Method and device for utilizing artificial intelligence and machine learning in wireless communication system

Abstract

According to an embodiment, a method performed by a UE in a communication system may include the steps of: receiving, from a base station, a radio resource control (RRC) message including configuration information related to report configuration of the UE for model training of the UE; and when it is determined, on the basis of the reception of the configuration information, that the UE prefers suspension of the configuration for model training, transmitting, to the base station, a UE assistance information message related to the suspension of the configuration for model training.

Description

Method and device for utilizing artificial intelligence and machine learning in wireless communication systems

[0001] The present disclosure relates to the operation of a terminal and a base station in a wireless communication system. More specifically, the present disclosure relates to a method and apparatus utilizing artificial intelligence and machine learning.

[0002] 5G mobile communication technology defines a wide frequency band to enable fast transmission speeds and new services, and can be implemented not only in frequency bands below 6 GHz ('Sub 6 GHz'), such as 3.5 gigahertz (3.5 GHz), but also in ultra-high frequency bands called millimeter waves (mmWave), such as 28 GHz and 39 GHz ('Above 6 GHz'). In addition, for 6G mobile communication technology, which is referred to as a system beyond 5G, implementation in the terahertz band (e.g., the 3 terahertz (3 THz) band at 95 GHz) is being considered to achieve transmission speeds 50 times faster and ultra-low latency reduced to one-tenth compared to 5G mobile communication technology.

[0003] In the early stages of 5G mobile communication technology, aiming to satisfy service support and performance requirements for enhanced Mobile BroadBand (eMBB), Ultra-Reliable Low-Latency Communications (URLLC), and massive Machine-Type Communications (mMTC), technologies such as beamforming and Massive MIMO to mitigate path loss and increase transmission distance in ultra-high frequency bands, support for various numerologies (such as the operation of multiple subcarrier spacings) and dynamic operation of slot formats for the efficient utilization of ultra-high frequency resources, initial access techniques to support multi-beam transmission and broadband, definition and operation of Band-Width Parts (BWP), Low Density Parity Check (LDPC) codes for high-volume data transmission, new channel coding methods such as Polar Codes for the reliable transmission of control information, and L2 pre-processing (L2 Standardization has been carried out for pre-processing, network slicing which provides a dedicated network specialized for specific services, and other methods.

[0004] Currently, discussions are underway to improve and enhance the performance of the initial 5G mobile communication technology, taking into account the services that the 5G mobile communication technology was intended to support. Additionally, physical layer standardization is in progress for technologies such as V2X (Vehicle-to-Everything), which helps autonomous vehicles make driving decisions and enhance user convenience based on their own location and status information transmitted by the vehicle; NR-U (New Radio Unlicensed), which aims for system operation in unlicensed bands that meets various regulatory requirements; NR terminal low power consumption technology (UE Power Saving); Non-Terrestrial Network (NTN), which is direct terminal-satellite communication for securing coverage in areas where communication with the terrestrial network is impossible; and positioning.

[0005] In addition, standardization is underway in the field of wireless interface architecture / protocols for technologies such as the Industrial Internet of Things (IIoT) to support new services through linkage and convergence with other industries, Integrated Access and Backhaul (IAB) which provides nodes to expand network service areas by integrating wireless backhaul links and access links, Mobility Enhancement including Conditional Handover and Dual Active Protocol Stack (DAPS) Handover, and 2-step Random Access (2-step RACH for NR) which simplifies random access procedures. Standardization is also underway in the field of system architecture / services for 5G baseline architectures (e.g., Service based Architecture, Service based Interface) to incorporate Network Functions Virtualization (NFV) and Software-Defined Networking (SDN) technologies, and Mobile Edge Computing (MEC), which provides services based on the location of the terminal.

[0006] When such 5G mobile communication systems are commercialized, connected devices, which are increasing explosively, will be connected to communication networks. Accordingly, it is expected that there will be a need to enhance the functionality and performance of 5G mobile communication systems and to integrate the operation of connected devices. To this end, new research is planned to be conducted on 5G performance improvement and complexity reduction, support for AI services, support for metaverse services, and drone communication using eXtended Reality (XR), Artificial Intelligence (AI), and Machine Learning (ML) to efficiently support Augmented Reality (AR), Virtual Reality (VR), and Mixed Reality (MR).

[0007] Furthermore, the advancement of these 5G mobile communication systems encompasses multi-antenna transmission technologies such as new waveforms to guarantee coverage in the terahertz band of 6G mobile communication technology, Full Dimensional MIMO (FD-MIMO), array antennas, and large-scale antennas; metamaterial-based lenses and antennas to improve terahertz band signal coverage; high-dimensional spatial multiplexing technology using OAM (Orbital Angular Momentum); and Reconfigurable Intelligent Surface (RIS) technology; as well as Full Duplex technology for enhancing frequency efficiency and system networks in 6G mobile communication technology; AI-based communication technologies that realize system optimization by utilizing satellites and AI from the design stage and internalizing end-to-end AI support functions; and the realization of services of complexity exceeding the limits of terminal computing capabilities by utilizing ultra-high-performance communication and computing resources. It could serve as a foundation for the development of next-generation distributed computing technologies.

[0008] The present disclosure proposes a method for receiving a report from a terminal for model learning in a communication system, and transmitting terminal auxiliary information in consideration of setting and / or stopping the setting for model learning.

[0009] The technical problems to be solved in the embodiments of the present disclosure are not limited to those mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art to which the present disclosure belongs from the description below.

[0010] The present disclosure, for solving these problems, proposes a method performed by a terminal of a communication system. More specifically, the method is characterized by comprising the steps of: receiving a Radio Resource Control (RRC) message from a base station containing configuration information related to a terminal's reporting settings for model learning; and, if the terminal determines, based on the reception of the configuration information, that it prefers to discontinue the settings for model learning, transmitting a Terminal Assistance Information (UE Assistance Information) message related to the discontinuation of the settings for model learning to the base station.

[0011] The present disclosure, for solving these problems, proposes a method performed by a base station of a communication system. More specifically, the method is characterized by comprising the steps of: transmitting a Radio Resource Control (RRC) message to a terminal that includes configuration information related to the terminal's reporting settings for model learning; and, if it is determined that the terminal prefers to discontinue the settings for model learning based on the reception of the configuration information, receiving a Terminal Assistance Information (UE Assistance Information) message from the terminal related to the discontinuation of the settings for model learning.

[0012] The present disclosure for solving these problems proposes a terminal in a communication system. The terminal comprises: at least one transceiver; at least one processor connected to the at least one transceiver so as to be able to communicate with the at least one transceiver; and a memory connected to the at least one processor so as to be able to communicate with the at least one processor and capable of executing the at least one processor individually or in any combination thereof, wherein the terminal receives a Radio Resource Control (RRC) message from a base station containing configuration information related to the terminal's reporting settings for model learning, and, based on the reception of the configuration information, determines that the terminal prefers to stop the settings for model learning, and stores a command to transmit a terminal assistance information (UE Assistance Information) message related to the stoppage of the settings for model learning to the base station.

[0013] The present disclosure for solving these problems proposes a base station of a communication system. The base station comprises: at least one transceiver; at least one processor connected to communicate with the at least one transceiver; and a memory connected to communicate with the at least one processor and capable of executing the at least one processor individually or in any combination thereof, which stores a command to transmit a Radio Resource Control (RRC) message containing configuration information related to the terminal's reporting settings for model learning to the terminal, and, when it is determined based on the reception of the configuration information that the terminal prefers to stop the settings for model learning, to receive a Terminal Assistance Information (UE Assistance Information) message related to the stoppage of the settings for model learning from the terminal.

[0014] According to one embodiment of the present disclosure, a terminal can efficiently set up and perform model learning by receiving a report from the terminal for model learning and transmitting terminal auxiliary information in consideration of setting up and / or stopping the setting for model learning.

[0015] The effects obtainable in the present disclosure are not limited to those mentioned in the various embodiments, and other unmentioned effects will be clearly understood by those skilled in the art to which the present disclosure pertains from the description below.

[0016] FIG. 1a is a drawing illustrating the structure of a mobile communication system according to one embodiment of the present disclosure.

[0017] FIG. 1b is a diagram illustrating a wireless connection state transition in a mobile communication system according to one embodiment of the present disclosure.

[0018] FIG. 1c is a diagram illustrating an Artificial Intelligence (AI) / Machine Learning (ML) model for predicting beam measurements for beam management according to one embodiment of the present disclosure.

[0019] FIG. 1d is a diagram illustrating a procedure in which a terminal performs a prediction using a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model for beam management according to one embodiment of the present disclosure.

[0020] FIG. 1e is a diagram illustrating a procedure in which a network performs a prediction using an NW-side Artificial Intelligence (AI) / Machine Learning (ML) model for beam management according to one embodiment of the present disclosure.

[0021] FIG. 1f is the first diagram illustrating a procedure in which a terminal or terminal server learns a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model and performs inference using it, according to one embodiment of the present disclosure.

[0022] FIG. 1g is the first drawing illustrating a procedure for a terminal or terminal server to learn a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model according to one embodiment of the present disclosure.

