Method and apparatus for reporting performance monitoring metrics for beam management in wireless communication system

By implementing a method for optimized performance metric reporting in mobile communication systems, the UE reduces power consumption and air overhead while enabling effective evaluation of intelligent beam management models.

KR1020260164488APending Publication Date: 2026-11-17ELECTRONICS & TELECOMM RES INST
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
KR1020260080418
Authority / Receiving Office
KR · KR
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-09
Filing Date
2026-05-04
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Current mobile communication systems lack precise methods for calculating performance monitoring indicators and updating the number of operations when a terminal is configured to use intelligent beam management, leading to inefficiencies in power consumption and air overhead.

Method used

A method for a UE to receive configuration information for reporting performance metrics, output prediction values using an intelligent model, and transmit a report when the number of calculations is reached, reducing the frequency of reporting and enabling the base station to evaluate the model's performance.

Benefits of technology

Reduces UE power consumption and air overhead by optimizing performance metric reporting frequency, allowing the base station to assess the intelligent model's performance effectively.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure PAT00012_ABST
    Figure PAT00012_ABST
Patent Text Reader

Abstract

A method of a UE according to one embodiment of the present disclosure may include: receiving an upper-level signaling message from a base station that includes configuration information for reporting a performance indicator (PM) of an intelligent model, wherein the configuration information includes the number of PM acquisitions and information related to the number of prediction successes of the intelligent model to be reported; receiving RSs from the base station until the number of PM acquisitions is reached; outputting one or more prediction values ​​by using the measurement result of each RSs as an input to the intelligent model for each RSs received; determining a PM using the measurement result of the RSs and one or more predicted values ​​predicted at a previous time point; and, when the number of PM calculations is reached, transmitting a PM report message to the base station that includes information related to the number of prediction successes based on the calculated PM value.
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] The present disclosure relates to a performance monitoring indicator reporting technology for beam management in a mobile communication system, and more specifically, to a performance monitoring indicator reporting technology for beam management in a mobile communication system using artificial intelligence (AI) / machine learning (ML). Background Technology

[0002] The 3rd Generation Partnership Project (3GPP), an international standardization organization that standardizes mobile communication systems, is discussing applications using artificial intelligence (AI) and machine learning (ML) as a work item (WI) in Release 19. In particular, the WIs of 3GPP Release 19 are establishing use cases for utilizing AI / ML in air interfaces and are proceeding with the specification work required for using AI / ML technology for each use case.

[0003] Currently, the fields presented as use cases for AI / ML technology at WI include beam management (BM), positioning accuracy enhancement, and channel state information (CSI) feedback enhancement.

[0004] However, to date, in mobile communication systems consisting of a base station and one or more terminals, when a terminal is configured to use the intelligent beam management function, precise procedures have not been presented regarding the method for calculating performance monitoring indicators and the procedure for updating the number of operations of performance indicators during performance monitoring. The problem to be solved

[0005] The objective of the present disclosure to address the above-mentioned requirements is to provide a method and apparatus for calculating a performance monitoring indicator when a terminal is configured to use an intelligent beam management function, and a method and apparatus for providing a procedure for updating the number of operations of a performance indicator. means of solving the problem

[0006] A method of a UE according to one embodiment of the present disclosure for achieving the above-mentioned purpose may include: receiving an upper-layer signaling message from a base station that includes configuration information for reporting a performance metric (PM) of an intelligent model, wherein the configuration information includes the number of PM acquisitions and information regarding the number of prediction successes of the intelligent model to be reported; receiving RSs from the base station until the number of PM acquisitions is reached; outputting one or more prediction values ​​by using the measurement result of each RSs as an input to the intelligent model whenever the RSs are received; determining a PM using the measurement result of the RSs and one or more predicted values ​​predicted at a previous time point; and, when the number of PM calculations is reached, transmitting a PM report message to the base station that includes information regarding the number of prediction successes based on the calculated PM value. Effects of the invention

[0007] According to one embodiment of the present disclosure, a method for configuring an upper-layer signaling message for reporting performance metrics of an intelligent model (e.g., an AI model) is provided. Based on the upper-layer signaling message, the UE can report performance metrics to a base station. The base station can configure the upper-layer signaling message so that performance metrics are reported based on the reception of reference signals for a predetermined schedule. Based on this configuration, the UE can reduce the frequency of reporting performance metrics of the intelligent model. Due to the reduction in the frequency of reporting performance metrics, there is an advantage of reducing the power consumption of the UE. In addition, there is an advantage of reducing overhead in the air.

[0008] Additionally, in the present disclosure, the UE may quantize and transmit information to be included in a performance indicator report based on an upper-layer signaling message. Furthermore, if the performance indicator calculation is not performed as many times as required to be reported, the UE may notify the base station of the number of performance indicator calculations. Accordingly, the base station may use the number of performance indicator calculations included in the performance indicator report message received from the UE as reference material for evaluating the performance of the intelligent model. Brief explanation of the drawing

[0009] FIG. 1 is a conceptual diagram illustrating an embodiment of a communication system. FIG. 2 is a block diagram illustrating an example of a communication node constituting a communication system. Figure 3 is a conceptual diagram illustrating a base station and a UE equipped with an AI model in a mobile communication system. FIG. 4a is a conceptual diagram according to a first embodiment for explaining the selection procedure between a base station beam and a UE beam in a communication system. FIG. 4b is a conceptual diagram according to a second embodiment for explaining the selection procedure between a base station beam and a UE beam in a communication system. FIG. 4c is a conceptual diagram according to a third embodiment for explaining the selection procedure between a base station beam and a UE beam in a communication system. Figure 5a is a conceptual diagram illustrating a positioning method based on direct AI / ML positioning techniques. Figure 5b is a conceptual diagram illustrating a positioning method using an AI / ML-assisted positioning technique. Figure 6 is a conceptual diagram of an autoencoder. Figure 7 is a conceptual diagram of a network configuration for inference and monitoring when the UE has an intelligent model in a communication system. Figure 8 is a conceptual diagram of a network configuration for inference and monitoring when the network in a communication system has an intelligent model. FIG. 9a is a conceptual diagram illustrating a first embodiment of spatial domain beam prediction in an intelligent beam management method. FIG. 9b is a conceptual diagram illustrating a second embodiment of spatial domain beam prediction in an intelligent beam management method. FIG. 9c is a conceptual diagram illustrating an example of time-domain beam prediction in an intelligent beam management method. Figure 10 is a flowchart illustrating the performance indicator reporting settings and performance indicator reporting for intelligent beam management. Specific details for implementing the invention

[0010] The present disclosure is capable of various modifications and may have various embodiments, and specific embodiments are illustrated in the drawings and described in detail. However, this is not intended to limit the present disclosure to specific embodiments, and it should be understood that it includes all modifications, equivalents, and substitutions that fall within the spirit and scope of the present disclosure.

[0011] Terms such as "first," "second," etc., may be used to describe various components, but said components should not be limited by said terms. Such terms are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the present disclosure, 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.

[0012] 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.

[0013] The terms used in this disclosure are used merely to describe specific embodiments and are not intended to limit this disclosure. The singular expression includes the plural expression unless the context clearly indicates otherwise. In this disclosure, terms such as “comprising” or “having” are intended to specify 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.

[0014] 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 this disclosure 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 disclosure.

[0015] A communication system to which embodiments according to the present disclosure are applied will be described. The communication system to which embodiments according to the present disclosure are applied is not limited to the details described below, and embodiments according to the present disclosure may be applied to various communication systems. Here, the term "communication system" may be used interchangeably with "communication network."

[0016] 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.

[0017] Throughout the specification, the term "terminal" may refer to a mobile station, mobile terminal, subscriber station, portable subscriber station, user equipment, access terminal, etc., and may include all or part of the functions of a terminal, mobile station, mobile terminal, subscriber station, portable subscriber station, user equipment, access terminal, etc.

[0018] 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.

[0019] Throughout the specification, the term "base station" may refer to an access point, a radio access station, a node B, an evolved node B, a base transceiver station, a mobile multihop relay (MMR)-BS, etc., and may include all or part of the functions of a base station, access point, radio access station, node B, eNodeB, base transceiver station, MMR-BS, etc.

