Artificial intelligence-based monitoring method and device

Abstract

The present disclosure relates to a 5G or 6G communication system for supporting higher data transmission rates than a 4G communication system, such as LTE. This method by a base station may comprise the operations of: while artificial intelligence(AI)-based channel state information (CSI) compression by a terminal (UE) is being deactivated, receiving a report message including a monitoring input related to monitoring of the AI-based CSI compression from the terminal; performing first monitoring of the AI-based CSI compression on the basis of the monitoring input; transmitting an indication message for activating the AI-based CSI compression, which is generated on the basis of a result of the first monitoring, to the terminal; receiving compressed CSI obtained by the AI-based CSI compression from the terminal after transmitting the indication message; and performing second monitoring of the AI-based CSI compression on the basis of the compressed CSI.

Description

Artificial intelligence-based monitoring method and device

[0001] The present disclosure relates to a method and device for performing artificial intelligence (AI)-based monitoring in a wireless communication system.

[0002] Looking back at the evolution of wireless communication over successive generations, technologies have primarily been developed for human-facing services such as voice, multimedia, and data. With the commercialization of 5G (5th-generation) communication systems, an explosive increase in connected devices is expected to be connected to communication networks. Examples of networked objects include vehicles, robots, drones, home appliances, displays, smart sensors installed in various infrastructures, construction equipment, and factory equipment. Mobile devices are expected to evolve into diverse form factors, including augmented reality glasses, virtual reality headsets, and holographic devices. In the 6th-generation (6G) era, efforts are being made to develop improved 6G communication systems to connect hundreds of billions of devices and objects and provide diverse services. For this reason, 6G communication systems are often referred to as "beyond 5G."

[0003] The 6G communication system, expected to be realized around 2030, will have a maximum transmission speed of terabytes per second (i.e., 1,000 gigabits per second) and a wireless latency of 100 microseconds (μsec). In other words, compared to 5G, the transmission speed in a 6G communication system will be 50 times faster, while the wireless latency will be reduced to one-tenth.

[0004] To achieve these high data rates and ultra-low latency, 6G communication systems are being considered for implementation in the terahertz band (e.g., from 95 gigahertz (GHz) to 3 terahertz (THz)). Compared to the millimeter wave (mmWave) band introduced in 5G, the terahertz band is expected to experience more severe path loss and atmospheric absorption, making it more crucial to ensure signal reach, or coverage, in this band. Key technologies to ensure coverage include radio frequency (RF) components, antennas, new waveforms that offer better coverage than OFDM (orthogonal frequency division multiplexing), beamforming, and multiple antenna transmission technologies such as massive multiple-input and multiple-output (MIMO), full-dimensional MIMO (FD-MIMO), array antennas, and large-scale antennas. In addition, new technologies such as metamaterial-based lenses and antennas, high-dimensional spatial multiplexing using orbital angular momentum (OAM), and reconfigurable intelligent surfaces (RIS) are being discussed to improve the coverage of terahertz band signals.

[0005] In addition, in order to improve frequency efficiency and system network, 6G communication systems are developing full duplex technology that utilizes the same frequency resources for uplink and downlink at the same time; network technology that integrates satellites and high-altitude platform stations (HAPS); network structure innovation technology that supports mobile base stations and enables optimization and automation of network operation; dynamic spectrum sharing technology through collision avoidance based on spectrum usage prediction; AI-based communication technology that utilizes artificial intelligence (AI) from the design stage and internalizes end-to-end AI support functions to realize system optimization; and next-generation distributed computing technology that realizes services with complexity that exceeds the limits of terminal computing capabilities by utilizing ultra-high-performance communication and computing resources (mobile edge computing (MEC), cloud, etc.). In addition, efforts are being made to further strengthen connectivity between devices, further optimize networks, promote softwareization of network entities, and increase the openness of wireless communications through the design of new protocols to be used in 6G communication systems, the implementation of hardware-based security environments, the development of mechanisms for the safe use of data, and the development of technologies for maintaining privacy.

[0006] Research and development of these 6G communication systems are expected to enable a new level of hyper-connected experience through the hyper-connectivity of 6G communication systems, which encompass not only connections between things but also connections between people and things. Specifically, 6G communication systems are expected to enable services such as truly immersive extended reality (Truly Immersive XR), high-fidelity mobile holograms, and digital replicas. Furthermore, services such as remote surgery, industrial automation, and emergency response, which are provided through enhanced security and reliability, will find application in diverse fields such as industry, healthcare, automotive, and home appliances.

[0007] For example, wireless communication systems, such as 6G communication systems, can utilize beam management and multiple frequency bands to support a variety of services. User equipment (UE) can perform channel estimation for various channel environments on a beam-by-beam and / or frequency-by-band basis and report channel state information (CSI) representing the estimated channel to the base station. CSI compression techniques can be used to reduce the resource consumption of CSI reporting by UEs.

[0008] In 5G NR systems, CSI reports can be generated using a codebook approach. The channel state recovered using the codebook can suffer from information loss depending on the quantization level of the codebook. Diversifying the codebook to reduce information loss increases the amount of data to be transmitted via CSI. Accordingly, in addition to codebook-based channel feedback in 5G NR, CSI compression technologies are being discussed. These technologies compress and transmit the estimated channel state at the terminal using an AI-based encoder, and the base station then decompresses the compressed channel state using an AI-based decoder.

[0009] The present disclosure provides a method and apparatus for monitoring loss of channel state information (CSI) compression in a wireless communication system.

[0010] The present disclosure provides a method and device for performing network-side monitoring for CSI compression in a wireless communication system.

[0011] The present disclosure provides a method and device for monitoring the performance of artificial intelligence (AI)-based CSI compression by a network.

[0012] The present disclosure provides a method and device for enabling disabled artificial intelligence (AI)-based CSI compression by a network.

[0013] A method by a base station according to one embodiment may include receiving a report message from a terminal (UE) including a monitoring input related to monitoring of artificial intelligence (AI)-based channel state information (CSI) compression while the AI-based CSI compression by the terminal is disabled, performing a first monitoring of the AI-based CSI compression based on the monitoring input, transmitting an instruction message to the terminal for activating the AI-based CSI compression generated based on a result of the first monitoring, receiving compressed CSI by the AI-based CSI compression from the terminal after transmitting the instruction message, and performing a second monitoring of the AI-based CSI compression based on the compressed CSI.

[0014] A method by a terminal (UE) according to one embodiment may include transmitting a report message including a monitoring input related to monitoring of AI-based channel state information (CSI) compression to a base station while the AI-based CSI compression is disabled, receiving an instruction message from the base station to activate the AI-based CSI compression after transmitting the report message, and transmitting compressed CSI by the AI-based CSI compression to the base station after receiving the instruction message.

