Wireless terminal and method

Abstract

To enable a network (e.g., a radio access network (RAN) node) to control whether a wireless terminal is permitted to take action on the basis of a prediction or decision made by a trained machine learning model.SOLUTION: A wireless terminal receives control information from a network indicating whether the wireless terminal is authorized to perform a specific action on the basis of the inference results of the machine learning-based artificial intelligence. The wireless terminal makes a prediction or decision using the trained machine learning model. When the received control information permits it, the wireless terminal performs a specific action triggered by the prediction or decision made by the machine learning model. The specific action may include, for example, transmitting assistance information to the network indicating the results of the prediction or decision made using the trained machine learning model.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present disclosure relates to a wireless communication network, and more particularly to the application of artificial intelligence (AI) to a wireless communication network.

Background Art

[0002] The 3rd Generation Partnership Project (3GPP (registered trademark)) is discussing the application or introduction of AI or machine learning (ML) to 5G. AI / ML can be considered for both internal network functions and the air interface (i.e., Uu). In 3GPP Release 17, Radio Access Network (RAN) Working Group #3 (RAN3) is discussing network-based AI / ML without User Equipment (UE) participation. Examples of targets here include energy saving, load balancing, and mobility optimization (see, for example, Non-Patent Documents 1 and 2). In network-based AI / ML, the network performs AI / ML inference. AI / ML inference means a prediction or decision based on a trained machine learning model. The AI / ML inference function may be located in the Next Generation Radio Access Network (NG-RAN) (e.g., gNB). The training of the machine learning model may be performed in the NG-RAN. Alternatively, Operation, Administration and Maintenance (OAM) may train the machine learning model and supply the trained machine learning model (i.e., trained parameters) to the NG-RAN (e.g., gNB).

[0003] Furthermore, for 3GPP Release 18, network-based AI / ML with UE involvement and UE-based AI / ML have been proposed (see, for example, Non-Patent Documents 3-5). In UE-based AI / ML, the UE performs AI / ML inference. Possible use cases for AI / ML for air interfaces include Channel State Information (CSI) feedback compression, beam management, positioning, Reference Signal (RS) overhead reduction, and mobility. In UE-based AI / ML, the UE runs an AI model (i.e., a trained machine learning model) and obtains the AI ​​inference results locally. For example, the UE can predict future events or measurements based on past measurements. The UE can then feed back the predicted results (e.g., mobility or beam predictions) to the network (e.g., gNB). [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] CMCC, "Revised SID: Study on enhancement for data collection for NR and ENDC", RP-201620, 3GPP TSG-RAN Meeting #89e, Electronic meeting, September 14-18, 2020 [Non-Patent Document 2] 3GPP TR 37.817 V0.3.0 (2021-08) "3rd Generation Partnership Project; Technical Specification Group RAN; Evolved Universal Terrestrial Radio Access (E-UTRA) and NR; Study on enhancement for Data Collection for NR and EN-DC (Release 17)", September 23, 2021 [Non-Patent Document 3] Ericsson, "Views on Rel-18 AI / ML on Air-Interface", RP-212346, 3GPP TSG-RAN Meeting#93e, Electronic meeting, September 13-17, 2021 [Non-Patent Document 4] ZTE, "Support of Artificial Intelligence Applications for 5G Advanced", RP-212383, 3GPP TSG-RAN Meeting#93e, Electronic meeting, September 13-17, 2021 [Non-Patent Document 5] Xiaomi, "Mobility enhancement by UE based AI", RP-211787, 3GPP TSG-RAN Meeting#93e, Electronic meeting, September 13-17, 2021 [Overview of the project] [Problems that the invention aims to solve]

[0005] The inventors of this invention have examined UE-based AI / ML and identified various challenges. One of these challenges concerns the execution conditions of UE-based AI / ML. For example, a UE performs some action based on predictions or decisions using a machine learning model. However, if this UE action results in interaction with the network or could affect the network's performance, it may not be appropriate for the UE to perform such action freely. Alternatively, it may not be appropriate for the UE to perform actions that are triggered by AI inference rather than by rules (or criteria or formulas) defined in the 3GPP specification. Actions triggered by rules (or criteria or formulas) defined in the 3GPP specification include, for example, sending measurement reports for handover, performing conditional mobility, beam selection, and cell (re)selection.

[0006] One of the objectives that the embodiments disclosed herein seek to achieve is to provide apparatus, methods, and programs that contribute to solving at least one of several problems relating to UE-based AI / ML, including the problems described above. It should be noted that this objective is only one of several objectives that the embodiments disclosed herein seek to achieve. Other objectives or problems and novel features will be evident from the description herein or from the accompanying drawings. [Means for solving the problem]

[0007] In a first embodiment, the wireless terminal includes at least one memory and at least one processor coupled to the at least one memory. The at least one processor is configured to receive control information from a network indicating whether the wireless terminal is permitted to perform a particular action based on the inference results of machine learning-based artificial intelligence. The at least one processor is configured to make predictions or decisions using a trained machine learning model. The at least one processor is configured to perform the particular action triggered by the prediction or decision if the control information permits the performance of the particular action based on the inference results.

[0008] In the second embodiment, the method performed by the wireless terminal includes the following steps: (a) Receiving control information from the network indicating whether the wireless terminal is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence, (b) making predictions or decisions using trained machine learning models, (c) If the control information permits the execution of the specific action based on the inference result, perform the specific action triggered by the prediction or decision.

[0009] In a third embodiment, a radio access network (RAN) node includes at least one memory and at least one processor coupled to the at least one memory. The at least one processor is configured to transmit control information to the radio terminal indicating whether the radio terminal is permitted to perform a particular action based on the inference results of machine learning-based artificial intelligence. If the control information permits the performance of the particular action based on the inference results, the control information causes the radio terminal to perform the particular action, which is triggered by a prediction or decision using a trained machine learning model.

[0010] In a fourth aspect, the method performed by the RAN node includes transmitting control information to the wireless terminal indicating whether the wireless terminal is permitted to perform a specific action based on the inference results of a machine learning-based artificial intelligence. If the control information permits the performance of the specific action based on the inference results, the control information causes the wireless terminal to perform the specific action, which is triggered by a prediction or decision using a trained machine learning model.

[0011] The fifth aspect is directed to a program, which, when loaded into a computer, includes a set of instructions (software code) for causing the computer to perform the methods described in the second or fourth aspect above. [Effects of the Invention]

[0012] According to the above-described embodiment, it is possible to provide an apparatus, method, and program that contribute to solving at least one of several problems related to UE-based AI / ML. [Brief explanation of the drawing]

[0013] [Figure 1] This figure shows an example configuration of a wireless communication system according to the embodiment. [Figure 2] This is a sequence diagram showing an example of the operation of a wireless terminal and a wireless access network node according to the embodiment. [Figure 3] This flowchart shows an example of the operation of a wireless terminal according to the embodiment. [Figure 4A] This figure shows a specific example of the control information format according to the embodiment. [Figure 4B] This figure shows a specific example of the control information format according to the embodiment. [Figure 5] This is a sequence diagram showing an example of the operation of a wireless terminal and a wireless access network node according to the embodiment. [Figure 6] This figure shows a specific example of the AI ​​interest display format according to the embodiment. [Figure 7]It is a flowchart showing an example of the operation of a radio access network node according to an embodiment. [Figure 8] It is a block diagram showing a configuration example of a wireless terminal according to an embodiment. [Figure 9] It is a block diagram showing a configuration example of a radio access network node according to an embodiment.