[0020] Hereinafter, preferred embodiments of the present disclosure will be described in more detail with reference to the attached drawings. In order to facilitate an overall understanding of the present disclosure, the same reference numerals are used for identical components in the drawings, and redundant descriptions of identical components are omitted.

[0021] FIG. 1 is a conceptual diagram illustrating an embodiment of a communication system.

[0022] Referring to FIG. 1, 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.

[0023] 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.

[0024] 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.

[0025] Meanwhile, each of the 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) constituting the communication system (100) may have the following structure.

[0026] FIG. 2 is a block diagram illustrating an example of a communication node constituting a communication system.

[0027] Referring to FIG. 2, 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.

[0028] 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).

[0029] 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 the methods according to embodiments of the present disclosure 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).

[0030] Referring again to FIG. 1, 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).

[0031] Here, each of the multiple base stations (110-1, 110-2, 110-3, 120-1, 120-2) may be referred to as Node B, evolved Node B, base transceiver station (BTS), radio base station, radio transceiver, access point, access node, road side unit (RSU), radio remote head (RRH), transmission point (TP), transmission and reception point (TRP), eNB, gNB, etc.

[0032] Each of the multiple terminals (130-1, 130-2, 130-3, 130-4, 130-5, 130-6) may be referred to as a UE (user equipment), terminal, access terminal, mobile terminal, station, subscriber station, mobile station, portable subscriber station, node, device, IoT (Internet of Thing) device, mounted module / device / terminal or on board device / terminal, etc.

[0033] 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.

[0034] In addition, each of the multiple base stations (110-1, 110-2, 110-3, 120-1, 120-2) can support MIMO transmission (e.g., SU (single user)-MIMO, MU (multi user)-MIMO, massive MIMO, etc.), CoMP (coordinated multipoint) transmission, CA (carrier aggregation) transmission, transmission in an unlicensed band, device-to-device communication (D2D) (or ProSe (proximity services)), etc. Here, each of the plurality of terminals (130-1, 130-2, 130-3, 130-4, 130-5, 130-6) can perform an operation corresponding to the base station (110-1, 110-2, 110-3, 120-1, 120-2) and an operation supported by the base station (110-1, 110-2, 110-3, 120-1, 120-2). For example, the second base station (110-2) can transmit a signal to the fourth terminal (130-4) based on the SU-MIMO method, and the fourth terminal (130-4) can receive a signal from the second base station (110-2) based on the SU-MIMO method. Alternatively, the second base station (110-2) can transmit a signal to the fourth terminal (130-4) and the fifth terminal (130-5) based on the MU-MIMO method, and each of the fourth terminal (130-4) and the fifth terminal (130-5) can receive a signal from the second base station (110-2) by the MU-MIMO method.

[0035] Each of the first base station (110-1), the second base station (110-2), and the third base station (110-3) can transmit a signal to the fourth terminal (130-4) based on the CoMP method, and the fourth terminal (130-4) can receive a signal from the first base station (110-1), the second base station (110-2), and the third base station (110-3) by the CoMP method. Each of the multiple base stations (110-1, 110-2, 110-3, 120-1, 120-2) can transmit and receive signals based on the CA method with terminals (130-1, 130-2, 130-3, 130-4, 130-5, 130-6) within its cell coverage area. Each of the first base station (110-1), the second base station (110-2), and the third base station (110-3) can control D2D between the fourth terminal (130-4) and the fifth terminal (130-5), and each of the fourth terminal (130-4) and the fifth terminal (130-5) can perform D2D by controlling each of the second base station (110-2) and the third base station (110-3).

[0036] Next, methods for configuring and managing wireless interfaces in a communication system will be described. Even when a method performed by a first communication node among the communication nodes (e.g., transmission or reception of a signal) is described, the corresponding second communication node may perform a method corresponding to the method performed by the first communication node (e.g., reception or transmission of a signal). That is, when the operation of a terminal is described, the corresponding base station may perform an operation corresponding to the operation of the terminal. Conversely, when the operation of a base station is described, the corresponding terminal may perform an operation corresponding to the operation of the base station.

[0037] Meanwhile, in a communication system, a base station can perform all functions of the communication protocol (e.g., remote radio transmission and reception functions, baseband processing functions). Alternatively, among all functions of the communication protocol, the remote radio transmission and reception function may be performed by a TRP (transmission reception point) (e.g., f(flexible)-TRP), and among all functions of the communication protocol, the baseband processing function may be performed by a BBU (baseband unit) block. The TRP may be an RRH (remote radio head), RU (radio unit), TP (transmission point), etc. A BBU block may include at least one BBU or at least one DU (digital unit). A BBU block may be referred to as a "BBU pool," "centralized BBU," etc. A TRP may be connected to a BBU block via a wired fronthaul link or a wireless fronthaul link. A communication system composed of backhaul links and fronthaul links may be as follows. When the function split method of the communication protocol is applied, the TRP can selectively perform some functions of the BBU or some functions of MAC (medium access control) / RLC (radio link control).

[0038] In the present disclosure, a phrase containing "~ case (e.g., when ~)" may be expressed as a phrase containing "~ based on (e.g., based on ~)" or a phrase containing "~ in response to (e.g., in response to ~)". In other words, a phrase containing "~ case" may be interpreted as identical or similar to a phrase containing "~ based on" or a phrase containing "~ in response to".

[0039] Figure 3 is a conceptual diagram illustrating a base station and a UE equipped with an AI model in a mobile communication system.

[0040] Referring to FIG. 3, the base station (310) may use an AI model (311). The first UE (320) and the second UE (330) may also each use AI models (321, 331). A one-sided AI / ML model may be a case where only the UEs (320, 330) use AI models (321, 331) or a case where only the base station (310) uses the AI ​​model (311).

[0041] In a cross-sectional model, when only UEs (320, 330) use AI models (321, 331) respectively, inference is performed entirely in the UE, and inference can be performed only in UE-sided AI / ML models (321, 331). In a cross-sectional model, when only the base station (310) uses AI model (311), inference is performed entirely in the network, and inference can be performed only in network-sided AI / ML models (311).

[0042] In the case of a two-sided AI / ML model, AI / ML models (311, 321) corresponding to a pair of base station (310) and a first UE (320) may jointly perform inference, and / or AI / ML models (311, 331) corresponding to a pair of base station (310) and a second UE (330) may jointly perform inference.

[0043] Co-inference may include AI / ML inference where inference is performed jointly across the UE and the network. For example, the first part of the inference may be performed by the UE's AI model (UE-part of two-sided model), and the remaining part by the base station's AI model (NW-part of two-sided model). Conversely, the first part of the inference may be performed by the base station's AI model, and the remaining part by the UE's AI model.

[0044] <지능형 기능 / 모델의 생애 주기 관리>

[0045] Since each of the AI ​​models (311, 321, 331) of the base station and / or UE is based on training data, it must be possible to perform life cycle management (LCM) for the creation and maintenance of intelligent models in response to changes in training data. Therefore, if one intends to apply functions based on AI / ML technology in a mobile communication system composed of a base station and / or UE in use cases as described above, the mobile communication system must be able to support LCM.

[0046] 3GPP is discussing the detailed steps of the LCM process, including 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, data for a specific intelligent model can be collected, and the AI ​​model can be trained using the collected data. The trained AI model can be deployed to UE(s) by the network, and the activation of the deployed AI model can be directed. The UE and / or base station can perform inference using the activated AI model and monitor the AI ​​model based on the inference results.

[0047] To use an AI model with intelligent functions in a mobile communication system, functional identification and model identification of the intelligent functions performed by the AI ​​model must be performed beforehand. For example, prior to performing lifecycle management of the AI ​​model, it must be identified whether the AI ​​model is supported within the network (e.g., base station and / or UE). Based on the identified intelligent functions and AI models, the network can then determine whether to enable specific intelligent functions and models. In a two-sided or one-sided model, if inference is performed entirely by the UE-side AI / ML model, the base station can instruct the UE to enable the UE-side AI / ML model.

[0048] Regarding the identification of intelligent functions and AI models, 3GPP is discussing the following two lifecycle management directions.

