inferred to be actually capable of being executed.

[0113] • Training configuration (e.g., UE settings, gNB settings), scenario, and condition indication: These indicate the configuration, scenario, and conditions under which the AI / ML-based models within the AI / ML-based features / functions / models supported by the UE are trained. These configurations, scenarios, and conditions include, but are not limited to: For AI / ML-based spatial beam prediction and AI / ML-based temporal beam prediction: Information items related to the scene: Inter-Site Distance (ISD) is used to indicate the cell range associated with AI / ML training. Supported values ​​for ISD can be, for example, UMa, UMi, 200 meters, and 500 meters.

[0114] The antenna height of the base station in the cell associated with AI / ML training. The supported values ​​for the antenna height of the base station can be 1 meter, 2 meters, 5 meters, 10 meters, etc.

[0115] Non-line-of-sight probability of NLOS wireless propagation associated with AI / ML training.

[0116] Indoor / Outdoor indicator, used to indicate whether the scene is indoor or outdoor and / or the indoor / outdoor ratio for AI / ML training.

[0117] Information items used for gNB settings: .DL Tx beamcodebook indicator, used to indicate the DL TX beamcodebook used for AI / ML model training.

[0118] An indication of the dimensions of the gNB antenna array associated with AI / ML model training.

[0119] Downtilt angle of the gNB antenna associated with AI / ML model training.

[0120] Information items used for training beam conditions and beam sets (set A can refer to the beam group containing the best K beams predicted by the AI / ML model / feature, set B can refer to the measured beams, and the corresponding measurements and / or associated beam IDs can be used as input to the AI / ML model / feature, while the complete beam set can refer to all beams, e.g., all 64 beams in a specific configuration): Set B Beam Pattern Indication: Used to indicate the beam pattern of set B relative to set A. For example, set B used for training can be a uniformly distributed 1 / 4 subset of set A. Alternatively, set B may not be a subset of set A. Another example is that set B can be SSB while set A can be CSI-RS, or vice versa.

[0121] Beam pattern of set A: Used to indicate the beam pattern of set A relative to the complete beam set (e.g., 64 beams). For example, beam set A can be a uniformly distributed 1 / 4 subset of the complete beam set, or beam set A can be the complete beam set, etc.

[0122] Information items related to UE settings associated with model training, such as: .UE Speed: Used to indicate UE speed information related to AI / ML features / functions / model training.

[0123] UE Orientation: Used to indicate the UE orientation associated with AI / ML feature / function / model training. In one implementation, it indicates the maximum value of the UE orientation variation of the associated AI / ML feature / function / model, such as 45 degrees, 90 degrees, 120 degrees, 180 degrees, etc.

[0124] .UE Rx codebook.

[0125] .UE antenna array dimension.

[0126] The parameter (set) explicitly includes one or more of the above information items with IDs.

[0127] The parameter set implicitly indicates one or more of the above information with an ID, such as a cell identifier (e.g., CGI (Cell Global Identity), PCI (Physical Cell Identity), etc.), which may include information items about NW-specific configurations / settings, cell scenarios, etc.

[0128] For AI / ML-based CSI feedback compression / decompression (e.g., two-sided models): Model ID or list of model IDs: A list of model IDs used to indicate the UE parts supported by the AI / ML models on both sides.

[0129] Training dataset ID or list of training data IDs: Dataset IDs used to indicate a list of datasets used to train the UE portion of the AI ​​model on both sides.

[0130] UE setting indication: used to indicate the mapping pattern of the transceiver unit (TxRU), such as [2,8,2], [4,4,2], etc.

[0131] Scene indication: A training scene used to indicate the UE portion supported by the UE on both sides of the function / model, which may include at least one of the following: Inter-site distance (ISD) is used to indicate the cell range associated with AI / ML training. Supported values ​​for ISD can be, for example, UMa, UMi, 200 meters, and 500 meters.

[0132] Outdoor / Indoor indicator, used to indicate indoor, outdoor, and indoor / outdoor ratios for training.

[0133] The frequency range or band of training.

[0134] For AI / ML-based temporal CSI prediction: Similar to the scenario-related information items in the list above for AI / ML-based temporal beam management.

[0135] Similar to the list above for AI / ML-based temporal beam management, this is a gNB settings information item.

[0136] Similar UE settings information items to the list above used for AI / ML-based temporal beam management.

[0137] In some example implementations, the signaling structure used in 608 for reporting UE capabilities based on AI / ML features / functions / models can be carried in UE-NR-Capability or RF-Parameters In supportedBandListNR The RRC information element. In some alternative implementations, UE capability reports for features, functions, and models can be separated or independent. As just one example, information from reports related to AI / ML-based features or functions can be carried within... UE-NR- Capability In the RRC information cells, the AI / ML functions corresponding to AI / ML-based features or the AI / ML models corresponding to AI / ML functions can be found in... RF-parameters In supportedBandListNR It is carried and transmitted in RRC cells.

[0138] In some example implementations, the signaling structure used in 608 for reporting UE capabilities based on AI / ML features / functions / models can be carried in the protocol signaling that terminates between the AI / ML logical layer and the NW logical entity / unit.

[0139] In one example implementation of coordinating AI / ML computing capabilities between the MN and SN in a DC (e.g., dual connectivity) scenario, inter-node information regarding AI / ML computing capabilities between the MN and SN can be introduced. The SN / MN can send information to the MN / SN containing the computing power currently used by the activated AI / ML features / functions / models, ensuring that it does not exceed the maximum computing power reported in the UE capabilities. In one example implementation, this information from the MN to the SN can be a parameter describing a quantified value of the computing power consumed by the MN. In another example implementation, this information from the SN to the MN can be a parameter describing a quantified value of the computing power consumed by the MN. In yet another example implementation, this information from the MN to the SN can be a parameter describing a quantified value of the maximum computing power that the SN can consume for AI / ML.

[0140] In another example implementation of coordinating AI / ML computing capabilities between MN and SN in a DC scenario, a UL MAC CE can be introduced. This UL MAC CE can be triggered by the MAC entity under at least one of the following conditions: • Activated or deactivated AI / ML features / functions / models for this MAC entity.

[0141] • The total computing power quantization value of the activated AI / ML features / functions / models in this MAC entity is greater than or equal to the pre-configured / predefined maximum value.

[0142] In one example implementation of generating a UL MAC CE, at least one of the following conditions may need to be met: • A UL authorization has been received for a MAC entity that is different from the MAC entity that triggered the UL MAC CE.

[0143] • UL authorization has the capability to accommodate this UL MAC CE according to the LCP (Logical Channel Prioritization) process.

[0144] • UL authorization has been received for any MAC entity.

[0145] In one example implementation, the UL MAC CE may include at least one of the following information: • Quantitative value of computing power used for AI / ML; • The maximum quantization value of computing power used for AI / ML; • Remaining quantized value of computing power used for AI / ML.

[0146] In an example implementation of coordinating AI / ML computing capabilities between the first NW and the second NW in a multi-SIM scenario, UAI (e.g., UE assistance information) messages from the UE to the NW can be used.

[0147] In an example implementation where the NW configures the UAI to the UE, the UAI configuration may include at least one of the following information: • Minimum quantization value for computing power used in AI / ML: The minimum quantization value used to indicate the remaining computing power.

[0148] • Maximum quantization value for AI / ML computing power: The maximum quantization value used to indicate the remaining computing power.

[0149] • Disable Timer: The duration of the timer, used to prevent UAI from being triggered while the timer is running.

[0150] In one example implementation of triggering UAI, at least one of the following conditions may need to be met: • The remaining quantization value for computing power used in AI / ML is less than or equal to the pre-configured minimum quantization value.

[0151] • No UAI has been sent since the remaining quantization value of the computing power became less than the pre-configured threshold.

[0152] • The remaining quantization value for computing power used in AI / ML is greater than or equal to the pre-configured maximum quantization value.

[0153] • No UAI has been sent since the remaining quantization value of computing power became greater than or equal to the pre-configured threshold.

[0154] • Prevent the timer from running.

[0155] • In one example implementation, the content of UAI may include the remaining quantized value of the computing power used for AI / ML on the UE side.

[0156] In one example implementation of sending UAI, the timer can be started / restarted to disable it.

[0157] UE-side AI / ML LCM – Applicable to the identification and reporting of AI / ML features / functions / models Back Figure 5 Step 2 can be performed to determine the applicable UE AI / ML features / functions / models among the supported UE AI / ML features / functions / models. This process can be implemented such that the NW further determines the applicable UE AI / ML features / functions / models based on network conditions and / or the actual available AI / ML features / functions / models stored at the UE (rather than all supported UE AI / ML features / functions / models). Such a process typically involves the NW sending a set of additional conditions (e.g., network conditions) and / or a request for reporting applicable UE AI / ML features / functions / models to the UE, and the UE determining a set of applicable UE AI / ML features / functions / models based on these additional conditions and / or the available AI / ML features / functions / models at the UE, and reporting the determined set of applicable UE AI / ML features / functions / models to the NW.