[0015] A base station according to one embodiment may include a transceiver and a processor connected to the transceiver. The processor may be configured to perform the following operations: receiving a report message including a monitoring input related to monitoring of AI-based channel state information (CSI) compression from a terminal (UE) while AI-based CSI compression by the terminal is disabled; performing a first monitoring of the AI-based CSI compression based on the monitoring input; transmitting an instruction message to the terminal for activating the AI-based CSI compression generated based on a result of the first monitoring; receiving compressed CSI by the AI-based CSI compression from the terminal after transmitting the instruction message; and performing a second monitoring of the AI-based CSI compression based on the compressed CSI.

[0016] A terminal (UE) according to one embodiment may include a transceiver and a processor connected to the transceiver. The processor may be configured to perform the following operations: while AI-based channel state information (CSI) compression is disabled, transmit a report message to a base station, the report message including a monitoring input related to monitoring of the AI-based CSI compression; after transmitting the report message, receive an instruction message from the base station to activate the AI-based CSI compression; and after receiving the instruction message, transmit compressed CSI by the AI-based CSI compression to the base station.

[0017] The present disclosure may provide a method for monitoring loss of channel state information (CSI) compression in a wireless communication system.

[0018] The above and other aspects, features and advantages of specific embodiments of the present disclosure will become more apparent from the following description taken in conjunction with the accompanying drawings.

[0019] FIG. 1 is a diagram illustrating a CSI reporting procedure according to one embodiment of the present disclosure.

[0020] FIG. 2 is a diagram for explaining non-AI-based network-side monitoring and terminal-side monitoring according to one embodiment of the present disclosure.

[0021] FIG. 3 is a diagram for explaining network-side monitoring according to one embodiment of the present disclosure.

[0022] FIG. 4 is a drawing for explaining a monitoring type instruction according to one embodiment of the present disclosure.

[0023] FIG. 5 is a sequence diagram for explaining a procedure for performing terminal-side monitoring according to one embodiment of the present disclosure.

[0024] FIG. 6 is a sequence diagram for explaining a procedure for performing network-side monitoring according to one embodiment of the present disclosure.

[0025] FIG. 7 is a sequence diagram illustrating a procedure for activating AI-based CSI compression according to one embodiment of the present disclosure.

[0026] FIG. 8 is a sequence diagram illustrating a procedure for performing terminal-side monitoring fallback according to one embodiment of the present disclosure.

[0027] FIG. 9 is a diagram showing an example configuration of a terminal in a wireless communication system according to one embodiment of the present disclosure.

[0028] FIG. 10 is a diagram showing an example configuration of a base station in a wireless communication system according to one embodiment of the present disclosure.

[0029] Hereinafter, embodiments of the present disclosure will be described with reference to the attached drawings. In the following description of the present disclosure, detailed descriptions of related known functions or configurations will be omitted if they are deemed to unnecessarily obscure the gist of the present disclosure. Furthermore, the terms described below are defined based on their functions in the present disclosure and may vary depending on the intent or custom of the user or operator. Therefore, their definitions should be based on the overall content of the present disclosure.

[0030] For the same reason, some components may be omitted or schematically depicted in the attached drawings. Furthermore, the dimensions of each component do not entirely reflect its actual size. Identical or corresponding components in each drawing may be assigned the same reference numbers.

[0031] The advantages and features of the present disclosure, and methods for achieving them, will become clearer with reference to the embodiments described below in detail with the accompanying drawings. However, the present disclosure is not limited to the embodiments disclosed below and may be implemented in various different forms. The various embodiments are provided to ensure that the present disclosure is complete and to fully convey the scope of the present disclosure to those skilled in the art, and the scope of the present disclosure is defined solely by the scope of the claims. Like reference numerals may refer to like elements throughout the specification.

[0032] It will be appreciated that each block of the flowchart drawings and combinations of the flowchart drawings can be implemented by computer program instructions. These computer program instructions can be installed in a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing equipment, such that the instructions, when executed by the processor of the computer or other programmable data processing equipment, create a means for performing the functions described in the flowchart block(s). These computer program instructions can also be stored in a computer-available or computer-readable memory that can direct a computer or other programmable data processing equipment to implement the functions in a particular manner, such that the instructions stored in the computer-available or computer-readable memory can produce an article of manufacture that includes instruction means for performing the functions described in the flowchart block(s). Since the computer program instructions may be installed on a computer or other programmable data processing device, a series of operational steps may be performed on the computer or other programmable data processing device to create a computer-executable process, and the instructions that cause the computer or other programmable data processing device to perform the steps for performing the functions described in the flowchart block(s) may also provide steps for performing the functions described in the flowchart block(s).

[0033] Each block may represent a module, segment, or portion of code that contains one or more executable instructions for performing a specific logical function(s). It should be noted that in some alternative implementation examples, the functions described in the blocks may occur out of order. For example, two blocks depicted in succession may actually be executed substantially concurrently, or the blocks may sometimes be executed in reverse order, depending on their respective functions.

[0034] The term "unit" used in the embodiments of the present disclosure refers to a software or hardware component, and the "unit" may perform certain roles. However, the "unit" is not limited to software or hardware. The "unit" may be configured to reside on an addressable storage medium and may be configured to trigger one or more processors. Thus, as an example, the "unit" includes components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables. The functionality provided within the components and "units" may be combined into a smaller number of components and "units" or further separated into additional components and "units." In addition, the components and "units" may be implemented to trigger one or more CPUs within a device or a secure multimedia card. Additionally, in various embodiments of the present disclosure, '~bu' may include one or more processors.

[0035] In this disclosure, phrases such as “A and / or B,” “A or B,” “at least one of A and B,” “at least one of A or B,” “A, B, or C,” “at least one of A, B, and C,” and “at least one of A, B, or C” can each include any one of the items listed together in that phrase, or all possible combinations thereof. Terms such as “first,” “second,” or “first” or “second” may be used merely to distinguish the corresponding component from other corresponding components and do not limit the corresponding components in any other respect (e.g., importance or order).

[0036] In embodiments of the present disclosure, a user equipment (UE) may be a terminal, a mobile station (MS), a cellular phone, a smartphone, a computer, or any other electronic device capable of performing a communication function. In embodiments of the present disclosure, a base station (BS) is a network entity that performs resource allocation to a UE, and may be at least one of a Node B, an eNB (eNode B), a gNB (gNode B), a wireless access unit, a base station controller, or a node on a network.

[0037] The embodiments of the present disclosure described below may also be applied to other communication systems with similar technical backgrounds or channel configurations. The embodiments of the present disclosure may be applied to other communication systems with some modifications, as determined by a person skilled in the art, without significantly departing from the scope of the present disclosure.

[0038] In specifically describing the embodiments of the present disclosure, the communication system may use a wireless communication system, for example, the LTE system proposed by 3GPP (3rd generation partnership project long term evolution), a wireless communication standard standardization organization, the 5G communication system based on the 5G communication standard (NR (New RAN)), or the communication system after 5G (e.g., 6G communication system). In addition, it may be applied to other communication systems with similar technical backgrounds with slight modifications within a range that does not significantly deviate from the scope of the present disclosure, and this may be possible at the discretion of a person skilled in the technical field of the present disclosure. For the convenience of the following description, some terms and names defined in the 3GPP standard may be used. However, the present disclosure is not limited by the above terms and names, and may be equally applied to systems following other standards.