[0023] FIG. 1h is a second figure illustrating a procedure for a terminal or terminal server to learn a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model according to one embodiment of the present disclosure.

[0024] FIG. 1i is a second figure illustrating a procedure in which a terminal or terminal server learns a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model and performs inference using it, according to one embodiment of the present disclosure.

[0025] FIG. 1j is the first diagram illustrating a procedure in which a terminal stops measurement and data collection for training a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model according to one embodiment of the present disclosure.

[0026] FIG. 1k is a second drawing illustrating a procedure in which a terminal stops measurement and data collection for training a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model according to one embodiment of the present disclosure.

[0027] FIG. 11 is a drawing illustrating the internal structure of a terminal according to one embodiment of the present disclosure.

[0028] FIG. 1m is a drawing illustrating the structure of a base station according to one embodiment of the present disclosure.

[0029] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the attached drawings.

[0030] In describing the embodiments, technical details that are well known in the technical field to which this disclosure belongs and are not directly related to this disclosure are omitted. This is intended to convey the essence of this disclosure more clearly without obscuring it by omitting unnecessary explanations.

[0031] For the same reason, some components in the attached drawings have been exaggerated, omitted, or schematically depicted. Additionally, the dimensions of each component do not entirely reflect their actual dimensions. Identical or corresponding components in each drawing have been assigned the same reference numbers.

[0032] The advantages and features of the present disclosure, and the methods for achieving them, will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present disclosure is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure is complete and to fully inform those skilled in the art of the scope of the disclosure, and the present disclosure is defined only by the scope of the claims. Throughout the specification, the same reference numerals refer to the same components. Furthermore, in describing the present disclosure, if it is determined that a detailed description of related functions or configurations might unnecessarily obscure the essence of the present disclosure, such detailed description is omitted. Additionally, the terms described below are defined considering their functions in the present disclosure, and these may vary depending on the intentions or conventions of the user or operator. Therefore, their definitions should be based on the content throughout the specification.

[0033] Hereinafter, the base station is an entity that performs resource allocation for terminals and may be at least one of a gNode B, eNode B, Node B, BS (Base Station), wireless access unit, base station controller, or a node on a network. The terminal may include a UE (User Equipment), MS (Mobile Station), cellular phone, smartphone, computer, or a multimedia system capable of performing communication functions. In this disclosure, the Downlink (DL) refers to the wireless transmission path of a signal transmitted by a base station to a terminal, and the Uplink (UL) refers to the wireless transmission path of a signal transmitted by a terminal to a base station. Furthermore, although an embodiment of this disclosure is described below using a 5G system as an example, the embodiment of this disclosure may be applied to other communication systems having similar technical backgrounds or channel types. For example, LTE or LTE-A mobile communication and mobile communication technologies developed after 5G may be included therein. In addition, the present disclosure may be applied to other communication systems with some modifications made at the discretion of a person with skilled technical knowledge, without significantly departing from the scope of the present disclosure. The contents of the present disclosure are applicable to FDD and TDD systems.

[0034] At this point, it will be understood that each block of the process flow diagrams and combinations of the flow diagrams can be executed by computer program instructions. Since these computer program instructions can be loaded into the processor of a general-purpose computer, a special-purpose computer, or other programmable data processing equipment, the instructions executed through the processor of the computer or other programmable data processing equipment create means to perform the functions described in the flow diagram block(s). Since these computer program instructions can also be stored in computer-available or computer-readable memory that can be directed toward the computer or other programmable data processing equipment to implement the function in a specific way, the instructions stored in computer-available or computer-readable memory can also produce a manufactured item containing instruction means to perform the function described in the flow diagram block(s). Since computer program instructions can be loaded onto a computer or other programmable data processing equipment, instructions that perform a series of operation steps on the computer or other programmable data processing equipment to create a process executed by the computer can also provide steps for executing the functions described in the flowchart block(s).

[0035] Additionally, each block may represent a module, segment, or part of code containing one or more executable instructions for executing a specific logical function(s). It should also be noted that in some alternative execution examples, the functions mentioned in the blocks may occur out of order. For example, two blocks described in succession may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order according to their corresponding functions.

[0036] In this embodiment, the term "part" refers to a software or hardware component such as an FPGA (Field Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit), and the "part" performs certain roles. However, the meaning of "part" is not limited to software or hardware. The "part" may be configured to reside in an addressable storage medium or may be configured to run one or more processors. Accordingly, as an example, the "part" includes components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables. The functions provided within the components and "parts" may be combined into a smaller number of components and "parts" or further separated into additional components and "parts." In addition, the components and '~parts' may be implemented to utilize one or more CPUs within the device or secure multimedia card. Also, in the embodiment, the '~part' may include one or more processors.

[0037] FIG. 1a is a drawing illustrating the structure of a mobile communication system according to one embodiment of the present disclosure.

[0038] Referring to FIG. 1a, the wireless access network of a mobile communication system (New Radio, NR) can be composed of a base station (next generation Node B, hereinafter gNB) (1a-10) and an AMF (1a-05, New Radio Core Network).

[0039] The user terminal (New Radio User Equipment, hereinafter NR UE or terminal) (1a-15) can connect to an external network through gNB (1a-10) and AMF (1a-05).

[0040] A mobile communication system can be a next-generation mobile communication system, and a base station can be a next-generation base station.

[0041] According to one embodiment, the gNB (1a-10) in FIG. 1a may correspond to the eNB (Evolved Node B) of an existing LTE system. The gNB (1a-10) is connected to an NR UE via a wireless channel and can provide superior service compared to the existing Node B (1a-20). In a next-generation mobile communication system according to one embodiment of the present disclosure, since all user traffic is serviced through a shared channel, a device for scheduling by collecting state information such as the buffer status, available transmission power status, and channel status of UEs may be required, and this can be performed by the gNB (1a-10). According to one embodiment, a single gNB (1a-10) can typically control multiple cells. To implement ultra-high-speed data transmission, it may have a bandwidth greater than the existing maximum bandwidth, and beamforming technology may be additionally incorporated by using Orthogonal Frequency Division Multiplexing (OFDM) as the wireless access technology. In addition, an Adaptive Modulation & Coding (AMC) scheme can be applied to determine the modulation scheme and channel coding rate according to the channel state of the terminal.

[0042] According to one embodiment, in FIG. 1a, the AMF (1a-05) can perform functions such as mobility support, bearer configuration, and quality of service (QoS) configuration. The AMF (1a-05) is a device responsible for various control functions as well as mobility management functions for terminals and can be connected to multiple base stations. In addition, the mobile communication system according to one embodiment of the present disclosure can be linked with an LTE system, and for example, the AMF (1a-05) can be connected to the MME (1a-25) through a network interface.

[0043] According to one embodiment, the MME (1a-25) can be connected to an existing base station eNB (1a-30). For example, in FIG. 1a, a terminal supporting LTE-NR Dual Connectivity can transmit and receive data while maintaining a connection to both the gNB (1a-10) and the eNB (1a-30) (1a-35).

[0044] FIG. 1b is a diagram illustrating a wireless connection state transition in a mobile communication system according to one embodiment of the present disclosure.

[0045] Referring to Fig. 1b, a mobile communication system may have three radio resource control (RRC) states or RRC modes.

[0046] Specifically, the connection mode (RRC_CONNECTED, 1b-05) may be a wireless connection state in which the terminal can transmit and receive data. The standby mode (RRC_IDLE, 1b-30) may correspond to a wireless connection state in which the terminal monitors whether paging is being transmitted to it. The above two modes are wireless connection states applicable to LTE systems, and the detailed description may be the same as that of LTE systems. A mobile communication system according to one embodiment of the present disclosure may be a next-generation mobile communication system.

[0047] According to one embodiment of the present disclosure, a new inactive (RRC_INACTIVE) radio access state (1b-15) may be defined in a mobile communication system. In the inactive radio access state (1b-15), a UE context may be maintained between the base station and the terminal, and RAN (radio access network) based paging may be supported. The features of the inactive radio access state (1b-15) may include at least one of the following:

[0048] - Cell re-selection mobility;

[0049] - CN - NR RAN connection (both C / U-planes (control plane / user plane)) has been established for UE;

[0050] - The UE AS(Access Stratum) context is stored in at least one gNB and the UE;

[0051] - Paging is initiated by NR RAN;

[0052] - RAN-based notification area is managed by NR RAN; or

[0053] - NR RAN knows the RAN-based notification area which the UE belongs to.

[0054] According to one embodiment of the present disclosure, a terminal in an INACTIVE wireless connection state (1b-15) may use a specific procedure and transition to a connection mode (1b-05) or a standby mode (1b-30). It may transition from the INACTIVE mode (1b-15) to the connection mode (1b-05) according to a Resume procedure, and may transition from the connection mode (1b-05) to the INACTIVE mode (1b-15) using a Release procedure including suspend setting information (1b-10). The procedure may be performed by transmitting and receiving one or more RRC messages between the terminal and the base station and may consist of one or more steps. Additionally, according to one embodiment, it may be possible to transition from the INACTIVE mode (1b-15) to the standby mode (1b-30) through a Release procedure after Resume (1b-20). The transition between the connection mode (1b-05) and the standby mode (1b-30) may follow LTE technology. Also, according to FIG. 1b, the transition between the above modes can be made through an establishment or release procedure (1b-25).

