Method and apparatus for performance monitoring for intelligent beam management in mobile communication system

Abstract

This specification proposes various embodiments related to a method and apparatus for performance monitoring of an intelligent model when utilizing intelligent beam management technology for wireless communication in a mobile communication system composed of a base station and one or more terminals.

Description

Technology Field

[0001] The present disclosure relates to mobile communication, and more specifically, to a method and apparatus for performance monitoring for intelligent beam management in a mobile communication system. Background Technology

[0002] Along with the advancement of information and communication technology, various wireless communication technologies can be developed. Representative wireless communication technologies include LTE (long term evolution), NR (new radio), and 6G (6th Generation), which are defined in the 3GPP (3rd generation partnership project) standards. LTE can be one of the wireless communication technologies among 4G (4th Generation) wireless communication technologies, and NR can be one of the wireless communication technologies among 5G (5th Generation) wireless communication technologies.

[0003] To handle the surge in wireless data following the commercialization of 4G communication systems (e.g., communication systems supporting LTE), not only the frequency bands of 4G communication systems (e.g., frequency bands below 6 GHz) but also 5G communication systems (e.g., communication systems supporting NR) that use frequency bands higher than those of 4G communication systems (e.g., frequency bands above 6 GHz) may be considered. 5G communication systems can support eMBB (enhanced Mobile BroadBand), URLLC (Ultra-Reliable and Low Latency Communication), and mMTC (massive Machine Type Communication).

[0004] Such communication systems can be designed considering various scenarios, service requirements, and potential system compatibility. In particular, discussions regarding beam-based communication in 5G NR communication systems may be active in order to perform broadband communication in high frequency bands. Accordingly, beam-based communication can be continuously used. In addition, communication systems can improve performance by utilizing artificial intelligence (AI) and machine learning (ML).

[0005] The international standardization organization 3GPP is discussing the application of intelligent technologies (or AI / ML models) for air interfaces, establishing use cases where intelligent technologies can be utilized in air interfaces, and discussing the specification work required for the use of intelligent technologies for each use case. Here, beam management, positioning accuracy enhancement, and CSI (Channel State Information) feedback enhancement are being considered as representative use cases.

[0006] The lifecycle management of AI / ML models (or intelligent functions / models) is described below.

[0007] Since intelligent technology is based on training data, it must be possible to perform Life Cycle Management (LCM) for the creation and maintenance of intelligent functions / models in response to changes in training data. Therefore, if one intends to apply functions based on intelligent technology in a mobile communication system composed of base stations and / or terminals, as in the aforementioned use cases, the mobile communication system must be able to support LCM. In this regard, detailed steps of the LCM process being considered include data collection, model training, model inference, model deployment, model activation, model deactivation, model selection, model switching, model fallback, and model monitoring. For example, in a mobile communication system, a specific intelligent model may be created through model training based on data collection, intelligent model-based operations may be performed through inference using the model after the model deployment and activation processes are carried out, and the model may be managed through the model monitoring process.

[0008] Additionally, for convenience of explanation in this specification, the types of AI / ML models may be classified as follows depending on the location of the network node where the inference operation of the AI / ML model is performed.

[0009] - Model Type 1: One-sided AI / ML Model

[0010] The above model may include an AI / ML model in which inference is performed entirely at a terminal (or UE) or a network. Here, the model may be classified as a UE-sided AI / ML model when inference is performed entirely at the terminal, and as a Network-sided AI / ML model when it is performed only at the network.

[0011] - Model Type 2: Two-sided AI / ML Model

[0012] The above model may include paired AI / ML model(s) on which joint inference is performed. Here, joint inference may include AI / ML inference performed jointly across the terminal and the network.

[0013] For example, if the first part of the inference is performed by the terminal (UE-part of two-sided model) and the remaining part is performed by the base station (NW-part of two-sided model), or vice versa, the model may be included in a two-sided AI / ML model.

[0014] Prior to utilizing intelligent functions and models and performing lifecycle management in a mobile communication system, it is first necessary to identify the intelligent functions (Functionality identification) and models supported within the network. For example, a base station identifies which intelligent functions and models a terminal can support and, based on the identified functions and models, can direct the terminal to activate specific functions and models. In relation to the identification of such intelligent functions and models, the following two directions for lifecycle management are being discussed.

[0015] - Functionality-based Life Cycle Management (Functionality-based LCM)

[0016] - Model-ID based Lifecycle Management (Model-ID based LCM)

[0017] Model-ID-based lifecycle management may refer to a lifecycle management process in which a base station (or base station server) and a terminal (or terminal server) share intelligent model information along with a Model-ID in advance, and subsequently, the base station and the terminal identify and manage the intelligent model through the Model-ID, etc. Functionality-based lifecycle management may refer to a lifecycle management process in which a base station (or base station server) and a terminal (or terminal server) share function information for intelligent functions in advance, and subsequently, the base station and the terminal identify and manage intelligent functions using the function name or function ID, etc. Here, the two lifecycle management functions may be configured independently, or they may be configured in a mixed form.

[0018] Intelligent beam management or AI / ML-based beam management is described below.

[0019] Figure 1 is a diagram illustrating intelligent beam management or AI / ML-based beam management.

[0020] Referring to Fig. 1, as the frequency band used in cellular mobile communication systems increases, millimeter wave (mmWave) bands (e.g., 30 GHz to 300 GHz) may be used. Using high frequency bands results in shorter wavelengths and increased path attenuation, which reduces the signal reach. To address this, antenna technology is used to generate a directional beam directed toward a specific area to improve reach. The power amplification of the transmitted signal provided by the base station beam is generally inversely proportional to the width or area covered by the base station beam. Therefore, to overcome the deterioration of received signal quality in high frequency bands, the base station may form multiple transmit beams to cover the entire cell area. Each base station beam provides coverage for an area narrower than the total coverage area, but provides sufficient transmit power amplification to overcome higher signal propagation loss. Here, when using such directional beams, a selection process may be required among the multiple beams to determine the base station beam (or transmit beam) and terminal beam (or receive beam) that demonstrate optimal performance for the base station and terminal, respectively, in order to ensure stable link performance. To select the optimal beam, methods such as obtaining measurement results for all available beams or performing measurement results for multiple beams and selecting the optimal beam based on the implementation method of the base station or terminal may be considered. However, overhead occurs during the process of finding the optimal beam in this way, and frequent overhead can lead to performance degradation and terminal power consumption.

[0021] Accordingly, the introduction of intelligent technologies into the beam management process is being considered. Beam management techniques utilizing intelligent technologies can maintain accuracy while reducing the overhead associated with conventional beam measurement. In AI / ML or intelligent technology-based beam management, beam prediction—a use case where intelligent technologies are applied to predict beams of unobserved resources in the spatial or temporal domains—is being discussed.

[0022] Below, intelligent positioning accuracy improvement or AI / ML-based positioning accuracy improvement is described.

[0023] Figure 2 is a diagram illustrating intelligent positioning accuracy improvement or AI / ML-based positioning accuracy improvement.