[0158] Two example alternatives can be used (such as) Figure 5 The instructions specify message passing for both Scheme 1 and Scheme 2. Figure 5 The process of step 2.

[0159] For example, in the first scheme, the message used to process the interaction between the UE and the NW for determining and reporting applicable UE AI / ML features / functions / models can be based on signaling or message formats, including but not limited to (1) RRC messages for conditions transmitted by the NW, and / or (2) message response RRC messages for the UE to send reports of applicable UE AI / ML features / functions / models in response to conditions or requests sent by the NW, such as... Figure 5Schemes 1-1 and 1-2 are shown in the diagram. In the second alternative, for example, messages used to process interactions between the UE and NW for determining and reporting applicable UE AI / ML features / functions / models can be based on RRC reconfiguration messages and corresponding RRC reconfiguration response messages, such as... Figure 5 Schemes 2-1 and 2-2 are shown in the figure.

[0160] In some example implementations of the transmission conditions of the first scheme, the NW can transmit the conditions to the UE via a message format, which includes, but is not limited to, (1) an RRC system information message for the NW to transmit the conditions and / or a request for a report on the applicable features / functions / models, and / or (2) a message called RequestApplicationReporting The (Request Applicable Report) message is used to transmit conditions from NW to UE.

[0161] In some example implementations, when additional conditions are provided for the UE to determine the applicable UE AI / M features / functions / models, the aforementioned RRC system information message may include or indicate at least one of the following information or information items: • AI / ML-based feature group indication: Used to indicate AI / ML-based feature groups that can provide NW additional conditions or request applicable feature / feature / model reports. In an example implementation, AI / ML-based feature groups can be hard-coded. For example, AI / ML-based features can be represented by some IDs. As an example only, an AI / ML-based feature with Id=0 can indicate an AI / ML-based beam management feature group. An AI / ML-based feature with Id=1 can indicate an AI / ML-based CSI feature group, etc.

[0162] • In one implementation, AI / ML-based features can be indicated by an enumerated type parameter, with example values ​​of {beam management, CSI, CSI prediction, spare 1...}.

[0163] • AI / ML-based feature indicators: Used to indicate AI / ML-based features that can provide additional NW conditions or request applicable functions / model reports for them: In one implementation, AI / ML-based features can be hard-coded. For example, AI / ML-based features can be represented by some IDs. As an example only, an AI / ML-based feature with Id=0 can indicate AI / ML-based spatial beam management, and an AI / ML-based feature with Id=1 can indicate AI / ML-based temporal beam management, etc.

[0164] In one implementation, AI / ML-based features can be indicated by an enumerated type parameter, with example values ​​of {spatial beam management, temporal beam management, CSI feedback compression / decompression, CSI prediction, spare 1...}.

[0165] • An indication of NW additional conditions, or an indication of NW additional conditions associated with the indicated AI / ML-based feature or feature group, used to indicate NW additional conditions regarding AI / ML-based features / feature groups. At least one of the following information items can be indicated as NW additional conditions in a system message: AI / ML-based feature indication: Used to indicate the AI / ML-based features that the message or the NW appendix targets.

[0166] ISD is used to indicate cell range, such as UMa, UMi, 200 meters, 500 meters.

[0167] Antenna height, used to indicate the antenna height of the base station in this cell.

[0168] NLOS probability: Used to indicate the probability of NLOS wireless propagation.

[0169] The indoor / outdoor indicator is used to indicate whether the current scene is indoors or outdoors, or the indoor / outdoor ratio.

[0170] DL Tx beamcodebook indicator: Used to indicate the DL Tx beamcodebook used for AI / ML model training.

[0171] gNB antenna array dimension indication.

[0172] Downtilt angle of the gNB antenna.

[0173] The beam pattern indicator for set B is used to indicate the beam pattern of set B relative to set A (see above). For example, set B could be a 1 / 4 subset of set A, etc.

[0174] A beam pattern for a set A of AI / ML-based functions, used to indicate the beam pattern of set A relative to the complete beam set (see above). For example, set A could be a 1 / 4 subset of the complete set, etc.

[0175] Similarly, in use RequestApplicationReporting In some example implementations of the first scheme of message transmission conditions described above, in the example sent from NW to UE RequestApplicationReportingThe message may indicate / include at least one of the following information or information items: • AI / ML support indication: Used to indicate that AI / ML-based features / feature groups (FG) are supported by NW on the UE side.

[0176] • AI / ML Feature Group Indication: Used to indicate the AI / ML feature group to which the NW additional conditions can be provided or the application of the AI / ML feature can be requested (e.g., AI / ML-based beam management, AI / ML-based CSI, AI / ML-based positioning).

[0177] • AI / ML-based feature indications: Used to indicate AI / ML-based features that can provide NW additional conditions or request applicable functions / model reports for them.

[0178] In one example implementation, AI / ML-based features / feature groups can be indicated by AI / ML-based feature indexes / identifiers that are already indicated by the UE capabilities.

[0179] In one example implementation, AI / ML-based features / feature groups can be indicated by AI / ML-based feature indices / identifiers, based on the order of entries in the list of supported AI / ML features in the UE capabilities. For example, AI / ML-based feature / FG index = 0 indicates the first entry in the list of supported AI / ML-based features / FGs. AI / ML-based feature / FG index = 1 indicates the second entry in the list of supported AI / ML-based features / FGs, and so on.

[0180] In a sample implementation, AI / ML-based features / feature groups can be indicated by an enumeration type parameter, with example values ​​of {spatial beam management, temporal beam management, CSI feedback compression / decompression, CSI prediction, spare 1...}.

[0181] In one example implementation, AI / ML-based features / feature groups can be indicated via a bit string type parameter based on the order of entries in the list of supported AI / ML features in the UE capabilities. For example, the leftmost or rightmost bit in the bit string represents the first entry in the list of supported AI / ML-based features / FGs, the second leftmost or rightmost bit in the bit string represents the second entry in the list of supported AI / ML-based features / FGs, and so on. If the corresponding bit is set to 1, an applicable report corresponding to the AI / ML-based feature / FG is requested.

[0182] • NW appended conditions, or indications of NW appended conditions associated with the indicated AI / ML-based feature or feature group, used to indicate NW appended conditions regarding the AI / ML-based feature / feature group. (Can be...) RequestApplicationReporting The message indicates at least one of the following information as an additional condition for NW: AI / ML-based feature indication: Used to indicate the AI / ML-based features that the message or the NW appendix targets.

[0183] ISD is used to indicate cell range, such as UMa, UMi, 200 meters, 500 meters.

[0184] Antenna height, used to indicate the antenna height of the base station in this cell.

[0185] NLOS probability is used to indicate the probability of NLOS wireless propagation.

[0186] The indoor / outdoor indicator is used to indicate whether the current scene is indoors or outdoors, or the indoor / outdoor ratio.

[0187] DL Tx beamcodebook indicator, used to indicate the DL Tx beamcodebook used for AI / ML model training.

[0188] gNB antenna array dimension indication.

[0189] Downtilt angle of the gNB antenna.

[0190] The beam pattern indicator for set B is used to indicate the beam pattern of set B relative to set A (see above). For example, set B could be a 1 / 4 subset of set A, etc.

[0191] A beam pattern for a set A of AI / ML-based functions, used to indicate the beam pattern of set A relative to the complete beam set. For example, set A could be a 1 / 4 subset of the complete set, etc.

[0192] In some example implementations, Figure 5 In schemes 1-2, the UE reports applicable UE features / functions / models to the NW, which can be achieved through at least one of the following UL RRC message formats or MAC protocol signaling: Option 1: UE Assistance Information (UAI) message.

[0193] Option 2: Dedicated RRC message, referred to as UEApplicableFunctionalityReporting (UE Applicable Function Report).

[0194] Option 3: UL MAC CE.

[0195] For option 1 above, for example, UAI messages used to report applicable functions / models for AI / ML-based features / feature groups or to report applicable features, can be triggered by meeting some predefined conditions, including but not limited to at least one or more of the following example triggering conditions: • Triggering condition 1: RRC configuration related to the applicable function / model report for AI / ML-based features / feature groups has been configured, and / or no applicable function / model for AI / ML-based features / feature groups has been sent before.