[0039] While the embodiments of the present disclosure will be described below with reference to AI-based CSI compression, the embodiments of the present disclosure can be extended and applied to various technologies, including model monitoring. Examples of technologies to which model monitoring according to embodiments of the present disclosure can be applied are as follows.

[0040] - When ground truth and AI model output exist in UE and NW separately; and / or

[0041] - Large overhead for groundtruth (or AI model output) transmission

[0042] FIG. 1 is a diagram illustrating a CSI reporting procedure according to one embodiment of the present disclosure.

[0043] Referring to FIG. 1, a base station (110) (e.g., gNB) may transmit a CSI-RS (102), which is a reference signal (RS) transmitted to identify a channel state of a terminal (UE) (100). Prior to transmitting the CSI-RS (102), the base station (110) may transmit (e.g., broadcast or unicast) CSI-RS configuration information that sets parameters related to transmission of the CSI-RS (102). The terminal (100) may receive the CSI-RS (102) based on the CSI-RS configuration information. In one embodiment, the CSI-RS configuration information may define at least one of a sequence, a frequency resource, a time resource, and / or a period included in the CSI-RS (102).

[0044] The terminal (100) can estimate a channel state based on the CSI-RS (102) and report CSI (104) indicating the estimated channel state to the base station (110). Prior to reporting the CSI (104), the base station (110) can transmit (e.g., broadcast or unicast) CSI configuration information that sets parameters related to transmission of the CSI (104). The terminal (100) can transmit the CSI (104) based on the CSI configuration information.

[0045] In one embodiment, the CSI configuration information may include at least one of a type of information included in the CSI (104), a frequency resource, a time resource, and / or a periodic / aperiodic configuration. In one embodiment, the CSI (104) may include at least one of a channel quality indicator (CQI), a precoding matrix indicator (PMI), a CSI-RS resource indicator, a synchronization signal / physical broadcast channel (SS / PBCH) block resource indicator, a layer indicator (LI), a rank indicator, a layer 1 reference signal received power (L1-RSRP), a layer 1 signal-to-noise and interference ratio (L1-SINR), or a capability index. As an example, the PMI may indicate a precoding matrix including weights to be applied to multi-layer signals.

[0046] The base station (110) can determine resource allocation for downlink or uplink transmission based on the CSI (104) reported from the terminal (100). In one embodiment, the base station (110) can transmit resource allocation for downlink transmission (e.g., physical downlink control channel (PDCCH)) and downlink data (e.g., physical data shred channel (PDSCH)) (106) to the terminal (100).

[0047] In the CSI reporting procedure, the terminal (100) can perform channel estimation based on the CSI (104) and determine the estimated channel H_hat. The terminal (100) can perform eigenvalue decomposition (EVD) based on the estimated channel to generate a PMI related to a precoding matrix to be used by the base station (110) for downlink transmission, and determine the matrix V as in <Mathematical Formula 1> below.

[0048] [Mathematical Formula 1]

[0049]

[0050] Here represents the estimated matrix (e.g. H_hat), and the superscript H represents the conjugate transpose, represents conjunction.

[0051] The terminal (100) can generate a PMI including a codebook index corresponding to the matrix V based on a pre-arranged codebook, for example, a codebook defined in 3GPP technical specification (TS) 38.214. The terminal (100) can include the PMI in the CSI (104) and report it to the base station (110).

[0052] The codebook-based PMI may differ from the V actually measured by the terminal (100) due to the granularity limitation of the codebook, which may result in inaccurate resource allocation determined by the base station (110), which may result in MIMO performance degradation. For example, since the performance of a multi-user MIMO system is greatly affected by the CSI (e.g., CSI (104)) fed back by each terminal (e.g., terminal (100)), the MU-MIMO system may require more accurate CSI feedback. Increasing the number of bits used for CSI reporting for more accurate CSI feedback may incur a very large overhead as the number of antennas, bandwidth, and / or codebook granularity of the base station (110) increases.

[0053] In 3GPP Rel-18 SI (study item), AI / ML-based CSI compression is being discussed, which compresses and reports the matrix V itself instead of indicating a codebook matrix similar to the matrix V. In the AI / ML-based CSI compression, the terminal (100) can utilize artificial intelligence (AI) and / or machine learning (ML) (hereinafter referred to as AI / ML) to compress the matrix V itself to generate a low-dimensional vector z, and include the vector z in the PMI to report to the base station (110). The base station (110) can restore the matrix V from the vector z and utilize the restored matrix V for channel adaptive scheduling (e.g., resource allocation).

[0054] AI modeling, such as AI / ML-based CSI compression, may include monitoring to verify that the AI ​​modeling is operating accurately. In one embodiment, monitoring in AI / ML-based CSI compression may include calculating a loss (e.g., similarity) between the ground truth (e.g., matrix V) obtained by the terminal (100) through measurement and the matrix (e.g., V_hat) reconstructed by the network (e.g., base station (110)). In this disclosure, monitoring of AI / ML-based CSI compression will be described as an example of monitoring of AI modeling. AI / ML-based CSI compression may experience performance degradation due to a bad training / validation dataset, imperfect model selection and switching, data distribution shift, or unexpected events, and such performance degradation may be calculated as Loss(V, V_hat).

[0055] Monitoring of AI / ML-based CSI compression can be categorized into terminal-side monitoring and network-side monitoring. In one embodiment, at least one of the terminal-side monitoring or the network-side monitoring can be performed based on AI. Terminal-side monitoring can include an operation in which the terminal (100) calculates a reconstructed matrix V_hat based on a vector z generated by compressing (e.g., auto-encoding) the matrix V, and calculates a loss between the original matrix V and the reconstructed V_hat. Terminal-side monitoring can depend on the AI ​​capability of the terminal (100). In one embodiment, the AI ​​capability of the terminal (100) can include AI computing-related overhead (e.g., input / output (I / O) memory bandwidth, and / or AI-driven floating point operations (FLOPs)) and / or AI model size-related overhead (e.g., memory storage for AI model structure / parameters).

[0056] Network-side monitoring may include an operation in which a network (e.g., a base station (110)) obtains monitoring information (e.g., a ground truth matrix V or a quantized matrix V_q) from a terminal (100), calculates a restored matrix V_hat based on a vector z fed back from the terminal (100), and calculates a loss between the obtained matrix V and the restored matrix V_hat. Network-side monitoring may be performed using ground truth (e.g., V or V_q) reported from the terminal (100) to the base station (110).

[0057] FIG. 2 is a diagram for explaining non-AI-based network-side monitoring and terminal-side monitoring according to one embodiment of the present disclosure.