[0024] Positioning refers to a technique for measuring the location of a specific terminal in a mobile communication system. 5G NR supports a method of transmitting a Positioning Reference Signal (PRS) to have the terminal report a Reference Signal Time Difference (RSTD), and then applying location measurement techniques such as Observed Time Difference Of Arrival (OTDOA). Recently, requirements for positioning accuracy in indoor environments have increased, and technologies to improve the accuracy of positioning measurements by applying intelligent technology are being discussed. In intelligent technology-based positioning, AI / ML assisted positioning, a use case that improves the accuracy of conventional positioning techniques by applying intelligent technology, and Direct AI / ML positioning, a use case that directly estimates the location of a terminal by applying intelligent technology, are being discussed.

[0025] Below, intelligent channel information feedback improvement or AI / ML-based CSI feedback improvement is described.

[0026] Channel State Information (CSI) feedback refers to a terminal reporting channel state information to support a base station in a mobile communication system in applying transmission techniques such as MIMO (Multiple Input Multiple Output) or precoding. 5G NR supports feedback information such as CQI (Channel Quality Indicator), PMI (Precoding Matrix Indicator), and RI (Rank Indicator) in relation to channel information feedback methods. In NR systems, discussions are ongoing regarding methods to improve CSI feedback techniques in order to effectively support transmission techniques such as MU-MIMO (Multi-user MIMO).

[0027] As one of the research directions for applying intelligent technology to channel information feedback, discussions are underway regarding channel compression (CSI compression) technology, which involves utilizing an auto-encoder—one of the intelligent techniques—to obtain a compressed latent representation of a MIMO channel. Figure 3 illustrates an example of the structure of an auto-encoder. Referring to Figure 3, an auto-encoder refers to a neural network structure that simply copies the input to the output, and is characterized by setting the number of neurons in the hidden layer between the encoder and the decoder to be smaller than that of the input layer to enable data compression (or dimensionality reduction). The problem to be solved

[0028] This specification proposes a performance monitoring method and apparatus for intelligent beam management in a mobile communication system. means of solving the problem

[0029] This specification proposes various embodiments related to a method and apparatus for performance monitoring of an intelligent model when utilizing intelligent beam management technology for wireless communication in a mobile communication system composed of a base station and one or more terminals.

[0030] According to one embodiment, in a mobile communication system composed of a base station and one or more terminals, a method is proposed to utilize a performance monitoring operation when the terminal is configured to use an intelligent beam management function. Specifically, a method for calculating performance indicators required during the performance monitoring process is proposed. Additionally, an operation process for linking a performance monitoring setting and an inference setting is proposed. Effects of the invention

[0031] By utilizing intelligent beam management technology that includes the performance monitoring method proposed in this specification, the network and the terminal can calculate performance indicators of supported intelligent functions or models and continuously track performance. Additionally, network performance can be improved by deactivating or switching an activated intelligent model when its performance deteriorates. Brief explanation of the drawing

[0032] Figure 1 is a diagram illustrating intelligent beam management or AI / ML-based beam management. Figure 2 is a diagram illustrating intelligent positioning accuracy improvement or AI / ML-based positioning accuracy improvement. Figure 3 illustrates an example of the structure of an autoencoder. Figure 4 illustrates an example of intelligent beam management. Figure 5 is a conceptual diagram of an example of the operation required when an intelligent model is located on a terminal. Figure 6 is a conceptual diagram of an example of operation when an intelligent model is located on the network side. Figure 7 illustrates examples of spatial domain beam prediction and temporal domain beam prediction. FIG. 8 illustrates an example of CSI information including performance indicators being transmitted. Figure 9 illustrates an example in which CSI information including performance indicators is transmitted during time domain beam prediction. FIG. 10 illustrates an example in which CSI information including monitoring window-based performance indicators is transmitted. FIG. 11 illustrates an example of a method for connecting a setting for performance monitoring and a setting for inference according to an embodiment of the present specification. FIG. 12 illustrates an example of a method for calculating and reporting performance indicators for time series beam forecasting and wide-to-narrow space beam forecasting according to one embodiment of the present specification. FIG. 13 illustrates an example of a structure in which a setting for one performance monitoring and a setting for a plurality of inference processes are connected according to one embodiment of the present specification. FIG. 14 is a flowchart of an example of a performance monitoring method for AI / ML-based beam management of a terminal according to one embodiment of the present specification. FIG. 15 is a conceptual diagram illustrating a first embodiment of a mobile communication system. FIG. 16 is a block diagram illustrating a first embodiment of a communication node constituting a communication system. Specific details for implementing the invention

[0033] The technical features, configurations, and proposed methods disclosed herein are subject to various modifications and may have various embodiments; therefore, specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the technical features, configurations, and proposed methods disclosed herein to specific embodiments, and it should be understood that they include all modifications, equivalents, and substitutions that fall within the scope of the technical features, configurations, and proposed methods disclosed herein. Similar reference numerals have been used for similar components in the description of each drawing.

[0034] Terms such as first, second, A, B, etc., may be used to describe various components, but said components should not be limited by said terms. These terms are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the rights of this specification, the first component may be named the second component, and similarly, the second component may be named the first component. The term “and / or” includes a combination of a plurality of related described items or any of a plurality of related described items.

[0035] In the embodiments of this specification, "at least one of A and B" may mean "at least one of A or B" or "at least one of one or more combinations of A and B". Additionally, in the embodiments of this specification, "at least one of A and B" may mean "at least one of A or B" or "at least one of one or more combinations of A and B".

[0036] In the embodiments of this specification, (re)transmission may mean “transmission,” “retransmission,” or “transmission and retransmission”; (re)setting may mean “setting,” “resetting,” or “setting and resetting”; (re)connection may mean “connection,” “reconnection,” or “connection and reconnection”; and (re)connection may mean “connection,” “reconnection,” or “connection and reconnection”.

[0037] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

[0038] The terms used herein are used merely to describe specific embodiments and are not intended to limit the technical features, configurations, or proposed methods disclosed herein. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as "comprising" or "having" are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0039] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the present invention pertains. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this specification.

[0040] Hereinafter, preferred embodiments of the technical features, configurations, and proposed methods disclosed in this specification will be described in more detail with reference to the attached drawings. In describing the technical features, configurations, and proposed methods disclosed in this specification, the same reference numerals are used for identical components in the drawings to facilitate overall understanding, and redundant descriptions of identical components are omitted.

[0041] The communication networks to which the embodiments disclosed herein are applied are not limited to those described below, and the embodiments disclosed herein may be applied to various communication networks. Here, the term "communication network" may be used interchangeably with "communication system."

[0042] Throughout the specification, a network may include, for example, wireless internet such as WiFi (wireless fidelity), mobile internet such as WiBro (wireless broadband internet) or WiMAX (world interoperability for microwave access), 2G mobile communication networks such as GSM (global system for mobile communication) or CDMA (code division multiple access), 3G mobile communication networks such as WCDMA (wideband code division multiple access) or CDMA2000, 3.5G mobile communication networks such as HSDPA (high speed downlink packet access) or HSUPA (high speed uplink packet access), 4G mobile communication networks such as LTE (long term evolution) networks or LTE-Advanced networks, and 5G mobile communication networks, and 6G mobile communication networks.

[0043] Throughout the specification, the terminal may be referred to as a terminal, access terminal, mobile terminal, station, subscriber station, mobile station, portable subscriber station, node, device, etc.