[0196] • Triggering condition 2: At least one AI / ML function / model for an AI / ML-based feature / feature group has been activated; • Triggering Condition 3: Since the most recent UE applicable function report, the AI / ML-based features / functions / models stored in the UE for the configured AI / ML-based features / feature groups have changed. In an example implementation of changes to AI / ML-based features / functions / models stored in the UE, this condition may include adding, removing, or modifying AI / ML-based features / functions / models.

[0197] • Triggering condition 4: An SIB has been received indicating that an AI / ML-based feature group is supported or that an associated applicable feature / function / model has been requested by NW, and no UAI related to an applicable AI / ML feature / function / model report has been sent previously.

[0198] Further regarding option 1, in some example implementations, the UAI message triggered and sent to the NW may contain various information about applicable functions / models for AI / ML-based features, and / or may contain various information about applicable AI / ML-based features. In some example implementations, the applicable functions / models for AI / ML-based features or the applicable AI / ML-based features for a group of AI / ML-based features may be a subset or the entire set of supported applicable features / functions / models reported in the UE capabilities. In some example implementations, the applicable functions / models for AI / ML-based features or the applicable AI / ML-based features for a group of AI / ML-based features may be features / functions / models other than those reported in the UE capabilities.

[0199] When a UAI message is triggered by a modification or update of an AI / ML-based feature / function / model, the UAI message may contain at least one of the following information regarding the NW additional conditions for the modified / updated AI / ML-based feature / function / model: • AI / ML-based feature indication: Used to indicate the AI / ML-based features that the message or the NW appended condition targets.

[0200] • ISD is used to indicate cell range, for example, UMa, UMi, 200 meters, 500 meters.

[0201] • Antenna height, used to indicate the antenna height of the base station in this cell.

[0202] • NLOS probability, used to indicate the probability of NLOS wireless propagation.

[0203] • Indoor / Outdoor indicator, used to indicate whether the current scene is indoors or outdoors, or the indoor / outdoor ratio.

[0204] • DL Tx beamcodebook indicator, used to indicate the DL Tx beamcodebook used for AI / ML model training.

[0205] • gNB antenna array dimension indication.

[0206] • Downtilt angle of the gNB antenna.

[0207] • Beam pattern indicator for set B, used to indicate the beam pattern of set B relative to set A (see above). For example, set B could be a 1 / 4 subset of set A, etc.

[0208] • A beam pattern for a set A of AI / ML-based functions, used to indicate the beam pattern of set A relative to the complete beam set. For example, set A could be a 1 / 4 subset of the complete set, etc.

[0209] When a UAI message is triggered by the addition of an AI / ML-based feature / function / model, the UAI message may contain at least one of the following information regarding the NW appended conditions for the newly added AI / ML-based feature / function / model: • AI / ML-based feature indication: Used to indicate the AI / ML-based features that the message or the NW appended condition targets.

[0210] • ISD is used to indicate cell range, for example, UMa, UMi, 200 meters, 500 meters.

[0211] • Antenna height, used to indicate the antenna height of the base station in this cell.

[0212] • NLOS probability, used to indicate the probability of NLOS wireless propagation.

[0213] • Indoor / Outdoor indicator, used to indicate whether the current scene is indoors or outdoors, or the indoor / outdoor ratio.

[0214] • DL Tx beamcodebook indicator, used to indicate the DL Tx beamcodebook used for AI / ML model training.

[0215] • gNB antenna array dimension indication.

[0216] • Downtilt angle of the gNB antenna.

[0217] • Beam pattern indicator for set B, used to indicate the beam pattern of set B relative to set A (see above). For example, set B could be a 1 / 4 subset of set A, etc.

[0218] • A beam pattern for a set A of AI / ML-based functions, used to indicate the beam pattern of set A relative to the complete beam set. For example, set A could be a 1 / 4 subset of the complete set, etc.

[0219] For option 2 (dedicated RRC message) above, example UEApplicableFunctionalityReporting The message can be related to the above. RequestApplicableFunctionality The response message RequestApplicableFunctionality The request contains NW additional conditions for at least one AI / ML-based feature / feature group and / or an applicable AI / ML-based feature / function / model report for at least one AI / ML-based feature / feature group. Example UEApplicableFunctionalityReporting The message may, for example, indicate / include applicable AI / ML features for an AI / ML-based feature or a group of AI / ML-based features. In some example implementations, the applicable AI / ML features for an AI / ML-based feature or a group of AI / ML-based features may be a subset or the entire set of supported applicable features / features / models reported in the UE capabilities. In some example implementations, the applicable AI / ML features for an AI / ML-based feature or a group of AI / ML-based features may be features / features / models other than those reported in the UE capabilities.

[0220] exist UEApplicableFunctionalityReporting In cases where the AI / ML features / functions / models included are updated or added compared to the supported AI / ML features / functions / models in the (most recently) reported UE capabilities, the UEApplic ableFunctionalityReportingIt may contain at least one of the following information regarding NW additional conditions for modified / updated / added AI / ML features / functions / models: AI / ML-based feature indication: Used to indicate the AI / ML-based features that the message or the NW appendix targets.

[0221] ISD is used to indicate cell range, such as UMa, UMi, 200 meters, 500 meters.

[0222] Antenna height, used to indicate the antenna height of the base station in this cell.

[0223] NLOS probability is used to indicate the probability of NLOS wireless propagation.

[0224] The indoor / outdoor indicator is used to indicate whether the current scene is indoors or outdoors, or the indoor / outdoor ratio.

[0225] DL Tx beamcodebook indicator, used to indicate the DL Tx beamcodebook used for AI / ML model training.

[0226] gNB antenna array dimension indication.

[0227] Downtilt angle of the gNB antenna.

[0228] The beam pattern indicator for set B is used to indicate the beam pattern of set B relative to set A (see above). For example, set B can be a uniformly distributed 1 / 4 subset of set A.

[0229] Beam pattern for a set A of AI / ML-based functions, used to indicate the beam pattern of set A relative to the full beam set (see above). For example, set A could be a uniformly distributed 1 / 4 subset of the full set.

[0230] In some example implementations, for options 1 and 2 above, the applicable AI / ML function / model for AI / ML-based features / feature groups can be reported / indicated via a BIT STRING type parameter (or bitmap type parameter), or the applicable AI / ML feature can be reported / indicated via a BIT STRING type parameter (or bitmap type parameter). As example syntax, the bit string type parameter is shown below for indicating the applicable AI / ML function for space beam management features:

[0231] In such an example bit string or bitmap, the first bit on the left (or right) can be mapped to the first entry in the list of functions reported by the UE capabilities for AI / ML-based features / feature groups (e.g., AI / ML-based spatial beam management), the second bit on the left (or right) can be mapped to the second entry in the list of functions reported by the UE capabilities for AI / ML-based features / FGs (e.g., AI / ML-based spatial beam management), and so on. If a specific bit is set to 1 in the report, the AI / ML function mapped to that specific bit can be considered applicable. Otherwise, the AI / ML function mapped to and associated with that specific bit cannot be considered applicable.

[0232] As another example syntax for applicable AI / ML-based features, the bit string type parameter is shown below to indicate the applicable AI / ML features for a beam-managed AI / ML feature set:

[0233] In such an example, a value of AI-ML-BeamManagement = 01 indicates that the features of AI / ML spatial beam management and AI / ML temporal beam management are applicable. A value of AI-ML-BeamManagement = 10 indicates that the features of AI / ML temporal beam management and AI / ML spatial beam management are applicable. A value of AI-ML-BeamManagement = 00 indicates that neither AI / ML temporal beam management nor AI / ML spatial beam management is applicable. A value of AI-ML-BeamManagement = 11 indicates that both AI / ML temporal beam management and AI / ML spatial beam management are applicable.

[0234] In some example implementations, for options 1 and 2 above, the applicable AI / ML functions / models for AI / ML-based features / feature groups can be reported / indicated by a list of AI / ML function indices / identifiers or AI / ML model indices / identifiers. As example syntax, a list of integer type parameters is shown below to indicate the applicable AI / ML-based spatial beam management:

[0235] or

[0236] In such an example AI / ML feature / model index or identifier, the feature or model index can be numbered according to the order of entries in the AI / ML feature list or model list for an AI / ML-based feature or function in the UE capability. For example, the first entry in the list is numbered as index=0, the second entry in the list is numbered as index=1, and so on.

[0237] In such an example AI / ML function / model index or identifier, the function or model index / identifier can be aligned with the function or model index / identifier reported in the UE capability.

[0238] In some example implementations, for options 1 and 2 above, the applicable AI / ML-based features can be reported / indicated by a list of AI / ML-based features. As example syntax, the list parameter for indicating the applicable AI / ML-based beam management is shown below:

[0239] In such an example AI / ML feature index or identifier, the feature index or identifier can be numbered according to the order of entries in the AI / ML feature list for an AI / ML-based feature group in the UE capability. For example, the first entry in the list is numbered as index=0, the second entry in the list is numbered as index=1, and so on. In such an example AI / ML feature index or identifier, the feature can be aligned with the AI / ML-based feature index / identifier reported in the UE capability for the AI / ML-based feature group.