Mode for Carrying Out the Invention

[0014] Hereinafter, specific embodiments will be described in detail with reference to the drawings. In each drawing, the same or corresponding elements are denoted by the same reference numerals, and redundant descriptions are omitted as necessary for clarity of explanation.

[0015] The plurality of embodiments described below can be implemented independently or in appropriate combination. These plurality of embodiments have different novel features. Therefore, these plurality of embodiments contribute to solving different purposes or problems and to achieving different effects.

[0016] The plurality of embodiments shown below are mainly described with respect to the 3GPP fifth-generation mobile communication system (5G system). However, these embodiments may be applied to other wireless communication systems.

[0017] As used in this specification, depending on the context, "(if)~then" may be interpreted to mean "when", "at or around the time", "after", "upon", "in response to determining", "in accordance with a determination", or "in response to detecting". These expressions may be interpreted to have the same meaning depending on the context.

[0018] First, the configuration and operation of several network elements common to multiple embodiments will be described. Figure 1 shows an example configuration of a wireless communication system according to several embodiments. In the example in Figure 1, the wireless communication system includes a wireless terminal (ie, UE) 1 and a wireless access network (RAN) node (e, gNB) 2. Each element (network function) shown in Figure 1 can be implemented, for example, as a network element on dedicated hardware, as a running software instance on dedicated hardware, or as an instantiated virtualization function on an application platform.

[0019] UE1 has at least one radio transceiver and is configured to communicate wirelessly with RAN node 2. UE1 is connected to RAN node 2 via air interface 101. RAN node 2 manages the cell and is configured to communicate wirelessly with multiple UEs, including UE1, using cellular communication technology (e.g., NR Radio Access Technology (RAT)). UE1 may be connected to multiple RAN nodes simultaneously for dual connectivity (DC).

[0020] RAN node 2 may be a Central Unit (e.g., gNB-CU) in a cloud RAN (C-RAN) deployment, or a combination of a CU and one or more Distributed Units (e.g., gNB-DUs). C-RAN is also called a CU / DU split. Furthermore, a CU may include a Control Plane (CP) Unit (e.g., gNB-CU-CP) and one or more User Plane (UP) Units (e.g., gNB-CU-UP). Therefore, RAN node 2 may be a CU-CP, or a combination of a CU-CP and a CU-UP. A CU may be a logical node hosting the gNB's Radio Resource Control (RRC), Service Data Adaptation Protocol (SDAP), and Packet Data Convergence Protocol (PDCP) protocols (or the gNB's RRC and PDCP protocols). A DU may be a logical node hosting the gNB's Radio Link Control (RLC), Medium Access Control (MAC), and Physical (PHY) layers.

[0021] UE1 may perform AI / ML inference locally. This AI / ML inference may also relate to the optimization of a wireless access network. UE1 may run AI inference on a trained machine learning model and take one or more actions according to the prediction or decision based on the AI ​​inference. The machine learning model may be any model known in the field of machine learning, including deep learning. The machine learning model may be, but is not limited to, a neural network model, a support vector machine model, a decision tree model, a random forest model, or a K-nearest neighbor model.

[0022] For example, the predictions or decisions based on AI inference by UE1 and one or more actions triggered thereon relate to beam management, mobility, or both. The one or more actions include, but are not limited to, at least one of the following: cell reselection, sending measurement reports, performing conditional mobility, and downlink beam selection. The downlink beam may be a Synchronization Signal (SS) / Physical Broadcast Channel (PBCH) block (SSB) beam. For example, the machine learning model may output predictions of candidate cells for cell reselection, target cells or nodes for handover, candidate cells or nodes for conditional mobility, candidate beams for beam selection, or UE trajectories. Furthermore, or alternatively, the machine learning model may predict or determine the timing of actions for mobility or beam management.

[0023] Cell reselection may occur when UE1 is RRC_IDLE or RRC_INACTIVE.

[0024] Mobility may occur when UE1 is RRC_CONNECTED. Mobility in RRC_CONNECTED may be a handover. The handover may be a Dual Active Protocol Stack (DAPS) handover. Furthermore, or alternatively, mobility in RRC_CONNECTED may relate to various mobilitys in the DC. Specifically, mobility may be a change of the primary cell in a Master Cell Group (MCG) in the DC, an inter-Master Node (MN) handover in the DC, a change of a secondary node in the DC, or an addition or change of the primary cell in a secondary cell group in the DC.

[0025] Conditional mobility may occur when UE1 is RRC_CONNECTED. Conditional mobility may also be a conditional handover. Furthermore, or alternatively, conditional mobility may relate to various mobilitys in the DC. Specifically, conditional mobility may be a change of the primary cell of a Master Cell Group (MCG) in the DC, an inter-Master Node (MN) handover in the DC, a change of a secondary node in the DC, or an addition or change of the primary cell of a secondary cell group in the DC.

[0026] The selection of the downlink beam may be performed during the beam failure recovery (BFR) procedure.

[0027] Training of machine learning models for AI / ML inference using UE1 may be performed by UE1 itself or by a network (e.g., OAM, RAN node 2). This training method may be offline learning, online learning, or a combination of both.

[0028] Similarly, RAN node 2 may perform AI / ML inference. This AI / ML inference may also relate to the optimization of the wireless access network. RAN node 2 may run AI inference on a trained machine learning model and take one or more actions according to the prediction or decision based on the AI ​​inference. The machine learning model may be any model known in the field of machine learning, including deep learning. The machine learning model may be, but is not limited to, a neural network model, a support vector machine model, a decision tree model, a random forest model, or a K-nearest neighbors model.

[0029] For example, and not limited to, predictions or decisions based on AI inference by RAN node 2 and one or more actions triggered thereby relate to at least one of the following: energy saving, load balancing, mobility optimization, CSI feedback enhancement, and positioning accuracy enhancement. For example, the predictions or decisions of the machine learning model relate to either or both energy saving strategies and mobility strategies. With respect to mobility strategies, the machine learning model may output predicted results for handover target cells or nodes, candidate cells or nodes for conditional mobility, or UE trajectories.

[0030] Training of machine learning models for AI / ML inference using RAN node 2 may be performed by RAN node 2 or by OAM. The training method may be offline learning, online learning, or a combination of both.