[0044] Here, a desktop computer, laptop computer, tablet PC, wireless phone, mobile phone, smartphone, smart watch, smart glass, e-book reader, PMP (portable multimedia player), portable game console, navigation device, digital camera, DMB (digital multimedia broadcasting) player, digital audio recorder, digital audio player, digital picture recorder, digital picture player, digital video recorder, digital video player, etc., capable of communicating with a terminal can be used.

[0045] Throughout the specification, base stations may be referred to as Node B, evolved Node B, base transceiver station (BTS), radio base station, radio transceiver, access point, access node, roadside unit (RSU), digital unit (DU), cloud digital unit (CDU), radio remote head (RRH), radio unit (RU), transmission point (TP), transmission and reception point (TRP), relay node, etc.

[0046] In the following, embodiments according to the present specification are described with reference to a 3GPP NR (New Radio) mobile communication system, and prior art documents defining the operation of a 3GPP NR mobile communication system may be referenced. The names of specific devices or specific signals / messages / fields described in the drawings are presented as examples, and therefore the technical features of the present specification are not limited to the specific names used in the drawings below.

[0047] Hereinafter, for the convenience of explanation, the performance monitoring method and apparatus for intelligent beam management proposed in this specification are mainly described from the perspective of the downlink of a wireless mobile communication system composed of a base station and a terminal; however, the proposed method of this specification can be extended and applied to any wireless mobile communication system composed of a transmitter and a receiver. Furthermore, the method proposed in this specification below is described from the perspective of use cases being discussed in 5G NR, such as an intelligent channel feedback enhancement method, an intelligent beam management method, and an intelligent positioning enhancement method; however, the proposed method of this specification can be extended and applied to other intelligent use cases.

[0048] Figure 4 illustrates an example of intelligent beam management.

[0049] In the inference process using intelligent beam management technology, measuring the performance of a specific beam may be required to select the optimal beam. For example, performance may include L1 (Layer 1)-RSRP (Reference Signal Received Power) or L1-SINR (Signal to Interference plus Noise Ratio) measured at the terminal. The measured beam performance can be input into an intelligent model (AI / ML model) to obtain outputs / results during the inference process. The results of the intelligent model may include the index of the beam with optimal performance, the predicted performance of all available beams (e.g., L1-RSRP), or information on the probability that the beam in question will be the best beam. An important aspect of the beam performance measurement process is determining the base station beam or beam pair (including both the base station beam and the terminal beam) to be used for the performance measurement. Using an exhaustive search method to obtain measurement results for all base station beams or beam pairs allows for the determination of the base station beam or beam pair exhibiting optimal performance in any situation, but it may result in wasted overhead during the measurement process. Generally, the resources used in the beam performance measurement process are Synchronization Signal Blocks (SSBs) or Channel State Information Reference Signals (CSI-RS), and beam performance measurements can be performed by transmitting a predetermined reference signal (RS). To balance the overhead of the measurement process with the performance of the finally selected beam, it may be necessary to reduce performance measurements for unnecessary beams. Intelligent models for beam selection can exist independently at both the base station and the terminal.If the intelligent model is located at the base station, an additional process may be required to transmit beam performance information (measurement report) measured by the terminal to the base station for use as input to the intelligent model. If the intelligent model is located at the terminal, a process may be required to transmit the results generated from the intelligent model's output (inference report) to the base station.

[0050] The intelligent model performing intelligent beam management technology may be located on the terminal side (UE-side) or on the network side (NW-side). Depending on the location of the intelligent model, the operations that the base station and the terminal must perform to support intelligent beam management technology may differ. The operations required for intelligent beam management may include LCM operations such as data collection (for training), inference, and model monitoring (performance monitoring). As used in this specification, the term “model monitoring” is used interchangeably with the term “performance monitoring” as it is intended to measure the performance of the intelligent model.

[0051] Figure 5 is a conceptual diagram of an example of the operation required when an intelligent model is located on a terminal.

[0052] For intelligent beam management technology to operate, it is essential for the base station to transmit a reference signal for beam measurement and for the terminal to measure that reference signal. When the intelligent model is located on the terminal side, the terminal is the entity performing the measurement; therefore, it may not transmit the beam measurement results to the base station during the data collection and inference processes for learning. However, even if the measurement results are not transmitted during the inference process, the terminal can still transmit the predicted results using the intelligent model to the base station. Additionally, when performing model monitoring operations, the calculation of performance metrics may be required. When the terminal performs the calculation of performance metrics, it can calculate them based on the measurement results. Furthermore, a process of transmitting the performance metrics calculated by the terminal to the base station may be necessary. If the performance metric calculation is performed within the network, it may be necessary for the terminal to transmit the beam measurement information and the output information of the intelligent model required for the calculation to the base station.

[0053] Figure 6 is a conceptual diagram of an example of operation when an intelligent model is located on the network side.

[0054] When an intelligent model is located on the network side, the terminal can report / transmit measurement results to the base station in response to a reference signal transmitted by the base station. The base station receives the measurement report from the terminal and can use it in the data collection or inference process for learning. Additionally, the calculation of performance indicators for performing model monitoring operations can be performed at the base station. To this end, if the terminal requires measurement information, it can transmit such information to the base station.

[0055] Figure 7 illustrates examples of spatial domain beam prediction and temporal domain beam prediction.

[0056] Intelligent beam management technology can be divided into spatial-domain beam prediction (in this specification, spatial-domain beam prediction is referred to as BM-case1) and temporal-domain beam prediction (in this specification, temporal-domain beam prediction is referred to as BM-case2).

[0057] Spatial domain beam prediction involves selecting the optimal beam for a specific point in time using beam performance results measured at that point. Spatial domain beam prediction can be subdivided into two categories based on the relationship between Set B, which is the set of beams measured during the inference process used as input to an intelligent model, and Set A, which is the output of the intelligent model and represents the entire predictable set of beams. The first case is when Set B is a subset of Set A, where the beams in Set B used during the inference process may consist of a portion of the entire set of available beams. The second case is when Set B is not a subset of Set A. For example, one might consider a scenario where Set B has the form of a wide beam, such as a Synchronization Signal Block (SSB), while Set A has the form of a narrow beam, such as CSI-RS. In other words, one might consider a case where the measurement results of a wide beam are used as input to the intelligent model, and a narrow beam is predicted to be the output of the intelligent model.

[0058] Time-domain beam prediction involves selecting the optimal beam for the present or future time point using beam performance results measured at the present and past time points. In the case of time-domain beam prediction, since intelligent models can predict the optimal beam for multiple time points, a single CSI report may include inference results for one or more future times.

[0059] The methods proposed in this specification are described below.

[0060] [Proposed Plan 01]

[0061] To perform performance monitoring for intelligent beam management technology, it is necessary to define performance metrics that represent the performance of the intelligent model. For example, beam prediction accuracy and L1-RSRP difference can be used as performance metrics for intelligent beam management. Regarding beam prediction accuracy, Top-1 or Top-K / 1 beam prediction accuracy may be considered. Top-1 prediction accuracy refers to the probability that the single best-performing beam predicted by the intelligent model matches the actual best-performing beam. Top-K / 1 prediction accuracy refers to the probability that the actual best-performing beam is included among the K beams predicted by the intelligent model for good performance. When using an intelligent model that predicts L1-RSRP values, the difference between the predicted L1-RSRP value and the measured L1-RSRP value can be used as a performance metric. Here, the L1-RSRP difference may be the difference value for the single best-performing or K beams, or it may be the average value of the difference values ​​for all measured beams.

