Method and device for ai model-based channel prediction in wireless communication system

Abstract

The present disclosure relates to a 5G or 6G communication system for supporting a data transmission rate higher than that of a 4G communication system such as LTE. According to various embodiments of the present disclosure, a method performed by a user equipment (UE) in a wireless communication system may comprise the steps of: receiving, from a base station, first information for configuring a set of candidate demodulation reference signal (DMRS) patterns; performing first channel prediction on the set of candidate DMRS patterns on the basis of an artificial intelligence (AI) model; identifying one or more DMRS patterns on the basis of results of the first channel prediction; transmitting, to the base station, information about the one or more DMRS patterns; and, on the basis of the one or more DMRS patterns, receiving, from the base station, second information for configuring at least one DMRS pattern.

Description

Method and device for predicting AI model-based channels in wireless communication systems

[0001] The present disclosure relates generally to wireless communication systems, and more particularly to a method and apparatus for performing channel prediction based on an artificial intelligence (AI) model 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 the 5G (5th Generation) communication system, 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, such as augmented reality glasses, virtual reality headsets, and holographic devices. In the 6G (6th Generation) 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 (1,000 gigabits) per second (bps) and a wireless latency of 100 microseconds (μsec). This means that compared to 5G, the transmission speed in a 6G communication system will be 50 times faster and 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 (THz) 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 have more severe path loss and atmospheric absorption, making it more important to develop technologies that can guarantee signal reach, or coverage. Key technologies to ensure coverage include Radio Frequency (RF) components, antennas, new waveforms that offer better coverage than Orthogonal Frequency Division Multiplexing (OFDM), 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 Surface (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 at the same time for uplink and downlink; network technology that integrates satellites and HAPS (High-Altitude Platform Stations); 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 AI (Artificial Intelligence) 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 (XR), high-fidelity mobile holograms, and digital replicas. Furthermore, services such as remote surgery, industrial automation, and emergency response, which are provided through 6G communication systems through enhanced security and reliability, will be applied in diverse fields such as industry, medicine, automobiles, and home appliances.

[0007] In wireless communication systems, artificial intelligence (AI) models can be used to perform channel prediction. To address the issue of transmitted demodulation reference signal (DMRS) resource overhead and enable more accurate channel measurement, specific AI-specific DMRS patterns can be established. Accordingly, a method is being considered for establishing DMRS patterns based on AI models that include both terminals and networks, and for measuring the channel accordingly.

[0008] According to various embodiments of the present disclosure, an object is to provide a device and method capable of effectively providing a service in a wireless communication system.

[0009] The technical problems to be achieved in this document are not limited to the technical problems mentioned above, and other technical problems not mentioned can be clearly understood by a person having ordinary skill in the technical field to which the present invention belongs from the description below.

[0010] According to various embodiments of the present disclosure, in a wireless communication system, a method performed by a user equipment (UE) may include the steps of: receiving, from a base station, first information for setting a set of candidate demodulation reference signal (DMRS) patterns; performing a first channel prediction for the set of candidate DMRS patterns based on an artificial intelligence (AI) model; identifying one or more DMRS patterns based on a result of the first channel prediction; transmitting, to the base station, information about the one or more DMRS patterns; and receiving, from the base station, second information for setting at least one DMRS pattern based on the one or more DMRS patterns.

[0011] According to various embodiments of the present disclosure, in a wireless communication system, a method performed by a base station may include the steps of transmitting, to a user equipment (UE), first information setting a set of candidate demodulation reference signal (DMRS) patterns, receiving, from the UE, information about one or more DMRS patterns, and transmitting, to the UE, second information setting at least one DMRS pattern based on the one or more DMRS patterns, wherein the one or more DMRS patterns may be identified based on a result of a first channel prediction for the set of candidate DMRS patterns based on an AI model of the UE.

[0012] According to various embodiments of the present disclosure, an object is to provide a device and method capable of effectively providing a service in a wireless communication system.