[0058] Referring to FIG. 2, the terminal (100) may include an encoder (204) that performs CSI compression. The terminal (100) may obtain a matrix V (202) for channel state feedback. In one embodiment, the matrix V (202) may be obtained by performing eigenvalue decomposition on a channel (e.g., H_hat) estimated by the terminal (100) based on a CSI-RS (e.g., CSI-RS (102)) received from a base station (110). The encoder (204) may encode (e.g., compress) the matrix V (202) based on a specified AI model and output a vector z (206). In one embodiment, the vector z (206) may include at least one of the elements of the matrix V (202) and / or information related thereto. The terminal (100) can include the vector z (206) in the CSI (e.g., CSI (104)) and feed it back to the base station (110).

[0059] The base station (110) may include a decoder (214) configured to operate in response to the encoder (204) of the terminal (100). The base station (110) may obtain a vector z (212) (e.g., vector z (206)) from CSI (e.g., CSI (104)) received from the terminal (100). The decoder (214) may decode (e.g., decompress) the vector z (212) based on a specified AI model and output a restored matrix V_hat (216).

[0060] In one embodiment, the base station (110) may include a monitoring unit (218) for non-AI-based network-side monitoring. The monitoring unit (218) of the base station (110) may receive the restored matrix V_hat (216) as input, and may also receive a ground truth matrix V (202) from the terminal (100). The monitoring unit (218) may compare the matrix V (202) and the restored matrix V_hat (216) to calculate a squared generalized cosine similarity (SGCS) (e.g., SGCS_true (220)) corresponding to the actual loss. The base station (110) may utilize the monitoring results by the monitoring unit (218) in various ways. In one embodiment, the base station (110) may re-configure AI-based CSI compression based on the monitoring results. In one embodiment, the base station (110) may instruct the terminal (100) to at least temporarily disable AI-based CSI compression based on a loss resulting from monitoring that falls outside a specified threshold range.

[0061] In one embodiment, the terminal (100) may include a monitoring unit (208) for terminal-side monitoring. The monitoring unit (208) of the terminal (100) may receive a vector z (206), which is an output of the encoder (204). The monitoring unit (208) may be configured to operate in response to the encoder (204). In one embodiment, the monitoring unit (208) may predict a loss based on the vector z (206) according to a specified AI model without restoring the matrix V_hat, and may calculate an SGCS (e.g., SGCS_pred (210)) corresponding to the predicted loss. Depending on the AI ​​model, SGCS_pred (210) may be similar to SGCS_true. The terminal (100) may utilize the monitoring results by the monitoring unit (208) in various ways. In one embodiment, the terminal (100) may report information indicating a monitoring result (e.g., a predicted loss) to the base station (110). In one embodiment, the terminal (100) may request the base station (110) to re-enable AI-based CSI compression based on the monitoring results. In one embodiment, the terminal (100) may request (or notify) the base station (110) to at least temporarily disable AI-based CSI compression based on the predicted loss based on the monitoring results falling outside a specified threshold range.

[0062] FIG. 3 is a diagram for explaining network-side monitoring according to one embodiment of the present disclosure.

[0063] Referring to FIG. 3, the terminal (100) may include an encoder (304) that performs CSI compression. The terminal (100) may obtain a matrix V (302) for channel state feedback. In one embodiment, the matrix V (302) may be obtained by performing eigenvalue decomposition on a channel (e.g., H_hat) estimated by the terminal (100) based on a CSI-RS (e.g., CSI-RS (102)) received from a base station (110). The encoder (304) may encode (e.g., compress) the matrix V (302) based on a specified AI model and output a vector z (306). In one embodiment, the vector z (306) may include at least one of the elements of the matrix V (302) and / or information related thereto. The terminal (100) can include the vector z (306) in the CSI (e.g., CSI (104)) and feed it back to the base station (110).

[0064] The base station (110) may include a decoder (314) configured to operate in response to the encoder (304) of the terminal (100). The base station (110) may obtain a vector z (312) (e.g., vector z (306)) from CSI (e.g., CSI (104)) received from the terminal (100). The decoder (314) may decode (e.g., decompress) the vector z (312) based on a specified AI model and output a restored matrix V_hat (316).

[0065] In one embodiment, the base station (110) may include a monitoring unit (318) for network-side monitoring. The monitoring unit (318) of the base station (110) may receive the received vector z (312). The monitoring unit (318) may be configured to operate in response to the encoder (204). In one embodiment, the monitoring unit (318) may use the vector z (312) based on a specified AI model to predict a loss between the matrix V generated by the terminal (100) and the matrix V_hat restored by the base station (110), and may calculate an SGCS (e.g., SGCS_pred) corresponding to the predicted loss. The base station (110) may utilize the monitoring results by the monitoring unit (318) in various ways. In one embodiment, the base station (110) may re-configure AI-based CSI compression based on the monitoring results. In one embodiment, the base station (110) may instruct the terminal (100) to at least temporarily disable AI-based CSI compression based on a loss resulting from monitoring that falls outside a specified threshold range.

[0066] When performing AI-based network-side monitoring by a monitoring unit (318) at a base station (110), AI-based CSI compression can be free from limitations of terminal capabilities, and can also be advantageous in terms of LCM (lifecycle management) because the base station (110) is responsible for both management and operation of the AI ​​model.

[0067] In one embodiment, the base station (110) may set the terminal (100) to one of non-AI-based network-side monitoring, AI-based terminal-side monitoring, or AI-based network-side monitoring based on UE capabilities and / or monitoring results. In one embodiment, the base station (110) may transmit information indicating a monitoring type (e.g., a monitoring type indication (400) of FIG. 4) to the terminal (100).

[0068] FIG. 4 is a drawing for explaining a monitoring type instruction according to one embodiment of the present disclosure.

[0069] Referring to FIG. 4, the monitoring type indication (400) may be set to a value indicating either Type 0, Type 1, or Type 2. In one embodiment, Type 0 may indicate non-AI-based network-side monitoring (402), Type 1 may indicate AI-based terminal-side monitoring (404), and Type 2 may indicate AI-based network-side monitoring (406).

[0070] In one embodiment, the base station (110) may determine whether to set the monitoring type indication (400) to Type 0 indicating non-AI based network-side monitoring (402) based on overhead and / or latency due to the operation of the terminal (100) reporting ground truth (e.g., matrix V) for non-AI based network-side monitoring (402). For example, the base station (110) may set the monitoring type indication (400) to Type 0 if the overhead of the ground truth is not greater than a specified threshold value and / or the latency for reporting the ground truth is not greater than a specified threshold value.

[0071] In one embodiment, the base station (110) may determine whether to set the monitoring type indication (400) to Type 1, which indicates AI-based terminal-side monitoring, based on the AI ​​capability (e.g., AI model inference-related capability) of the UE capability report reported by the terminal (100) (e.g., the UE capability report of operation 502 of FIG. 5 ). For example, if the UE capability report includes information indicating AI operation availability and / or the AI ​​computing capability and AI model size-related capability indicated by the UE capability report are determined to be appropriate for performing AI-based terminal-side monitoring, the base station (110) may set the monitoring type indication (400) to Type 1.

[0072] In one embodiment, the base station (110) may set the monitoring type indication (400) to Type 2 to indicate network-side monitoring (406) if it is not determined to be non-AI based network-side monitoring (402) or AI based terminal-side monitoring (404).