[0049] (1) Functionality-based life cycle management (functionality-based LCM)

[0050] (2) Model ID-based Lifecycle Management (Model-ID based LCM)

[0051] For functionality-based lifecycle management, the base station (or base station-side server) and the UE (or UE-side server) can share functional information of intelligent functions in advance. Additionally, the base station and the UE can identify intelligent functions using functional identification information, such as function names or function identifiers (IDs), and manage the lifecycle of the corresponding intelligent functions, in other words, AI models.

[0052] For Model-ID-based lifecycle management, the base station (or base station-side server) and the UE (or UE-side server) can share intelligent model information along with the Model-ID in advance, as previously described. Additionally, the base station and the UE can use the Model-ID to identify and manage the lifecycle of intelligent models.

[0053] The two life cycle management functions mentioned above may not be configured independently but may be configured and operate in a mixed form.

[0054] <지능형 빔 관리(AI / ML based beam management)>

[0055] The frequency bands used to provide higher transmission rates in mobile communication systems are also increasing. For example, the millimeter wave (mmWave) band (e.g., 30 GHz to 300 GHz) is used in cellular mobile communication systems to provide higher transmission rates. Using high frequency bands results in shorter wavelengths and increased path attenuation, which reduces the signal's range. To address this, antenna technology is being used to generate a directional beam that directs the signal toward a specific direction (or area), thereby improving the range.

[0056] Meanwhile, generally, the power of a transmission signal transmitted by a base station must increase in proportion to the beam width or beam area covered by the base station beam. Therefore, when transmitting a signal with the same power, the signal transmission distance is inversely proportional to the beam width or beam area covered by the base station beam. Accordingly, to overcome signal quality degradation in high frequency bands and to cover the entire cell area, the base station can form multiple transmission beams (multiple base station beams). Each base station beam provides coverage for an area narrower than the total coverage area, but sufficient transmission power amplification can be achieved to overcome signal propagation loss.

[0057] In cases where directional beams are used as described above, an additional selection process may be required among multiple beams to ensure stable link performance, specifically selecting the base station beam (e.g., transmit beam (Tx beam)) and the UE beam (e.g., receive beam (Rx beam)) that provide optimal performance between the base station and the UE.

[0058] FIG. 4a is a conceptual diagram according to a first embodiment for explaining the selection procedure between a base station beam and a UE beam in a communication system.

[0059] Referring to FIG. 4a, only a base station (410) and one UE (420) are illustrated. However, in a cellular mobile communication system, there may be multiple UEs communicating with the base station. FIG. 4a describes the procedure for selecting the base station beam and the UE beam with the base station (410) and one UE (420), and this beam selection procedure can be applied in the same way when other UEs select a beam with the base station.

[0060] In FIG. 4a, for convenience of explanation, a base station (410) is exemplified with four base station beams (411, 412, 413, 414), and a UE (420) is exemplified with two UE beams (421, 422). Additionally, for convenience of explanation below, it is assumed that the base station (410) transmits four beams (411, 412, 413, 414). In other words, it is assumed that the base station beams (411, 412, 413, 414) are transmission beams. And it is assumed that the two UE beams (421, 422) are reception beams.

[0061] The base station (410) can transmit by beam sweeping four base station beams (411, 412, 413, 414). The UE (420) can receive each of the four base station beams (411, 412, 413, 414) by beam sweeping two UE beams (421, 422) and measure the received signal strength. For example, the UE (420) can receive each of the four base station beams (411, 412, 413, 414) through the UE beam (421) and measure the received signal strength of each of the received beams. Subsequently, the UE (420) can receive each of the four base station beams (411, 412, 413, 414) through the UE beam (422) and measure the received signal strength of each of the received beams.

[0062] In the example of FIG. 4a, the UE (420) is illustrated through hatching as a case where the UE beam (421) is selected as the optimal UE beam for reception among the UE beams (421, 422). Additionally, the base station beam (412) is selected as the optimal base station beam corresponding to the optimal UE beam (421) is illustrated through hatching.

[0063] Depending on the beam management method of the UE and the base station, the beams selected in the procedure of Fig. 4a may be used, and additional procedures may be performed thereafter.

[0064] FIG. 4b is a conceptual diagram according to a second embodiment for explaining the selection procedure between a base station beam and a UE beam in a communication system.

[0065] FIG. 4b may be an additional procedure for selecting a better beam by adjusting the beam width of a selected base station beam among the base station beams after the procedure of FIG. 4a. Accordingly, when the optimal beam (412) for use in communication with the UE (420) is selected among the base station beams (411, 412, 413, 414), the base station (410) may further subdivide the optimal base station beam (412) into subdivided base station beams (412a, 412b, 412c). FIG. 4b illustrates a case where the optimal base station beam (412) is subdivided into three subdivided base station beams (412a, 412b, 412c) for convenience of explanation, but the present disclosure is not limited thereto. In other words, the number of subdivided base station beams may be two or four or more.

[0066] The base station (410) can transmit the subdivided base station beams (412a, 412b, 412c) to the UE (420) by beam sweeping them again. At this time, the UE (420) can measure the received signal strength again for each of the subdivided base station beams (412a, 412b, 412c) through the optimal UE beam (421) among the UE beams (421, 422), and select the optimal subdivided beam based on the measured received signal strength. For example, the UE (420) can select the optimal subdivided base station beam (412b) among the three subdivided base station beams (412a, 412b, 412c).

[0067] As described above, when selecting the optimally segmented base station beam (412b), the base station can transmit signals to the UE (420) with better signal quality even when using the same transmission power.

[0068] FIG. 4c is a conceptual diagram according to a third embodiment for explaining the selection procedure between a base station beam and a UE beam in a communication system.

[0069] FIG. 4b may be an additional procedure for selecting a better beam by adjusting the beam width of a selected UE beam among the UE beams after the procedure of FIG. 4a or after the procedures of FIG. 4a and FIG. 4b. Previously, in the procedure of FIG. 4a, if the optimal base station beam (412) is selected and the optimal UE beam (421) is also selected, the UE (420) may subdivide the selected UE beam (421). FIG. 4c illustrates a case where the selected UE beam (421) is further subdivided into three subdivided UE beams (421a, 421b, 421c). As previously described in FIG. 4b, the number of subdivided UE beams is not limited to the three subdivided beams exemplified in FIG. 4c. In other words, the number of subdivided UE beams may be two, or four or more.

[0070] The UE (420) can measure the signal strength received from the optimal base station beam (412) using each of the three subdivided UE beams (421a, 421b, 421c). Although FIG. 4c illustrates a case where the optimal base station beam (412) is not subdivided, the signal may be received through the subdivided base station beam. As another example, if FIG. 4c is performed first and the procedure of FIG. 4b is performed, the optimal base station beam (412) may be in an unsubdivided state. As yet another example, if the procedure of FIG. 4b is not performed, the optimal base station beam (412) may not be subdivided.

[0071] The UE (42) can select the optimal subdivided UE beam based on the signal strength measurement results using each of the three subdivided UE beams (421a, 421b, 421c).

[0072] As described above, when selecting an optimally segmented UE beam, the UE (420) has the effect of being able to transmit a signal to the base station (410) using less power when transmitting a signal to the base station (410).

[0073] As described above, for optimal beam selection, measurement results for all available beams must be obtained, or depending on the implementation method, the base station or UE must measure multiple beams and select the optimal beam using the measured results. However, this process of finding the optimal beam incurs significant overhead. Furthermore, time delays caused by this high overhead can lead to a degradation in communication quality and / or power consumption by the UE.

[0074] Intelligent technology can be introduced into the beam management process to resolve issues arising from the procedure of finding the optimal beam. Beam management technology using intelligent technology can be a technique that maintains accuracy while reducing the overhead used in conventional beam measurement.

[0075] Regarding AI / ML intelligent technology-based beam management methods, beam prediction, a use case for predicting beams of unobserved resources in the spatial or temporal domain by applying intelligent technology, is being further discussed.

[0076] <지능형 포지셔닝 향상(AI / ML based positioning accuracy enhancement)>

[0077] Positioning can refer to a positioning technique used to measure the location of a specific UE in a mobile communication system. The 5G NR standard defined by 3GPP stipulates that base stations transmit a positioning reference signal (PRS) to the UE. The 5G NR standard also stipulates that the UE receives the PRS transmitted by the base station and reports the reference signal time difference (RSTD) for the received PRS to the base station. Furthermore, 5G NR specifies that base stations must support methods for applying positioning techniques, such as the observed time difference of arrival (OTDOA), based on the RSTD received from the UE.