[0240] For option 3 above, a UL MAC CE (e.g., Applicable AI / ML Reporting MAC CE) can be triggered for reporting applicable UE AI / ML features / functions / models. In some example implementations, such an Applicable Function Reporting MAC CE can be triggered by any one or more conditions similar to options 1 and 2: • Triggering condition 1: RRC configuration has been configured for applicable feature / function / model reports for AI / ML-based feature groups / features / functions, and no applicable feature / function / model reports have been sent before.

[0241] • Triggering condition 2: At least one AI / ML feature / function / model for an AI / ML-based feature group / feature / function has been activated; • Triggering Condition 3: Since the most recent transmission of the UE applicable feature / function / model report, the AI / ML-based feature / function / model stored in the UE has changed. In an example implementation of changes to the AI / ML-based feature / function / model stored in the UE, this condition may include adding, removing, or modifying AI / ML-based features / functions / models.

[0242] In some example implementations, if no available UL authorization is received for transmitting such a UL MAC CE, the aforementioned UL MAC CE can trigger a scheduling request (SR). In some example implementations, for applicable feature reports, only one SR configuration may be applied in a cell group.

[0243] In some example implementations, the UL MAC CE described above may include at least one of the following information or information items: • Serving cell indicator: Used to indicate the serving cell for applications based on AI / ML features.

[0244] • AI / ML-based feature group indication: Used to indicate AI / ML-based feature groups of interest for applicable AI / ML features / features / models.

[0245] • AI / ML-based feature indication: Used to indicate AI / ML-based features for which there is interest in applicable AI / ML features / functions / models.

[0246] • AI / ML Function Indicator: Used to indicate AI / ML-based functions that are relevant to the applicable AI / ML model.

[0247] • AI / ML Applicable Feature / Function / Model Indication: Used to indicate the currently applicable AI / ML feature / function / model at the UE for the indicated AI / ML-based feature.

[0248] In some example implementations, the above UL MAC CE can follow Figure 7 The example structure / format is shown below. Figure 7 In the first eight bits of UL MAC CE, the "F" i "Indicates that the i-th AI / ML-based feature contains Fu" i,j Does the 8-bit value exist (or is applicable)? If Fi = 1, then Fu is included. i,j The eight bits exist, otherwise it contains Fu. i,j The first eight bits do not exist. Therefore, the number n of the eight bits in Fu will be equal to the number of F bits in the first eight bits.i The number of 1s.

[0249] Furthermore, in Figure 7 In the example, if the indication exists, then "Fu" i,j "Eight-bit instruction at the UE for the corresponding "F" i "Applicable functions based on AI / ML features." i,j Each bit in “” corresponds to and indicates the capability of the UE (e.g., Figure 5 In Step 1, the reported UE capabilities list indicates whether the j-th AI / ML feature is applicable. i,j If set to 1, the j-th AI / ML feature in the AI / ML-based features is applicable. Otherwise, the AI / ML feature is not applicable.

[0250] Back Figure 5 The second alternative (Solution 2) in step 2, which uses the RRC reconfiguration mechanism to determine and report the available AI / ML features / functions / models for the UE, may include: sending an RRC Reconfiguration message from the NW to the UE to provide the aforementioned conditions. Figure 5 Scheme 2-1), and the UE sends a response message to the NW to report applicable AI / ML features / functions / models, RRC reconfiguration complete (RRCReconfigurationComplete). Figure 5 Scheme 2-2).

[0251] Specifically, the NW can request the UE to report the applicable AI / ML functions for AI / ML-based features / feature groups through the RRCReconfiguration message, and the UE can correspondingly respond to the NW with the RRCReconfigurationComplete message regarding the applicable AI / ML functions for AI / ML-based features / feature groups.

[0252] In some example implementations, parameters, such as ENUMERATED type parameters per AI / ML-based feature / feature group, may be introduced or included in the RRCReconfiguration message used to request applicable UE AI / ML functions / models for AI / ML-based features, as shown in the example below:

[0253] In some example implementations, for requesting applicable AI / ML-based features / functions / models for AIML-based feature groups of UEs... RRCReconfigurationMessages can introduce or include parameters, such as ENUMERATED type parameters for each AI / ML-based feature / feature group, as shown in the example below:

[0254] In one example implementation of the above parameters, these parameters may be provided per UE, per cell group, per frequency band, or per cell.

[0255] In the context of requesting applicable features / functions / models RRCReconfiguration In an example implementation of the message, one or more information elements may be introduced for requesting a report of applicable AI / ML features / functions / models. These elements may include at least one of the following information items regarding NW additional conditions, which the UE may consider for reporting AI / ML features / functions / models: • AI / ML-based feature indication: Used to indicate the AI / ML-based features that the message or the NW appended condition targets.

[0256] • ISD is used to indicate cell range, for example, UMa, UMi, 200 meters, 500 meters.

[0257] • Antenna height, used to indicate the antenna height of the base station in this cell.

[0258] • NLOS probability: Used to indicate the probability of NLOS wireless propagation.

[0259] • Indoor / Outdoor indicator, used to indicate whether the current scene is indoors or outdoors, or the indoor / outdoor ratio.

[0260] • DL Tx beamcodebook indicator: Used to indicate the DL Tx beamcodebook used for AI / ML model training.

[0261] • gNB antenna array dimension indication.

[0262] • Downtilt angle of the gNB antenna.

[0263] • Beam pattern indicator for set B, used to indicate the beam pattern of set B relative to set A (see above). For example, set B could be a 1 / 4 subset of set A, etc.

[0264] • A beam pattern for a set A of AI / ML-based functions, used to indicate the beam pattern of set A relative to the complete beam set (see above). For example, set A could be a 1 / 4 subset of the complete set, etc.

[0265] In response to the request in the aforementioned RRCReconfiguration, the UE... RRCReconfigurationCompleteIn some example implementations of the included parameters, the applicable AI / ML-based features / functions / models for AI / ML-based feature groups / features can be indicated by example BIT STRING or bitmap type parameters. In such example bitstrings or bitmaps, the first bit on the left (or right) can map to the first entry in the list of features / functions / models (e.g., AI / ML-based beam management or AI / ML-based spatial beam management) reported by the UE capability, the second bit on the left (or right) can map to the second entry in the list of features / functions / models (e.g., AI / ML-based beam management or AI / ML-based spatial beam management) reported by the UE capability, and so on. If a specific bit is set to 1 in the report, the AI / ML feature / function / model mapped to that specific bit can be considered applicable. Otherwise, the AI / ML feature / function / model mapped to and associated with that specific bit cannot be considered applicable. An example bitstring is shown below:

[0266] In one example implementation of the parameters described above that apply to AI / ML features / functions / models, these parameters may be provided per UE, per cell group, per frequency band, or per cell.

[0267] UE-side AI / ML LCM – Testing applicable to UE AI / ML features / functions / models Turn Figure 5 Step 3, regarding preparation for the activation of AI / ML features / functions / models that may follow the UE's report of applicable AI / ML functions in Step 2, can involve testing the applicable AI / ML features / functions / models to further consider their adoption and activation. For example, the conditions provided to the UE by the NW in Step 2 may be limited / restricted so that the NW avoids exposing sensitive NW additional conditions as much as possible (e.g., conditions that may be analyzed and expose sensitive user information should not be disseminated). The available AI / ML functions that the UE determines to report to the NW in Step 2 may be executed under the limited / restricted conditions provided by the NW. Therefore, it may be desirable for the UE and NW to test the applicable AI / ML functions reported by the UE in Step 3 to determine their predictive performance / accuracy, thereby further determining whether they can be adopted / activated in predictions (e.g., gNB settings, deployment scenarios, etc.).

[0268] Figure 8 The document illustrates an example of such a test procedure for UE-side AI / ML functions performed by UE 802 and NW 804, which includes the following example steps: Step 3-1: NW 804 can send message A to test one or more applicable AI / ML features / functions / models.

[0269] Step 3-2: UE 802 can input the list of reference data received from message A into one or more applicable AI / ML features / functions / models to obtain the corresponding inference output or value.

[0270] Step 3-3: UE 802 can send the inferred output or value to NW 804 via message B, so that NW can compare the inferred output or value with the true value known to NW 804.

[0271] In some example implementations of step 3-1 above, message A can be implemented as a DL RRC message, DL MAC CE, or PDCCH signaling. In another example implementation, message A can be a protocol signaling termination between the AI / ML logical layer and the NW logical entity / unit.