[0073] Figure 5 is a sequence diagram illustrating a procedure for performing terminal-side monitoring according to one embodiment of the present disclosure. At least one of the operations described below may be omitted, modified, or executed in a different order.

[0074] Referring to FIG. 5, in operation 502, the terminal (100) may transmit a UE capability report indicating AI model inference related capabilities to a network (e.g., a base station (110)). In one embodiment, the UE capability report may include at least one of the following information.

[0075] - Information indicating whether AI capabilities are supported

[0076] - AI computing capability information (e.g., input / output memory bandwidth and / or AI-driven FLOPS)

[0077] - AI model size-related capability information (e.g. AI model parameters, memory, and / or storage)

[0078] - Supported model representation formats (MRF) (e.g. ONNX (open neural network exchange))

[0079] - Information indicating the occurrence or resolution of a resource restriction (e.g., heat, power, or battery-related event).

[0080] In one embodiment, a UE capability report may include information indicating whether AI is enabled or disabled. If AI is disabled, existing CSI reporting methods (e.g., codebook-based CSI reporting) may be used.

[0081] In one embodiment, a UE capability report may include a value indicating at least one of the AI ​​capability categories in Table 1 below.

[0082] [Table 1]

[0083]

[0084] The terminal (100) may include at least one category value in [Table 1] in the UE capability report.

[0085] In one embodiment, a UE capability report may be transmitted based on the occurrence of an event related to heat (e.g., CPU temperature), battery status (e.g., battery level), or power status (e.g., charging). For example, based on identifying that the terminal (100) has entered a power saving mode, the terminal (100) may include information indicating AI incapability in the UE capability report and transmit it. For example, based on identifying that the terminal (100) has exited the power saving mode, the terminal (100) may include information indicating AI capability in the UE capability report and transmit it. For example, based on identifying that the terminal (100) is connected to a charging power source, the terminal (100) may include information indicating AI capability in the UE capability report and transmit it. For example, based on identifying that the terminal (100) is disconnected from a charging power source, the terminal (100) may include information indicating AI incapability in the UE capability report and transmit it.

[0086] In operation 504, the base station (110) may transmit monitoring configuration information that sets monitoring parameters related to the AI-based CSI compression to the terminal (100). In one embodiment, the monitoring configuration information may include a monitoring type indication (e.g., a monitoring type indication (400) of FIG. 4). The monitoring type indication (400) may include a value indicating any one of non-AI-based network-side monitoring (402), AI-based terminal-side monitoring (404), or AI-based network-side monitoring (406). In one embodiment, the base station (110) may set the monitoring type indication (400) included in the monitoring configuration information to Type 1 indicating AI-based terminal-side monitoring (404).

[0087] In one embodiment, the monitoring configuration information may further include configuration information for monitoring result reporting (e.g., monitoring result report configuration information). The monitoring result report configuration information may be used by the terminal (100) to transmit a monitoring result report (e.g., operation 508) to the base station (110).

[0088] In operation 506, the terminal (100) can perform AI-based monitoring using a terminal-side monitoring model based on the monitoring configuration information, thereby confirming and verifying the validity of the AI-based CSI compression. In one embodiment, the terminal (100) can perform the monitoring based on identifying that the monitoring configuration information includes a monitoring type indication (400) set to type 1. In one embodiment, the terminal (100) can perform the monitoring to determine the validity of the compressed CSI while performing an AI-based CSI compression procedure (not shown) (e.g., operation 606)). In one embodiment, the AI-based CSI compression procedure can include receiving a CSI-RS (102) from a base station (110) and reporting compressed CSI (e.g., CSI (104)) based on AI-based CSI compression to the base station (110).

[0089] In operation 508, the terminal (100) may transmit a monitoring result report based on the result of performing the monitoring to the base station (110). In one embodiment, the terminal (100) may transmit the monitoring result report according to a method defined by the monitoring configuration information (e.g., time and / or frequency resources). In one embodiment, the terminal (100) may include information defined by the monitoring configuration information in the monitoring result report. In one embodiment, the monitoring result report may include information (e.g., SGCS_pred) indicating a predicted loss obtained by the terminal (100) through AI-based terminal-side monitoring. In one embodiment, the monitoring result report may include information requesting activation or deactivation of AI-based CSI compression. The base station (110) may reset, activate, or deactivate the AI-based CSI compression of the terminal (100) based on the monitoring result report.

[0090] Figure 6 is a sequence diagram illustrating a procedure for performing network-side monitoring according to one embodiment of the present disclosure. At least one of the operations described below may be omitted, modified, or executed in a different order.

[0091] Referring to FIG. 6, in operation 602, the terminal (100) may transmit a UE capability report to a network (e.g., a base station (110)). In one embodiment, the UE capability report may be similar to the UE capability report of operation 502.

[0092] In operation 604, the base station (110) may transmit monitoring configuration information to the terminal (100). In one embodiment, the monitoring configuration information may include a monitoring type indication (e.g., a monitoring type indication (400) of FIG. 4). In one embodiment, the base station (110) may set the monitoring type indication (400) included in the monitoring configuration information to Type 2, which indicates AI-based network-side monitoring (406).

[0093] In one embodiment, when the monitoring configuration information includes a monitoring type instruction (400) indicating AI-based network-side monitoring (406) of type 2, the monitoring configuration information may further include at least one of the following information:

[0094] - Information indicating whether to perform terminal-side monitoring fallback (e.g., a 1-bit field, or the presence of a specified field)

[0095] - Setting information for monitoring input reporting and / or monitoring result reporting (e.g., monitoring input reporting setting information and / or monitoring result reporting setting information)

[0096] In one embodiment, the monitoring input report setting information may be used by the terminal (100) to transmit a monitoring input report (e.g., operation 708) to the base station (110), and / or by the terminal (100) to transmit a monitoring result report (e.g., operation 810) to the base station (110).

[0097] In operation 606, the terminal (100) may perform AI-based CSI compression. In one embodiment, the terminal (100) may determine not to perform terminal-side monitoring based on identifying that the monitoring configuration information includes a monitoring type indication (400) set to Type 2. Operation 606 may include operations 610, 612, 614, and 616.

[0098] In operation 610, the base station (110) may transmit a CSI-RS (e.g., a CSI-RS (102)). The terminal (100) may receive the CSI-RS (102) and obtain a matrix V from a channel H_hat estimated based on the CSI-RS (102). In operation 612, the terminal (100) may encode (e.g., compress) the matrix V to generate an encoder output (e.g., a vector z).

[0099] In operation 614, the terminal (100) may report the CSI (104) including the encoder output (e.g., vector z) to the base station (110). In operation 616, the base station (110) may decode (e.g., decompress) the vector z included in the CSI (104) to obtain a restored matrix V_hat. The restored matrix V_hat may be used for resource allocation of the base station (110).