[0062] To calculate the aforementioned performance indicators, both measurement results for the beams included in Set B, which corresponds to the input of the intelligent model, and measurement results for the beams included in Set A, which corresponds to the output of the intelligent model, may be required. To measure beam performance, the terminal may receive resource set information containing resource information composed of the relevant beams from the base station. Generally, the transmission of such resource set information utilizes the CSI framework and can be transmitted via Radio Resource Control (RRC) signals, etc., by being included in CSI-related configuration information such as CSI-ReportConfig Information Elements (IE). In this case, since the Set B beam set information is used as input to the intelligent model during the inference process, it can be delivered by being included in CSI-related configuration information, such as CSI Report configuration for inference operations. On the other hand, since the Set A beam set information is not used during the inference process but is required during the performance monitoring process to calculate performance indicators, it can be included in CSI-related configuration information, such as CSI Report configuration for performance monitoring operations. In this case, the included resource set information of Set A may include the entire Set A or, if necessary, a part of Set A.

[0063] Performance indicators calculated through performance measurements of beams included in Set A and Set B can be included in a CSI report and transmitted to the base station. Information regarding the CSI report containing performance indicators (e.g., report resource cycle, timing, etc.) may be included in the CSI report settings for performance monitoring or in the CSI report settings for inference operations. The timeline of the CSI report containing performance indicators may occur after the Set B signals are measured first and then the Set A signals are measured, or after the Set A signals are measured first and then the Set B signals are measured. FIG. 8 illustrates an example of CSI information including performance indicators being transmitted.

[0064] In the case of time domain beam prediction, Set B signals corresponding to past beam measurement results may be transmitted, followed by Set A signals corresponding to future beam prediction results. Therefore, a CSI report including performance indicators may occur after the measurement of Set A signals. However, if the transmission of Set A and Set B signals is set to be repeated at specific intervals or transmitted non-periodically, a CSI report including performance indicators may occur / transmit after the measurement of Set B signals as needed. Figure 9 illustrates an example of CSI information including performance indicators being transmitted during time domain beam prediction.

[0065] In the process of reporting performance indicators, reporting performance indicator measurements through a single calculation can cause frequent overhead issues and may actually hinder stable performance monitoring operations. Therefore, a method of reporting performance indicators measured during a monitoring window that includes a preset period or number of times may be considered. FIG. 10 illustrates an example of CSI information including monitoring window-based performance indicators being transmitted.

[0066] Information for the monitoring window can be included in the CSI reporting settings for performance monitoring operations. For example, if time information is included in the information for the monitoring window, the minimum time size of the monitoring window (in units such as milliseconds) may be included. The terminal can calculate or infer how many performance metric calculations are possible within a single monitoring window through this information and the reference signal transmission interval. As another example, a direct value indicating the minimum number of performance metric calculations required within a single monitoring window may be included in the information for the monitoring window.

[0067] When beam prediction accuracy is used as a performance metric, the performance metric included in the CSI report may take the following form. For example, assuming that N performance metric calculations are required within the aforementioned single monitoring window, beam prediction accuracy can be expressed as N_s / N, using the count (N_s) indicating how many predictions were successful in those N calculations. Here, when considering the Top-1 beam, the prediction may be considered successful if the beam with the best performance predicted using the Set B measurement results as input to the intelligent model is identical to the actual Set A measurement result. Additionally, when considering the Top-K beam, the prediction may be considered successful if the beam with the best performance in the actual Set A measurement result is included among the K high-performance beams predicted using the Set B measurement results as input to the intelligent model. Regarding the method of conveying the performance metric, methods such as including the number of successes N_s in the CSI report or quantizing the success probability N_s / N into a specific interval and conveying it may be considered.

[0068] The calculation of the aforementioned performance indicators assumes that Set A for performance monitoring is the entire Set A. However, in the intelligent beam management operation, if the number of beams in the beam set is large, transmitting the entire Set A during the performance monitoring process at regular intervals or as needed can result in significant overhead. Therefore, a method of transmitting only a subset of the entire Set A beam set and using it to perform the performance monitoring process may be considered.

[0069] However, in this case, the accuracy of the intelligent model's performance estimation may decrease compared to when the entire Set A beam set is used. Therefore, additional explanation regarding performance metrics may be required when using a subset of Set A. For example, if beam prediction accuracy is used as a performance metric, the Top-1 or Top-K beams predicted to have good performance using measurement results from Set B as input to the intelligent model may not be included in the beam set corresponding to the part of Set A used for performance monitoring.

[0070] In such cases, to determine the success of beam prediction, which is a performance metric, R (Number of beams in Set A / Number of beams in subset of Set A), which is the ratio of the total number of beams in Set A to the number of beams in the subset of Set A used for performance monitoring, may be defined and used. Regarding beam prediction, the performance metric can be measured only when K high-performance beams are predicted and at least an integer number of beams greater than or equal to K / R is included in the subset of Set A used for performance monitoring. In this case, if the best-performing beam in the measurement results using the subset of Set A is included among the predicted K beams, the prediction may be considered successful. In other words, to fairly evaluate performance, the performance metric measurement result may be used if at least K / R of the K beams predicted by the intelligent model are included in the subset of Set A used for performance monitoring; however, if fewer than K / R beams are included in the subset of Set A used for performance monitoring, the performance metric measurement result may not be used. Through this process, N' valid performance indicator calculations can be determined within the monitoring window. Depending on the number of successful predictions (N_s) during the N' calculation process, the performance indicator can be expressed in a form such as N_s / N'. Regarding the delivery of the performance indicator, the number of successes N_s and the number of valid measurements N' may be included in the CSI report, or the result of the success probability N_s / N' being quantized into a certain interval may be delivered.

[0071] [Proposed Plan 02]

[0072] [Proposed Method 01] assumes the case where beam prediction accuracy is used as a performance indicator. When the L1-RSRP difference is used as a performance indicator, the specific form of the performance indicator is as follows. For example, if an intelligent model can predict L1-RSRP values, the difference between the L1-RSRP value of the predicted Top-1 beam or Top-K beam, obtained by using the Set B measurement results as input to the intelligent model, and the L1-RSRP value of the same beam obtained through the actual measurement results for Set A during the performance monitoring process can be used as a performance indicator.

[0073] Assuming that N performance metric calculations are required within a single monitoring window, the L1-RSRP difference can be expressed as the average of the L1-RSRP difference values ​​calculated over those N calculations. For the Top-1 beam, the L1-RSRP difference between the predicted value and the actual measured value for the corresponding single beam can be calculated in a single performance metric calculation. For the Top-K beam, the L1-RSRP difference between the predicted value and the actual measured value can be calculated for the K predicted beams. These difference values ​​can be used by calculating the average for the K beams. The calculated performance metric can be conveyed by including the average L1-RSRP difference in the CSI report, or by quantizing the average L1-RSRP difference into specific intervals.

[0074] When using only a subset of Set A during the performance monitoring process, the Top-1 or Top-K beams predicted to have good performance—based on the measurement results for Set B used as input for an intelligent model—may not be included in the beam set corresponding to the part of Set A used for performance monitoring. In such cases, to calculate the L1-RSRP difference, a performance metric, R (Number of beams in Set A / Number of beams in subset of Set A), which is the ratio of the total number of beams in Set A to the number of beams in the subset of Set A used for performance monitoring, can be defined and used. The performance metric can be calculated and measured only when, when K high-performance beams are predicted, a number of beams corresponding to an integer greater than or equal to K / R is included in the subset of Set A used for performance monitoring.