[0055] FIG. 1c is a diagram illustrating an Artificial Intelligence (AI) / Machine Learning (ML) model for predicting beam measurements for beam management according to one embodiment of the present disclosure.

[0056] In one embodiment of the present disclosure, beam management may be one use case in which an AI / ML (machine learning) model can be utilized. For example, a base station may utilize an AI / ML model for beam management of a downlink (DL) transmission beam.

[0057] In one embodiment of the present disclosure, positioning accuracy enhancements may be one use case in which an AI / ML model can be utilized.

[0058] As one embodiment of the present disclosure, CSI (Channel state information) feedback enhancement may be one use case in which an AI / ML model can be utilized.

[0059] In one embodiment of the present disclosure, a beam management use case may include sub-use cases of spatial prediction and temporal prediction.

[0060] In one embodiment of the present disclosure, a set of beams used as input to an AI / ML model for a beam management use case may be referred to as SET B. In one embodiment of the present disclosure, a set of beams derived as output to an AI / ML model for a beam management use case may be referred to as SET A.

[0061] In one embodiment of the present disclosure, during spatial prediction (1c-05) for beam management, the input of an AI / ML model is at a specific time point (e.g., t K One or more beams in ) (e.g., B i )(i=1, 2, …, N), measurement value for (e.g., Reference Signal Received Power (RSRP) and / or Reference Signal Received Quality (RSRQ) and / or Signal-to-Interference-plus-Noise Ratio (SINR)) (e.g., Meas(Bi ,t K )) (1c-15) It may be.

[0062] In one embodiment of the present disclosure, the measurement value for the beam may be one of the following:

[0063] - RSRP and / or RSRQ and / or SINR measured at Layer 1;

[0064] - RSRP and / or RSRQ and / or SINR measured / acquired at Layer 3;

[0065] - The filtered RSRP and / or RSRQ and / or SINR values ​​measured at Layer 1 (e.g., a (weighted) average value using measurements over a specified period); or

[0066] - Values ​​obtained by filtering RSRP and / or RSRQ and / or SINR measured / acquired at Layer 3 (e.g., (weighted) average values ​​using measurements over a specified period).

[0067] In one embodiment of the present disclosure, when performing spatial prediction (1c-05) for beam management, the following information may be considered / used as input to an AI / ML model:

[0068] - SET B related information (e.g., beam-specific ID);

[0069] - Measurement time information;

[0070] - L1-RSRP measurement based on Set B;

[0071] - L1-RSRP measurement based on Set B and assistance information;

[0072] - CIR (Channel Impulse Response) based on Set B; and / or

[0073] - L1-RSRP measurement based on Set B and the corresponding DL Tx and / or Rx beam ID.

[0074] In one embodiment of the present disclosure, during spatial prediction (1c-05) for beam management, the output of the AI / ML model is at a specific time point (e.g., t K One or more beams in ) (e.g., B i Predicted value for ) (e.g., RSRP and / or RSRQ and / or SINR) (e.g., P_Meas(B i ,t K )(i=N+1, N+2,...,N+M)) (1c-20) may be.

[0075] In one embodiment of the present disclosure, the predicted value for the beam may be one of the following:

[0076] - RSRP and / or RSRQ and / or SINR predicted to be measured at Layer 1;

[0077] - RSRP and / or RSRQ and / or SINR predicted to be measured / acquired at Layer 3;

[0078] - Predicted value (or filtered value) of the filtered RSRP and / or RSRQ and / or SINR measured (or predicted to be measured) at Layer 1 (e.g., a (weighted) average value using measured / predicted values ​​over a specified period); or

[0079] - Predicted value (or filtered value) of the filtered RSRP and / or RSRQ and / or SINR measured / acquired (or predicted to be measured / acquired) at Layer 3 (e.g., a (weighted) average value using measured / predicted values ​​over a specified period).

[0080] In one embodiment of the present disclosure, when performing spatial prediction (1c-05) for beam management, the following information may be considered / used as the output of an AI / ML model:

[0081] - SET A related information (e.g., ID per beam);

[0082] - One beam predicted to have the highest (best) measurement value (e.g., Top-1 beam);

[0083] - N (≥1) beams predicted to have the highest (best) measurements (e.g., Top-N beam); and / or

[0084] - Probability that each beam in SET A is a Top-1 or Top-N beam.

[0085] In one embodiment of the present disclosure, when performing a temporal prediction (1c-10) for beam management, the input to an AI / ML model is one or more (past) time points (e.g., t i A single beam at (i=1, 2,…,N)) (e.g., B) K ), measured values ​​for (e.g., RSRP and / or RSRQ and / or SINR) (e.g., Meas(B K ,t i ))(1c-25) may be. In one embodiment of the present disclosure, the measurement value for the beam may be one of the following:

[0086] - RSRP and / or RSRQ and / or SINR measured at Layer 1;

[0087] - RSRP and / or RSRQ and / or SINR measured / acquired at Layer 3;

[0088] - The filtered RSRP and / or RSRQ and / or SINR values ​​measured at Layer 1 (e.g., a (weighted) average value using measurements over a specified period); or

[0089] - Values ​​obtained by filtering RSRP and / or RSRQ and / or SINR measured / acquired at Layer 3 (e.g., (weighted) average values ​​using measurements over a specified period).

[0090] In one embodiment of the present disclosure, when performing temporal prediction (1c-10) for beam management, the following information may be considered / used as input to an AI / ML model:

[0091] - SET B related information (e.g., beam-specific ID);

[0092] - Measurement time information;

[0093] - L1-RSRP measurement based on Set B;

[0094] - L1-RSRP measurement based on Set B and assistance information;

[0095] CIR (Channel Impulse Response) based on Set B; and / or

[0096] - L1-RSRP measurement based on Set B and the corresponding DL Tx and / or Rx beam ID.

[0097] In one embodiment of the present disclosure, when performing a temporal prediction (1c-10) for beam management, the output of an AI / ML model is at one or more (future) points in time (e.g., t i A single beam in (i=N+1, N+2,…,N+M)) (e.g. B K Predicted value for ) (e.g., RSRP and / or RSRQ and / or SINR) (e.g., P_Meas(B K ,t i ))(1c-30) may be.

[0098] In one embodiment of the present disclosure, the predicted value for the beam may be one of the following:

[0099] - RSRP and / or RSRQ and / or SINR predicted to be measured at Layer 1;

[0100] - RSRP and / or RSRQ and / or SINR predicted to be measured / acquired at Layer 3;

[0101] - Predicted value (or filtered value) of the filtered RSRP and / or RSRQ and / or SINR measured (or predicted to be measured) at Layer 1 (e.g., a (weighted) average value using measured / predicted values ​​over a specified period); or

[0102] - Predicted value (or filtered value) of the filtered RSRP and / or RSRQ and / or SINR measured / acquired (or predicted to be measured / acquired) at Layer 3 (e.g., a (weighted) average value using measured / predicted values ​​over a specified period).

[0103] In one embodiment of the present disclosure, when performing temporal prediction (1c-10) for beam management, the following information may be considered / used as the output of an AI / ML model:

[0104] - SET A related information (e.g., beam-specific ID);

[0105] - Prediction time point information;

[0106] - The point in time when the measurement is predicted to be highest (best); and / or

[0107] - N (≥1) time points where the measurement value is predicted to be highest (best).

[0108] In one embodiment of the present disclosure, an AI / ML model for beam management may be an AI / ML model that simultaneously performs the aforementioned spatial prediction and temporal prediction. In the case of an AI / ML model that simultaneously performs spatial prediction and temporal prediction, the input of the AI / ML model may be a measurement value at one or more time points for each beam for one or more beams. Refer to the foregoing for the input information in this regard. Additionally, the output of the AI / ML model may be a prediction value at one or more time points for each beam for one or more beams. Refer to the foregoing for the output information in this regard.

[0109] In one embodiment of the present disclosure, a terminal (UE) may run an AI / ML model to derive predicted values ​​and related information. The AI / ML model run by the terminal may be referred to as a terminal-side AI / ML model or a UE-side AI / ML model. The terminal may transmit the derived information and related information to a base station. For example, the base station may utilize the received information for downlink transmission beam management, location accuracy improvement, and CSI feedback improvement.

[0110] In one embodiment of the present disclosure, a network (NW) (base station or LMF (e.g., location management function)) may drive an AI / ML model to derive predicted values ​​and related information. The AI / ML model driven by the network may be referred to as a network-side AI / ML model or an NW-side AI / ML model. For example, a base station may perform predictions based on measurement information received from a terminal and utilize them for downlink transmission beam management, improving location accuracy, and improving CSI feedback.

[0111] FIG. 1d is a diagram illustrating a procedure in which a terminal performs a prediction using a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model for beam management according to one embodiment of the present disclosure.