[0013] The effects that can be obtained from the present disclosure are not limited to the effects mentioned above, and other effects that are not mentioned can be clearly understood by a person having ordinary skill in the art to which the present disclosure belongs from the description below.

[0014] FIG. 1 illustrates a wireless communication system according to various embodiments of the present disclosure.

[0015] FIG. 2 illustrates a configuration of a base station in a wireless communication system according to various embodiments of the present disclosure.

[0016] FIG. 3 illustrates a configuration of a terminal in a wireless communication system according to various embodiments of the present disclosure.

[0017] FIG. 4 illustrates an example of a wireless resource region in a wireless communication system according to various embodiments of the present disclosure.

[0018] FIG. 5 illustrates an example of resources per DMRS (demodulation reference signal) port in a wireless communication system according to various embodiments of the present disclosure.

[0019] FIG. 6 illustrates a block diagram and signal flow for predicting a channel based on an artificial intelligence (AI) model according to various embodiments of the present disclosure.

[0020] FIG. 7 illustrates an example for training an AI model that performs channel prediction according to various embodiments of the present disclosure.

[0021] FIG. 8A and FIG. 8B illustrate an example for setting a DMRS pattern within a slot according to various embodiments of the present disclosure.

[0022] FIG. 9 illustrates a flow of operations for selecting a DMRS pattern based on channel measurements and AI models according to various embodiments of the present disclosure.

[0023] FIG. 10 illustrates an example for performing channel prediction for frequency domain and time domain based on an AI model according to various embodiments of the present disclosure.

[0024] FIG. 11 illustrates a signal flow for measuring a channel based on an AI model of a network according to various embodiments of the present disclosure.

[0025] FIG. 12 illustrates a signal flow for measuring a channel based on an AI model of a terminal according to various embodiments of the present disclosure.

[0026] The terms used in this disclosure are used only to describe specific embodiments and may not be intended to limit the scope of other embodiments. The singular expression may include plural expressions unless the context clearly indicates otherwise. Terms used herein, including technical or scientific terms, may have the same meaning as commonly understood by those of ordinary skill in the art described in this disclosure. Terms defined in general dictionaries among the terms used in this disclosure may be interpreted as having the same or similar meaning in the context of the relevant technology, and shall not be interpreted in an idealized or overly formal sense unless explicitly defined in this disclosure. In some cases, even if a term is defined in this disclosure, it cannot be interpreted to exclude embodiments of the present disclosure.

[0027] The various embodiments of the present disclosure described below illustrate a hardware-based approach as an example. However, since the various embodiments of the present disclosure include techniques utilizing both hardware and software, the various embodiments of the present disclosure do not exclude a software-based approach.

[0028] In the following description, terms referring to components of the device (control unit, controller, processor, artificial intelligence (AI) model, neural network (NN) model, etc.) and terms referring to data (signal, feedback, report, reporting, information, parameter, value, bit, codeword, etc.) are examples for convenience of explanation. Therefore, the present disclosure is not limited to the terms described below, and other terms having similar or equivalent technical meanings may be used.

[0029] Additionally, while this disclosure describes various embodiments using terminology used in certain communication standards (e.g., 3rd Generation Partnership Project (3GPP)), these are merely illustrative examples. The various embodiments of this disclosure can be easily modified and applied to other communication systems.

[0030] FIG. 1 illustrates a wireless communication system according to various embodiments of the present disclosure. FIG. 1 illustrates a base station (110), a terminal (120), and a terminal (130) as some of the nodes utilizing a wireless channel in the wireless communication system. While FIG. 1 illustrates only one base station, other base stations identical to or similar to base station (110) may be included.

[0031] The base station (110) is a network infrastructure that provides wireless access to terminals (120, 130). The base station (110) has coverage defined as a certain geographical area based on the distance at which a signal can be transmitted. In addition to the base station, the base station (110) may be referred to as an 'access point (AP)', 'eNodeB (eNB)', 'gNodeB (gNB)', '5th generation node', '6th generation node', 'wireless point', 'transmission / reception point (TRP)', or other terms having equivalent technical meanings.