[0078] Recently, requirements for indoor positioning accuracy have been increasing, and technologies that apply intelligent technologies to improve the accuracy of positioning measurements are being discussed in light of these requirements. Intelligent technology-based positioning is being discussed in two use cases: AI / ML-assisted positioning, which applies intelligent technologies to enhance the accuracy of conventional positioning methods, and direct AI / ML positioning, which applies intelligent technologies to directly estimate the location of a UE.

[0079] Figure 5a is a conceptual diagram illustrating a positioning method based on direct AI / ML positioning techniques.

[0080] The AI / ML model (510) can take a measured signal as input and output an estimated position. In this case, the measured signal may be a channel impulse response (CIR) and / or a reference signal received power (RSRP).

[0081] When CIR is used as an input to an AI / ML model (510), it may be a case where signal arrival strength information is used. The AI / ML model (510) can infer the estimated location (511) of the UE based on the signal arrival strength information, that is, the signal arrival profile. For example, in the case of a line of sight (LOS) path, the CIR may be strong and clean. On the other hand, in the case of a non-line of sight (NLOS) path, since the CIR is transmitted over the non-line of sight path, not only is a delay occurring but signal spreading may also be detected. Therefore, the AI / ML model (510) can infer the estimated location (511) of the UE based on the signal arrival profile of the measured signal.

[0082] When an RSRP value is used as an input to an AI / ML model (510), the RSRP value may vary depending on the distance over which the signal is transmitted, that is, the distance between a signal transmitting device (e.g., base station) and a signal receiving device (e.g., UE), and whether there are obstacles between the signal transmitting device and the receiving device. Accordingly, the AI / ML model (510) can output the estimated location (511) of the UE based on the measured RSRP value.

[0083] Figure 5b is a conceptual diagram illustrating a positioning method using an AI / ML-assisted positioning technique.

[0084] The AI / ML auxiliary positioning technique may be a case where an AI model (520) is used as an auxiliary means for positioning. The AI ​​model (520) can take measured signal values, such as CIR and / or RSRP values, as input and infer and output time of arrival (TOA) and / or non-line of sight (NLOS) identification values. The AI ​​model (520) used in the AI / ML auxiliary positioning technique may have different functionality from the AI ​​model (510) used in the direct AI / ML positioning technique.

[0085] When TOA is used as input to a non-AI processor (530), the non-AI processor (530) can estimate the position of the UE using an accurate distance (e.g., a value based on LOS) and / or an overestimated distance (e.g., a value based on NLOS) based on the TOA output from the AI / ML model (520). Then, the non-AI processor (530) can output the estimated position (531) value of the UE. At this time, the estimated position (531) of the UE may be determined using, for example, a triangulation method.

[0086] As explained above, positioning methods using AI / ML can be used indoors. If used outdoors, positioning can also be performed by utilizing satellite signals received from satellites.

[0087] <지능형 채널 정보 피드백 향상(AI / ML based CSI feedback enhancement)>

[0088] Channel state information (CSI) feedback refers to the UE reporting channel state information to support a base station in a mobile communication system in applying transmission techniques such as multiple input multiple output (MIMO) or precoding. The 5G NR standard defined by 3GPP specifies procedures for transmitting feedback information, such as the channel quality indicator (CQI), precoding matrix indicator (PMI), and rank indicator (RI), from the UE to the base station in relation to channel information feedback methods. Furthermore, discussions are ongoing regarding the improvement of CSI feedback techniques to effectively support transmission techniques such as multi-user MIMO (MU-MIMO) in NR systems.

[0089] Meanwhile, intelligent techniques are one of the research directions for incorporating intelligent technology into channel information feedback. One method of channel information feedback using intelligent techniques may be a technique utilizing an auto-encoder. At 3GPP, discussions are underway regarding channel compression (CSI compression) technology, which obtains a compressed latent space for MIMO channels using an auto-encoder.

[0090] Figure 6 is a conceptual diagram of an autoencoder.

[0091] An autoencoder can be divided into an encoder portion (610) and a decoder portion (620). The encoder portion (610) and the decoder portion (620) can be shared among communication nodes of a mobile communication system. For example, if a base station uses the encoder portion (610) of the autoencoder, the UE can use the decoder portion (620) of the autoencoder. Conversely, if a base station uses the decoder portion (620) of the autoencoder, the UE can use the encoder portion (610) of the autoencoder. Generally, when the UE transmits a compressed latent space for a MIMO channel to a base station via channel compression using an autoencoder, the UE can use the encoder portion (610) of the autoencoder, and the base station can use the decoder portion (620) of the autoencoder.

[0092] The encoder part (610) may include an input layer (611), a hidden layer (612), and an output layer (613). The input layer (611) of the encoder part (610) may include a plurality of input nodes (611a, …, 611n). Here, each of the nodes may mean a neuron. The plurality of input nodes (611a, …, 611n) receive input signals and may be connected to each node (612a, …, 612k) of the hidden layer (612) through a connecting neuron as exemplified in FIG. 6 to have a weight (or coefficient) obtained through learning. Each hidden node (612a, …, 612k) of the hidden layer (612) can also be connected to have weights learned between each node (613a, …, 613j) of the output layer. These weights may be the same or different values ​​for each connection of nodes.

[0093] In order to allow the UE to report a small amount of data to the base station through channel compression, as exemplified in FIG. 6, the number of multiple input nodes (611a, …, 611n) may be greater than the number of hidden nodes (612a, …, 612k), and the number of hidden nodes (612a, …, 612k) may be greater than the number of output nodes (613a, …, 613j). In other words, the number of output nodes (613a, …, 613j) included in the encoder part (610) of the autoencoder may be smaller than the number of multiple input nodes (611a, …, 611n) of the encoder part (610).

[0094] If the encoder part (610) and the decoder part (620) are not operated at the UE and the base station, respectively, the output layer (613) of the encoder part (610) may also be understood as a hidden layer. In the following description, the case where the encoder part (610) and the decoder part (620) are divided and operated at the UE and the base station, respectively, is assumed.

[0095] Information output from the output nodes (613a, …, 613j) of the encoder part (610) can be transmitted to the base station by the UE. The base station can provide the information received from the UE to the decoder part (620) of the autoencoder. In other words, the information received from the UE (information output from the output nodes (613a, …, 613j) of the encoder part) can be transmitted to the input layer of the decoder part (620) of the base station. In FIG. 6, the output layer (613) of the encoder part (610) can correspond to the input layer of the decoder part (620). Each input node (613a, …, 613j) included in the input layer of the decoder part (620) can be connected to each hidden node (622a, …, 622k) of the hidden layer (622) to have weight values ​​based on the learning of the AI ​​model. Each hidden node (622a, …, 622k) can also output decoded information through the output layer (623) of the decoder part (620). Ideally, the output information output from each output node (623a, …, 623n) included in the output layer (623) of the decoder part (620) may be the same information as the input information input to each of the input nodes (611a, …, 611n) included in the input layer of the encoder part (610).

[0096] As described above, an autoencoder may refer to a neural network structure capable of copying inputs to outputs. As exemplified in FIG. 6, by setting the number of neurons in the hidden layer in the encoder part (610) to be less than the number of neurons in the input layer and configuring the number of neurons in the output layer to be less than the number of neurons in the hidden layer, the amount of data transmitted between the UE and the base station can be reduced through data compression (or dimension compression).

[0097] The present disclosure described below describes a case in which a UE is configured by a base station to use an intelligent beam management function in a mobile communication system composed of a base station and one or more UEs. Furthermore, when the UE is configured to use an intelligent beam management function, performance monitoring may be performed on a model, such as an AI model, that performs the operation of the intelligent beam management function. Performance indicators may be calculated for the performance monitoring of the AI ​​model. The performance indicators calculated for performance monitoring may have a form according to the present disclosure. The present disclosure described below describes the form of the performance indicators for performance monitoring. Additionally, performance indicators may be updated. Accordingly, the present disclosure described below describes a procedure for updating performance indicators.