[0112] In step 1d-05, a terminal (e.g., terminal 1) (1d-10) may transmit the terminal's capability information to a connected base station (e.g., base station 1 or network 1) (1d-15). The terminal capability information may be transmitted via a UECapabilityInformation message. To receive this, the base station may first request the terminal to transmit the terminal's capability information. The request for the terminal capability information may be transmitted via a UECapabilityEnquiry message. The capability information transmitted by the terminal may include whether the terminal supports AI / ML-related functions (e.g., UE-side model-related functions / operations) (by AI / ML functionality, by use case, or by sub-use case).

[0113] In step 1d-25, the base station (1d-15) may provide the terminal (1d-10) with a measurement / reporting configuration (e.g., training configuration) to collect data necessary for UE-side model training. Prior to this, the base station (1d-15) may receive configuration information regarding model training or a configuration request regarding model training from the terminal (1d-10) or a training object (1d-20).

[0114] In one embodiment of the present disclosure, the learning object (1d-20) may be a terminal or a terminal server. The terminal server may be connected to the terminal via a 3GPP network (e.g., a server within the 3GPP network) or connected to the terminal via an external network (e.g., Wi-Fi) (e.g., a server outside the 3GPP network).

[0115] In step 1d-30, the terminal (1d-10) may perform measurements to generate data necessary for training the UE-side model, and then report the measurement results (e.g., training report) to the base station (1d-15) or the training object (1d-20). If the terminal reports the measurement results to the base station, the base station may reprocess the received report and deliver it to the training object.

[0116] In step 1d-35, the learning object (1d-20) can learn the UE-side model using the received report.

[0117] In step 1d-50, the terminal (e.g., terminal 2) (1d-40) can receive the trained model from the training object (1d-20). For example, the trained model can be delivered to the terminal via a base station (e.g., base station 2) (1d-45).

[0118] Before receiving the model, terminal 2 (1d-40) may transmit the terminal's capability information to the connected base station 2 (1d-45). The terminal capability information may be transmitted via a UECapabilityInformation message. To receive this, the base station may first request the terminal to transmit the terminal's capability information. The request for the terminal capability information may be transmitted via a UECapabilityEnquiry message. The capability information transmitted by the terminal may include whether the terminal supports AI / ML related functions (e.g., UE-side model related functions / operations) (by AI / ML functionality, by use case, or by sub-use case).

[0119] In step 1d-55, terminal 2 (1d-40) can perform inference or prediction using the received model. For example, in the case of beam management, the terminal can perform spatial or temporal beam prediction. This step can be performed after the terminal receives inference-related settings from the base station.

[0120] In step 1d-60, terminal 2 (1d-40) can report the result of the prediction or inference to base station 2 (1d-45).

[0121] In step 1d-65, the base station can perform downlink beam management for the terminal based on the prediction / inference results received from the terminal and select an appropriate beam to service the terminal.

[0122] FIG. 1e is a diagram illustrating a procedure in which a network performs a prediction using an NW-side Artificial Intelligence (AI) / Machine Learning (ML) model for beam management according to one embodiment of the present disclosure.

[0123] In step 1e-05, a terminal (e.g., terminal 1) (1e-10) may transmit the terminal's capability information to a connected base station (e.g., base station 1 or network 1) (1e-15). The terminal capability information may be transmitted via a UECapabilityInformation message. To receive this, the base station may first request the terminal to transmit the terminal's capability information. The request for the terminal capability information may be transmitted via a UECapabilityEnquiry message. The capability information transmitted by the terminal may include whether the terminal supports AI / ML related functions (e.g., NW-side model related functions / operations) (e.g., inference and / or learning) (by AI / ML functionality, by use case, or by sub-use case).

[0124] In step 1e-25, the base station (1e-15) may provide a measurement / reporting configuration (e.g., training configuration) to the terminal for data collection necessary for NW-side model training. Prior to this, the base station (1e-15) may receive configuration information regarding model training or a configuration request regarding model training from the terminal (1e-10) or the training object (1e-20).

[0125] In one embodiment of the present disclosure, the learning object (1e-20) may be a base station (e.g., 1e-15 or another base station) or an AMF or UPF or OAM or TCE (Trace collection entity) or MCE (Measurement collection entity). Or it may be a server connected to them. It may also be an external server (outside of 3GPP).

[0126] In step 1e-30, the terminal (1e-10) may perform measurements to generate data necessary for NW-side model training, and then report the measurement results (e.g., training report) to the base station (1e-15) or the training object (1e-20). When the terminal reports the measurement results to the base station, the base station may reprocess the received report and deliver it to the training object.

[0127] In step 1e-35, the learning object (1e-20) can learn the NW-side model using the received report. Alternatively, the base station (1e-15) can learn the NW-side model using the measurement results received from 1e-30 (e.g., without transmitting them to the learning object).

[0128] In step 1e-50, the base station (e.g., base station 2) (1e-45) may receive a trained model from the training object (1e-20). Alternatively, the base station may possess a model that it has trained itself.

[0129] Before or after step 1e-50, a terminal (e.g., terminal 2) (1e-40) may transmit the terminal's capability information to a connected base station 2 (1e-45). The terminal capability information may be transmitted via a UECapabilityInformation message. To receive this, the base station may first request the terminal to transmit the terminal's capability information. The request for the terminal capability information may be transmitted via a UECapabilityEnquiry message. The capability information transmitted by the terminal may include whether the terminal supports AI / ML related functions (e.g., NW-side model related functions / operations) (by AI / ML functionality, by use case, or by sub-use case).

[0130] In step 1e-55, terminal 2 (1e-40) can perform a measurement (e.g., after receiving measurement settings for the NW-side model from the base station) and then report the measurement result to base station 2 (1e-45).

[0131] In step 1e-60, base station 2 (1e-45) can perform inference or prediction using the measurement report received from terminal 2 (1e-40) and the learned model. For example, in the case of beam management, spatial or temporal beam prediction can be performed.

[0132] In step 1e-65, the base station 2 (1e-45) can perform downlink beam management for terminal 2 (1e-40) based on the prediction / inference result and select an appropriate beam to service the terminal.

[0133] In one embodiment of the present disclosure, the content of the invention related to the UE-side model and / or NW-side model has been described assuming a downlink beam management use case; however, it can be used / applied in the same way to various use cases using the UE-side model and / or NW-side model regardless of the use case.

[0134] As an example of one embodiment of the present disclosure, the contents of the present invention may be used in a UE-side model and / or an NW-side model.

[0135] FIG. 1f is the first diagram illustrating a procedure in which a terminal or terminal server learns a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model and performs inference using it, according to one embodiment of the present disclosure.

[0136] In step 1f-20, the terminal (e.g., 1f-05) may report the terminal's AI / ML-related capability information to a base station (gNB) or a network (NW) (e.g., 1f-10). The report may be transmitted via a UE Capability Information message. Before the terminal makes the report, the base station may transmit a UE Capability Enquiry message to the terminal to request the terminal to report its capability or to request the transmission of the report. The AI / ML-related capability information may include at least one of the following information.

[0137] - Whether the device supports AI / ML model operations (e.g., training and / or inference) (by use case or sub-use case)

[0138] - Whether the terminal supports UE-sided AI / ML model-related operations (e.g., training and / or inference) (by use case or sub-use case)

[0139] - Information regarding use cases, sub-use cases, and / or functionality that the terminal supports operations related to UE-sided AI / ML models

[0140] - Network setting information (to be described later) and / or network setting ID (to be described later) supported by the terminal

[0141] - Information related to the network implementation (to be described later) supported by the terminal and / or the network implementation ID (to be described later)

[0142] - Network configuration and implementation information (to be described later) and / or network configuration / implementation ID (to be described later) supported by the terminal

[0143] In step 1f-25, the terminal (1f-05) may receive a setting regarding permission or authorization for a model learning (initiation) request from the base station (1f-10). The setting may be transmitted via an RRC Reconfiguration message, an RRC Resume message, a UE Information Request message, or System Information (SIB).

[0144] The above settings may be transmitted as UE Assistance Information settings. The above settings may include use cases and / or sub-use cases and / or functionality that allow or permit model learning (initiation) requests to the base station and the terminal.

[0145] In step 1f-40, the terminal (1f-05) may transmit a request to the base station (1f-10) regarding model learning (initiation). The request may be transmitted via a UE Assistance Information message or a UE Information Response message.

[0146] The above request may include use cases and / or sub-use cases and / or functionality of model learning that the terminal requests from the base station. If the terminal has received inference (inference or prediction) related settings from the base station but is unable to perform inference (e.g., if the inference settings are not applicable, i.e., inapplicable), the terminal may transmit the above request to the base station (e.g., by indicating that the inference settings are inapplicable, or separately from the inapplicable indication).

[0147] In step 1f-45, the base station (1f-10) may provide model training related settings to the terminal (1f-05). The base station may provide AI / ML related settings to the terminal. The AI / ML related settings may include configuration information for the terminal or terminal server to train an AI / ML model. The settings may be provided via an RRC Reconfiguration message or an RRC Resume message. For example, the settings may include measurement settings for SET A as well as SET B. This is because the information may be necessary for the terminal or terminal OTT to train an AI / ML model that derives a specific output (e.g., a measurement value for SET A) for a specific input (e.g., a measurement value for SET B). The AI / ML related settings or model training related settings may include at least one of the following information.