[0031] <First Embodiment> The configuration example of the wireless communication system according to this embodiment may be the same as the example shown in Figure 1. Figure 2 shows an example of the operation of UE1 and RAN node 2 for UE-based AI / ML, that is, for UE1 to perform AI inference on a machine learning model. In step 201, RAN node 2 transmits control information to UE1. This control information may be transmitted via broadcast (e.g., system information). The system information may be any system information block. Alternatively, this control information may be transmitted for UE1 via dedicated signaling. The dedicated signaling may be a dedicated RRC signaling, such as an RRC Reconfiguration message, an RRC Reestablishment message, an RRC Resume message, or an RRC Setup message. This control information may be an information element contained within an SIB or dedicated RRC signaling. The name of this information element is not limited, but may be, for example, "AI configuration".

[0032] The control information indicates whether UE1 is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence. In other words, the control information indicates whether UE1 is permitted to perform a specific action triggered by a prediction or decision using a trained machine learning model. If the control information indicates that it is permitted, then the control information causes UE1 to perform the specific action triggered by a prediction or decision using a trained machine learning model. RAN node 2 may send the control information (step 201) only if it is permitted for UE1 to perform a specific action based on the AI / ML inference results. In this case, sending the control information can implicitly indicate permission for UE1 to perform a specific action based on the AI / ML inference results.

[0033] The term "machine learning-based artificial intelligence inference results" may be rephrased as "inference results by artificial intelligence," "self-learning estimation results (by UE)," "self-correction estimation results," or "optimization (algorithm) estimation results." The expression "perform a specific action based on machine learning-based artificial intelligence inference results" may be rephrased as "apply artificial intelligence inference results to a specific action," "perform self-optimization (by UE) to a specific action," or "apply optimization (algorithm) to a specific action." The term "wireless access network optimization" may mean, for example, the optimization of the functionality or processing of a wireless network (e.g., one or more RAN nodes), or the optimization of the values ​​or settings of wireless parameters that the wireless network sets for wireless terminals (e.g., one or more UEs).

[0034] Specific examples of predictions or decisions based on AI inference by UE1 and one or more actions triggered thereby are as described above. Examples, but not limited to, of specific actions triggered by predictions or decisions using machine learning models in UE1 relate to at least one of beam management, mobility, CSI feedback, and positioning (or location estimation). Specific actions include, but are not limited to, at least one of cell reselection, sending measurement reports, performing conditional mobility, and selecting downlink beams. Sending measurement reports may trigger a decision by the network (e.g., RAN node 2) to initiate mobility (e.g., handover).

[0035] For example, and not an exhaustive list, a specific action may relate to CSI feedback. Another specific action may be adjusting the timing (e.g., period) of the CSI report. Alternatively, a specific action may be reducing or compressing the amount of information in the CSI report (e.g., beam information, channel matrix, precoding matrix).

[0036] As an example, though not an exhaustive one, a specific action relates to the positioning (or position estimation) of UE1. A specific action may also involve correcting the information used for position estimation, or correcting the position estimation result.

[0037] Furthermore, or alternatively, certain actions may include transmitting UE assistance information (e.g., UE assistance information) to the network (e.g., RAN node 2) indicating the results of predictions or decisions made using a first machine learning model at UE1. This assistance information may be used to train a second machine learning model for optimizing the wireless access network. This assistance information may be used for inference execution on the second trained machine learning model for optimizing the wireless access network. The second machine learning model and the second trained machine learning model may be located at RAN node 2 or OAM. Furthermore, or alternatively, this assistance information may be used to train a first machine learning model located at UE1 (e.g., offline reinforcement learning).

[0038] The control information may indicate whether, in lieu of, or in addition to, performing a specific action based on a predefined rule (or criterion or formula), the execution of that specific action based on the AI ​​inference result is permitted. The predefined rule may be a rule specified in the 3GPP specification. Alternatively, the predefined rule may be a rule pre-configured by the network (e.g., RAN node 2). More specifically, the predefined rule may be a cell reselection rule (or criterion or formula), measurement reporting rule (or criterion or formula), conditional mobility execution rule (or criterion or formula), or beam selection rule (or criterion or formula) specified in the 3GPP specification or configured by the network. These rules (or criterion or formula) are generally based on a comparison of cell or beam quality (e.g., Reference Signal Received Power (RSRP)) with a threshold.

[0039] The control information may collectively indicate to UE1 permission to use machine learning-based artificial intelligence (i.e., UE-based AI / ML) for multiple actions, including a specific action. Alternatively, the control information may individually indicate to UE1 whether each of the multiple actions, including the specific action described above, is permitted. The control information may indicate the actions or categories (or groups) of actions for which UE-based AI / ML is permitted. The control information may indicate the functions or features (e.g., mobility, power saving) for which UE-based AI / ML is permitted. The control information may indicate subfunctions or subfeatures (e.g., cell selection, handover, beam management) for which UE-based AI / ML is permitted. The control information may indicate the procedures (e.g., RRC re-establishment, beam fault recovery (BFR)) for which UE-based AI / ML is permitted. The control information may indicate the RRC configuration (e.g., at the level of information elements or fields in Abstract Syntax Notation One (ASN.1)) for which UE-based AI / ML is permitted.

[0040] The control information may indicate the conditions under which a specific action based on AI / ML inference results is permitted. In this case, UE1 may perform the specific action triggered by the prediction or decision of the machine learning model (i.e., UE-based AI / ML) only if the indicated conditions are met. The conditions under which the specific action is permitted may indicate constraints relating to at least one of frequency band, location, and time.

[0041] For example, the condition may indicate a location (e.g., geographical area, cell, or set of cells) where a specific action based on AI / ML inference results is permitted. The condition may indicate a time or period during which a specific action based on AI / ML inference results is permitted. The condition may indicate one or more frequency bands during which a specific action based on AI / ML inference results is permitted.

[0042] Furthermore, or alternatively, the condition may be that a consistent fault has been detected. For example, the condition may allow UE1 to perform a specific action based on AI / ML inference results if a given fault is repeatedly detected in the same or similar circumstances (e.g., location, cell, cell pair, time, frequency band, time). The given fault may be, for example, a handover fault or a beam fault. The circumstances may be at least one of location, cell, cell pair, time, frequency band, and time.

[0043] Furthermore, or alternatively, the condition may be that UE1 receives a signal from the network (e.g., RAN node 2) indicating a predetermined identifier. The predetermined identifier may be, but is not limited to, a Config Set Number or ID, an AI Set Number or ID, or a Combination Set Number or ID. The predetermined identifier may be set by the network to represent the same or similar situation (e.g., the transmit power of RAN node 2, antenna tilt, number of downlink beams, or other physical layer settings).