[0098] In the following, the performance monitoring method for intelligent beam management according to the present disclosure is described from the perspective of the downlink of a mobile communication system. However, this is merely to aid in understanding the present disclosure, and performance monitoring for intelligent beam management can be performed in the same or similar form from the perspective of the uplink. Furthermore, it can be extended to any other wireless communication system having a transmitting device and a receiving device. However, for the sake of understanding, the present disclosure described below will be described from the perspective of use cases discussed in the 5G NR standard, which is one of the 3GPP standards, namely intelligent channel feedback enhancement, intelligent beam management, and / or intelligent positioning enhancement.

[0099] In order to select the optimal beam during the inference process using intelligent beam management technology, a process of measuring the performance of a specific beam is required. For example, performance can be the Layer 1 (L1)-RSRP (L1-RSRP) or the L1-signal-to-interference plus noise ratio (L1-SINR) measured by the UE. The measured beam performance can be used as input to an intelligent model (AI / ML model). The intelligent model can output an inferred result based on learning from the input. The inferred result of the intelligent model can be any one of the following: the index of the beam with optimal performance, the predicted performance of each of the total available beams (e.g., the predicted L1-RSRP for each beam), or information on the probability that a specific beam will be the best beam.

[0100] In the beam performance measurement process, a critical aspect may be determining the base station beam or the beam pair consisting of a base station beam and a UE beam to be used for the measurement. By using an exhaustive search method that obtains measurement results for all base station beams or beam pairs of base station beams and UE beams, it is possible to determine the base station beam or beam pair with optimal performance in any situation. However, as discussed earlier, performing the exhaustive search method presents a problem of waste caused by significant overhead during the measurement process.

[0101] Generally, resources used in the beam performance measurement process can perform beam performance measurements by transmitting a predetermined reference signal (RS), such as a synchronization signal block (SSB) or a channel state information-reference signal (CSI-RS). In this case, to balance the overhead of the beam performance measurement process with the performance when using the finally selected beam, a method to reduce performance measurements for unnecessary beams may be required.

[0102] As previously described, an intelligent model for beam selection may exist as either a one-sided AI / ML model or a two-sided AI / ML model. In other words, the intelligent model may be provided only at the base station, only at the UE, or at both the base station and the UE.

[0103] In a cross-sectional AI / ML model where only the base station is equipped with an intelligent model (AI / ML model), the UE can transmit measured beam performance information (measurement report) to the base station. The base station can use the beam performance information received from the UE as input to the intelligent model. As such, when only the base station possesses an intelligent model among cross-sectional AI / ML models, a procedure may be required for the UE to transmit beam performance information to the base station.

[0104] In cases where only the UE among the cross-sectional AI / ML models is equipped with an intelligent model (AI / ML model), the UE can obtain an inference report from the AI / ML model by using the measured beam performance information as input. The inference result must be reported to the base station. Therefore, the UE may require a procedure to transmit the inference result to the base station.

[0105] As explained above, in the case of a cross-sectional AI / ML model, the intelligent model performing intelligent beam management may be located on the UE side (UE-side) or on the network side (e.g., base station). The reason it is necessary to distinguish the location of the intelligent model is that the actions required by the base station and the UE to support intelligent beam management technology may differ depending on the location of the intelligent model.

[0106] In addition, the operations required for intelligent beam management can be composed of LCM operations such as data collection (for training), inference, and model monitoring (performance monitoring).

[0107] Figure 7 is a conceptual diagram of a network configuration for inference and monitoring when a UE has an intelligent model in a mobile communication system.

[0108] Referring to FIG. 7, a network (710) and a UE (720) are illustrated. The network (700) illustrated in FIG. 7 may include a base station. The network (700) may include all or part of the configuration described earlier in FIG. 2. The network (700) may include a part of a core network. The network (700) may further include an interface with an external network (not shown in FIG. 7) that provides an AI model or provides data associated with the AI ​​model. It should be noted that a network (700) having such a configuration can be understood as being substituted for a base station.

[0109] The UE (720) may also include all or part of the configuration described above in FIG. 2. The UE (720) may have additional configurations in addition to the configurations exemplified in FIG. 2. For example, it may further include various sensors, cameras, and / or satellite communication devices for user convenience. Additionally, the UE (720) may further include the configuration exemplified in FIG. 7 according to the present disclosure. The configuration of the UE (720) exemplified in FIG. 7 may be driven by the processor (210) exemplified in FIG. 2 or implemented as a separate hardware device.

[0110] The network (710) can transmit a reference signal (RS) (730) for beam measurement to the UE (720) for beam management. In this case, each beam can be transmitted using a beam sweeping method as previously described, and an RS (730) can be transmitted for each beam. The UE (720) can receive the RS (730) for beam measurement transmitted by the network.

[0111] The measuring device (721) of the UE (720) can receive RS (730) through each beam transmitted by the network (710) beam sweeping, and can measure the received RS (730). The measured value of RS (730) may be L1-RSRP or L1-SINR as previously described. In the example of FIG. 7, since the intelligent model (AI model) is performed only on the UE (720) side, the measuring device (721) of the UE (720) can provide the measured value of RS (730) to the data collection device (722), the AI ​​model (723), and the monitoring device (725).

[0112] The example in FIG. 7 is a cross-sectional AI model in which the AI ​​model (723) is located in the UE (720), so the UE (720) does not need to transmit the measured values ​​to the network (710) for data collection. However, the UE (720) must transmit the inference results using the AI ​​model (723) to the network (710).

[0113] The data collection device (722) can perform data preprocessing so that the measured RS (730) received from the measurement device (721) can be used later in the AI ​​model (723), and can store the data after performing labeling and / or structuring for the learning and / or inference of the AI ​​model (723). The data collection device (722) can provide the labeled and / or structured data to the AI ​​model (723) for the learning and / or inference of the AI ​​model (723).

[0114] The AI ​​model (723) can perform inference using the measured value of RS (730) received from the measuring device (721) and / or structured data from the data collection device (722) that can be used as input to the AI ​​model (723). The AI ​​model (723) can output an inference result through inference. The inference result of the AI ​​model (723) for beam management may be a predicted value in a future time instance. The predicted value may be, for example, a predicted RSRP or a predicted SINR. The AI ​​model (723) may provide the inference result to an inference information transmission device (724). Additionally, the AI ​​model (723) may provide the inference result to a monitoring device (725) that monitors the performance of the AI ​​model (723).

[0115] The inference information transmission device (724) can transmit the inference result (740) to the network (720) through a first reporting resource (e.g., time resource and frequency resource) set to be used for reporting inference information by the network (710). Thus, the network (720) can receive the result (740) inferred by the AI ​​model (723) from the inference information transmission device (724) of the UE (720).

[0116] The monitoring device (725) can calculate a performance metric for monitoring the performance of the AI ​​model (723) based on the inference result of the AI ​​model (723) and the measured value of the RS (730) of the measuring device (721). The monitoring device (725) can transmit the calculated performance metric (750) to the network (710) through a preset second reporting resource (e.g., time resource and frequency resource).

[0117] If the performance indicator of the AI ​​model (723) is calculated in the network (710), the monitoring device (725) can transmit the result (750) of measuring the RS (730) in the measuring device (721) to the network (710) through a preset third reporting resource (e.g., time resource and frequency resource).

[0118] Based on what has been described above, the network (710) can pre-configure RS transmission resources (time and frequency resources) through which RS (730) is transmitted to the UE (720) in order to transmit RS (730) to the UE. Additionally, it can pre-configure a first reporting resource to the UE (720) and pre-configure a second reporting resource or a third reporting resource.

[0119] The RS transmission resource described above may be set by higher layer signaling (e.g., radio resource control (RRC) signaling). As another example, the RS transmission resource may be set and directed by higher layer signaling and dynamic signaling (e.g., medium access control-control element (MAC-CE) and / or downlink control information (DCI)). As yet another example, the RS transmission resource may be transmitted by a request from the UE (720).

[0120] Reporting resources may be set by upper-level signaling (e.g., RRC signaling), or may be set and directed by upper-level signaling and dynamic signaling (e.g., MAC-CE and / or DCI).