[0148] - Information 1. (AI / ML Related) Configuration ID

[0149] The terminal can receive multiple AI / ML-related settings from the base station, and an AI / ML-related setting ID may be assigned for each AI / ML-related setting to distinguish each AI / ML-related setting.

[0150] - Information 2. Information regarding use cases, sub-use cases, and / or functionality utilizing AI / ML

[0151] - Information 3. Information regarding network configuration and / or information regarding network implementation

[0152] Network configuration information may refer to configuration information provided by the network to the terminal. Network configuration information may include at least one of the information in the sub-paragraph.

[0153] >> Information 3-1. Measurement-related configuration information for the cells or beams required for the terminal or the server responsible for and storing the training of the AI / ML model connected to the terminal (e.g., terminal OTT (over-the-top)) (e.g., 1f-15) to train the AI / ML model (e.g., beam-related information used as input, beam-related information to be inferred / derived as output, measurement cell information, measurement beam information, measurement beam ID, SET A / B related information, measurement cycle, measurement resource settings for SET A / B)

[0154] >> Information 3-2. Information regarding input and / or output and related configuration information required when the terminal performs inference using an AI / ML model (e.g., a trained AI / ML model) (e.g., values / information to be used as input by the terminal, measurement beam information for acquiring the measurement values ​​to be used as input by the terminal, measurement beam ID, measurement resource settings, measurement period, SET B related information, values / information to be derived as output by the terminal, prediction period of the predicted values ​​to be derived as output by the terminal, SET A related information, prediction beam ID, time interval subject to prediction)

[0155] >> Information 3-3. Network Configuration ID. Each network configuration-related information can be distinguished / indicated by a single ID (e.g., network configuration ID). A terminal may receive multiple network configuration-related information from a base station, and a network configuration ID may be indicated for each network configuration-related information to distinguish each network configuration-related information.

[0156] Network implementation information may refer to network implementation information other than network settings that the network sets for the terminal, and may be referred to as NW-side additional conditions. For example, it may include size / order / indexing / pattern information of SET A and / or SET B, downlink spatial domain transmission filter, QCL (Quasi colocation) information, TCI (Transmission configuration indication) information, input and output order of the model, downlink transmission beam information (direction, number), downlink transmission beam operation settings, beam codebook, antenna operation settings, transmission power information, terminal distribution information, antenna height information, cell or base station installation environment information (e.g., indoor, outdoor, urban, rural, road), etc.

[0157] In one embodiment of the present disclosure, each network implementation-related information may be distinguished / indicated by a single ID (e.g., network implementation ID or associated ID). A terminal may receive a plurality of network implementation-related information from a base station, and a network implementation ID may be indicated for each network implementation-related information to distinguish each network implementation-related information.

[0158] In one embodiment of the present disclosure, each network setting and implementation related information may be distinguished / indicated by a single ID (e.g., network setting / implementation ID or associated ID). A terminal may receive a plurality of network setting / implementation related information from a base station, and a network setting / implementation ID may be indicated for each network setting / implementation related information in order to distinguish each network setting / implementation related information.

[0159] In one embodiment of the present disclosure, if the network configuration and implementation information (of a base station (e.g., 1f-10)) for which the terminal performs learning (measurement for learning) and the network configuration and implementation information (of a base station (e.g., 1f-75)) for which the terminal performs inference are different, the terminal may not be able to perform inference using an AI / ML model and may not be able to perform inference. In one embodiment of the present disclosure, if the network configuration and implementation information (of a base station (e.g., 1f-10)) for which the terminal performs learning (measurement for learning) matches the network configuration and implementation information (of a base station (e.g., 1f-75)) for which the terminal performs inference, the terminal may be able to perform inference using an AI / ML model and may perform inference.

[0160] - Information 4. PLMN Information and / or Area Information

[0161] Some or all of the AI / ML-related settings received by the terminal may indicate a valid PLMN or Area. In one embodiment of the present disclosure, some or all of the AI / ML-related settings received by the terminal may be values ​​used within a specific PLMN(s) and / or area(s). That is, if the terminal remains within (or moves within) a specific PLMN(s) and / or area(s), it may maintain / use existing setting information, and if the terminal moves out of the specific PLMN(s) and / or area(s) to a new PLMN(s) and / or area(s), it may disable or delete the existing settings. If the terminal moves out of the specific PLMN(s) and / or area(s) to a new PLMN(s) and / or area(s), it may update the settings or receive new settings (by requesting new settings from the base station or by the base station providing new settings).

[0162] In step 1f-50, the terminal (1f-05) can perform a measurement (e.g., a measurement for SET A and / or SET B) (based on the model learning settings received from the base station).

[0163] In step 1f-55, the terminal (1f-05) may transmit the result or data regarding the measurement to the terminal server (1f-15). For the transmission, the terminal may first transmit the result or data regarding the measurement to the base station. The result or data regarding the measurement may be transmitted to the base station via a Measurement report message, a UE Assistance Information message, a UE Information Response message, a new RRC message, via Layer 1 signaling, via MAC CE, or via an RRC Reconfiguration complete message. In one embodiment of the present disclosure, when a UE Information Response message is used for the transmission, the terminal may first receive a request for a measurement report from the base station via a UE Information Request message, and then transmit the measurement result to the base station using the UE Information Response message. Upon receiving this, the base station may transmit the received information (after reprocessing) to the terminal server. The terminal may transmit the aforementioned AI / ML-related setting ID and / or network setting ID and / or network implementation ID and / or network setting / implementation ID along with the measurement result to indicate which AI / ML-related setting the result was measured under. In addition, the terminal may transmit information necessary for the terminal server to train the AI / ML model.

[0164] In step 1f-60, the terminal server (1f-15) can perform model training (based on measurement data required for received model training). The terminal server can perform separate model training for each AI / ML related setting ID and / or network setting ID and / or network implementation ID and / or network setting / implementation ID.

[0165] In step 1f-75, the terminal server (1f-15) may transmit the (trained) model (and related information) to the terminal (e.g., terminal 2) (1f-65). To transmit the model, the terminal server may first transmit the model to a base station (e.g., base station 2) (1f-70). Upon receiving this, the base station may transmit the received model and related information (after reprocessing) to the terminal.

[0166] In step 1f-80, base station 2 (1f-70) may provide inference-related settings to terminal 2 (1f-65). The inference-related settings may be transmitted to terminal 2 (1f-65) via an RRC Reconfiguration message or a UE information Request message, via Layer 1 signaling, via MAC CE, or via Layer 1 and / or RRC CSI reporting settings. The inference-related settings may refer to the description in step 1f-45 and may include at least one of the information in step 1f-45.

[0167] In step 1f-85, terminal 2 (1f-65) can determine applicability or inapplicability (for the received inference setting).

[0168] For example, a terminal may determine that it is applicable if it can infer with a received inference setting. Cases where inference with a received inference setting is applicable may include cases where it possesses a suitable or trained model, or cases where inference for SET A is valid / possible based on measurements for SET B included in the inference setting.

[0169] For example, if a terminal cannot infer a received inference setting, it may be determined to be inapplicable. Cases where inference is impossible for a received inference setting may include cases where a suitable or trained model is not possessed.

[0170] For example, the terminal may determine applicability or inapplicability based on a network configuration / implementation ID (or associated ID) within a received inference configuration. For example, the terminal may determine that it is applicable if it can infer the received network configuration / implementation ID. Cases where the terminal can infer the received network configuration / implementation ID may include cases where it possesses a suitable or trained model for the network configuration / implementation ID, or cases where inference for SET A is valid / possible based on measurements for SET B included in the inference configuration regarding the network configuration / implementation ID included in the inference configuration.

[0171] For example, if a terminal cannot infer a received network configuration / implementation ID, it may be determined to be inapplicable. A case where a terminal cannot infer a received network configuration / implementation ID may include a case where it does not possess a suitable or trained model for the network configuration / implementation ID.

[0172] In step 1f-90, terminal 2 (1f-65) may indicate to the base station whether the applicability or inapplicability (of the received inference configuration) is applicable or inapplicable, or report relevant information to base station 2 (1f-70). This information may be reported to base station 2 via a Measurement report message, a UE Assistance Information message, a UE Information Response message, a new RRC message, via Layer 1 signaling, via MAC CE, or via an RRC Reconfiguration complete message.

[0173] In step 1f-95, (if applicability is reported) terminal 2 (1f-65) may perform measurements (e.g., measurements for SET B) and inference (e.g., inference for SET A) based on the inference settings. Prior to step 1f-95, the base station may (explicitly) instruct the terminal to perform inference regarding the inference settings. In one embodiment of the present disclosure, the terminal may initiate inference using an AI / ML model on its own after instructing the base station on applicability or transmitting applicable information (e.g., without an explicit activation instruction from the base station).

[0174] In step 1f-100, terminal 2 (1f-65) may transmit inferred measurement values, predicted values, predicted information and / or output information to base station 2 (1f-70) using an AI / ML model. The inferred measurement values, predicted values, predicted information and / or output information may be transmitted to base station 2 via a Measurement report message, a UE Assistance Information message, a UE Information Response message, a new RRC message, via Layer 1 signaling (PUCCH / PUSCH), via MAC CE, or via an RRC Reconfiguration complete message. At this time, the terminal may transmit the aforementioned AI / ML related setting ID, network setting ID, network implementation ID, and / or network setting / implementation ID along with the measurement result to indicate which AI / ML related setting the result was measured according to. At this time, the terminal may also transmit the information or measurement value information used as input.