[0272] Furthermore, in some example implementations of step 3-1 above, message A may include / indicate at least one of the following information or information items: • Applicable AI / ML Features / Functions / Models Indicator: Used to indicate the applicable AI / ML features / functions / models that need to be tested. In some example implementations, the applicable AI / ML features / functions / models may come from the above. Figure 5 The applicable AI / ML report in step 2.

[0273] • Supported AI / ML Features / Functions / Models Indicator: Used to indicate the supported AI / ML features / functions / models that need to be tested. In some example implementations, the supported AI / ML features / functions / models may be from the above. Figure 5 The (most recent) reported UE capabilities in step 1.

[0274] • Reference Data List (Input Data List for AI / ML Functions to be Tested): A list of data that will be used as input data for the AI / ML features / functions / models that need to be tested.

[0275] Reference Data List (Benchmark List): A list of data that will be used as benchmarks to evaluate the output data from the AI / ML features / functions / models that need to be tested.

[0276] In some example implementations of step 3-2 above, the UE can use one or more applicable AI / ML features / functions / models to process the input reference data list and / or baseline reference data list received from message A to obtain the corresponding inferred output value (which can be used by the NW to check the performance and effectiveness of these AI / ML features / functions / models), or to obtain performance metrics (which can be used by the UE to check the performance and effectiveness of these AI / ML features / functions / models).

[0277] In some example implementations of step 3-3 above, the protocol format of message B can be implemented as a UL RRC message (e.g., UAI), UL MAC CE, or via PUCCH signaling.

[0278] Furthermore, in some example implementations of steps 3-3 above, message B may contain / indicate at least one of the following information or information items: • Applicable AI / ML features / functions / models indication: Used to indicate the applicable AI / ML features / functions / models that have been tested.

[0279] • Output data list: Includes output data obtained from inferences from each applicable AI / ML feature / function / model.

[0280] • UE Additional Conditions: Includes additional UE conditions to assist NW in determining the validity or performance of the tested AI / ML features / functions / models.

[0281] • Available AI / ML features / functions / models: Indicates the valid AI / ML features / functions / models that have been tested.

[0282] • KPI (Key Performance Indicator) value indication: Used to indicate the KPI value for each AI / ML feature / function / model. In some example implementations, it is used to indicate the KPI value for each AI / ML feature / function / model considered available.

[0283] In some example implementations that test an alternative to UE AI / ML features / functions / models, the following steps can be performed: • Step 1: NW can configure RRC signaling for the applicable AI / ML features / functions / models on the UE side.

[0284] Step 2: The UE can send a message to the NW to report available AI / ML features / functions / models.

[0285] In some example implementations of the RRC signaling in step 1, the RRC signaling may include an RRC configuration, which may be one or more CSI-ResourceConfigs for one or more applicable AI / ML features / functions / models reported in step 2 (e.g., applicable AI / ML feature / model reports). The RRC configuration may also be one or more CSI-ResourceConfigs for one or more supported AI / ML features / functions / models reported in step 1 (e.g., UE capabilities). In some example implementations, the CSI-ResourceConfig may be a configuration of a reference signal used by the UE to perform measurements to obtain inputs to the AI / ML function / model and / or to obtain benchmark values ​​(e.g., measurement results) for calculating performance metrics by comparing actual measurement results with inferred values ​​(e.g., the output of the tested AI / ML function / model).

[0286] In some example implementations of the RRC signaling in step 1, the RRC signaling may include RRC configurations, which can be one or more RRC configurations used to indicate the AI / ML features / functions / models to be tested. In some example implementations, the AI / ML features / functions / models to be tested can... Figure 5 In step 2, the UE report should include one of the applicable AI / ML features / functions / models. In some example implementations, the AI / ML features / functions / models to be tested could be... Figure 5 Step 1 is one of the supported AI / ML features / functions / models in the UE capabilities.

[0287] In some example implementations, the above message can be a UL RRC signaling, a UL MAC CE, or a UL RRC message / UL MAC CE. It may include / indicate at least one of the following information: • Available AI / ML-based feature indicators.

[0288] • Available AI / ML-based feature indicators.

[0289] • Available AI / ML-based model indicators.

[0290] • Performance KPI values ​​for each available AI / ML-based feature / function / model.

[0291] UE-side AI / ML LCM – UE AI / ML configuration Turn now Figure 5In step 4, the NW can now determine the configuration of the AI / ML function based on the available UE AI / ML functions, test output, and / or UE additional conditions received from the UE in step 3, and further transmit such configuration to the UE. For example, the NW can determine which models among the applicable AI / ML models being tested provide acceptable prediction accuracy or performance and the configuration of those models.

[0292] In one example implementation of step 4 above, the configuration of the UE AI / ML functions determined by the NW can be sent to the UE through several alternative RRC configuration structures.

[0293] In the first alternative or alternative example, it is possible to construct as follows: Figures 9A to 9E The RRC configuration structure for UE AI / ML related configurations is shown, and it is sent from the NW to the UE. Figures 9A to 9E As shown in the example: • 1: AI / ML (AIML) related configurations (ALMLConfig, ALML configuration) can be configured in the serving cell configuration (“ServingCellConfig”), such as Figure 9A As shown. In another example implementation, as... Figure 9B As shown, AI / ML (AIML) related configurations (“AIMLConfig”) can be configured in the cell group configuration (CellGroupConfig), or as follows: Figure 9C As shown, it is configured in the Physical Cell Group Configuration (PhysicalCellGroupConfig), or as... Figure 9D As shown, it is configured in the MAC-CellGroupConfig. In another implementation, the AI / ML (AIML) related configuration (“AIMLConfig”) can be configured per UE, i.e., as shown... Figure 9E As shown, AIMLConfig and IE CellGroupConfig are at the same level.

[0294] ·2: AIML-related configurations (such as AIMLConfig) can contain a list of configurations based on AI / ML features (e.g., “AIMLBasedFeatureConfig (AIML-based feature configuration) #1” to “AIMLBasedFeatureConfig #4”).

[0295] ·3: Each AI / ML-based feature configuration may contain a list of configurations for AI / ML functions (e.g., “AIMLFunctionalityAddModList”, including “AIMLFunctionalityConfig (AIML Function Configuration) #1” through “AIMLFunctionalityConfig #3”), and / or an index / identifier of the AI / ML-based feature configuration, and / or an indication of the AI / ML-based features for such configuration.

[0296] •4: AIML feature configuration can include at least one of the following information items: The index / identifier of the AI / ML function or the AI / ML function ID (e.g., "AIMLFunctionalityID").

[0297] A list of AI / ML models for AI / ML functionality (e.g., “AIMLModelToAddModList”). Each AI / ML model in this list may include a model ID representing that AI / ML model, and / or a quantized value of the computing power consumed when the corresponding AI / ML model is activated.

[0298] Infer the relevant configuration (list) (e.g., “InferenceConfig” or “InferenceConfigToAddModList”).

[0299] Performance monitoring related configurations (list) (e.g., “MonitoringConfig” or “MonitoringToAddModList”).

[0300] Traditional configurations used to enable or disable corresponding non-AI-based features.

[0301] The quantified value of computing power consumption when the corresponding AI / ML feature / function / model is activated.

[0302] Furthermore, in Figures 9A to 9E In this example RRC configuration structure, all AI / ML functions / features / models can be configured together to facilitate future AI development. It allows for easy addition of new AI / ML functions / features / models and their configurations. In this example, AI / ML-related functions are configured by cell, by cell group, or by UE.

[0303] In some example implementations, AI / ML features can be indexed to ensure consistent understanding between the user experience (UE) and the new workspace (NW) regarding the correct identification of AI / ML features / models. For example: • In some example implementations, AI / ML function indexes / identifiers can be based on UE capabilities (such as...) Figure 5 The location of the AI / ML function for the AI / ML feature-based feature in the supported AI / ML function list (as reported in step 1) is constructed. For example, an AI / ML function index for the AI / ML feature-based feature in the RRC configuration = 0 can indicate the first entry in the list of supported AI / ML functions for the AI / ML feature-based feature in the UE capability, an AI / ML function index for the AI / ML feature-based feature in the RRC configuration = 1 can indicate the second entry in the list of supported AI / ML functions for the AI / ML feature-based feature in the UE capability, and so on.

[0304] • In some example implementations, the AI / ML functions represented by the AI / ML function index / identifier in the above RRC configuration can be accessed through UE capabilities (such as... Figure 5 The same AI / ML feature index / identifier as reported in step 1 is used to build the index.

[0305] • In some example implementations, the AI / ML feature index / identifier in the RRC configuration can be based on the applicable feature report (e.g., Figure 5 The AI / ML functions are built according to the location of the applicable function list in step 2 (reported in step 2).