[0044] Figure 3 shows an example of UE1's operation. In step 301, UE1 receives the control information described above. In step 302, UE1 makes a prediction or decision using the trained machine learning model. In step 303, if the control information indicates that it is permitted, UE1 performs the specific action triggered by the prediction or decision based on the machine learning model. In step 303, UE1 may consider the conditions indicated in the control information. As described above, these conditions may be conditions for which UE1 is permitted to perform a specific action based on the AI / ML inference result. In this case, UE1 may perform the specific action triggered by the prediction or decision of the machine learning model (i.e., UE-based AI / ML) only if the indicated conditions are met.

[0045] Figure 4A shows an example of the format of control information (steps 201, 301). The Ai-configuration Information Element (IE) 401 shown in Figure 4A may be sent from RAN node 2 to UE1 in an SIB or individual RRC message. The Ai-configuration IE 401 may contain an AI-ML-ConfigList IE 402. The AI-ML-ConfigList IE 402 contains one or more AI-ML-Config IEs 403. Each AI-ML-Config IE 403 may contain one or any combination of targetFeature IE 404, targetArea IE 405, and condition IE 406.

[0046] The targetFeature IE404 may indicate one or more categories (or groups) of actions for which UE-based AI / ML is permitted. The action categories may include, for example, one or more of the following: “mobilityConnected”, “mobilityIdleInactive”, “beamManagement”, “energySaving”, “rrmMeasurement”, “csiFeedback”, and “positioningAccuracy”.

[0047] If “mobilityConnected” is set for targetFeature IE404, this indicates that, for example, UE1 is permitted to apply UE-based AI / ML to mobility functions in the RRC_Connected state. The application of UE-based AI / ML to mobility functions in the RRC_Connected state may be adjustments related to handover or adjustments related to conditional handover (CHO). Adjustments related to handover may be adjustments to the timing of measurement report (MR) reporting, or adjustments to the offset value or threshold of MR events. Adjustments related to CHO processing may be adjustments to the offset value or threshold of the CHO execution conditions. Furthermore, or alternatively, the application of UE-based AI / ML to mobility functions in the RRC_Connected state may be adjustments related to the addition or modification of Multi-Radio Dual Connectivity (MR-DC) PSCells, or adjustments related to conditional PSCell addition / change (CPAC). PSCell is an abbreviation for Primary SCell or Primary Secondary Cell Group (SCG) Cell.

[0048] If “mobilityIdleInactive” is set for targetFeature IE404, this indicates that UE1 is permitted to apply UE-based AI / ML to mobility functions in either the RRC_IDLE state, the RRC_INACTIVE state, or both. Applying UE-based AI / ML to mobility functions in the RRC_IDLE or RRC_INACTIVE state may mean adjustments related to cell reselection. Adjustments related to cell reselection may be adjustments to parameters used in the cell reselection process. For example, this may be adjustments to offset values ​​or thresholds, adjustments to priority per frequency or between frequencies, or adjustments to priority per network slice or between network slices.

[0049] If “beamManagement” is set in targetFeature IE404, this indicates, for example, that UE1 is permitted to apply UE-based AI / ML to beam management functions. Applying UE-based AI / ML to beam management functions may mean adjustments related to beam selection by UE1, adjustments related to beam failure detection (BFD), or adjustments related to beam failure recovery (BFR). Adjustments related to beam selection may be adjustments to the offset value or threshold of radio quality (e.g., RSRP, Reference Signal Received Quality (RSRQ), Received Signal Strength Indicator (RSSI)), or adjustments to beam-specific or inter-beam priority. Adjustments related to beam failure detection may be adjustments to the threshold for determining BFD. Adjustments related to beam failure recovery may be adjustments to the beam selection criteria for BFR (e.g., threshold or type of reference signal (RS) (e.g., SSB or CSI-RS)).

[0050] If “energySaving” is set in targetFeature IE404, this indicates, for example, that UE1 is permitted to apply UE-based AI / ML to power saving functions. Applying UE-based AI / ML to power saving functions may involve adjusting the timing (e.g., period, target frequency) of measurements in either or both the serving cell and / or neighbor cell to lead to (or be expected to lead to) a reduction in power consumption. Such measurements may be either or both Radio Link Monitoring (RLM) and / or Radio Resource Management (RRM) measurements.

[0051] If “rrmMeasurement” is set in targetFeature IE404, this indicates, for example, that UE1 is permitted to apply UE-based AI / ML to the RRM measurement function. Applying UE-based AI / ML to the RRM measurement function may involve adjusting the timing of measurements to reduce power consumption as described above. Alternatively, this may involve adjusting the frequency of RRM measurements or the criteria for determining the target relaxation (i.e., RRM relaxation). For example, the threshold for determining not-at-cell edge or low-mobility may be adjusted. For example, if UE1 determines itself to be not-at-cell edge or low-mobility, UE1 may be permitted to lengthen the RRM measurement period or reduce the target frequency.

[0052] If “csiFeedback” is set in targetFeature IE404, this indicates, for example, that UE1 is permitted to apply UE-based AI / ML to the CSI feedback function. Applying UE-based AI / ML to the CSI feedback function may involve adjusting the timing (e.g., period) of the CSI report to RAN node 2. Alternatively, applying UE-based AI / ML to the CSI feedback function may involve reducing or compressing the amount of information in the CSI report (e.g., beam information, channel matrix, precoding matrix).

[0053] If “positioningAccuracy” is set for targetFeature IE404, this indicates, for example, that UE1 is allowed to apply UE-based AI / ML to features related to improving positioning accuracy. Applying UE-based AI / ML to features related to improving positioning accuracy may mean correcting the information used for position estimation, or correcting the position estimation results.

[0054] The targetArea IE405 may indicate one or more areas (i.e., functional areas to which UE-based AI / ML is permitted) where application is allowed. The areas include, for example, one or more of “intraFreq”, “interFreq”, “intraAndInterFreq”, “interRAT”, and “any”. “intraFreq” means cell reselection or handover within the same frequency band. “interFreq” means cell reselection or handover between different frequency bands. “intraAndInterFreq” means cell reselection or handover within the same frequency band and between different frequency bands. “interRAT” means cell selection or handover between different radio access technologies. “any” means no area restrictions. Alternatively, the fact that targetArea IE405 is not included in AI-ML-Config IE403 may mean no area restrictions. Furthermore, targetArea may be defined as targetFunc (targetFunction), targetScope, or applyableTarget, which may indicate one or more target functions, one or more scopes, or one or more applicable targets, respectively, where UE-based AI / ML is permitted to be applied.

[0055] Condition IE406 may indicate a situation in which the application of UE-based AI / ML is permitted. For example, the situation may include one or more of the following: “consistentFailure”, “predictable”, “lowBattery”, “gnssAvailable”, and “nlos”. “consistentFailure” means a situation in which a given fault is repeatedly detected in the same or similar circumstances. “predictable” means a situation that can be predicted based on history. “lowBattery” means a situation in which the remaining battery power of UE1 is low. “gnssAvailable” means a situation in which location information can be obtained by the Global Navigation Satellite System (GNSS). GNSS may be, for example, the Global Positioning System (GPS), Galileo, and Global Navigation Satellite System (GLONASS). “nlos” means that the radio environment (or communication environment) of UE1 is a non-line of sight (NLOS) environment.