[0075] In this case, the L1-RSRP difference can be calculated only for the Top K beams (or a number of beams corresponding to an integer greater than or equal to K / R) included in the subset of Set A used for performance monitoring. At this time, the average of these values ​​can be calculated and used. In other words, to fairly evaluate performance, if at least K / R of the K beams predicted by the intelligent model are included in the subset of Set A used for performance monitoring, the corresponding performance metric measurement result can be used; if fewer than K / R of the beams are included in the subset of Set A used for performance monitoring, the corresponding performance metric measurement result may not be used. Through this process, N' valid measurements out of N performance metric calculations within the monitoring window can be determined. The average L1-RSRP difference for these N' calculation processes can be calculated. The performance metric can be conveyed in the CSI report as the number of successful valid measurements N' and the average L1-RSRP difference, or the result of the average L1-RSRP difference quantized into a specific range can be conveyed.

[0076] If the intelligent model cannot predict L1-RSRP values ​​and only predicts the optimal beam index, the performance metric for the L1-RSRP difference may be defined differently from the above. That is, the alternative L1-RSRP difference can be used as a performance metric by measuring the L1-RSRP values ​​for the predicted Top-1 or Top-K beams during the performance monitoring process using the measurement results for Set B as input to the intelligent model, and comparing them with the single largest L1-RSRP value or the K largest L1-RSRP values ​​among the L1-RSRP measurement results corresponding to the entire Set A. In other words, the aforementioned L1-RSRP difference is the difference between the predicted L1-RSRP value and the actual measured L1-RSRP value for the same beam (or beam index), whereas the alternative L1-RSRP difference may be the difference between the actual measured L1-RSRP value for the predicted beam (or beam index) and the actual measured L1-RSRP value with the highest value among the entire Set A. Therefore, in the latter case, an L1-RSRP comparison is performed for different beams (or beam indices).

[0077] Assuming that N performance metric calculations are required within a single monitoring window, as described above, the L1-RSRP difference can be expressed as the average of the L1-RSRP difference values ​​calculated in those N calculations. For the Top-1 beam, the L1-RSRP difference between the measured value for a single predicted beam and the highest measured value in Set A can be calculated in a single performance metric calculation. For the Top-K beam, the L1-RSRP difference between the measured values ​​for K predicted beams and the K highest measured values ​​in Set A can be calculated. These values ​​can be used by calculating the average for the K beams. The calculated performance metric can be transmitted either by including the average L1-RSRP difference in the CSI report or by quantizing the average L1-RSRP difference into specific intervals.

[0078] When using only a subset of the beam set in Set A instead of the entire set during the performance monitoring process, R (Number of beams in Set A / Number of beams in subset of Set A), which is the ratio of the total number of beams in Set A to the number of beams in the subset of Set A used for performance monitoring, can be defined and used. Here, when K high-performance beams are predicted, a performance metric can be measured only if the number of beams included in the subset of Set A used for performance monitoring is an integer greater than or equal to K / R. In this case, the L1-RSRP difference can be calculated only for the top K beams (or the number of beams corresponding to an integer greater than or equal to K / R) included in the subset of Set A used for performance monitoring. These values ​​can be used by calculating the average.

[0079] [Proposed Plan 03]

[0080] In the performance monitoring process, since information related to the inputs and outputs of the intelligent model is required to calculate performance indicators, measurement results for both Set B and Set A may be necessary. To measure beam performance, the terminal may receive resource set information containing resource details for the corresponding beams from the base station. This process can be carried out using the CSI framework, included in CSI-related configuration information such as CSI-ReportConfig IE (Information Element), and transmitted via RRC (Radio Resource Control) signals. In this case, since the Set B beam set information is used as input to the intelligent model during its inference process, it may be included in the CSI reporting settings for the inference operation and transmitted. On the other hand, since the Set A beam set information is not used during the inference process but is required during the performance monitoring process to calculate performance indicators, it may be included in the CSI reporting settings for the performance monitoring operation. Therefore, as information for Set A and Set B may be included in different CSI reporting settings, it may be necessary to link them to calculate performance indicators when multiple reporting settings exist.

[0081] For example, a method of adding a linkage ID within CSI-ReportConfig can be considered. A configuration for performance monitoring and a configuration for the inference process having the same linkage ID can be considered to be linked to each other. FIG. 11 illustrates an example of a method for linking a configuration for performance monitoring and a configuration for inference according to an embodiment of the present specification. As another example, a method of linking a configuration for performance monitoring and a configuration for the inference process by reusing an existing CSI-ReportConfigID within CSI-ReportConfig without introducing a new linkage ID can be considered.

[0082] Performance indicators calculated based on the connected settings can be included in a CSI report and transmitted from the terminal to the base station. In this process, multiple performance indicators, not just a single one, may be included and transmitted within the CSI report. FIG. 12 illustrates an example of a method for calculating and reporting performance indicators for time-series beam prediction and wide-to-narrow spatial beam prediction according to an embodiment of the present specification. For example, performance indicators for different beam prediction use cases may be included within a single CSI report. Specifically, performance monitoring for time-series (Temporal) beam prediction and wide-to-narrow spatial beam prediction, which predicts a narrow beam based on a wide beam in the spatial domain, may be performed simultaneously. In this case, Set A may have the same beam set, but Set B may have a different beam set depending on the use case. Therefore, when performing a CSI report according to the reporting settings for performance monitoring, performance indicators for the two use cases may be included.

[0083] To distinguish performance metrics for two use cases included in a single CSI report, the aforementioned linkage ID may be included and used within the CSI report. Thus, a structure in which a configuration for one performance monitoring and a configuration for multiple inference processes are linked may be created. Conversely, a structure in which a configuration for one inference process and a configuration for multiple performance monitoring are linked may also be created. Thus, the configuration for performance monitoring and the configuration for inference processes may have a multiple-to-multiple linkage structure. FIG. 13 illustrates an example of a structure in which a configuration for one performance monitoring and a configuration for multiple inference processes are linked according to an embodiment of the present specification.

[0084] [Proposed Plan 04]

[0085] As mentioned above, only a subset of Set A, rather than the entire Set A, may be used during the performance monitoring process. In this case, the subset of Set A used for performance monitoring is not determined based on explicit rules but can be determined by configuring resources according to the needs of the base station. Since the beam set for performance monitoring can be utilized identically by multiple terminals using intelligent beam management operations within the base station rather than by a single terminal, it may be difficult to configure it in a form optimized for a specific terminal. Additionally, if necessary, the base station may use methods such as randomly selecting a subset of Set A by rotating through the entire Set A (like Round Robin), or determining the subset of Set A by quantizing beams heading in similar directions to reduce the number of beams. However, since the beam configuration of the entire Set A may vary depending on the hardware structure of the base station, it may be impossible to enforce it using simple and specific rules. Therefore, the selection of the subset of Set A for performance monitoring from the entire Set A can be left to the autonomy and / or implementation of the base station.