[0306] In some other implementations, the AI / ML function represented by the AI / ML function index / identifier in the RRC configuration can be associated with the applicable function report (such as...). Figure 5 The AI / ML feature index / identifier is the same as that reported in step 2. For example, AI / ML feature index = 0 for AI / ML-based features in the RRC configuration can indicate the first entry in the list of AI / ML features supported for AI / ML-based features in the applicable AI / ML feature report, AI / ML feature index = 1 for AI / ML-based features in the RRC configuration can indicate the second entry in the list of AI / ML features supported for AI / ML-based features in the AI / ML feature report, and so on.

[0307] In some example implementations, the above Figures 9A to 9E of InferenceConfigurations (Inferred configuration) may include at least one of the following information or information items: • A list of measurement configurations used to obtain the input data for inference, for example, MeasurementForInferenceConfig or MeasurementForInferenceConfigToAddmodList .

[0308] • Output data, inference results, report configuration (list) (e.g.: inferenceReporting or inferenceReportingToAddModList For example, in one implementation, Inference Reporting It may include instructions regarding the content to be reported. For example, in one implementation of the information regarding the reported content, for beam management, it could be at least one of the following information about the content: Based on the inference results of the first N DL beams, N>0; Based on the inference results of the first N DL beams and their corresponding RSRP values, N>0; Based on the inference results of the first N DL beams and the RSRP values ​​of the M best DL beams among these N DL beams, N>M>=0; In one implementation, Inference Reporting Instructions regarding the report type may be included: Semi-persistent periodicity; Periodicity; Event triggered: Based on the inference, the first beam has been changed.

[0309] Based on the inference results, the current DL Tx beam is no longer among the first N beams.

[0310] Based on the inference results, the current RSRP value of the DL Tx beam is below the threshold.

[0311] Based on the inference results, the RSRP value of the current DL Tx beam is below the threshold, and at least one beam has an RSRP value above the threshold.

[0312] In some example implementations of performance monitoring-related configurations, at least one of the following information may be included: • Measurement Configuration: Measurement configuration for performance monitoring • Measurement report configuration: Used for configuring performance metrics or performance result reports.

[0313] • In one implementation of a traditional related configuration, at least one of the following information may be included: • CSI-ReportingConfigId: Used to indicate the CSI reporting configuration for non-AI-based beam management. In one implementation, the reporting type configured in the CSI-ReportingConfig indicated by such CSI-ReportingConfigId is semi-persistent on PUSCH, semi-persistent on PUCCH, or aperiodic.

[0314] In one implementation of Option 1, assuming the AI / ML model (list) exists, each entry of InferenceToAddModList is associated with each entry of the AI / ML model (list) in ascending order, and / or each entry of MonitoringToAddModList is associated with each entry of the AI / ML model (list) in ascending order.

[0315] In a second optional or alternative example of the RRC configuration structure for AI / ML features / features / models, AI / ML-based features and / or AI / ML-based features / models can be associated with CSI-ReportConfig. An example CSI-ReportConfig structure is shown below, which includes AI / ML-based feature / feature / model configurations:

[0316]

[0317]

[0318]

[0319] Based on the example RRC structure above, AI / ML functions / features / models can be implicitly activated / deactivated by activating / deactivating semi-persistent CSI reporting. In some example implementations for configuring AI / ML features / functions / models to the UE, AI / ML related information can be configured in the RRC configuration for CSI reporting (e.g., CSI-ReportConfig), as shown above.

[0320] In some example implementations, the RRC configuration for AI / ML features / functions / models may include and / or indicate at least one of the following information or information items: • AI / ML Feature / Function / Model Indicator: Used to indicate the corresponding AI / ML feature / function / model configured in the CSI report.

[0321] • Inference Measurement Resource Configuration Indicator: Used to indicate the measurement resource configuration associated with the AI / ML function / model inference inputs related to this CSI report configuration. This indicator can be accessed via... CSI-ResourceConfigIdconduct.

[0322] • Monitoring Measurement Resource Configuration Indicator: Used to indicate the measurement resource configuration for monitoring AI / ML functions / model performance related to this CSI report configuration. This indicator can be... CSI-ResourceConfigId In one example implementation, monitoring measurement resource configuration indicators can be mutually exclusive with inferring measurement resource configuration indicators.

[0323] • AI / ML Report Indicators: Reporting information used to indicate relevant AI / ML features / functions / models. In one example implementation, it can be included in the information elements of the CSI report configuration. ReportQuantity middle.

[0324] • AI / ML Report Type Indicator: Used to indicate the report type of the relevant AI / ML feature / function / model. In one example implementation, to introduce event-triggered type reports, this will be described in detail below.

[0325] In some example implementations, for configuring CSI reports based on AI / ML features / functions / models, if the CSI report configuration is related to the measurement configuration used for inference, the CSI report type can be configured as a semi-persistent CSI report on PUCCH, a semi-persistent CSI report on PUSCH, or an event-triggered CSI report. If the CSI report configuration is related to the measurement configuration used for monitoring, the CSI report type can be configured as a semi-persistent CSI report on PUCCH, a semi-persistent CSI report on PUSCH, an aperiodic CSI report, or an event-triggered CSI report.

[0326] Turn now Figure 5 In step 5, the UE can perform, for example, activation and / or deactivation of one or more AI / ML features / functions / models as instructed by the NW to the UE.

[0327] In an example implementation of step 5, when using Figures 9A to 9E In the example RRC configuration structure, activation of AI / ML functions can be achieved through separate signaling from the NW to the UE. Such signaling may contain at least one of the following information or information items: • Serving cell indication: Used to indicate the serving cell for which AI / ML features / functions / models are activated / deactivated.

[0328] • AI / ML-based feature indication: Used to indicate the AI / ML-based feature to which the AI / ML function belongs.

[0329] • AI / ML Function Indicator: Used to indicate which AI / ML functions need to be activated / deactivated.

[0330] • AI / ML function indicator: Used to indicate which AI / ML models need to be activated / deactivated.

[0331] • Activation / Deactivation Indicator: Used to indicate the activation / deactivation of the indicated AI / ML feature / function / model.

[0332] This signaling can be implemented, for example, as DL MAC CE. Figure 10 The example DL MAC CE for signaling activation / deactivation of AI / ML functions is shown in the figure.

[0333] exist Figure 10 In the example DL MAC CE, the "C" in the first eight bits i "Indicates the serving cell servingCellId=i, where the AI / ML function should be activated / deactivated if the octet is present (when Ci=1). Ci can be 0 or 1. The 'F' in the second octet..." i "Indicates how many octets will exist after the second octet. For example, if C..." i =1, then F i =1 means that for C i The indicated serving cell (e.g., for AI / ML-based spatial beam management, AI / ML-based temporal beam management) will have two octets, while Ci=1, Fi=0 means that for the serving cell indicated by Ci, there will be one octet. “Feat IDi,j” indicates the AI / ML-based feature ID to which the j-th function to be activated / deactivated belongs for the serving cell servingCellId=i. “Functionality IDi,j” indicates the AI / ML function indicated as to be activated / deactivated for the serving cell servingCellId=i. Figure 10 The R bit indicated in the text represents a reserved bit.

[0334] exist Figure 5 In another example implementation of step 5, when using the example CSI-ReportConfig RRC structure, the signaling for activating / deactivating AI / ML features / functions / models can be accomplished through associated semi-persistent CSI report activation / deactivation, the configuration of which can be inferred (e.g., ReportQuantity is set to...). Inference-SSB-Index-RSRP For example, a DL MAC CE can be used, which contains at least one of the following information or information items: •1: Serving Cell ID: Used to indicate the serving cell reported by the CSI report.

[0335] ·2: BWP Id: Used to indicate the UL BWP reported in the CSI report.

[0336] •3: CSI Report Configuration Instructions: Used to indicate the activation / deactivation of CSI report configurations for AI / ML features / functions / models.

[0337] UE-side AI / ML LCM – UE AI / ML activation / deactivation Turn Figure 5 In step 6, the UE can perform AI / ML feature / function / model activation and / or deactivation based on the indication / signaling from the NW in step 5 above. In some example implementations, the UE can perform one or more of the following AI / ML feature / function / model activation operations when it receives DL MAC CE signaling, or in another example implementation, when it receives protocol signaling terminated between the AI / ML logic layer at the UE and the AI / ML logic entity / unit at the NW in step 5: • Send instructions to lower levels regarding the activation / deactivation of the AI / ML function MAC CE.

[0338] • Instruct lower layers to activate AI / ML functions associated with traditional beam management. CSI- ReportConfig .