[0056] Figure 4B shows another example of the format of the control information (steps 201, 301). The Ai-configuration Information Element (IE) 421 shown in Figure 4B may be sent from RAN node 2 to UE1 in an SIB or individual RRC message. The Ai-configuration IE 421 may contain an AI-ML-ConfigList IE 422. The AI-ML-ConfigList IE 422 contains one or more AI-ML-Config IEs 423. Each AI-ML-Config IE 423 may contain one or any combination of mobility IE 424, energySaving IE 425, and positioningAccuracy IE 426.

[0057] The mobility IE424 specifies a mobility-related target function (targetFuncMob IE) and target scope (targetScopeMob IE). The target function may indicate one or more categories (or groups) of actions to which UE-based AI / ML is permitted to be applied. The action categories may include one or more of the following, for example: “mobilityConnected”, “mobilityIdleInactive”, “beamManagement”, and “rrmMeasurement”. The target scope may indicate one or more scopes to which UE-based AI / ML is permitted to be applied. The target scope may include one or more of the following, for example: “intraFreq”, “interFreq”, “intraAndInterFreq”, “interRAT”, and “any”.

[0058] The energySaving IE425 specifies target functions (targetFuncES IE) and target scopes (targetScopeES IE) related to power saving. A target function may indicate one or more categories (or groups) of actions to which UE-based AI / ML is permitted. The action categories may include, for example, one or more of “rrmMeasurement” and “csiFeedback”. A target scope may indicate one or more scopes to which UE-based AI / ML is permitted. The target scopes may include, for example, one or more of “intraFreq”, “interFreq”, “intraAndInterFreq”, “interRAT”, “any”, “pCell”, “sCell”, and “servCell”. “pCell” means the primary cell of a carrier aggregation or the MCG or SCG primary cell of a DC. “sCell” means the secondary cell of a DC or carrier aggregation. “servCell” means the serving cell of UE1.

[0059] The positioningAccuracy IE426 specifies the target function (targetFuncPosi IE) and target scope (targetScopePosi IE) related to positioning accuracy. The target function may indicate one or more categories (or groups) of actions for which UE-based AI / ML is permitted. The action categories may include one or more of the following, for example: “nlosMitigation”, “multipathMitigation”, and “propagationDelayCompensation”. “nlosMitigation” means correcting the positioning result (or UL transmission timing) when UE1 is in an NLOS environment. “multipathMitigation” means correcting the positioning result (or UL transmission timing) when UE1 is in a multipath environment. “propagationDelayCompensation” means compensating for the effects of propagation delay (e.g., UL transmission timing). The target scope may indicate one or more scopes for which UE-based AI / ML is permitted. The target scope includes, for example, one or more of "positioning" and "timingAdvance". "positioning" means the acquisition of positional information by the positioning function. "timingAdvance" means the calculation of the Timing Advance (TA) transmission timing adjustment value that UE1 applies when uplink transmitting. The result of UE1's TA calculation may be reflected in the UL transmission timing of UE1 in the Non-Terrestrial Network (NTN), for example, or the result of said TA calculation may be reported from UE1 to RAN node 2.

[0060] As described with reference to Figures 2 to 4B, the operation of UE1 and RAN node 2 allows RAN node 2 to control whether UE1 is permitted to take action based on predictions or decisions made by UE-based AI / ML. For example, if an action by UE1 triggered by AI inference may result in interaction with the network or affect network performance, it may not be appropriate for UE1 to freely perform such an action. Alternatively, it may not be appropriate for UE1 to freely perform an action triggered by AI inference instead of being triggered by rules (or criteria or formulas) defined in the 3GPP specification. The operation of UE1 and RAN node 2 described with reference to Figures 2 to 4B can contribute to solving these problems.

[0061] <Second Embodiment> The configuration example of the wireless communication system according to this embodiment may be the same as the example shown in Figure 1. Figure 5 shows an example of the operation of UE1 and RAN node 2 for UE-based AI / ML, that is, for UE1 to perform AI inference on a machine learning model. In step 501, UE1 sends an AI interest indication to RAN node 2. In one example, UE1 may send an AI interest indication in the procedure of setting up, resuming, or re-establishing an RRC connection. RAN node 2 may request UE1 to send an AI interest indication.

[0062] The AI ​​interest indicator may show RAN node 2 that UE1 supports machine learning-based artificial intelligence. The AI ​​interest indicator may show RAN node 2 that UE1 is interested in running UE-based AI / ML. The AI ​​interest indicator may show RAN node 2 that UE1 is requesting permission to run UE-based AI / ML. The AI ​​interest indicator may be included in UE Capability Information.

[0063] Furthermore, or alternatively, the AI ​​interest indicator may indicate to RAN node 2 that UE1 wishes to perform a specific action based on the inference results of machine learning-based artificial intelligence. In other words, the AI ​​interest indicator may indicate to RAN node 2 that it is expected that UE1's execution of UE-based AI / ML will be beneficial for performance improvement.

[0064] The AI ​​interest indicator may indicate the category of AI / ML estimation that UE1 wants to perform. In other words, the AI ​​interest indicator may indicate the action or category of action that UE1 wishes to be allowed to apply UE-based AI / ML to. The AI ​​interest indicator may indicate the function or feature (e.g., mobility, power saving) that UE1 wishes to be allowed to apply UE-based AI / ML to. The AI ​​interest indicator may indicate the subfunction or subfeature (e.g., cell selection, handover, beam management) that UE1 wishes to be allowed to apply UE-based AI / ML to. The AI ​​interest indicator may indicate the procedure (e.g., RRC re-establishment, beam fault recovery (BFR)) that UE1 wishes to be allowed to apply UE-based AI / ML to. The AI ​​interest indicator may indicate the RRC configuration (e.g., at the level of information elements or fields in Abstract Syntax Notation One (ASN.1)) that UE1 wishes to be allowed to apply UE-based AI / ML to.

[0065] The AI ​​interest indicator may show the magnitude or level of the effect expected from performing AI / ML estimation by UE1.

[0066] Step 502 is the same as step 201 in Figure 2. RAN node 2 transmits control information to UE1. Details regarding the control information (e.g., transmission method, content) are the same as those described in the first embodiment, so a description of the control information is omitted here. RAN node 2 may decide whether or not to transmit control information (502) based on the AI ​​interest indication (501). RAN node 2 may also decide the content of the control information (502) based on the AI ​​interest indication (501).

[0067] Figure 6 shows an example of the format for an AI interest indication. The ai-ML-Assistance IE602 shown in Figure 6 corresponds to an AI interest indication. ai-ML-Assistance IE602 may be included in UEAssistanceInformation IE601. UEAssistanceInformation IE601 may be sent to RAN node 2 via RRC messages (e.g., RRC Setup Request, RRC Setup Complete, RRC Resume Request, RRC Resume Complete, RRC Re-establishment Request, RRC Re-establishment Complete). ai-ML-Assistance IE602 may also include ai-ML-InterestIndication IE603 or 604.