[0086] However, it is possible for the terminal to transmit some information to the base station to assist in selecting a subset of Set A to be used for base station performance monitoring. The base station may refer to the information transmitted by the terminal to determine the beam set for performance monitoring as necessary. As previously mentioned, when using a subset for performance monitoring, valid performance indicators can only be calculated if the subset contains a certain number of beams corresponding to the Top-K predicted by the intelligent model. Therefore, the terminal may transmit beam set information consisting of M preferred beams to the base station by accumulating the results predicted by the Top-K beams during the existing inference process. Conversely, the terminal may also transmit beam set information consisting of M non-preferred beams to the base station.

[0087] Alternatively, instead of the terminal directly transmitting information about its preferred beam to the base station, a method may be considered in which it conveys whether the value of R (Number of beams in Set A / Number of beams in subset of Set A), which is set for current performance monitoring, is sufficient. If it is determined that the size of the subset is too small to calculate valid performance metrics, the terminal may convey to the base station information that the current R value should be increased or convey a preferred R value. Conversely, if the terminal determines that the size of the subset is too large, it may convey to the base station information that the current R value should be decreased or convey a preferred R value.

[0088] When using a subset of Set A instead of the entire Set A during the performance monitoring process, a method may be required for the resources (or beams) defined in the resource set(s) for that performance monitoring to be connected to the resources belonging to the entire Set A. In other words, the terminal must be able to identify which resources among the beam resources of the entire Set A were selected to form the beam set. To this end, the terminal can distinguish beams using the TCI state possessed by each beam resource. Resources used as reference signals may include information indicating the TCI state (e.g., TCI-state information), and this information indicating the TCI state may include Quasi-CoLocation (QCL) information. Through QCL information, the terminal can recognize and distinguish the directionality of the corresponding beam resource.

[0089] [Proposed Plan 05]

[0090] In the aforementioned proposed method, the calculated performance indicator may be transmitted from the terminal to the base station whenever the performance indicator is calculated within a specific monitoring window. Alternatively, the terminal may transmit the calculated performance indicator to the base station only when a specific event occurs while monitoring the calculated performance indicator. Here, for example, the specific event may include cases where the calculated performance indicator fails to satisfy a preset performance condition more than a certain number of times. In other words, the terminal sets a counter and increments the counter when the calculated performance indicator falls below the expected performance indicator. When the counter reaches a certain number of times, a performance monitoring event occurs. At this time, the performance condition for the performance indicator may be set by the base station to the terminal via an RRC signal or selected by the terminal itself.

[0091] When beam prediction accuracy is used as a performance metric, Top-1 or Top-K beam prediction accuracy can be calculated in the form of probabilities within the monitoring window, as described in [Proposed Method 1]. The terminal can increment a counter if the calculated performance metric is smaller than a preset threshold value. When the count of the counter exceeds a preset count condition, an event occurs, and the terminal can report the result to the base station. During this process, the counter may be reset when an event occurs, or reset or decremented at regular intervals.

[0092] When using L1-RSRP difference as a performance metric, the top-1 or top-K L1-RSRP difference values ​​can be calculated within the monitoring window as described in [Proposed Method 2]. If the calculated performance metric is greater than a preset threshold value, the terminal increments the counter. When the counter count exceeds a preset count condition, an event occurs, and the terminal can report the result to the base station. During this process, the counter may be reset when an event occurs, or reset or decremented at regular intervals.

[0093] [Proposed Plan 06]

[0094] In relation to the aforementioned proposed methods, information regarding both Set B, used as the input to the intelligent model, and Set A, used as the output of the intelligent model, may be required to calculate performance indicators. As another method for performance monitoring, a method may be considered to calculate the performance indicators of the intelligent model and perform performance monitoring operations using only the measurement results of Set B, without measuring Set A.

[0095] If the intelligent model used for intelligent beam management is capable of outputting predicted L1-RSRP values, the terminal can obtain predicted L1-RSRP values ​​for the entire Set A by using the measured values ​​for Set B as input to the intelligent model. Unless Set B is a use case such as a wide-to-narrow configuration that is completely exclusive to Set A, the output predicted L1-RSRP values ​​may include values ​​belonging to the Set B beam set. Therefore, a performance metric can be calculated using the difference between the predicted L1-RSRP value and the actual measured L1-RSRP value for the corresponding Set B beam. A smaller difference in the calculated L1-RSRP values ​​may indicate better performance of the current intelligent model.

[0096] In addition, to verify the performance of the intelligent model, a portion of the beams of Set B, rather than the entire Set B, may be used as input, and a performance metric may be calculated using the difference between the predicted L1-RSRP value corresponding to the unused Set B beams and the actual measured L1-RSRP value. For example, assuming that Set B consists of 8 beams, only 6 beams are used as input to the intelligent model to calculate the L1-RSRP output value of the intelligent model. Then, the L1-RSRP difference between the predicted value and the actual measured value for 2 of the unused Set B beams can be calculated.

[0097] The L1-RSRP difference calculated in this proposed method can be used by calculating the average based on the number of beams. The calculated performance indicator can be transmitted by including the average L1-RSRP difference in the CSI report or by quantizing the average L1-RSRP difference into specific intervals.

[0098] The above [Proposed Plan 01] through [Proposed Plan 06] may be applied together to the extent that they do not conflict with other proposed plan(s) of this specification. Each of the multiple proposed plans described so far is a mutually related plan rather than an independent one; for example, a technical component belonging to one proposed plan may be combined with a technical component belonging to another proposed plan to form a new proposed plan, and such a new proposed plan is also included in the various embodiments proposed in this specification. In this way, a technical component selected from one proposed plan and a technical component selected from another proposed plan may be combined to form the following embodiments.

[0099] FIG. 14 is a flowchart of an example of a performance monitoring method for AI / ML-based beam management of a terminal according to one embodiment of the present specification.

[0100] Referring to FIG. 14, the terminal receives beam set information from a base station (S1410). Here, the beam set information may include at least one of first information about a first beam set inferred based on an AI / ML model used for AI / ML-based beam management and second information about a second beam set used as input information for inferring the first beam set.

[0101] The terminal acquires measurement information based on the beam set information (S1420). Here, the measurement information may be generated / acquired by the terminal performing a measurement on at least one of the first beam set and the second beam set based on resource information included in the beam set information. For example, if the beam set information includes the first information and the second information, and the first information includes first resource information for the first beam set and the second information includes second resource information for the second beam set, the terminal may perform a measurement on the first beam set based on the first resource information and perform a measurement on the second beam set based on the second resource information. Subsequently, the terminal may generate / acquire measurement information including a measurement result for the first beam set and a measurement result for the second beam set.

[0102] Alternatively, if the beam set information includes one of the first information and the second information, and the one information includes resource information for a corresponding beam set (i.e., a beam set for one of the first information and the second information), the terminal may perform a measurement on the corresponding beam set based on the resource information and generate / acquire measurement information including a measurement result for the corresponding beam set. In this case, the aforementioned proposed method 06 may be applied.

[0103] Alternatively, to determine whether beam prediction is successful as a performance indicator, a ratio R can be defined and used between the number of beams in the first beam set and the number of beams in the subset of the first beam set used for performance monitoring. Regarding beam prediction, when K high-performance beams are predicted, a performance indicator can be measured only if a number of beams corresponding to an integer greater than or equal to K / R is included in a part of the first beam set used for performance monitoring. In this case, if the best-performing beam in the measurement result using a part of the first beam set is included among the predicted K beams, the prediction may be successful. In this regard, at least some of the aforementioned proposed methods 01 to 06 may be applied.