[0175] In step 1f-105, base station 2 (1f-70) can perform network management (e.g., beam management) using information received from terminal 2 (1f-65).

[0176] In one embodiment of the present disclosure, the network configuration / implementation ID (or associated ID) for which the terminal (1f-10) or terminal server (1f-15) wishes to perform model learning may not match the network configuration / implementation ID (or associated ID) used or supported by the base station (1g-10). For example, even if supported by the base station, a learning procedure (e.g., 1f-45, 1f-50, 1f-55) for a network configuration / implementation ID (or associated ID) for which the terminal or terminal server does not wish to perform model learning may not be necessary and may result in a waste of energy or wireless resources of the terminal or base station. Additionally, for example, even if the terminal or terminal server wishes to perform model learning, the base station may not be able to provide a measurement configuration (e.g., 1f-45) for a network configuration / implementation ID (or associated ID) for which the base station does not use or support. Therefore, it may be necessary to match the network configuration / implementation ID (or associated ID) that the terminal or terminal server wants to train a model with the network configuration / implementation ID (or associated ID) that the base station (1g-10) uses or supports.

[0177] In one embodiment of the present disclosure, a single network configuration / implementation ID (or associated ID) may be information indicating a plurality of beams. If the Associated ID is 1, there may be 64 corresponding beams ranging from 1 to 64, and if the Associated ID is 2, there may be 32 corresponding beams ranging from 1 to 32. However, a terminal or terminal server may not learn all beams (e.g., 64 beams) for a single associated ID (e.g., Associated ID=1) (e.g., learning completed for only 10 beams). Additionally, during the learning process, the base station may not use or support all beams (e.g., 64 beams) for a single associated ID (e.g., Associated ID=1) for the terminal (e.g., support for only 15 beams).

[0178] In one embodiment of the present disclosure, for each network configuration / implementation ID (or associated ID), the beam (one or more) that the terminal (1f-10) or terminal server (1f-15) wants to model learn and the beam (one or more) that the base station (1g-10) uses or supports may not match. For example, even if the base station supports it, a learning procedure (e.g., 1f-45, 1f-50, 1f-55) for a beam (or beam ID for identifying it) that the terminal or terminal server does not want to model learn may not be necessary and may result in a waste of energy or wireless resources of the terminal or base station. Additionally, for example, even if the terminal or terminal server wants to model learn, the base station may not be able to provide a measurement configuration (e.g., 1f-45) for a beam (or beam ID for identifying it) that the base station does not use or support. Therefore, for each network configuration / implementation ID (or associated ID), it may be necessary to match the beam (or beam ID for identifying the beam) that the terminal or terminal server wants to train a model with the beam (or beam ID for identifying the beam) that the base station (1g-10) uses or supports.

[0179] FIG. 1g is the first drawing illustrating a procedure for a terminal or terminal server to learn a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model according to one embodiment of the present disclosure.

[0180] Steps 1g-05 through 1g-60 mentioned above may cross-refer to the procedures or descriptions from steps 1f-05 through 1f-60.

[0181] In step 1g-25, the base station (1g-10) may instruct the terminal (1g-05) to allow or permit (and information regarding therein) a request for model learning (initiation). At this time, the base station may instruct the terminal to the network configuration / implementation ID (or associated ID) (one or more) that the base station supports / allows / uses. Additionally, the base station may instruct the terminal to the beam ID (one or more) that the base station supports / allows / uses for each network configuration / implementation ID (or associated ID).

[0182] In one embodiment of the present disclosure, an associated ID and a beam ID may be transmitted separately (through a separate procedure or a separate message) from an allow or permission instruction for a model learning (initiation) request transmitted by a base station to a terminal.

[0183] In step 1g-27, the terminal (or the terminal AS layer or the terminal RRC layer) may transmit permission or authorization (and information regarding such permission) for a model learning (initiation) request (received from the base station) to the upper layer. The transmitted information may include a network configuration / implementation ID (or associated ID) and / or a beam ID by network configuration / implementation ID (or associated ID).

[0184] In step 1g-30, the terminal (1g-05) may transmit an allow or permission instruction to the terminal server (1g-15) for a model learning (initiation) request. Along with this, or separately (through a separate procedure or a separate message), a network configuration / implementation ID (or associated ID) (one or more) supported / allowed / used by the base station and / or a beam ID (one or more) supported / allowed / used by the base station for each network configuration / implementation ID (or associated ID) may be transmitted.

[0185] In step 1g-32, the terminal server (1g-15) can determine / select a network configuration / implementation ID (or associated ID) and / or a beam ID (one or more) per network configuration / implementation ID (or associated ID) for which it wishes to perform model training (among the received information).

[0186] In step 1g-35, the terminal server (1g-15) may transmit / instruct a model learning (initiation) request to the terminal (1g-05). Along with this, the terminal server may transmit to the terminal the determined / selected (one or more) associated IDs and / or (one or more) beam IDs for each associated ID (for model learning).

[0187] In step 1g-37, the terminal upper layer may forward the received model learning (initiation) request and information to the lower layer (terminal AS layer or terminal RRC layer). If the terminal lower layer (terminal AS layer or terminal RRC layer) receives the model learning (initiation) request and / or (one or more) associated IDs and / or (one or more) beam IDs for each associated ID from the upper layer, it may transmit the model learning (initiation) request (1g-40) to the base station.

[0188] In step 1g-40, the terminal (1g-05) may transmit a request for (initiation) of model learning to the base station. At this time, the terminal may include (received) (one or more) associated IDs and / or (one or more) beam IDs for each associated ID.

[0189] In step 1g-45, the base station (1g-10) may provide the terminal with measurement settings for model learning. The base station may include an associated ID (associated with the setting) and / or one or more beam IDs for each associated ID. The base station may instruct the terminal to have a corresponding associated ID and / or beam ID for each measurement resource (e.g., CSI resource (set) (ID), SSB resource (set) (ID), CRI, SSBRI).

[0190] In step 1g-50, the terminal (1g-05) can perform measurements for model learning.

[0191] In step 1g-52, the terminal (1g-05) can transmit the results of the measurement and related data for model learning to an upper layer.

[0192] In step 1g-55, the terminal (1g-05) may transmit the measurement results and related data for model learning to a base station (1g-10) or a terminal server (1g-15) (or a terminal server via the base station). At this time, the terminal (or base station) may include a measurement resource ID that performed the measurement or a corresponding associated ID and / or a beam ID (per associated ID) for each measurement result.

[0193] In step 1g-60, the terminal server (1g-15) can perform model training based on the measurement result and the corresponding associated ID and / or beam ID (per associated ID).

[0194] FIG. 1h is a second figure illustrating a procedure for a terminal or terminal server to learn a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model according to one embodiment of the present disclosure.

[0195] Steps 1h-05 through 1h-60 may cross-refer to the procedures or descriptions of steps 1f-05 through 1f-60 described above.

[0196] Step 1h-25 may cross-refer to Step 1f-25 and / or Step 1g-25.

[0197] In step 1h-35, the terminal server (1h-15) may transmit a request to the terminal (1h-05) to start model training (via the base station (1h-10)). At this time, the terminal server may transmit the request including a network configuration / implementation ID (or associated ID) (for which model training is desired) and / or a beam ID (one or more) per network configuration / implementation ID (or associated ID).

[0198] In step 1h-37, the terminal upper layer may forward the received model learning (initiation) request and information to the lower layer (terminal AS layer or terminal RRC layer). When the terminal lower layer (terminal AS layer or terminal RRC layer) receives the model learning (initiation) request and / or (one or more) associated IDs and / or (one or more) beam IDs per associated ID from the upper layer, it may select (one or more) associated IDs and / or (one or more) beam IDs per associated ID that the base station has instructed to support / use from among these. That is, the terminal may select the associated ID and / or beam ID per associated ID that the (currently connected) base station supports / uses after receiving a request from the terminal server.

[0199] In step 1h-40, the terminal (1h-05) may request the base station (1h-10) to initiate model learning, including the selected information. This may cross-refer to step 1f-40 and / or step 1g-40.

[0200] In step 1h-45, the base station (1h-10) may provide learning-related settings to the terminal (1h-05). This may cross-reference steps 1f-45 and / or 1g-45.

[0201] FIG. 1i is a second figure illustrating a procedure in which a terminal or terminal server learns a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model and performs inference using it, according to one embodiment of the present disclosure.

[0202] Steps 1i-05 through 1i-60 may cross-reference the procedures or descriptions of the aforementioned steps 1f-05 through 1f-60 and / or the procedures or descriptions of the aforementioned steps 1g-05 through 1g-60 and / or the procedures or descriptions of the aforementioned steps 1h-05 through 1h-60.

[0203] Steps 1i-65 through 1i-105 may cross-refer to the procedures or descriptions of steps 1f-65 through 1f-105 described above.