[0068] The ai-ML-InterestIndication IE603 indicates, for example, that UE1 is interested in running UE-based AI / ML. If UE1 wishes to allow the execution of UE-based AI / ML, UE1 may include the ai-ML-InterestIndication IE603 in the UEAssistanceInformation IE601.

[0069] On the other hand, ai-ML-InterestIndication IE604 indicates the category of AI / ML estimation that UE1 wants to perform. In other words, ai-ML-InterestIndication IE604 indicates the action or category of action that UE1 wants to be allowed to apply UE-based AI / ML to. The category may include one or more of the following, for example: “mobilityConnected”, “mobilityIdleInactive”, “beamManagement”, “energySaving”, “rrmMeasurement”, “csiFeedback”, and “positioningAccuracy”. What each of these means (indicates) may be as described above.

[0070] Figure 7 is a flowchart illustrating an example of the operation of RAN node 2. In step 701, RAN node 2 receives an AI interest indication from UE1. In step 702, RAN node 2 determines whether UE1 is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence. In step 703, RAN node 2 sends control information to UE1. The control information in step 703 indicates whether UE1 is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence. Details regarding the control information (e.g., transmission method, content) are the same as those described in the first embodiment, so a description of the control information is omitted here.

[0071] According to the operation of UE1 and RAN node 2 as described with reference to Figures 5 to 7, RAN node 2 can receive an AI interest indication (steps 501, 701) from UE1 before sending control information (steps 502, 703) for authorizing the execution of UE-based AI / ML. Therefore, for example, RAN node 2 can determine whether or not to authorize UE1 to execute UE-based AI / ML based on the AI ​​interest indication received from UE1. Alternatively, RAN node 2 can determine one or more actions (or categories of actions) that UE1 is permitted to apply UE-based AI / ML to based on the AI ​​interest indication received from UE1.

[0072] Next, the following describes configuration examples of UE1 and RAN node 2 according to the above-described multiple embodiments. Figure 8 is a block diagram showing a configuration example of UE1. The Radio Frequency (RF) transceiver 801 performs analog RF signal processing to communicate with the RAN node. The RF transceiver 801 may include multiple transceivers. The analog RF signal processing performed by the RF transceiver 801 includes frequency up-conversion, frequency down-conversion, and amplification. The RF transceiver 801 is coupled with the antenna array 802 and the baseband processor 803. The RF transceiver 801 receives modulation symbol data (or OFDM symbol data) from the baseband processor 803, generates a transmit RF signal, and supplies the transmit RF signal to the antenna array 802. The RF transceiver 801 also generates a baseband receive signal based on the received RF signal received by the antenna array 802 and supplies it to the baseband processor 803. The RF transceiver 801 may include an analog beamformer circuit for beamforming. The analog beamformer circuit includes, for example, multiple phase shifters and multiple power amplifiers.

[0073] The baseband processor 803 performs digital baseband signal processing (data plane processing) and control plane processing for wireless communication. Digital baseband signal processing includes (a) data compression / decompression, (b) data segmentation / concatenation, (c) generation / decomposition of transmission format (transmission frame), (d) transmission path coding / decoding, (e) modulation (symbol mapping) / demodulation, and (f) generation of OFDM symbol data (baseband OFDM signal) by Inverse Fast Fourier Transform (IFFT). Control plane processing, on the other hand, includes communication management at Layer 1 (e.g., transmit power control), Layer 2 (e.g., radio resource management and hybrid automatic repeat request (HARQ) processing), and Layer 3 (e.g., signaling related to attach, mobility, and call management).

[0074] For example, the digital baseband signal processing by the baseband processor 803 may include signal processing for the Service Data Adaptation Protocol (SDAP) layer, Packet Data Convergence Protocol (PDCP) layer, Radio Link Control (RLC) layer, Medium Access Control (MAC) layer, and Physical (PHY) layer. Furthermore, the control plane processing by the baseband processor 803 may include processing for the Non-Access Stratum (NAS) protocol, Radio Resource Control (RRC) protocol, MAC Control Elements (CEs), and Downlink Control Information (DCIs).

[0075] The baseband processor 803 may perform Multiple Input Multiple Output (MIMO) encoding and precoding for beamforming.

[0076] The baseband processor 803 may include a modem processor (e.g., Digital Signal Processor (DSP)) for performing digital baseband signal processing and a protocol stack processor (e.g., Central Processing Unit (CPU) or Micro Processing Unit (MPU)) for performing control plane processing. In this case, the protocol stack processor for performing control plane processing may be shared with the application processor 804 described later.

[0077] The application processor 804 is also called a CPU, MPU, microprocessor, or processor core. The application processor 804 may include multiple processors (multiple processor cores). The application processor 804 implements various functions of the UE1 by executing system software programs (Operating System (OS)) and various application programs (e.g., calling applications, web browsers, mail clients, camera operation applications, music playback applications) read from memory 806 or memory not shown.

[0078] In some implementations, the baseband processor 803 and the application processor 804 may be integrated on a single chip, as shown by the dashed line (805) in Figure 8. In other words, the baseband processor 803 and the application processor 804 may be implemented as a single System on Chip (SoC) device 805. An SoC device is sometimes called a System Large Scale Integration (LSI) or chipset.

[0079] Memory 806 is volatile memory, non-volatile memory, or a combination thereof. Memory 806 may include multiple physically independent memory devices. Volatile memory is, for example, Static Random Access Memory (SRAM) or Dynamic RAM (DRAM), or a combination thereof. Non-volatile memory is Mask Read Only Memory (MROM), Electrically Erasable Programmable ROM (EEPROM), flash memory, or hard disk drive, or any combination thereof. For example, memory 806 may include an external memory device accessible from the baseband processor 803, the application processor 804, and the SoC 805. Memory 806 may also include an internal memory device integrated within the baseband processor 803, the application processor 804, or the SoC 805. Furthermore, memory 806 may include memory within a Universal Integrated Circuit Card (UICC).

[0080] The memory 806 may store one or more software modules (computer programs) 807 containing instruction sets and data for performing the processing by the UE1 as described in the above embodiments. In some implementations, the baseband processor 803 or application processor 804 may be configured to read and execute the software modules 807 from the memory 806 to perform the processing of the UE1 as described in the above embodiments with reference to the drawings.

[0081] Furthermore, the control plane processing and operation performed by the UE1 described in the above embodiment can be realized by other elements other than the RF transceiver 801 and antenna array 802, namely at least one of the baseband processor 803 and application processor 804 and the memory 806 storing the software module 807.