[0104] The terminal obtains performance indicators for the AI / ML-based beam management based on the measurement information (S1430). Here, regarding the generation / acquisition of the performance indicators, at least some of the aforementioned proposed methods 01 to 06 may be applied.

[0105] The terminal transmits the performance indicator to the base station (S1440). Here, at least some of the aforementioned proposed methods 01 to 06 may be applied in step S1440.

[0106] FIG. 15 is a conceptual diagram illustrating a first embodiment of a mobile communication system.

[0107] Referring to FIG. 15, the communication system (100) may include a plurality of communication nodes (110-1, 110-2, 110-3, 120-1, 120-2, 130-1, 130-2, 130-3, 130-4, 130-5, 130-6). The plurality of communication nodes may support 4G communication (e.g., LTE (long term evolution), LTE-A (advanced)), 5G communication (e.g., NR (new radio)), etc., as defined in the 3GPP (3rd generation partnership project) standard. 4G communication may be performed in a frequency band of 6 GHz or lower, and 5G communication may be performed not only in a frequency band of 6 GHz or lower but also in a frequency band of 6 GHz or higher.

[0108] For example, for 4G communication and 5G communication, multiple communication nodes can support communication protocols based on CDMA (code division multiple access), WCDMA (wideband CDMA), TDMA (time division multiple access), FDMA (frequency division multiple access), OFDM (orthogonal frequency division multiplexing), Filtered OFDM, CP (cyclic prefix)-OFDM, DFT-s-OFDM (discrete Fourier transform-spread-OFDM), OFDMA (orthogonal frequency division multiple access), SC (single carrier)-FDMA, NOMA (Non-orthogonal Multiple Access), GFDM (generalized frequency division multiplexing), FBMC (filter bank multi-carrier) based communication protocol, UFMC (universal filtered multi-carrier) based communication protocol, SDMA (Space Division Multiple Access) based communication protocol, etc.

[0109] Additionally, the communication system (100) may further include a core network. If the communication system (100) supports 4G communication, the core network may include an S-GW (serving-gateway), a P-GW (PDN (packet data network)-gateway), an MME (mobility management entity), etc. If the communication system (100) supports 5G communication, the core network may include a UPF (user plane function), an SMF (session management function), an AMF (access and mobility management function), etc.

[0110] Meanwhile, each of the multiple communication nodes (110-1, 110-2, 110-3, 120-1, 120-2, 130-1, 130-2, 130-3, 130-4, 130-5, 130-6) constituting the communication system (100) may have the following structure. In addition, as an example, the communication system (100) described above can be applied not only to 5G communication but also to subsequent next-generation communication systems (e.g., 6G), and is not limited to a specific form.

[0111] FIG. 16 is a block diagram illustrating a first embodiment of a communication node constituting a communication system.

[0112] Referring to FIG. 16, the communication node (200) may include at least one processor (210), a memory (220), and a transceiver (230) that is connected to a network to perform communication. Additionally, the communication node (200) may further include an input interface device (240), an output interface device (250), a storage device (260), etc. Each component included in the communication node (200) may be connected by a bus (270) to communicate with one another.

[0113] However, each component included in the communication node (200) may be connected via individual interfaces or individual buses centered around the processor (210), rather than via a common bus (270). For example, the processor (210) may be connected via a dedicated interface to at least one of a memory (220), a transmission / reception device (230), an input interface device (240), an output interface device (250), and a storage device (260).

[0114] The processor (210) can execute a program command stored in at least one of the memory (220) and the storage device (260). The processor (210) may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to embodiments of the present invention are performed. Each of the memory (220) and the storage device (260) may be composed of at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory (220) may be composed of at least one of read-only memory (ROM) and random access memory (RAM).

[0115] Referring again to FIG. 15, the communication system (100) may include a plurality of base stations (110-1, 110-2, 110-3, 120-1, 120-2) and a plurality of terminals (130-1, 130-2, 130-3, 130-4, 130-5, 130-6). The communication system (100) including the base stations (110-1, 110-2, 110-3, 120-1, 120-2) and terminals (130-1, 130-2, 130-3, 130-4, 130-5, 130-6) may be referred to as an "access network". Each of the first base station (110-1), the second base station (110-2), and the third base station (110-3) can form a macro cell. Each of the fourth base station (120-1) and the fifth base station (120-2) can form a small cell. The fourth base station (120-1), the third terminal (130-3), and the fourth terminal (130-4) may be located within the cell coverage of the first base station (110-1). The second terminal (130-2), the fourth terminal (130-4), and the fifth terminal (130-5) may be located within the cell coverage of the second base station (110-2). The fifth base station (120-2), the fourth terminal (130-4), the fifth terminal (130-5), and the sixth terminal (130-6) may be located within the cell coverage of the third base station (110-3). The first terminal (130-1) may be located within the cell coverage of the fourth base station (120-1). The sixth terminal (130-6) may be located within the cell coverage of the fifth base station (120-2).

[0116] Here, each of the plurality of base stations (110-1, 110-2, 110-3, 120-1, 120-2) may be referred to as Node B, evolved Node B, gNB, base transceiver station (BTS), radio base station, radio transceiver, access point, access node, etc. Each of the plurality of terminals (130-1, 130-2, 130-3, 130-4, 130-5, 130-6) may be referred to as User Equipment (UE), terminal, access terminal, mobile terminal, station, subscriber station, mobile station, portable subscriber station, node, device, etc.

[0117] Meanwhile, each of the multiple base stations (110-1, 110-2, 110-3, 120-1, 120-2) may operate in different frequency bands or in the same frequency band. Each of the multiple base stations (110-1, 110-2, 110-3, 120-1, 120-2) may be connected to each other via an ideal backhaul link or a non-ideal backhaul link, and may exchange information with each other via an ideal backhaul link or a non-ideal backhaul link. Each of the multiple base stations (110-1, 110-2, 110-3, 120-1, 120-2) may be connected to a core network via an ideal backhaul link or a non-ideal backhaul link. Each of the multiple base stations (110-1, 110-2, 110-3, 120-1, 120-2) can transmit a signal received from the core network to the corresponding terminal (130-1, 130-2, 130-3, 130-4, 130-5, 130-6), and can transmit a signal received from the corresponding terminal (130-1, 130-2, 130-3, 130-4, 130-5, 130-6) to the core network.

[0118] Various embodiments of the present invention may be implemented by hardware, firmware, software, or a combination thereof. In the case of implementation by hardware, it may be implemented by one or more ASICs (Application Specific Integrated Circuits), DSPs (Digital Signal Processors), DSPDs (Digital Signal Processing Devices), PLDs (Programmable Logic Devices), FPGAs (Field Programmable Gate Arrays), general processors, controllers, microcontrollers, microprocessors, etc.

[0119] The scope of the present invention includes software or machine-executable instructions (e.g., operating systems, applications, firmware, programs, etc.) that enable operations according to the methods of various embodiments to be executed on a device or computer, and a non-transitory computer-readable medium on which such software or instructions, etc. are stored and which are executable on a device or computer. Examples of computer-readable media include hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, etc. Examples of program instructions include machine code, such as that produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The hardware devices described above may be configured to operate as at least one software module to perform the operations of the present invention, and vice versa.