[0339] In some example implementations, the UE can perform one or more of the following AI / ML functions to activate the operation: • Send instructions to lower levels regarding the activation / deactivation of the AI / ML function MAC CE.

[0340] • Instruct lower layers to activate and deactivate AI / ML functions related to traditional beam management. CSI- ReportConfig .

[0341] UE-side AI / ML LCM – UE AI / ML Inference Report Turn now Figure 5 Step 7: After the UE performs activation and / or deactivation of AI / ML functions and / or actual measurements, an inference and / or measurement report can be provided from the UE to the NW. Such inference / measurement reports may contain predictions or inferences made by the UE based on the activated AI / ML functions and models and / or actual measurements, which the NW can use for network configuration and provisioning.

[0342] In some example implementations of step 7, when the associated AI / ML function / model is activated, the inference / measurement report can be implemented as PUCCH signaling or PUSCH signaling.

[0343] For example, for AI / ML-based space beam management, the inference / measurement report may include at least one of the following information or information items: • 1: The first N beam IDs and their corresponding N values ​​of Reference Signal Received Power (RSRP), where N >= 1; ·2: The Id of the first N beams and the RSRP values ​​of M beams among the N beams, where N>M>=0; ·3: The first N beams with RSRP values ​​greater than the predefined RSRP threshold.

[0344] • 4: Flag indicating the RSRP value of a beam. If this flag is present, its value can indicate whether the RSRP value of the corresponding beam with beam ID is inferred through AI / ML functions / models or actually measured.

[0345] Furthermore, such example inference / measurement reports can be triggered by at least one of the following events: • When the UE determines, based on the inference results, that the current DL Tx beam is not among the first N beams, it is necessary to switch beams.

[0346] • When the UE determines, based on the inference results, that the RSRP value of the current DL Tx beam is lower than a predefined threshold RSRP value, and there exists at least one DL Tx beam with an RSRP value higher than a predefined threshold RSRP value. These two predefined thresholds may be the same or different.

[0347] For example, in AI / ML-based temporal beam management, the inference / measurement report described in step 7 can be reported as the measurement or inference result at each configured reporting time. Such a report may include and / or indicate one or more of the following information or information items: • Each inference report contains N beam IDs with the best N RSRP values, where N>=1.

[0348] • Each inference report contains N beam Ids with the best M RSRP values, where N>M>=1.

[0349] • Indication, used to indicate whether the report contains inferences or actual measurements.

[0350] NW-side AI / ML LCM – Overview In some example implementations, AI / ML functions may reside on the NW side rather than the UE side. For AI / ML LCM functions on the NW side, while the NW may be able to control the selection, activation, and deactivation of AI / ML functions and models, it may still require UE assistance to provide parameters and information so that the NW can apply appropriate AI / ML functions. Such UE assistance information can be provided to the NW as one or more UE-addressed condition reports. Figure 11 The document illustrates the overall LCM procedure for NW-side AI / ML functions performed by UE 1102 and NW 1104, which includes the following example steps: • Step 1: NW 1104 can send a request message to UE 1102 for UE additional condition reporting.

[0351] • Step 2: UE 1102 may determine whether to trigger a UE Additional Condition Report based on one or more criteria. If the one or more criteria are met, UE 1102 proceeds to Step 3 below. Otherwise, the process ends.

[0352] Step 3: UE 1102 can then send a UE additional conditions report to NW 1104.

[0353] In some example implementations, the request message for reporting additional conditions to the UE in step 1 above (e.g., RequestUEAdditionalConditions) may be a DL RRC message (e.g., RRCReconfiguration). In another example implementation, it may be protocol signaling terminated between the AI / ML logical layer and the NW logical entity / unit. This request message may include and / or indicate at least one of the following information or information items: • UE speed. For example, UE speed can be provided as an enumeration type parameter with group values ​​of {15, 30, 60, 90, 120, etc.}. Alternatively, UE speed can be provided as an enumeration type parameter with group values ​​of {static, low, medium, high, etc.}, where each value can represent a speed range. For example, the value "static" can represent a speed range of [0, 5 km / h), the value "low" can represent a speed range of [5 km / h, 15 km / h), the value "medium" can represent a speed range of [15 km / h, 60 km / h), and the value "high" can represent a speed range of [60 km / h, 120 km / h).

[0354] • UE rotation, for example, can be an enumerated type parameter with values ​​in {30, 60, 90, 120, 180}.

[0355] In some example implementations, a disable timer (with a predefined initial timer value) can be introduced to control the reporting of additional conditions by the UE. For example, if the disable timer is still running, the report can be disabled and therefore cannot be triggered.

[0356] In some example implementations of step 2 above, at least one of the following conditions may need to be met to trigger the UE additional condition report: • The aforementioned timer for prohibiting UE additional condition reporting (if set or configured) is not running.

[0357] • The UE's additional conditions have changed since the most recent UE additional conditions report. For example: Since the last report, the UE speed has changed, for example, from static to low, from low to high, etc.

[0358] Since the most recent report, the UE speed has changed to exceed the configured speed range. For example, the UE speed has reached a value greater than 15 km / h, or the UE speed has exceeded 30 km / h, etc.

[0359] The UE rotation reaches a value greater than a threshold. For example, the UE has rotated more than 30 degrees or more than 60 degrees since the most recent report, etc.

[0360] In some example implementations, the condition for the first UE additional condition report to NW can be: since the last time the above-mentioned conditions were included. RequestUEAdditionalConditions of RRCReconfiguration Since the message was sent, no additional conditions have been reported by the UE.

[0361] In some example implementations of step 3 above, the report can be included RRCReconfigurationComplete The message, or the protocol signaling terminated between the AI / ML logical layer and the NW logical entity / unit, may contain at least one of the following information or information items regarding additional conditions for the UE: • 1: UE speed information: used to indicate when generated RRCReconfigurationComplete The UE's current speed information at the time of the message.

[0362] ·2: UE codebook information: Used to indicate UE codebook information about the RX beam.

[0363] •3: UE antenna array distribution information: used to indicate the antenna array distribution of the UE.

[0364] In some other example implementations of step 3 above, the report can be included UEAssistanceInformationThe (UE Assistance Information) message, or the protocol signaling terminated between the AI / ML logical layer and the NW logical entity / unit, may also contain at least one of the following information or information items regarding additional UE conditions: •1: UE speed information: Used to indicate the current speed information of the UE when the UE supplement report is generated.

[0365] ·2: UE rotation information: used for indication and reception RequestUEAdditionalConditions The UE's orientation / direction information is compared with the UE's current rotation information.

[0366] The above description and accompanying drawings provide specific example embodiments and implementations. However, the subject matter can be implemented in a variety of different forms, and therefore, the covered or claimed subject matter is intended to be construed as not being limited to any of the example embodiments described herein. The claimed or covered subject matter is intended to have a reasonably broad scope. In particular, for example, the subject matter can be implemented as a method, apparatus, component, system, or non-transitory computer-readable medium for storing computer code. Thus, embodiments can take the form of, for example, hardware, software, firmware, storage medium, or any combination thereof. For example, the above-described method embodiments can be implemented by components, apparatus, or systems that include memory and a processor executing computer code stored in the memory.

[0367] Throughout the specification and claims, terms may have nuanced meanings beyond their explicitly stated meanings, implied or suggested by the context. Similarly, the phrase "in one embodiment / implementation" as used herein does not necessarily refer to the same embodiment, and the phrase "in another embodiment / implementation" as used herein does not necessarily refer to a different embodiment. For example, it is intended that the claimed subject matter encompasses, in whole or in part, combinations of exemplary embodiments.

[0368] References to features, advantages, or similar language throughout this specification do not imply that all such features and advantages achievable using this solution should be, or be included in, any single implementation thereof. Rather, the language used to refer to these features and advantages is to be understood as indicating that a particular feature, advantage, or characteristic described in connection with an embodiment is included in at least one embodiment of this solution. Therefore, descriptions of features and advantages and similar language throughout this specification may refer to the same embodiment, but not necessarily the same embodiment.

[0369] Furthermore, in one or more embodiments, the described features, advantages, and characteristics of this solution can be combined in any suitable manner. Based on the description herein, those skilled in the art will recognize that this solution can be practiced without one or more of the specific features or advantages of a particular embodiment. In other instances, additional features and advantages may be recognized in certain embodiments that may not be present in all embodiments of this solution.

Claims

1. A method performed by a user equipment (UE) communicating with a network (NW) in a wireless communication network, comprising: Perform a capability reporting process to transmit the UE's AIML capabilities to the NW in an Artificial Intelligence or Machine Learning (AIML) capability report; Perform an applicable AIML reporting process to transmit a set of applicable AIML features, functions, or models of the UE to the NW; Receive AIML configuration for one or more applicable AIML features, functions, or models selected by the NW; and Based on the AIML configuration, predictions are made using one or more of the selected applicable AIML features, functions, or models.