[0082] Figure 9 is a block diagram showing an example configuration of RAN node 2 according to the embodiment described above. Referring to Figure 9, RAN node 2 includes a Radio Frequency transceiver 901, a network interface 903, a processor 904, and a memory 905. The RF transceiver 901 performs analog RF signal processing to communicate with UEs, including UE1. The RF transceiver 901 may include multiple transceivers. The RF transceiver 901 is coupled with an antenna array 902 and a processor 904. The RF transceiver 901 receives modulation symbol data from the processor 904, generates a transmit RF signal, and supplies the transmit RF signal to the antenna array 902. The RF transceiver 901 also generates a baseband receive signal based on the received RF signal received by the antenna array 902 and supplies it to the processor 904. The RF transceiver 901 may include an analog beamformer circuit for beamforming. The analog beamformer circuit includes, for example, multiple phase shifters and multiple power amplifiers.

[0083] The network interface 903 is used to communicate with network nodes (e.g., other RAN nodes, as well as control and forwarding nodes of the core network). The network interface 903 may include, for example, a network interface card (NIC) compliant with the IEEE 802.3 series.

[0084] Processor 904 performs digital baseband signal processing (data plane processing) and control plane processing for wireless communication. Processor 904 may include multiple processors. For example, processor 904 may include a modem processor (e.g., Digital Signal Processor (DSP)) for digital baseband signal processing and a protocol stack processor (e.g., Central Processing Unit (CPU) or Micro Processing Unit (MPU)) for control plane processing. Processor 904 may also include a digital beamformer module for beamforming. The digital beamformer module may include a Multiple Input Multiple Output (MIMO) encoder and precoder.

[0085] Memory 905 is comprised of a combination of volatile and non-volatile memory. Volatile memory may be, for example, Static Random Access Memory (SRAM) or Dynamic RAM (DRAM), or a combination thereof. Non-volatile memory may be Mask Read Only Memory (MROM), Electrically Erasable Programmable ROM (EEPROM), flash memory, or a hard disk drive, or any combination thereof. Memory 905 may also include storage located away from the processor 904. In this case, the processor 904 may access memory 905 via a network interface 903 or an I / O interface not shown.

[0086] The memory 905 may store one or more software modules (computer programs) 906 containing instruction sets and data for processing by the RAN node 2 as described in the above embodiments. In some implementations, the processor 904 may be configured to read the software modules 906 from the memory 905 and execute them to perform the processing of the RAN node 2 as described in the above embodiments.

[0087] Furthermore, if RAN node 2 is a CU (e.g., gNB-CU) or CU-CP (e.g., gNB-CU-CP), RAN node 2 does not need to include RF transceiver 901 (and antenna array 902).

[0088] As illustrated with reference to Figures 8 and 9, each of the processors in the UE1 and RAN node 2 according to the above embodiment can execute one or more programs, each containing a set of instructions for causing a computer to perform the algorithms described with reference to the drawings. The program, when loaded into a computer, contains a set of instructions (or software code) for causing the computer to perform one or more of the functions described in the embodiment. The program may be stored on a non-temporary computer-readable medium or a physical storage medium. Examples, but not limited to, include random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD) or other memory technologies, CD-ROM, digital versatile disk (DVD), Blu-ray® disc or other optical disc storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices. The program may be transmitted over a temporary computer-readable medium or a communication medium. Examples, but not limited to, include temporary computer-readable medium or a communication medium containing electrical, optical, acoustic or other forms of propagating signals.

[0089] <Other Embodiments> The embodiments described above may be applied to Non-Terrestrial Networks (NTNs). The embodiments described above may also be applied to various new applications such as vehicle-to-everything (V2X), high-speed trains (HSTs), unmanned aerial vehicles (UAVs), uncrewed aerial vehicles (UAVs), and urban air mobility (UAMs).

[0090] The above-described embodiment may also be applied in a secondary cell group (SCG) in Dual Connectivity (e.g., MR-DC). In this case, RAN node 2 may be a secondary node (SN). In this case, SN may send and receive RRC messages with UE1 directly in the SCG using a signaling radio bearer (SRB3), or via a master node (MN). Alternatively, RAN node 2 may be a master node (MN) and may be responsible for forwarding RRC messages between SN and UE1.

[0091] In the above-described embodiment, RAN node 2 may be implemented in a C-RAN configuration. For example, RAN node 2 may include a CU (e.g., gNB-CU) and a DU (e.g., gNB-DU). The CU may decide whether or not to allow UE1 to perform a specific action based on the inference results of machine learning-based artificial intelligence, and may transmit control information indicating the result of that decision to UE1 (via the DU). Alternatively, the DU may decide whether or not to allow the wireless terminal to perform a specific action based on the inference results of machine learning-based artificial intelligence, and may transmit control information indicating the result of that decision to the CU, which may then transmit to UE1. Depending on the target, type, and content of the action permitted to UE1, the control by the CU and DU described above may be selected or combined.

[0092] Furthermore, the embodiments described above are merely examples of how the technical concept obtained by the present inventor can be applied. In other words, the technical concept is not limited to the embodiments described above, and various modifications are certainly possible.

[0093] For example, some or all of the above embodiments may also be described as follows, but are not limited to the following.