[0204] In step 1i-80, base station 2 (1i-70) may provide a setting for inference or an inference setting to the terminal (1i-65). The setting may include an associated ID and / or beam ID used for measurement / inference. For example, the setting may include a measurement setting for SET B and / or an inference setting for SET A. The base station may instruct the terminal to have an associated ID and / or beam ID corresponding to each measurement resource (e.g., CSI resource (set) (ID), SSB resource (set) (ID), CRI, SSBRI) for SET B. The base station may instruct the terminal to have an associated ID and / or beam ID (of the target of inference) for SET A.

[0205] In step 1i-85, terminal 2 (1i-65) may determine applicability or inapplicability (for the received inference setting). For example, the terminal may determine that it is applicable if it is inferable for the received inference setting. The case where the terminal is inferable for the received inference setting may include having a learned model for the associated ID and / or beam ID indicated for SET A and / or SET B.

[0206] For example, if the terminal is unable to infer a received inference setting, it may be determined to be inapplicable. Cases where the terminal is unable to infer a received inference setting may include cases where it does not possess a learned model for the associated ID and / or beam ID instructed for SET A and / or SET B.

[0207] In step 1i-90, terminal 2 (1i-65) may indicate to the base station the applicability or inapplicability (of the received inference setting) or report related information to base station 2 (1f-70). This may cross-refer to step 1f-90.

[0208] In step 1i-95, (if applicability is reported) terminal 2 (1i-65) can perform measurements (e.g., measurements for SET B) and inferences (e.g., inferences for SET A) based on the inference setup. This can cross-reference step 1f-95.

[0209] In step 1i-100, terminal 2 (1i-65) can transmit measurements, predicted values, predicted information and / or output information inferred using an AI / ML model to base station 2 (1i-70). This can cross-reference step 1f-100.

[0210] In step 1i-105, base station 2 (1i-70) can perform network management (e.g., beam management) using information received from terminal 2 (1i-65). This may cross-refer to step 1f-105.

[0211] In one embodiment of the present disclosure, a base station may provide a terminal with settings related to model learning (e.g., 1f-45). However, the base station may not know how long the relevant settings should be maintained. This is because the base station may not know when the terminal or terminal server wants to end model learning or when model learning has been completed. Therefore, relevant information exchange between the terminal or terminal server and the base station may be necessary.

[0212] FIG. 1j is the first diagram illustrating a procedure in which a terminal stops measurement and data collection for training a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model according to one embodiment of the present disclosure.

[0213] Steps 1j-05 through 1j-60 may cross-reference the procedures or descriptions of steps 1f-05 through 1f-60 described above and / or the procedures or descriptions of steps 1g-05 through 1g-60 described above and / or the procedures or descriptions of steps 1h-05 through 1h-60 described above and / or the procedures or descriptions of steps 1i-05 through 1i-60 described above.

[0214] In step 1j-35, the terminal server (1j-15) may send a request to the terminal (1j-05) to start model training. At this time, the terminal server may instruct the terminal to maintain the measurement settings (e.g., time 1) required (at least) for model training and measurement data collection.

[0215] In step 1j-40, the terminal (1j-05) may transmit a request to the base station (1j-10) to start model learning. At this time, the terminal may instruct the base station to maintain the measurement settings (e.g., time 1) required (at least) for model learning and measurement data collection.

[0216] In step 1j-45, the base station (1j-10) can provide measurement settings (for model learning) to the terminal (1j-05).

[0217] In step 1j-46, the base station (1j-10) may maintain the measurement settings (for model training) without releasing them to the terminal for (at least) a directed time of 1. During said time, the terminal may collect / report data (e.g., 1j-55) through measurements (e.g., 1j-50), and the terminal server may train the model (e.g., 1j-60).

[0218] In step 1j-120, the base station (1j-10) can release the measurement settings (for model learning) set for the terminal (1j-05).

[0219] FIG. 1k is a second drawing illustrating a procedure in which a terminal stops measurement and data collection for training a UE-side Artificial Intelligence (AI) / Machine Learning (ML) model according to one embodiment of the present disclosure.

[0220] Steps 1k-05 through 1k-60 may cross-reference the procedures or descriptions of steps 1f-05 through 1f-60 described above and / or the procedures or descriptions of steps 1g-05 through 1g-60 described above and / or the procedures or descriptions of steps 1h-05 through 1h-60 described above and / or the procedures or descriptions of steps 1i-05 through 1i-60 described above.

[0221] In step 1k-25, the base station (1k-15) may instruct the terminal (1k-05) to allow or permit a request for model learning (initiation and / or termination).

[0222] Step 1k-35 can cross-reference Step 1j-35.

[0223] In step 1k-45, the terminal (1k-05) can receive model learning related settings from the base station (1k-10).

[0224] In step 1k-46, the terminal (1k-10) may maintain the model learning-related settings for time 1 and perform measurements (e.g., 1k-50), data collection, and reporting (e.g., 1k-55). To do this, the terminal may start one timer (e.g., Timer 1) upon reception in the model learning-related settings (1k-45). Timer 1 may run for time 1. When the base station releases the model learning-related settings to the terminal, the terminal may stop running Timer 1.

[0225] In step 1k-117, the terminal (1k-05) may request the base station (1k-10) to disable the model learning-related settings or to stop the measurement for model learning. The request to disable the model learning-related settings or to stop the measurement for model learning may be made via a UE Assistance information message or a UE Information Response message. For example, the terminal may make the request when timer 1 expires (or when time 1 has passed since step 1k-46, or when the measurement has been performed for time 1).

[0226] In one embodiment of the present disclosure, if the terminal receives a request from the terminal server to disable a setting related to model learning or to stop a measurement for model learning (e.g., 1k-115), the terminal may perform said request.

[0227] In step 1k-120, the base station (1k-10) can release the measurement settings (for model learning) set for the terminal (1k-05).

[0228] Referring to FIG. 11, the terminal may include an RF (Radio Frequency) processing unit (11-10), a baseband processing unit (11-20), a storage unit (11-30), and a control unit (11-40).

[0229] The RF processing unit (1l-10) can perform functions for transmitting and receiving signals through a wireless channel, such as signal band conversion and amplification. For example, the RF processing unit (1l-10) can up-convert a baseband signal provided by the baseband processing unit (1l-20) into an RF band signal and transmit it through an antenna, and can down-convert an RF band signal received through the antenna into a baseband signal. For example, the RF processing unit (1l-10) may include a transmit filter, a receive filter, an amplifier, a mixer, an oscillator, a DAC (digital to analog converter), an ADC (analog to digital converter), etc. Although only one antenna is shown in FIG. 1l, the terminal may be equipped with multiple antennas. Additionally, the RF processing unit (1l-10) may include multiple RF chains. Furthermore, the RF processing unit (1l-10) may perform beamforming. For the above beamforming, the RF processing unit (1l-10) can adjust the phase and magnitude of each of the signals transmitted and received through multiple antennas or antenna elements. In addition, the RF processing unit can perform MIMO (multi-input multi-output) and can receive multiple layers when performing MIMO operation.

[0230] The baseband processing unit (1l-20) can perform a conversion function between a baseband signal and a bit sequence according to the physical layer specifications of the system. For example, when transmitting data, the baseband processing unit (1l-20) can generate complex symbols by encoding and modulating the transmitted bit sequence. Additionally, when receiving data, the baseband processing unit (1l-20) can restore the received bit sequence by demodulating and decoding the baseband signal provided by the RF processing unit (1l-10). For example, in the case of following the orthogonal frequency division multiplexing (OFDM) method, when transmitting data, the baseband processing unit (1l-20) can generate complex symbols by encoding and modulating the transmitted bit sequence, and after mapping the complex symbols to subcarriers, can construct OFDM symbols through inverse fast Fourier transform (IFFT) operations and cyclic prefix (CP) insertion. Additionally, upon receiving data, the baseband processing unit (1l-20) can divide the baseband signal provided from the RF processing unit (1l-10) into OFDM symbol units, restore the signals mapped to subcarriers through a fast Fourier transform (FFT) operation, and then restore the received bit sequence through demodulation and decoding.

[0231] The baseband processing unit (1l-20) and the RF processing unit (1l-10) can transmit and receive signals as described above. Accordingly, the baseband processing unit (1l-20) and the RF processing unit (1l-10) may be referred to as a transmitting unit, a receiving unit, a transmitting and receiving unit, or a communication unit. Furthermore, at least one of the baseband processing unit (1l-20) and the RF processing unit (1l-10) may include a plurality of communication modules to support a plurality of different wireless access technologies. Additionally, at least one of the baseband processing unit (1l-20) and the RF processing unit (1l-10) may include different communication modules to process signals of different frequency bands. For example, the different wireless access technologies may include wireless LAN (e.g., IEEE 802.11), cellular network (e.g., LTE), etc. In addition, the above different frequency bands may include super high frequency (SHF) bands (e.g., 2 NRHz, NRHz) and millimeter wave (e.g., 60 GHz) bands.

[0232] The storage unit (1l-30) can store data such as basic programs, application programs, and setting information for the operation of the terminal. In particular, the storage unit (1l-30) can store information related to a second connection node that performs wireless communication using wireless connection technology. Additionally, the storage unit (1l-30) can provide the stored data upon a request from the control unit (1l-40).