2. The method according to claim 1, wherein, The capability reporting process includes: the UE sending an AIML capability report request to the NW, and sending the AIML capability report to the NW after receiving an AIML capability query from the NW.

3. The method according to claim 2, wherein, The AIML capability report request sent by the UE to the NW is triggered by at least one of the following: At least one of the AIML features, functions, or models stored at the UE has been changed; The computational resource requirements for at least one of the AIML features, functions, or models stored at the UE have changed; or At least one AIML feature, function, or model has been added to the UE or configured for the UE by the NW.

4. The method according to claim 2, wherein, The AIML capability report request includes uplink RRC signaling, which includes uplink UE Assistance Information (UAI) messages.

5. The method according to claim 2, wherein, The AIML capability report request includes at least one of the following: One or more indications are provided to indicate the AIML features, functions, or models that have been changed at the UE; or One or more AIML change indicators are used to indicate the type of change to the AIML feature, function, or model that has been modified at the UE.

6. The method according to claim 1, wherein, The AIML capability report includes at least one of the following: A list of one or more AIML-based features supported by the UE; One or more AIML-based functions associated with the one or more AIML-based features and supported by the UE; A list of one or more AIML-based models supported by the UE and associated with the one or more AIML-based functions or the one or more AIML-based features; Indication of the supported wireless frequency bands for each of the aforementioned AIML-based features, functions, or models; The consumption of computing resources associated with one or more of the AIML-based features, functions, or models; The UE can provide the maximum computing resources to support the AIML-based features, functions, or models; or One or more configurations, scenarios, contexts, or conditions associated with the training of the AIML-based features, functions, or models.

7. The method according to claim 1, wherein, The set of applicable AIML features, functionalities, or models includes at least one of the following: Space beam prediction based on AIML; AIML-based temporal beam prediction; AIML-based channel state information (CSI) feedback compression or decompression; or AIML-based temporal CSI prediction.

8. The method according to claim 1, wherein, Executing the applicable AIML report includes: receiving an applicable AIML report request from the NW and sending the applicable AIML report to the NW.

9. The method according to claim 8, wherein, The applicable AIML report request includes a system information message or a special downlink RRC message constructed for an AIML report request for applicable AIML features, functions, or models at the UE.

10. The method according to claim 8, wherein, The applicable AIML report is included in a UE auxiliary information message, an uplink MAC CE, or a special uplink RRC message for reporting applicable AIML features, functions, or models at the UE.

11. The method according to claim 10, wherein, The applicable AIML report includes a bit string indicating the applicability of AIML features, functions, or models supported within the UE capabilities.

12. The method according to claim 8, wherein, The applicable AIML report request includes an RRC reconfiguration message, and the applicable AIML report is included in the RRC reconfiguration complete message.

13. The method according to claim 12, wherein, The applicable AIML report includes a bit string indicating the applicability of AIML features, functions, or models supported within the AIML capabilities of the UE.

14. The method according to claim 8, wherein, The applicable AIML report request includes at least one of the following: A list of additional conditions associated with at least one of the AIML features, functions, or models; One or more indicators are used to identify the at least one AIML feature, function, or model associated with the additional condition; or One or more indicators are used to identify at least one of the AIML features, functions, or models, wherein the applicable AIML report is requested for the at least one AIML feature, function, or model.

15. The method according to claim 14, wherein, The one or more indicators are provided by cell group, by frequency band, or by cell.

16. The method of claim 14, wherein, The list of additional conditions includes at least one of the following: Indicators indicating the boundaries of the residential area; The antenna height of the base station in the current cell; Non-line-of-sight probability; Indoor / outdoor conditions; Downlink transmission beamcodebook indication for AIML model training; The antenna array dimensions of the base station; The downtilt angle of the base station's antenna; or At least one beam pattern.

17. The method according to claim 8, wherein, The sending of the applicable AIML report by the UE to the NW is triggered by at least one of the following: The RRC configuration associated with the applicable AIML report has been established, and no other indications of applicable AIML features, functions, or models have been previously sent to the NW; At least one AI / ML function or model for AIML-based features has been activated; or Since the UE last made an applicable AIML report, at least one AI / ML function or model stored in the UE for the configured AIML-based features has been changed.

18. A method performed by a network node communicating with a user equipment (UE) in a wireless communication network, comprising: Receive an AIML capability report from the UE, the AIML capability report indicating the AIML capabilities of the UE; Obtain an applicable AIML report from the UE, the applicable AIML report indicating a set of applicable AIML features, functions or models of the UE; Generate AIML configurations for one or more selected applicable AIML features, functions, or models; and The AIML configuration is sent to the UE so that the UE can make predictions using one or more of the selected applicable AIML features, functions, or models according to the AIML configuration.

19. The method of claim 18, further comprising: Before receiving the AIML capability report from the UE, receive the AIML capability report request from the UE.

20. The method of claim 19, further comprising: Send an AIML capability query to the UE so that the UE can respond using the AIML capability report.

21. The method according to claim 19, wherein, The AIML capability report request includes uplink RRC signaling, which includes uplink UE Assistance Information (UAI) messages.

22. The method according to claim 19, wherein, The AIML capability report request includes at least one of the following: One or more indications are provided to indicate the AIML features, functions, or models that have been changed at the UE; or One or more AIM change indicators are used to indicate the type of change to the AIML feature, function, or model that has been modified at the UE.

23. The method according to claim 18, wherein, The AIML capability report includes at least one of the following: A list of one or more AIML-based features supported by the UE; One or more AIML-based functions associated with the one or more AIML-based features and supported by the UE; A list of one or more AIML-based models supported by the UE and associated with the one or more AIML-based functions or the one or more AIML-based features; Indication of the supported wireless frequency bands for each of the aforementioned AIML-based features, functions, or models; The consumption of computing resources associated with one or more of the AIML-based features, functions, or models; The UE can provide the maximum computing resources to support the AIML-based features, functions, or models; or One or more configurations, scenarios, contexts, or conditions associated with the training of the AIML-based features, functions, or models.

24. The method according to claim 18, wherein, The set of applicable AIML features, functionalities, or models includes at least one of the following: Space beam prediction based on AIML; AIML-based temporal beam prediction; AIML-based channel state information (CSI) feedback compression or decompression; or AIML-based temporal CSI prediction.

25. The method according to claim 18, further comprising: Send an AIML report request to the UE, so that the UE can send the AIML report in response.

26. The method of claim 25, wherein, The applicable AIML report request includes a system information message or a special downlink RRC message constructed for an AIML report request for applicable AIML features, functions, or models at the UE.

27. The method according to claim 25, wherein, The applicable AIML report is included in a UE auxiliary information message, an uplink MAC CE, or a special uplink RRC message for reporting applicable AIML features, functions, or models at the UE.

28. The method according to claim 27, wherein, The applicable AIML report includes a bit string indicating the applicability of AIML features, functions, or models supported within the UE capabilities.

29. The method according to claim 25, wherein, The applicable AIML report request includes an RRC reconfiguration message, and the applicable AIML report is included in the RRC reconfiguration complete message.

30. The method according to claim 29, wherein, The applicable AIML report includes a bit string indicating the applicability of AIML features, functions, or models supported within the AIML capabilities of the UE.

31. The method according to claim 25, wherein, The applicable AIML report request includes at least one of the following: A list of additional conditions associated with at least one of the AIML features, functions, or models; One or more indicators are used to identify the at least one AIML feature, function, or model associated with the additional condition; or One or more indicators are used to identify at least one of the AIML features, functions, or models, wherein the applicable AIML report is requested for the at least one AIML feature, function, or model.

32. The method according to claim 31, wherein, The one or more indicators are provided by cell group, by frequency band, or by cell.

33. The method according to claim 31, wherein, The list of additional conditions includes at least one of the following: Indicators indicating the boundaries of the residential area; The antenna height of the base station in the current cell; Non-line-of-sight probability; Indoor / outdoor conditions; Downlink transmission beamcodebook indication for AIML model training; The antenna array dimensions of the base station; The downtilt angle of the base station's antenna; or At least one beam pattern.

34. A UE or network node according to any one of claims 1 to 33, wherein the UE or network node comprises a processor and a memory, wherein, The processor is configured to read computer code from the memory to cause the UE or the network node to perform the method according to any one of claims 1 to 33.

35. A computer program product comprising a non-transitory computer-readable program medium storing computer code thereon, the computer code causing the processor to implement the method according to any one of claims 1 to 33 when executed by a processor of a UE or a network node according to any one of claims 1 to 33.

Images