[0094] (Note 1) A wireless terminal, At least one memory, The system comprises at least one processor coupled to the at least one memory, The aforementioned at least one processor is The wireless terminal receives control information from the network indicating whether or not it is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence. To make predictions or decisions using trained machine learning models, and If the control information permits the execution of the specific action based on the inference result, then the specific action triggered by the prediction or decision is performed. A wireless terminal configured in such a way. (Note 2) The control information indicates whether, in lieu of or in addition to performing the specific action based on a predefined rule, the execution of the specific action based on the inference result is permitted. The wireless terminal described in Appendix 1. (Note 3) The aforementioned predefined rules are rules specified in the Third Generation Partnership Project (3GPP) specifications or rules pre-configured by the network. The wireless terminal described in Appendix 2. (Note 4) The control information individually indicates whether each of the multiple actions, including the specific action, is permitted. A wireless terminal as described in any one of the following appendices 1 to 3. (Note 5) The control information collectively indicates permission to use the machine learning-based artificial intelligence for multiple actions, including the specific action. A wireless terminal as described in any one of the following appendices 1 to 3. (Note 6) The control information indicates the conditions under which the execution of a specific action based on the inference result is permitted. The at least one processor is configured to perform the specific action triggered by the prediction or decision only if the conditions are met. A wireless terminal as described in any one of the items 1 to 5 in the appendix. (Note 7) The above conditions impose constraints relating to at least one of frequency band, location, and time. The wireless terminal described in Appendix 6. (Note 8) The control information is transmitted by the network when the execution of the specific action based on the inference result is permitted. A wireless terminal as described in any one of the items 1 to 7 of the appendix. (Note 9) The at least one processor is configured to inform the network that the wireless terminal supports machine learning-based artificial intelligence before it receives the control information. A wireless terminal as described in any one of the items 1 to 8 of the appendix. (Note 10) The at least one processor is configured to inform the network that the wireless terminal wishes to perform the specific action based on the inference results of machine learning-based artificial intelligence, before receiving the control information. A wireless terminal as described in any one of the items 1 to 9 in the appendix. (Note 11) The aforementioned specific actions relate to beam management or mobility. A wireless terminal as described in any one of the appendices 1 to 10. (Note 12) The aforementioned specific actions include at least one of the following: cell reselection, sending measurement reports, performing conditional mobility, and selecting downlink beams. A wireless terminal as described in any one of the items 1 to 11 of the appendices. (Note 13) The aforementioned specific action includes transmitting supporting information to the network that indicates the result of a prediction or decision using the trained machine learning model. A wireless terminal as described in any one of the appendices 1 to 12. (Note 14) The aforementioned support information is used for training a second machine learning model for optimizing wireless access networks or for performing inference on the second trained machine learning model for optimizing wireless access networks. The wireless terminal described in Appendix 13. (Note 15) A method performed by a wireless terminal, Receiving control information from the network indicating whether or not the wireless terminal is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence, Making predictions or decisions using trained machine learning models, and If the control information permits the execution of the specific action based on the inference result, then perform the specific action triggered by the prediction or decision. A method for providing this. (Note 16) A program for causing a computer to perform a method for wireless terminals, The aforementioned method, Receiving control information from the network indicating whether or not the wireless terminal is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence, Making predictions or decisions using trained machine learning models, and If the control information permits the execution of the specific action based on the inference result, then perform the specific action triggered by the prediction or decision. A program that includes the following features. (Note 17) At least one memory, At least one processor coupled to the at least one memory and configured to transmit control information to the wireless terminal indicating whether or not the wireless terminal is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence, Equipped with, If the control information permits the execution of the specific action based on the inference result, the control information causes the wireless terminal to perform the specific action triggered by a prediction or decision using a trained machine learning model. Wireless access network node. (Note 18) The control information indicates whether, in lieu of or in addition to performing the specific action based on a predefined rule, the execution of the specific action based on the inference result is permitted. The wireless access network node described in Appendix 17. (Note 19) The aforementioned predefined rules are rules specified in the Third Generation Partnership Project (3GPP) specifications or rules pre-configured by the network. The wireless access network node described in Appendix 18. (Note 20) The control information individually indicates whether each of the multiple actions, including the specific action, is permitted. A wireless access network node as described in any one of the items 17-19 of the appendix. (Note 21) The control information collectively indicates permission to use the machine learning-based artificial intelligence for multiple actions, including the specific action. A wireless access network node as described in any one of the items 17-19 of the appendix. (Note 22) The control information indicates the conditions under which the execution of a specific action based on the inference result is permitted. The control information causes the wireless terminal to perform the specific action triggered by the prediction or decision only if the conditions are met. A wireless access network node as described in any one of the items 17 to 21 of the appendices. (Note 23) The above conditions impose constraints relating to at least one of frequency band, location, and time. The wireless access network node described in Appendix 22. (Note 24) The at least one processor is configured to transmit the control information when the execution of the specific action based on the inference result is permitted. A wireless access network node as described in any one of the items 17 to 23 of the appendix. (Note 25) The at least one processor is configured to be notified by the wireless terminal that the wireless terminal supports machine learning-based artificial intelligence before transmitting the control information. A wireless access network node as described in any one of the items 17-24 of the appendix. (Note 26) The at least one processor is configured to be notified by the wireless terminal that it desires to perform the specific action based on the inference results of machine learning-based artificial intelligence, before transmitting the control information. A wireless access network node as described in any one of the items 17 to 25 of the appendix. (Note 27) The aforementioned specific actions relate to beam management or mobility. A wireless access network node as described in any one of the items 17 to 26 of the appendix. (Note 28) The aforementioned specific actions include at least one of the following: cell reselection, sending measurement reports, performing conditional mobility, and selecting downlink beams. A wireless access network node as described in any one of the items 17 to 27 of the appendix. (Note 29) The aforementioned specific action includes transmitting supporting information indicating the results of a prediction or decision using the trained machine learning model to the wireless access network node. A wireless access network node as described in any one of the items 17 to 28 of the appendix. (Note 30) The aforementioned support information is used for training a second machine learning model for optimizing wireless access networks or for performing inference on the second trained machine learning model for optimizing wireless access networks. The wireless access network node described in Appendix 29. (Note 31) The system includes transmitting control information to the wireless terminal indicating whether or not it is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence. If the control information permits the execution of the specific action based on the inference result, the control information causes the wireless terminal to perform the specific action triggered by a prediction or decision using a trained machine learning model. A method performed by a wireless access network node. (Note 32) A program for causing a computer to perform a method for a wireless access network node, The method comprises transmitting control information to the wireless terminal indicating whether or not the wireless terminal is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence. If the control information permits the execution of the specific action based on the inference result, the control information causes the wireless terminal to perform the specific action triggered by a prediction or decision using a trained machine learning model. program.

[0095] This application claims priority based on Japanese Patent Application No. 2021-182104, filed on 8 November 2021, and incorporates all of its disclosures herein. [Explanation of Symbols]

[0096] 1 UE 2 RANNodes 803 Baseband Processor 804 Application Processor 806 memory 807 Modules 904 Processor 905 memory 906 modules

Claims

1. A wireless terminal, A means for receiving control information from a network indicating whether or not the wireless terminal is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence, A means for making predictions or decisions using a trained machine learning model, If the control information permits the execution of the specific action based on the inference result, means for performing the specific action triggered by the prediction or decision, Equipped with, The aforementioned specific action includes transmitting supporting information to the network that indicates the result of a prediction or decision using the trained machine learning model. Wireless terminal.

2. The control information collectively indicates permission to use the machine learning-based artificial intelligence for multiple actions, including the specific action. The wireless terminal according to claim 1.

3. The control information is transmitted by the network when the execution of the specific action based on the inference result is permitted. The wireless terminal according to claim 1.

4. The aforementioned support information is used for training a second machine learning model for optimizing wireless access networks or for performing inference on the second trained machine learning model for optimizing wireless access networks. The wireless terminal according to claim 1.

5. A method performed by a wireless terminal, Receiving control information from the network indicating whether or not the wireless terminal is permitted to perform a specific action based on the inference results of machine learning-based artificial intelligence, Making predictions or decisions using trained machine learning models, and If the control information permits the execution of the specific action based on the inference result, the specific action triggered by the prediction or decision is performed, The aforementioned specific action includes transmitting supporting information to the network that indicates the result of a prediction or decision using the trained machine learning model. method.

6. The control information collectively indicates permission to use the machine learning-based artificial intelligence for multiple actions, including the specific action. The method according to claim 5.

7. The control information is transmitted by the network when the execution of the specific action based on the inference result is permitted. The method according to claim 5.

8. The aforementioned support information is used for training a second machine learning model for optimizing wireless access networks or for performing inference on the second trained machine learning model for optimizing wireless access networks. The method according to claim 5.

Images