[0121] Figure 8 is a conceptual diagram of a network configuration for inference and monitoring when the network in a mobile communication system has an intelligent model.

[0122] Referring to FIG. 8, a network (810) and a UE (820) are illustrated. The network (800) illustrated in FIG. 8 may include a base station. The network (800) may include all or part of the configuration described above in FIG. 2. The network (800) may include a part of a core network. The network (800) may further include an interface with an external network (not shown in FIG. 8) that provides an AI model or provides data associated with the AI ​​model. The configuration of the network (810) illustrated in FIG. 8 may be run on the processor (210) illustrated in FIG. 2 or implemented as a separate hardware device. Additionally, the network (810) may further include the configuration illustrated in FIG. 8 according to the present disclosure. It should be noted that a network (800) having such a configuration may be understood as being substituted for a base station.

[0123] The UE (820) may also include all or part of the configuration described above in FIG. 2. The UE (820) may have additional configurations in addition to the configurations exemplified in FIG. 2. For example, it may further include various sensors, cameras, and / or satellite communication devices for user convenience.

[0124] The network (810) can transmit a reference signal (RS) (830) for beam measurement to the UE (820) for beam management. At this time, each beam can be transmitted using a beam sweeping method as previously described, and an RS (830) can be transmitted for each beam. The UE (820) can receive the RS (830) for beam measurement transmitted by the network (810).

[0125] The measuring device (821) of the UE (820) can receive RS (830) through each beam transmitted by the network (810) beam sweeping, and can measure the received RS (830). The measured value of RS (830) may be L1-RSRP or L1-SINR as previously described. In the example of FIG. 8, since the intelligent model (AI model) is performed only on the network (820) side, the measuring device (821) of the UE (820) can transmit the measurement result (840) of the measured RS (830) to the network (810) through a pre-set fourth reporting resource (e.g., time resource and frequency resource).

[0126] The network (810) can receive measurement results (840) of RS (830) measured from the UE (820). The measurement results (840) can be provided to a data collection device (811), an AI model (812), and a monitoring device (813).

[0127] The data collection device (811) can perform data preprocessing on the measurement results (840) received from the UE (820) so that they can be used later in the AI ​​model (812), and can store them after performing labeling and / or structuring for the training and / or inference of the AI ​​model (812). The data collection device (811) can provide the labeled and / or structured data to the AI ​​model (823) for training and / or inference in the AI ​​model (812).

[0128] The AI ​​model (812) can perform inference using structured data from the measurement result (840) and / or the data collection device (811) as input to the AI ​​model (812). The AI ​​model (812) can output an inference result through inference. The inference result of the AI ​​model (812) for beam management may be a predicted value in a future time instance. The predicted value may be, for example, a predicted RSRP or a predicted SINR. The AI ​​model (812) may provide the inference result to an inference result processing device (814) and a monitoring device (813).

[0129] The inference information processing device (814) can determine the beam to be used for communication with the UE (820) and / or the beam to transmit the RS signal (830) based on the beam inference result of the AI ​​model (812).

[0130] The monitoring device (813) can calculate a performance metric for monitoring the performance of the AI ​​model (812) based on the inference results and measurement results (840) of the AI ​​model (812).

[0131] Based on what has been described above, the network (810) can pre-configure an RS transmission resource (time and frequency resource) to which the RS (830) is transmitted to the UE (820) in order to transmit the RS (830) to the UE. Additionally, the network (810) can pre-configure a fourth reporting resource to the UE (820). The fourth reporting resource may be configured to be the same resource as the third reporting resource of FIG. 7, or it may be configured to be a different resource.

[0132] The RS transmission resources described above may be set by upper-layer signaling (e.g., RRC signaling). As another example, RS transmission resources may be set and directed by upper-layer signaling and dynamic signaling (e.g., MAC-CE and / or DCI). As yet another example, RS transmission resources may be transmitted by a request from the UE (720). Additionally, reporting resources may be set by upper-layer signaling (e.g., RRC signaling) and may be set and directed by upper-layer signaling and dynamic signaling (e.g., MAC-CE and / or DCI).

[0133] Meanwhile, intelligent beam management technology can be classified into spatial-domain beam prediction (BM-case1) and temporal-domain beam prediction (BM-case2). Spatial-domain beam prediction may be a method of selecting the optimal beam for a specific point in time (or a specific time instance) by utilizing beam measurement results obtained at that specific point in time. On the other hand, temporal-domain beam prediction may be a method of selecting the optimal beam for a current or future time instance by utilizing beam measurement results obtained at the present and past points in time.

[0134] FIG. 9a is a conceptual diagram illustrating a first embodiment of spatial domain beam prediction in an intelligent beam management method.

[0135] Referring to FIG. 9a, beam set B (900) and beam set A (910) are illustrated. Beam set B (900) may be a set of beams used as input to an AI model, which is an intelligent model for spatial domain beam prediction. The beam set B (900) illustrated in FIG. 9a is illustrated as a set having three different beams (901, 902, 903). FIG. 9a illustrates a case where beam set B (900) includes three beams (901, 902, 903) for convenience of explanation and simplification of the drawings, but the present disclosure is not limited thereto. For example, beam set B (900) may include only two beams, or it may include four or more beams.

[0136] The base station can pre-configure the configuration information of beam set B (900) to the UE. The base station can transmit RS to the UE through each of the beams (901, 902, 903) belonging to beam set B (900). The UE can measure the RS received through the beams (901, 902, 903) for measurement based on the pre-configured information of beam set B (900). The measurement result of the RS can be used as input to an AI model. The information of beam set A (910) to be output from the AI ​​model can also be pre-configured.

[0137] In step S910, the AI ​​model can perform inference using measurement results of the RS transmitted through the beams (901, 902, 903) belonging to beam set B (900) as input. The AI ​​model can output an inference result for beam set A (910). In the example of FIG. 9a, the beams (901, 902, 903) included in beam set B (900) may be a subset of the beams (911, 912, 913, 914, 915, 916, 817) included in beam set A (910). In other words, the first beam (901) belonging to beam set B may be a beam corresponding to (or identical to) the second beam (912) belonging to beam set A, the second beam (902) belonging to beam set B may be a beam corresponding to (or identical to) the fourth beam (914) belonging to beam set A, and the third beam (903) belonging to beam set B may be a beam corresponding to (or identical to) the sixth beam (916) belonging to beam set A. Here, a corresponding beam may mean a beam with identical beam characteristics, such as beam direction and beam width.

[0138] In other words, the AI ​​model can infer all beams (911, 912, 913, 914, 915, 916, 817) included in beam set A (910) through the inference of step S910 using the measurement results of RS received through beam set B (910), which is a subset of beam set A (910).

[0139] FIG. 9b is a conceptual diagram illustrating a second embodiment of spatial domain beam prediction in an intelligent beam management method.

[0140] Referring to FIG. 9b, beam set B (920) and beam set A (940) are illustrated. Beam set B (920) may be a set of beams used as input to an AI model, which is an intelligent model for spatial domain beam prediction. The beam set B (920) illustrated in FIG. 9b is illustrated as a set having three different beams (921, 922, 923). FIG. 9b illustrates a case where beam set B (920) includes three beams (921, 922, 923) for convenience of explanation and simplification of the drawings, but the present disclosure is not limited thereto. For example, beam set B (920) may include only two beams or may include four or more beams.

[0141] The base station can pre-configure the configuration information of beam set B (920) to the UE. The base station can transmit RS to the UE through each of the beams (921, 922, 923) belonging to beam set B (920). Based on the pre-configured information of beam set B (920), the UE can measure the received RS through the beams (921, 922, 923) for measurement. The measurement result of the RS can be used as input to an AI model. The information of beam set A (930) to be output from the AI ​​model can also be pre-configured.

[0142] In step S920, the AI ​​model can perform inference using measurement results of the RS transmitted through the beams (921, 922, 923) belonging to beam set B (920) as input. The AI ​​model can output an inference result for beam set A (930). In the example of FIG. 9b, the beams (921, 922, 923) included in beam set B (920) may not be a subset of the beams (931, 932, 933, 934, 935) included in beam set A (930), unlike FIG. 9a. In other words, the beams (921, 922, 923) included in beam set B (920) exemplified in FIG. 9b may be beams having a wider beam width than the beams included in beam set A (930). Alternatively, the beams (921, 922, 923) included in beam set B (920) may not have the same beam direction as the beams included in beam set A (930).