[0120] The methods according to the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. A computer-readable medium may include program instructions, data files, data structures, etc., either individually or in combination. The program instructions recorded on the computer-readable medium may be those specifically designed and configured for the present invention, or they may be those known and available to those skilled in the art of computer software. The operation of the method according to an embodiment of the present invention may be implemented as a computer-readable program or code on a computer-readable recording medium. A computer-readable recording medium includes any type of recording device in which information that can be read by a computer system is stored. Additionally, the computer-readable recording medium may be distributed across networked computer systems, allowing computer-readable programs or code to be stored and executed in a distributed manner.

[0121] Some aspects of the invention have been described in the context of a device, but may also be described according to a corresponding method, wherein a block or device corresponds to a method step or a feature of a method step. Similarly, aspects described in the context of a method may also be described according to a corresponding block or item or a feature of a corresponding device. Some or all of the method steps may be performed by (or using) a hardware device, such as, for example, a microprocessor, a programmable computer, or an electronic circuit. In some embodiments, at least one of the most important method steps may be performed by such a device.

[0122] In the embodiments, a programmable logic device (e.g., a field-programmable gate array) may be used to perform some or all of the functions of the methods described herein. In the embodiments, the field-programmable gate array may operate with a microprocessor to perform one of the methods described herein. Generally, it is preferable that the methods be performed by some hardware device.

[0123] The exemplary methods of the present invention are described as a series of operations for clarity of description, but this is not intended to limit the order in which the steps are performed, and if necessary, each step may be performed simultaneously or in a different order. To implement the method according to the present invention, additional steps may be included in addition to the steps exemplified, steps excluding some steps and including the remaining steps, or steps excluding some steps and including additional steps.

[0124] The various embodiments of the present invention are not intended to list all possible combinations but to explain representative aspects of the invention, and the matters described in the various embodiments may be applied independently or in combination of two or more.

[0125] Although the present invention has been described with reference to preferred embodiments, those skilled in the art will understand that various modifications and changes can be made to the invention without departing from the spirit and scope of the invention as described in the following claims.

Claims

Claim 1 A method performed by a terminal in a wireless communication system, wherein beam set information is received from a base station, the beam set information includes at least one of first information regarding a first beam set inferred based on an AI / ML model used for AI / ML (Artificial Intelligence / Machine Learning)-based beam management and second information regarding a second beam set used as input information for inferring the first beam set; measurement information is obtained based on the beam set information, wherein the measurement information is obtained by the terminal performing a measurement on at least one of the first beam set and the second beam set based on resource information included in the beam set information; and a performance indicator for the AI / ML-based beam management is obtained based on the measurement information, wherein the performance indicator includes at least one of beam prediction accuracy and L1-RSRP difference (Layer 1-Reference Signal Received Power Difference); and the performance indicator is transmitted to the base station. Claim 2 A method according to claim 1, wherein the beam set information comprises the first information and the second information, and the first information comprises first resource information for the first beam set and the second information comprises second resource information for the second beam set, wherein the terminal performs a measurement for the first beam set based on the first resource information and performs a measurement for the second beam set based on the second resource information. Claim 3 A method according to claim 1, wherein the beam set information includes one of the first information and the second information, and the first information and the second information include resource information for a corresponding beam set, wherein the terminal performs a measurement for the corresponding beam set based on the resource information. Claim 4 A method according to claim 1, wherein the terminal transmits Channel State Information (CSI) reporting information including the performance indicator to the base station. Claim 5 A method according to claim 1, wherein the terminal transmits the performance indicator to the base station based on the occurrence of a performance indicator transmission event. Claim 6 A method according to claim 5, wherein the performance indicator transmission event occurs in at least one of the cases where the performance indicator is lower than the performance condition and where the number of times the performance indicator lower than the performance condition is acquired reaches a specific counter value. Claim 7 A method according to claim 1, wherein the first information is transmitted through the first setting information, the second information is transmitted through the second setting information, and each of the first setting information and the second setting information includes a setting information identifier. Claim 8 A method according to claim 7, wherein the terminal obtains the performance indicator based on the measurement information, based on the fact that the above-mentioned setting information identifier is the same. Claim 9 A method according to claim 1, wherein the terminal obtains measurement information for a subset containing one or more beams among a plurality of beams included in the first beam set. Claim 10 In claim 9, the method wherein the terminal transmits auxiliary information to the base station, and the auxiliary information includes information related to the subset. Claim 11 A method according to claim 9, wherein the terminal obtains the performance indicator based on the ratio of a first value, which is the number of one or more beams included in the subset, and a second value, which is the number of beams included in the first beam set. Claim 12 A method according to claim 11, wherein the terminal predicts K beams with the best performance, K is an integer greater than or equal to 1, and the terminal obtains the performance indicator based on the fact that among the K beams, beams greater than or equal to the ratio are included in the first beam set. Claim 13 A method according to claim 1, wherein the measurement information is obtained by the terminal performing a measurement within a monitoring window. Claim 14 A method according to claim 1, wherein the beam prediction accuracy is the probability that one or more beams with the highest performance among a plurality of beams included in the first beam set match one or more beams with the highest measured performance, and the L1-RSRP difference is the difference between the L1-RSRP for one or more beams with the highest performance among a plurality of beams included in the first beam set and the L1-RSRP of one or more beams with the highest measured performance. Claim 15 A terminal comprises: one or more memories for storing instructions; one or more transceivers; and one or more processors connecting the one or more memories and the one or more transceivers, wherein the one or more processors execute the instructions to receive beam set information from a base station, wherein the beam set information includes at least one of first information regarding a first beam set inferred based on an AI / ML model used for AI / ML (Artificial Intelligence / Machine Learning) based beam management and second information regarding a second beam set used as input information for inferring the first beam set, wherein measurement information is obtained based on the beam set information, wherein the measurement information is obtained by the terminal performing a measurement on at least one of the first beam set and the second beam set based on resource information included in the beam set information, wherein a performance indicator for the AI / ML based beam management is obtained based on the measurement information, wherein the performance indicator includes at least one of beam prediction accuracy and L1-RSRP difference (Layer 1-Reference Signal Received Power Difference), and wherein the terminal transmits the performance indicator to the base station. Claim 16 In paragraph 15, the terminal, based on the fact that the beam set information includes the first information and the second information, and the first information includes first resource information for the first beam set and the second information includes second resource information for the second beam set, performs a measurement for the first beam set based on the first resource information and performs a measurement for the second beam set based on the second resource information. Claim 17 In paragraph 15, the beam set information includes one of the first information and the second information, and when one of the first information and the second information includes resource information for a corresponding beam set, the terminal performs a measurement for the corresponding beam set based on the resource information. Claim 18 In paragraph 15, the terminal transmits Channel State Information (CSI) reporting information including the performance indicators to the base station. Claim 19 In paragraph 15, the terminal transmits the performance indicator to the base station based on the occurrence of a performance indicator transmission event. Claim 20 In paragraph 19, the terminal, wherein the performance indicator transmission event occurs in at least one of the cases where the performance indicator is lower than the performance condition and where the number of times the performance indicator lower than the performance condition is acquired reaches a specific counter value.

Images