[0233] The control unit (1l-40) can control the overall operations of the terminal. For example, the control unit (1l-40) can control the terminal to perform the embodiments and / or methods of the present disclosure described above. For example, the control unit (1l-40) can transmit and receive signals through the baseband processing unit (1l-20) and the RF processing unit (1l-10). Additionally, the control unit (1l-40) can write and read data to and from the storage unit (1l-30). To this end, the control unit (1l-40) may include at least one processor. For example, the control unit (1l-40) may include a communication processor (CP) that performs control for communication and an application processor (AP) that controls upper layers such as applications, and may include a multiple connection processing unit (1l-42) as illustrated in the drawing.

[0234] FIG. 1m is a drawing illustrating the structure of a base station according to one embodiment of the present disclosure.

[0235] Referring to FIG. 1m, according to one example of the present disclosure, a base station may be configured to include an RF processing unit (1m-10), a baseband processing unit (1m-20), a backhaul communication unit (1m-30), a storage unit (1m-40), and a control unit (1m-50).

[0236] The RF processing unit (1m-10) can perform functions for transmitting and receiving signals through a wireless channel, such as signal band conversion and amplification. That is, the RF processing unit (1m-10) can up-convert a baseband signal provided by the baseband processing unit (1m-20) into an RF band signal and transmit it through an antenna, and can down-convert an RF band signal received through the antenna into a baseband signal. For example, the RF processing unit (1m-10) may include a transmit filter, a receive filter, an amplifier, a mixer, an oscillator, a DAC, an ADC, etc. Although only one antenna is shown in the drawing, the base station may be equipped with multiple antennas. Additionally, the RF processing unit (1m-10) may include multiple RF chains. Furthermore, the RF processing unit (1m-10) may perform beamforming. For the above beamforming, the RF processing unit (1m-10) can adjust the phase and magnitude of each of the signals transmitted and received through a plurality of antennas or antenna elements. The RF processing unit can perform down-to-down MIMO operation by transmitting one or more layers.

[0237] The baseband processing unit (1m-20) can perform a conversion function between a baseband signal and a bit sequence according to the physical layer specifications of the wireless access technology. For example, when transmitting data, the baseband processing unit (1m-20) can generate complex symbols by encoding and modulating the transmitted bit sequence. Additionally, when receiving data, the baseband processing unit (1m-20) can restore the received bit sequence by demodulating and decoding the baseband signal provided by the RF processing unit (1m-10). For example, in the case of an OFDM method, when transmitting data, the baseband processing unit (1m-20) can generate complex symbols by encoding and modulating the transmitted bit sequence, and after mapping the complex symbols to subcarriers, can construct OFDM symbols through IFFT operation and CP insertion. Additionally, upon receiving data, the baseband processing unit (1m-20) can divide the baseband signal provided by the RF processing unit (1m-10) into OFDM symbol units, restore the signals mapped to subcarriers through FFT operations, and then restore the received bit sequence through demodulation and decoding. The baseband processing unit (1m-20) and the RF processing unit (1m-10) can transmit and receive signals as described above. Accordingly, the baseband processing unit (1m-20) and the RF processing unit (1m-10) may be referred to as a transmitting unit, a receiving unit, a transceiver unit, a communication unit, or a wireless communication unit.

[0238] The backhaul communication unit (1m-30) can provide an interface for communicating with other nodes within the network. That is, the backhaul communication unit (1m-30) can convert a bit sequence transmitted from the main base station to another node, e.g., an auxiliary base station, a core network, etc., into a physical signal, and can convert a physical signal received from the other node into a bit sequence.

[0239] The storage unit (1m-40) can store data such as basic programs, application programs, and configuration information for the operation of the main station. In particular, the storage unit (1m-40) can store information regarding bearers assigned to connected terminals, measurement results reported from connected terminals, etc. Additionally, the storage unit (1m-40) can store information that serves as a criterion for determining whether to provide or disconnect multiple connections to the terminals. Furthermore, the storage unit (1m-40) can provide the stored data upon a request from the control unit (1m-50).

[0240] The control unit (1m-50) can control the overall operations of the main base station. For example, the control unit (1m-50) can control the base station to perform the embodiments and / or methods of the present disclosure described above. For example, the control unit (1m-50) can transmit and receive signals through the baseband processing unit (1m-20) and the RF processing unit (1m-10) or through the backhaul communication unit (1m-30). Additionally, the control unit (1m-50) can write and read data to and from the storage unit (1m-40). To this end, the control unit (1m-50) may include at least one processor and may include a multiple connection processing unit (1m-52) as illustrated in the drawing.

[0241] Meanwhile, the embodiments of the present disclosure disclosed in this specification and drawings are merely specific examples provided to facilitate the explanation of the technical content of the present disclosure and to aid in understanding the present disclosure, and are not intended to limit the scope of the present disclosure. That is, it is obvious to those skilled in the art that other variations based on the technical concept of the present disclosure are possible.

[0242] In addition, each of the above embodiments may be combined and operated as needed. For example, parts of one embodiment of the present disclosure and another embodiment may be combined to operate a base station and a terminal.

[0243] In addition, the embodiments of the present disclosure are applicable to other communication systems, and other variations based on the technical concept of the embodiments may also be implemented. For example, the embodiments may be applied to LTE systems, 5G, NR systems, or 6G systems.

[0244] Therefore, the scope of the present disclosure should not be limited to the described embodiments, but should be defined by the claims set forth below as well as equivalents thereof.

Claims

1. A method performed by a terminal in a communication system, A step of receiving a Radio Resource Control (RRC) message from a base station containing configuration information related to the terminal's reporting settings for model learning of the terminal; A method characterized by including the step of transmitting a terminal assistance information (UE Assistance Information) message related to the suspension of the model learning setting to the base station when, based on the reception of the above setting information, the terminal determines that it prefers to suspend the setting for model learning.

2. In Paragraph 1, A method characterized by including the step of transmitting a terminal assistance information (UE Assistance Information) message regarding the settings for model learning to the base station when, based on the reception of the above setting information, it is determined that the terminal prefers the settings for model learning.

3. In Paragraph 1, A method characterized by the above setting information including at least one of setting information for measurement or setting information for prediction.

4. In Paragraph 1, A method characterized by the above-mentioned configuration information further including at least one implementation ID (associated ID).

5. In a method performed by a base station in a communication system, A step of transmitting a Radio Resource Control (RRC) message to a terminal, the message including configuration information related to the terminal's reporting settings for model training of the terminal; A method characterized by including the step of receiving a terminal assistance information (UE Assistance Information) message related to the suspension of the model training setting from the terminal when it is determined that the terminal prefers to suspend the setting for model training based on the reception of the above setting information.

6. In Paragraph 5, A method characterized by including the step of receiving a terminal assistance information (UE Assistance Information) message regarding the settings for model learning from the terminal when it is determined that the terminal prefers the settings for model learning based on the reception of the above setting information.

7. In Paragraph 5, A method characterized by the above setting information including at least one of setting information for measurement or setting information for prediction.

8. In Paragraph 5, A method characterized by the above-mentioned configuration information further including at least one implementation ID (associated ID).

9. In a terminal in a communication system, At least one transceiver; At least one processor connected to the above at least one transceiver so as to be able to communicate; and The terminal is connected to communicate with at least one processor and is capable of executing individually or in any combination of the at least one processor, so that the terminal, Receive a Radio Resource Control (RRC) message from a base station containing configuration information related to the terminal's reporting settings for model training of the terminal, and A memory storing a command to transmit a terminal assistance information (UE Assistance Information) message related to the suspension of the model learning setting to the base station when, based on the reception of the above setting information, the terminal determines that it prefers to suspend the setting for model learning; A terminal including 10. In Clause 9, the above command is that the terminal, A terminal characterized by transmitting a terminal assistance information (UE Assistance Information) message regarding the settings for model learning to the base station when the terminal determines that it prefers the settings for model learning based on the reception of the above setting information.

11. In Paragraph 9, A terminal characterized by including at least one of the above-mentioned setting information for measurement or setting information for prediction.

12. In Paragraph 9, A terminal characterized by the above-mentioned configuration information further including at least one implementation ID (associated ID).

13. In a base station of a communication system, At least one transceiver; At least one processor connected to the above at least one transceiver so as to be able to communicate; and The terminal is connected to communicate with at least one processor and is capable of executing individually or in any combination of the at least one processor, so that the terminal, Transmit a Radio Resource Control (RRC) message to a terminal that includes configuration information related to the terminal's reporting settings for training the terminal's model, and A memory storing a command to receive a terminal assistance information (UE Assistance Information) message related to the suspension of the model training setting from the terminal when it is determined that the terminal prefers to suspend the model training setting based on the reception of the above setting information; Base station including 14. In Paragraph 13, the above command is that the base station, A base station characterized by receiving a terminal assistance information (UE Assistance Information) message regarding the settings for model learning from the terminal when it is determined that the terminal prefers the settings for model learning based on the reception of the above setting information.

15. In Paragraph 5, A method characterized by the above-mentioned setting information including at least one of setting information for measurement, setting information for prediction, and implementation ID (associated ID).

Images