[0143] For example, in the case of FIG. 9b, the AI ​​model may infer the beams (931, 932, 933, 934, 935) of beam set A (940) having a narrow beam width using beams having a wide beam width included in beam set B (930).

[0144] FIG. 9c is a conceptual diagram illustrating an example of time-domain beam prediction in an intelligent beam management method.

[0145] Referring to FIG. 9c, the measurement history (940) of beams belonging to beam set B and the inference result (950) of beam set A at a future point in time are illustrated. The measurement history (940) of beams belonging to beam set B may include the measurement result (941) of beams belonging to beam set B at time P2 and the measurement result (942) of beams belonging to beam set B at time P1. If time P1 is the current time, time P2 may be the measurement result of beams belonging to beam set B immediately before the current time.

[0146] The AI ​​model takes the measurement history (940) of the beams belonging to beam set B as input and can predict the beams belonging to beam set A at a future point in time through inference in step S940. At this time, the prediction may be for one or more future points in time. If the AI ​​model equipped in the UE performs a prediction for one or more future points in time using the measurement history (940) of the beams belonging to beam set B, the UE may transmit a report message containing the predicted result (the inference result of beam set A at time F1) to the base station. At this time, the resource to which the report message is transmitted may be pre-configured by the base station. Additionally, the prediction result included in the report message may include a prediction result for at least one point in time. If the AI ​​model predicts for two or more future points in time, the UE may configure the report message to include only one prediction result or to include two or more predicted results based on the base station's reporting settings.

[0147] Meanwhile, for performance monitoring of beam management of an AI model (or intelligent model or ML model), it may be necessary to calculate the performance metrics of the AI ​​model. In the following description, "model monitoring" may have the same meaning as "performance monitoring." Additionally, "model" may refer to an AI model, an intelligent model, or an ML model.

[0148] Beam prediction accuracy may be used as a performance metric for intelligent beam management. When a UE reports performance metrics to a base station, if the UE reports the performance metric measurement value (or performance metric calculation value) to the base station every time the UE calculates the performance metric, overhead issues may arise due to frequent reporting of performance metrics. Furthermore, frequent reporting of performance metrics may actually hinder stable performance monitoring operations. The present disclosure, described below, provides a method for stable performance monitoring while reducing overhead.

[0149] Figure 10 is a flowchart illustrating the performance indicator reporting settings and performance indicator reporting for intelligent beam management.

[0150] Referring to FIG. 10, the performance indicators for intelligent beam management may be a diagram assuming the case where the UE reports to the base station. In other words, when a cross-sectional AI model is used, the base station and the UE can be configured as described earlier in FIG. 7.

[0151] In step S1000, the base station may transmit to the UE via higher layer signaling messages, such as RRC signaling. The higher layer signaling messages transmitted in step S1000 may include reference signal configuration information, for example, CSI-RS resource configuration information, CSI measurement configuration information, and / or CSI report configuration information.

[0152] CSI-RS resource configuration information may include information on the periodicity (e.g., periodic, semi-persistent, aperiodic) of the transmission of CSI-RS, resource information to which CSI-RS is transmitted, etc. CSI measurement configuration information may include configuration information referenced by the UE through one or more CSI-RS resource set identifiers (IDs) to which the CSI-RS to be measured is transmitted. Additionally, CSI reporting configuration information may include information to be included in the CSI report, and information on the periodicity (e.g., periodic, semi-persistent, or aperiodic) of the CSI report, etc.

[0153] Information to be included in the CSI report may include performance indicator reporting setting information for reporting the performance monitoring of an AI model in accordance with the present disclosure. The performance indicator reporting setting information may include, for example, one or more of the number of performance indicator calculations to be performed by the time of the performance indicator report, the number of beam prediction successes of the calculated performance indicator (or quantized values ​​for success probabilities), or the time of the performance indicator report (e.g., in the case of periodic or semi-permanent reporting).

[0154] Additionally, CSI reporting configuration information may require measurement information for monitoring resources used to calculate performance metrics, as well as inference report information linked to those monitoring resources. Such linked monitoring resources and inference reports can be referred to as "linked pairs."

[0155] In step S1000, the UE can receive an upper-layer signaling message transmitted by the base station, and based on the received upper-layer signaling message, can check the resource settings, CSI measurement settings, and CSI reporting settings of the reference signal (e.g., CSI-RS).

[0156] In step S1002, the base station can transmit RS by beam sweeping the beams of beam set B. Beam set B may be the beams described earlier in FIGS. 9a through 9c. In step S1002, the UE can receive RS transmitted by the base station through the beams included in beam set B.

[0157] In step S1004, the UE measures the RS received through the beams included in beam set B and can perform inference using an intelligent model (e.g., an AI model) based on the measurement results of the RS. The inference information can be provided to a monitoring device (725) as described above in FIG. 7. The UE's monitoring device (725) can calculate performance indicators using the value(s) predicted through inference and the measured values.

[0158] Additionally, CSI report configuration information may include information on the number of performance metric calculations. The number of performance metric calculations may be a value representing the number of times such performance metric calculations are performed. For example, if the number of performance metric calculations is set to N (where N is a natural number greater than or equal to 1), the UE can calculate the performance metric by repeating N times.

[0159] In other words, the procedure from step S1002 to step S1004, whereby the UE receives RS through the beams of beam set B (step S1012) and uses the measured RS value to perform inference and calculate performance indicators through an AI model (step S1014), can be performed N times.

[0160] Subsequently, the UE may generate a performance indicator report message in step S1016. The performance indicator report message may include information based on the calculation of performance indicators. The method for calculating performance indicators according to the present disclosure is described in more detail below.

[0161] As previously explained, the calculation of performance metrics may vary depending on whether the intelligent model (e.g., AI model) performs beam prediction in the spatial domain (BM-case1) or beam prediction in the temporal domain (BM-case2).

[0162] <공간 도메인의 빔 예측 시 성능 지표 계산(또는 획득)>

[0163] In the case where the performance indicator calculation is a beam prediction (BM-case 1) in the spatial domain described in FIGS. 9a and 9b, the UE can calculate (or obtain) the performance indicator by using inference results associated with the transmission occasion(s) of N (or N times) of the latest monitoring resources. In other words, in the example of FIG. 10, the performance indicators and measurements calculated during the most recent N times, such as in steps S1004 and S1014, can be used based on the measurements of RS received through the beams included in beam set B, such as in steps S1002 and S1012. In the following description, the performance indicator calculation can be understood as being replaced by the performance indicator acquisition and performance indicator derivation.

[0164] When calculating performance metrics, inference reports or transmission occasions of monitoring resources at a later (or subsequent) point in time than the CSI reference resource corresponding to the CSI report for monitoring cannot be used.

[0165] In addition, an inference report associated with the measurement result of the transmission occasion of reference signal resources for monitoring (e.g., CSI-RS and / or SSB) may have a CSI reference resource of the inference report that has a minimal slot offset from the transmission occasion for monitoring.

[0166] Let t2 be the time point for an arbitrary inference report approval. An arbitrary inference report approval may be associated with the measurement results of the transmission approval of reference signal resources transmitted prior to the inference report. The reference signal resources associated with the inference report of the arbitrary inference report approval may be monitoring reference signal resources. If t1 is the time point for the transmission approval of the reference signal resources for monitoring, the interval between t1 and t2 may be within the slot offset described earlier. If the interval between time points t1 and t2 exceeds the slot offset, accurate measurement of performance metrics may be impossible.

[0167] The threshold slot offset X value can be set from the base station to the UE via RRC signaling. For example, the threshold slot offset X value can be configured with CSI report configuration information and transmitted from the base station to the UE. This ensures that the interval between the transmission approval and the inference report of monitoring resources does not exceed the set X slot offset. If the time interval between time t1 and time t2 exceeds the preset slot offset X value, the UE can determine that there is no inference result linked to that transmission approval.