[0100] In operation 608, the base station (110) can perform AI-based monitoring using a network-side monitoring model based on the monitoring configuration information of operation 604, thereby verifying and confirming the validity of AI-based CSI compression (e.g., operation 606). The vector z received in operation 614 can be used as an input of the network-side monitoring model. In one embodiment, the base station (100) can perform the monitoring based on identifying that the monitoring configuration information includes a monitoring type indication (400) set to type 2. The base station (110) can reset, activate, or deactivate the AI-based CSI compression of the terminal (100) based on the monitoring result (e.g., SGCS_pred).

[0101] FIG. 7 is a sequence diagram illustrating a procedure for activating AI-based CSI compression according to one embodiment of the present disclosure. At least one of the operations described below may be omitted, modified, or executed in a different order.

[0102] Referring to FIG. 7, AI-based CSI compression (e.g., operation 606) may be performed between a terminal (100) and a base station (110) at operation 702. At operation 704, the base station (110) may transmit a message to the terminal (100) instructing to disable AI-based CSI compression (e.g., an AI-based CSI compression disablement instruction). In one embodiment, the base station (110) may determine to disable AI-based CSI compression of the terminal (100) based on a monitoring result report received from the terminal (100) (e.g., operation 508) or a monitoring result by the base station (110) (e.g., operation 608). In one embodiment, the base station (110) may determine to disable AI-based CSI compression of the terminal (100) based on receiving information requesting to disable AI-based CSI compression (e.g., an AI-based CSI compression disablement request) from the terminal (100). In one embodiment, the terminal (100) may determine to deactivate AI-based CSI compression of the terminal (100) based on a monitoring result (e.g., operation 508) by the terminal (100) and request deactivation of AI-based CSI compression to the base station (110).

[0103] Although not illustrated, the base station (110) may provide information (e.g., monitoring input report configuration information) for configuring monitoring input reporting to the terminal (100) when or before AI-based CSI compression is deactivated. In one embodiment, the monitoring input report configuration information may be included and transmitted in the monitoring configuration information of operation 604. In one embodiment, the base station (110) may transmit the monitoring input report configuration information (or monitoring configuration information including the monitoring input report configuration information) to the terminal (100) together with or after an AI-based CSI compression deactivation instruction.

[0104] In one embodiment, the monitoring input report setting information may include at least one of the following information:

[0105] - Period for reporting monitoring input: For example, 1 second, 2 seconds, or 5 seconds. For example, the period may be set longer than the period for transmitting CSI-RS by the base station (110).

[0106] - Conditions for sending monitoring input reports (e.g., monitoring input report triggering events)

[0107] - Format of information to be included in the monitoring input report: For example, encoder output of the terminal (100) (e.g. vector z), or compressed (e.g. quantized) vector z_q (vector z_q may have a smaller size than vector z)

[0108] In operation 706, the base station (110) may transmit information requesting a monitoring input (e.g., a monitoring input request) to the terminal (100). In one embodiment, the monitoring input request may include monitoring input report setting information (e.g., a format of information to be included in a monitoring input report).

[0109] In operation 708, the terminal (100) may report a monitoring input based on the monitoring input report setting information to the base station (110). In one embodiment, the monitoring input may include information of vector z or vector z_q according to the monitoring input report setting information.

[0110] In operation 710, the base station (110) may perform AI-based monitoring based on the monitoring input. In one embodiment, the base station (110) may perform monitoring using the vector z or vector z_q received through the monitoring input report as input, rather than the vector z reported through the AI-based CSI compression procedure. The base station (110) may determine whether to activate AI-based CSI compression based on the monitoring result (e.g., SGCS_pred).

[0111] In operation 712, the base station (110) may transmit a message to the terminal (100) instructing to activate AI-based CSI compression (e.g., an AI-based CSI compression activation instruction) based on a determination to activate AI-based CSI compression based on the monitoring result.

[0112] At operation 714, the terminal (100) may perform AI-based CSI compression (e.g., operation 606) based on receiving an AI-based CSI compression activation instruction. In one embodiment, after AI-based CSI compression is deactivated at operation 704 and before receiving an AI-based CSI compression activation instruction at operation 712, the terminal (100) may perform CSI reporting according to a conventional CSI reporting method (e.g., codebook-based CSI reporting).

[0113] In operation 716, the base station (110) may perform AI-based monitoring using a network-side monitoring model based on the vector z reported by the terminal (100) through AI-based CSI compression. Operation 716 may be similar to operation 608.

[0114] Figure 8 is a sequence diagram illustrating a procedure for performing terminal-side monitoring fallback according to one embodiment of the present disclosure. At least one of the operations described below may be omitted, modified, or executed in a different order.

[0115] Referring to FIG. 8, AI-based CSI compression (e.g., operation 606) may be performed between a terminal (100) and a base station (110) at operation 802. At operation 804, the base station (110) may transmit a message (e.g., an AI-based CSI compression deactivation instruction) to the terminal (100) instructing to deactivate AI-based CSI compression. Operation 804 may be similar to operation 704.

[0116] Although not shown, the base station (110) may provide information (e.g., a terminal-side monitoring fallback instruction) to the terminal (100) instructing terminal-side monitoring fallback when or before AI-based CSI compression is deactivated. In one embodiment, the terminal-side monitoring fallback instruction may be expressed by a value of a 1-bit field or the presence of a designated field. In one embodiment, the terminal-side monitoring fallback instruction may be transmitted in the monitoring configuration information of operation 604. In one embodiment, the base station (110) may transmit the terminal-side monitoring fallback instruction (or monitoring configuration information including the terminal-side monitoring fallback instruction) to the terminal (100) together with or after the AI-based CSI compression deactivation instruction.

[0117] In operation 806, the terminal (100) may perform monitoring using the terminal-side monitoring model based on the terminal-side monitoring fallback instruction while AI-based CSI compression is disabled. In one embodiment, the terminal (100) may perform encoding, which is part of AI-based CSI compression, to generate vector z. The terminal (100) may not report the vector z by including it in the CSI.

[0118] In operation 808, the base station (110) may transmit a monitoring result request to the terminal (100). In one embodiment, the base station (110) may transmit the monitoring result request to the terminal (100) to determine whether to enable AI-based CSI compression while AI-based CSI compression is disabled.

[0119] In operation 810, the terminal (100) may transmit a monitoring result report to the base station (110). In one embodiment, operation 810 may be similar to operation 508. In one embodiment, the monitoring result report may include information (e.g., SGCS_pred) indicating a predicted loss obtained by the terminal (100) in operation 806. In one embodiment, the monitoring result report may be transmitted by the terminal (100) in response to receiving a monitoring result request in operation 808.

[0120] In one embodiment, operation 808 may be omitted, and the terminal (100) may transmit the monitoring result report based on the monitoring result report configuration information included in the monitoring configuration information. In one embodiment, the monitoring result report configuration information may include at least one of a transmission cycle for reporting the monitoring result, information on an event that triggers the monitoring result report (e.g., a monitoring result report triggering event), and a format of information to be included in the monitoring result report.