[0168] <시간 도메인의 빔 예측 시 성능 지표 계산>

[0169] If the performance indicator calculation is a time domain beam prediction (BM-case 2) as described in FIG. 9c, multiple time instances may exist within a single inference result report message transmitted by the UE in step S1016. If multiple time instances exist within a single inference result report message, the UE may need to determine which point in time (or time instance) the performance indicator should be calculated for. If the performance indicator calculation is a time domain beam prediction, the base station may set one or more time instances for the performance indicator calculation in an RRC signaling message (e.g., CSI report setting).

[0170] When a time instance for calculating a single performance metric is set via an RRC signaling message, the UE may include an indicator indicating that the performance metric transmitted via an inference result report message is the f-th time instance (a time instance set by the base station), based on CSI report setting information received from the base station. As described above, one or more time instances indicating the performance metric included in the inference result report message may be set. The performance metric for the f-th time instance set by the base station to the UE via the CSI report setting may be calculated through an inference report associated with the transmission occasion(s) of the latest N monitoring resources. Even in this case, an inference report or a transmission occasion of monitoring resources at a later point in time than the CSI reference resource corresponding to the CSI report for monitoring cannot be used.

[0171] The f-th time instance in an inference report message associated with the measurement result of the transmission ok time of reference signal resources for monitoring (e.g., CSI-RS and / or SSB) may have a minimal slot offset with the transmission ok time for monitoring.

[0172] As explained earlier in the case of the spatial domain, if the interval between two times exceeds the time interval set by the slot offset, it may be impossible to measure performance metrics.

[0173] The threshold slot offset X value can be set from the base station to the UE via RRC signaling. For example, the threshold slot offset X value can be configured with CSI report configuration information and transmitted from the base station to the UE. This ensures that the interval between the transmission approval and inference report of monitoring resources does not exceed the set X slot offset. As previously described, if the time interval between the two points in time exceeds the preset slot offset X value, the UE may determine that there is no inference result linked to that transmission approval.

[0174] The performance metric report message may include information on the number of successful beam predictions. A successful beam prediction may be defined as the difference between the predicted beam and the beam measured at a later predicted point being within a preset threshold. If the performance metric report message includes the number of successful beam predictions itself, an amount of bits equal to that shown in Equation 1 below may be required.

[0175]

[0176] As another example, if the performance metric report message is configured to include beam prediction success rate information, information such as the following Equation 2 may be included based on the number of times beam prediction was successful (N_s) and the total number of beam predictions (N).

[0177]

[0178] As shown in Equation 2, the number of successful beam predictions (N_s) as the numerator of the performance indicator can be defined as follows for each transmission time of the monitoring resources. The monitoring beam set may be, for example, beam set A described earlier in FIGS. 9a to 9c. In the measurement results for the monitoring beam set, at least one of the M beams (where M is a natural number greater than or equal to 1) with the best performance may be included among the K beams (where K is a natural number greater than or equal to 1) with the highest quality predicted using the measurement results of beam set B described in FIGS. 9a to 9c as input to the AI ​​model. The M beams with the best performance may refer to the M beams sequentially starting from the beam with the highest measurement value (e.g., L1-RSRP or L1-SINR) among the beams of beam set A, which is the monitoring beam set. Additionally, the value of K may refer to the number of predicted beams included in the inference report message. The value of M can be configured in the UE through CSI reporting configuration information for monitoring by the network, and the value of M can have a value such as {1, 2}. If M is '1', it may mean that the beam with the highest L1-RSRP measurement value among the set of monitoring beams is included within the K beams predicted by the intelligent model.

[0179] To calculate the performance indicators described above, it may be necessary to have a transmission approval of the performance monitoring resource corresponding to the denominator of the performance indicator and a linked pair of N inference results. In particular, the value of N, which represents the number of performance indicator calculations, can be included in the CSI report configuration information as previously explained and transmitted from the base station to the UE in the form of an RRC signaling message.

[0180] Therefore, the base station must transmit a sufficient number of monitoring resources, at least N times, for the UE to calculate performance metrics. The RS (or monitoring resources) for monitoring can be transmitted periodicly, semi-persistently, or aperiodisically.

[0181] Meanwhile, as shown below, there may be cases where the UE fails to secure N linked pairs for calculating performance metrics.

[0182] For example, if there is no inference result associated with the transmission OK of a monitoring resource, a connected pair may not be generated.

[0183] As another example, if the monitoring resource is a subset of beam set A rather than the entire beam set A, it may be impossible to calculate valid performance metrics because the predicted beams based on the inference results using the monitoring beam set are not sufficiently included.

[0184] As another example, depending on the UE's CSI reporting and / or the capability of the AI ​​model, connected pairs for calculating performance metrics may not be generated because the AI ​​model fails to perform operations or measurement actions at the time monitoring resources are delivered or at the time inference results need to be inferred.

[0185] Since beam prediction accuracy, a performance metric, is expressed as the number of successes (N_s), which indicates how many predictions were successful in N calculations, it may be necessary to transmit the number of connected pairs, N', back to the base station when the number of actual valid performance metrics is small.

[0186] If only the number of successes N_s is included in the performance indicator report message at step S1016, the UE may also include N', the number of updated connected pairs, in the performance indicator report message.

[0187] Meanwhile, in the case of Equation 2, a decimal form may occur due to the division calculation. Therefore, to reduce the amount of information, the success probability may be transmitted by mapping it into a quantized form within a fixed interval. In this case, the quantized mapping table for the success probability may need to be shared between the UE and the base station. Therefore, the quantized mapping table may be agreed upon in the standard or transmitted through CSI report configuration information.

[0188] If, as previously explained, the number of actual valid performance indicators is less than the value N set by the base station, the success probability can be newly calculated as shown in Equation 3 below based on the number of connected pairs value N'.

[0189]

[0190] The success probability calculated by Equation 3 can be replaced with a value mapped to a quantized table and then included in a performance indicator report message and transmitted from the UE to the base station. At this time, the UE may also include an updated number of linked pairs, N', in the performance indicator report message to indicate that the success probability is based on N' linked pairs that are measurable in the performance indicator report message.

[0191] Alternatively, if the UE fails to secure N connected pairs set by the base station, it may stop generating the performance indicator report message. In this case, steps S1016 and S1018 may not be performed.

[0192] As another example, if the UE fails to secure N connected pairs set by the base station, it can be configured to include information and a reason (failure to secure N connected pairs) indicating that the performance indicator calculation failed in the performance indicator report message at step S1016.

[0193] If a performance indicator report message is generated using any of the methods described above, in step S1018, the UE can transmit the generated performance indicator report message to the base station. The performance indicator report message may be transmitted through the same resources at the time of transmission as the CSI report message used in 5G NR, or through dedicated resources configured for separate performance indicator reporting. In step S1018, the base station can receive the performance indicator report message from the UE based on the CSI report configuration information.

[0194] Meanwhile, the embodiment of FIG. 10 may assume a case where an AI model is used in the previously described UE and performance monitoring of the AI ​​model is also performed in the AI ​​model. However, based on the above description, even if performance monitoring of the AI ​​model is performed at a base station or the AI ​​model itself is included in the base station, it can be applied through a variation of the method described above.

[0195] The operation of the method according to an embodiment of the present disclosure can be implemented as a computer-readable program or code on a computer-readable recording medium. A computer-readable recording medium includes all types of recording devices 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 the computer-readable program or code to be stored and executed in a distributed manner.

[0196] In addition, computer-readable recording media may include hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Program instructions may include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc.

[0197] Some aspects of the present disclosure 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.

[0198] 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.

[0199] Although the present disclosure 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 present disclosure without departing from the spirit and scope of the present disclosure as set forth in the following claims.

Claims

Claim 1 A method of user equipment (UE), comprising: receiving a higher-level signaling message from a base station containing configuration information for reporting a performance metric (PM) of an intelligent model, wherein the configuration information includes the number of PM acquisitions and information regarding the number of prediction successes of the intelligent model to be reported; receiving reference signals (RSs) from the base station until the number of PM acquisitions is reached; outputting one or more prediction values ​​by using the measurement result of each RSs as an input to the intelligent model whenever the RSs are received; determining the PM using the measurement result of the RSs and one or more prediction values ​​predicted at a previous time point; and, when the number of PM calculations is reached, transmitting a PM report message to the base station containing information regarding the number of prediction successes based on the calculated PM value.