[0121] In one embodiment, the terminal (100) may transmit the monitoring result report periodically (e.g., at a period of 1 second, 2 seconds, or 5 seconds) according to the transmission cycle. In one embodiment, the terminal (100) may transmit the monitoring result report according to a specified condition (e.g., the monitoring result report triggering event) related to a loss (e.g., SGCS_pred) obtained through monitoring. For example, the terminal (100) may transmit the monitoring result report including specified information (e.g., SGCS_pred) when SGCS_pred is less than a specified reference value (e.g., -0.7) (SGCS_pred < -0.7). For example, the terminal (100) may transmit the monitoring result report including information requesting activation of AI-based CSI compression (e.g., AI-based CSI compression activation request) when SGCS_pred < -0.7.

[0122] The base station (110) may determine whether to activate AI-based CSI compression based on the monitoring result report. In one embodiment, the base station (110) may determine to activate AI-based CSI compression based on whether SGCS_pred included in the monitoring result report is less than a specified threshold value (e.g., -0.7). In one embodiment, the base station (110) may determine to activate AI-based CSI compression based on whether the monitoring result report includes a request to activate AI-based CSI compression.

[0123] In operation 812, the base station (110) may transmit a message to the terminal (100) instructing to activate AI-based CSI compression (e.g., an AI-based CSI compression activation instruction) based on a determination to activate AI-based CSI compression based on the monitoring result.

[0124] At operation 814, the terminal (100) may perform AI-based CSI compression (e.g., operation 606) based on receiving an AI-based CSI compression activation instruction. In one embodiment, after AI-based CSI compression is deactivated at operation 804 and before receiving an AI-based CSI compression activation instruction at operation 812, the terminal (100) may perform CSI reporting according to a conventional CSI reporting method (e.g., codebook-based CSI reporting).

[0125] In operation 816, the base station (110) may perform AI-based monitoring using a network-side monitoring model based on the vector z reported by the terminal (100) through AI-based CSI compression. Operation 816 may be similar to operation 608.

[0126] FIG. 9 is a diagram illustrating an example configuration of a terminal in a wireless communication system according to an embodiment of the present disclosure. The terminal (100) of FIG. 9 may operate through each of the embodiments of FIGS. 2 to 8 described above, as well as a combination of two or more embodiments. The terminal (100) may include a processor (902), a transceiver (904), and a memory (906). The memory (906) may include instructions that can be executed by the processor (902). The instructions, when executed by the processor (902), may be configured to cause the terminal (100) to operate according to the embodiments of FIGS. 2 to 8. However, the components of the terminal (100) are not limited to the examples described above. For example, the terminal (100) may include more or fewer components than the components described above. In addition, one or more of the processor (902), the transceiver (904), and the memory (906) may be implemented in the form of a single chip.

[0127] The transceiver (904) is a general term for the receiver and transmitter of the terminal (100) and can transmit and receive wireless signals with the base station (110). The transmitted and received signals may include at least one of control information and data. The transceiver (904) can receive a signal through an antenna and output it to the processor (902), and transmit a signal output from the processor (902) through the antenna. The transceiver (904) may include an RF (radio frequency) transmitter that up-converts and amplifies the frequency of a transmitted signal, and an RF receiver that low-noise amplifies and frequency-converts a received signal.

[0128] The memory (906) can store programs and data necessary for the operation of the terminal (100) according to at least one of the embodiments of FIGS. 2 to 8. The memory (906) can store control information or data included in a signal acquired from the terminal (100). The memory (906) can be configured as a storage medium or a combination of storage media, such as a read-only memory (ROM), a random access memory (RAM), a hard disk, a compact disk ROM (CD-ROM), and a digital video disk (DVD).

[0129] The processor (902) may control a series of operations so that the terminal (100) may operate according to at least one of the embodiments of FIGS. 2 to 8. The processor (902) may include at least one processing circuit. The processor (902) may control the transceiver (904) to perform, for example, an operation of identifying, by the terminal (100), that artificial intelligence (AI)-based channel state information (CSI) compression is disabled, an operation of receiving a message requesting monitoring input, an operation of transmitting a monitoring input report, an operation of receiving a message instructing activation of AI-based CSI compression based on a monitoring result, and / or an operation of reporting compressed CSI by AI-based CSI compression.

[0130] FIG. 10 is a diagram illustrating an example configuration of a base station in a wireless communication system according to an embodiment of the present disclosure. The base station (110) of FIG. 10 may operate according to each of the embodiments of FIGS. 2 to 8 described above, as well as a combination of two or more embodiments. The base station (110) may include a processor (1002), a transceiver (1004), and a memory (1006). The memory (1006) may include instructions executable by the processor (1002). The instructions, when executed by the processor (1002), may be configured to cause the base station (110) to operate according to the embodiments of FIGS. 2 to 8. However, the components of the base station (110) are not limited to the examples described above. For example, the base station (110) may include more or fewer components than the components described above. In addition, one or more of the processor (1002), the transceiver (1004), and the memory (1006) may be implemented in the form of a single chip.

[0131] The transceiver (1004) is a general term for the receiver and transmitter of the base station (110) and can transmit and receive wireless signals with the terminal (100). The transmitted and received signals can include at least one of control information and data. The transceiver (1004) can receive a signal through an antenna and output it to the processor (1002), and transmit a signal output from the processor (1002) through the antenna. The transceiver (1004) can include an RF transmitter that up-converts and amplifies the frequency of a transmitted signal, and an RF receiver that low-noise amplifies and frequency-converts a received signal.

[0132] The memory (1006) can store programs and data necessary for the operation of the base station (110) according to at least one of the embodiments of FIGS. 2 to 8. The memory (1006) can store control information or data included in a signal acquired from the base station (110). The memory (1006) can be configured as a storage medium or a combination of storage media, such as a ROM, a RAM, a hard disk, a CD-ROM, and a DVD.

[0133] The processor (1002) may control a series of operations to enable the base station (110) to operate according to at least one of the embodiments of FIGS. 2 to 8. The processor (1002) may include at least one processing circuit. For example, the processor (1002) may control the transceiver (1004) to perform, by the base station (110), an operation of identifying that artificial intelligence (AI)-based channel state information (CSI) compression is disabled, an operation of transmitting a message requesting monitoring input, an operation of receiving a monitoring input report, an operation of determining activation of AI-based CSI compression based on a monitoring result, an operation of transmitting a message indicating activation of AI-based CSI compression, and / or an operation of receiving compressed CSI by AI-based CSI compression.

[0134] Embodiments of the present disclosure can solve problems of transmission overhead and latency caused by the terminal (100) reporting ground truth.

[0135] The embodiments of the present disclosure enable monitoring the performance of AI-based CSI compression by allowing a base station (110) with relatively higher capabilities than a terminal (100) to perform network-side monitoring without limitations on terminal capabilities.

[0136] Embodiments of the present disclosure can increase the freshness of monitoring results by having the base station (110) directly utilize the monitoring results obtained through network-side monitoring.

[0137] Embodiments of the present disclosure may provide a procedure for resuming monitoring when AI-based CSI compression is disabled.

[0138] The methods according to the embodiments described in the claims or specification of the present disclosure may be implemented in the form of hardware, software, or a combination of hardware and software. If implemented in software, a computer-readable storage medium storing one or more programs (software modules) may be provided. The one or more programs stored in the computer-readable storage medium are configured for execution by one or more processors within an electronic device. The one or more programs include instructions that cause the electronic device to execute the methods according to the embodiments described in the claims or specification of the present disclosure.

[0139] These programs (software modules, software) may be stored in random access memory, non-volatile memory including flash memory, read only memory (ROM), electrically erasable programmable read only memory (EEPROM), magnetic disc storage devices, compact disc-ROMs (CD-ROMs), digital versatile discs (DVDs) or other forms of optical storage devices, magnetic cassettes, or may be stored in memories formed by a combination of some or all of these. In addition, each configuration memory may include multiple copies. The above program may be stored on an attachable storage device that is accessible via a communication network such as the Internet, an intranet, a local area network (LAN), a wide area network (WAN), a storage area network (SAN), or a combination thereof. This storage device may be connected to a device performing embodiments of the present disclosure via an external port. Additionally, a separate storage device on the communication network may be connected to a device performing embodiments of the present disclosure.

[0140] In the embodiments of the present disclosure described above, components included in the present disclosure are expressed singularly or plurally, depending on the specific embodiments presented. However, the singular or plural expressions are selected to suit the presented situation for convenience of explanation, and the present disclosure is not limited to singular or plural components. Components expressed in plural may be composed of singular elements, or components expressed in singular may be composed of plural elements.

[0141] The embodiments of the present disclosure disclosed in the detailed description and drawings are merely specific examples to easily explain the technical content of the present disclosure and facilitate understanding of the present disclosure, and are not intended to limit the scope of the present disclosure. In other words, it will be apparent to those skilled in the art to which the present disclosure pertains that other modifications based on the technical concepts of the present disclosure are possible. Furthermore, the embodiments described in the present disclosure can be combined and operated as needed.

Claims

1. In the method by the base station, An operation of receiving a report message from a terminal (UE) including a monitoring input related to monitoring of artificial intelligence (AI)-based channel state information (CSI) compression while AI-based CSI compression by the terminal is disabled; An operation of performing first monitoring of the AI-based CSI compression based on the above monitoring input; An action of transmitting an instruction message to the terminal to activate the AI-based CSI compression generated based on the results of the first monitoring; An operation of receiving compressed CSI by the AI-based CSI compression from the terminal after transmitting the above instruction message, and A method characterized by comprising an action of performing second monitoring of the AI-based CSI compression based on the compressed CSI.

2. In the first paragraph, the monitoring input includes at least one of a vector z or a quantized vector z, A method characterized in that the above vector z is generated by estimating a channel based on a channel state information reference signal (CSI-RS) received by the terminal from the base station, decomposing the estimated channel into an eigenvalue to obtain a matrix V, and compressing the matrix V.

3. In paragraph 1, A method further comprising: transmitting a request message of the monitoring input to the terminal based on identifying that the AI-based CSI compression is disabled.

4. In paragraph 1, Further comprising an action of transmitting monitoring setting information related to reporting of monitoring input related to the AI-based CSI compression to the terminal, A method characterized in that the above monitoring setting information indicates at least one of a transmission cycle for reporting the monitoring input, or a format of information included in the monitoring input.

5. In paragraph 4, the monitoring setting information is: A method characterized by including a monitoring type instruction that instructs either non-AI based network-side monitoring, AI based terminal-side monitoring, or AI based network-side monitoring.

6. In paragraph 4, the monitoring setting information is: A method characterized by including at least one of a transmission cycle for reporting the results of the second monitoring, information of an event for reporting the results of the second monitoring, or a format of information included in a report of the results of the second monitoring.

7. In paragraph 1, An operation of receiving a monitoring result report of the AI-based CSI compression from the terminal while the AI-based CSI compression is disabled; An action for determining whether to activate the AI-based CSI compression based on the above monitoring result report, and A method characterized by further comprising an action of transmitting an instruction message to the terminal for activating the AI-based CSI compression according to the judgment result.

8. In the method by terminal (UE), An operation of transmitting a report message to a base station including monitoring inputs related to monitoring of said AI-based CSI compression while the AI-based CSI compression is disabled; After transmitting the above report message, the operation of receiving an instruction message for activating the AI-based CSI compression from the base station, and A method characterized by including an action of transmitting compressed CSI by the AI-based CSI compression to the base station after receiving the above instruction message.

9. In the 8th paragraph, the monitoring input includes at least one of a vector z or a quantized vector z, A method characterized in that the above vector z is generated by estimating a channel based on a channel state information reference signal (CSI-RS) received by the terminal from the base station, decomposing the estimated channel into an eigenvalue to obtain a matrix V, and compressing the matrix V.

10. In paragraph 8, A method further comprising the action of receiving a request message for said monitoring input from said base station while said AI-based CSI compression is disabled.

11. In paragraph 8, Further comprising an action of receiving monitoring setting information related to reporting of monitoring input related to the AI-based CSI compression from the base station, The above monitoring setting information indicates at least one of a transmission cycle for reporting the monitoring input, a format of information included in the monitoring input, or a monitoring type indication. A method characterized in that the above monitoring type instruction indicates either non-AI based network-side monitoring, AI based terminal-side monitoring, or AI based network-side monitoring.

12. In paragraph 11, the monitoring setting information is: A method characterized by including at least one of a transmission cycle for reporting monitoring results, information on an event triggering a monitoring result report, or a format of information included in a monitoring result report.

13. In paragraph 8, An operation of performing monitoring of the AI-based CSI compression while the AI-based CSI compression is disabled; An action of transmitting a monitoring result report of the above AI-based CSI compression to the above base station; A method characterized by further comprising the action of receiving an instruction message for activating the AI-based CSI compression from the base station after transmitting the monitoring result report.

14. At the base station, Transmitter and receiver, and A processor connected to the transceiver, the processor comprising: An operation of receiving a report message from a terminal (UE) including a monitoring input related to monitoring of artificial intelligence (AI)-based channel state information (CSI) compression while the AI-based CSI compression by the terminal is disabled; An operation of performing first monitoring of the AI-based CSI compression based on the above monitoring input; An action of transmitting an instruction message to the terminal to activate the AI-based CSI compression generated based on the results of the first monitoring; An operation of receiving compressed CSI by the AI-based CSI compression from the terminal after transmitting the above instruction message, and A base station characterized in that it is configured to perform an operation of performing second monitoring of the AI-based CSI compression based on the compressed CSI.

15. In the terminal (UE), Transmitter and receiver, and A processor connected to the transceiver, the processor comprising: An operation of transmitting a report message to a base station including monitoring inputs related to monitoring of said AI-based CSI compression while the AI-based CSI compression is disabled; After transmitting the above report message, an action of receiving an instruction message for activating the AI-based CSI compression from the base station; A terminal characterized in that it is configured to perform an operation of transmitting compressed CSI by the AI-based CSI compression to the base station after receiving the above instruction message.

Images