[0032] Each of the terminals (120) and (130) is a device used by a user and communicates with the base station (110) via a wireless channel. In some cases, at least one of the terminals (120) and (130) may be operated without the user's intervention. That is, at least one of the terminals (120) and (130) is a device that performs machine type communication (MTC) and may not be carried by the user. Each of the terminal (120) and the terminal (130) may be referred to as a 'terminal', 'user equipment (UE),' 'mobile station,' 'subscriber station,' 'customer premises equipment (CPE),' 'remote terminal,' 'wireless terminal,' 'electronic device,' or 'user device,' or other terms having similar or equivalent technical meanings thereto.

[0033] The base station (110), the terminal (120), and the terminal (130) can transmit and receive wireless signals in an upper-mid band (e.g., 10 GHz, 13 GHz) or a millimeter wave (mmWave) band (e.g., 28 GHz, 30 GHz, 38 GHz, 60 GHz, over 60 GHz, etc.). At this time, in order to improve channel gain, the base station (110), the terminal (120), and the terminal (130) can perform beamforming. Here, the beamforming can include transmission beamforming and reception beamforming. That is, the base station (110), the terminal (120), and the terminal (130) can provide directionality to a transmission signal or a reception signal. To this end, the base station (110) and terminals (120, 130) can select serving beams (112, 113, 121, 131) through a beam search or beam management procedure. After the serving beams (112, 113, 121, 131) are selected, subsequent communication can be performed through resources that are in a quasi-co-located (QCL) relationship with the resources that transmitted the serving beams (112, 113, 121, 131).

[0034] FIG. 2 illustrates a configuration of a base station in a wireless communication system according to various embodiments of the present disclosure. According to various embodiments of the present disclosure, the base station (110) may be conveniently referred to as a network. The configuration illustrated in FIG. 2 may be understood as the configuration of the base station (110). Terms such as "... unit" and "... unit" used hereinafter refer to a unit that processes at least one function or operation, and this may be implemented by hardware, software, or a combination of hardware and software.

[0035] Referring to FIG. 2, the base station (110) may include a wireless communication unit (210), a backhaul communication unit (220), a storage unit (230), and a control unit (240).

[0036] The wireless communication unit (210) performs functions for transmitting and receiving signals through a wireless channel. For example, the wireless communication unit (210) performs a conversion function between a baseband signal and a bit stream according to the physical layer specifications of the system. For example, when transmitting data, the wireless communication unit (210) generates complex symbols by encoding and modulating a transmission bit stream. In addition, when receiving data, the wireless communication unit (210) restores a reception bit stream by demodulating and decoding a baseband signal. In addition, the wireless communication unit (210) up-converts a baseband signal into an RF (radio frequency) band signal and transmits it through an antenna, and down-converts an RF band signal received through the antenna into a baseband signal.

[0037] To this end, the wireless communication unit (210) may include a transmission filter, a reception filter, an amplifier, a mixer, an oscillator, a digital to analog convertor (DAC), an analog to digital convertor (ADC), etc. In addition, the wireless communication unit (210) may include a plurality of transmission and reception paths. Furthermore, the wireless communication unit (210) may include at least one antenna array composed of a plurality of antenna elements. In terms of hardware, the wireless communication unit (210) may be composed of a digital unit and an analog unit, and the analog unit may be composed of a plurality of sub-units according to operating power, operating frequency, etc.

[0038] The wireless communication unit (210) can transmit and receive signals. To this end, the wireless communication unit (210) may include at least one transceiver. For example, the wireless communication unit (210) may transmit a synchronization signal, a reference signal, system information, messages, control information, or data. In addition, the wireless communication unit (210) may perform beamforming.

[0039] The wireless communication unit (210) transmits and receives signals as described above. Accordingly, all or part of the wireless communication unit (210) may be referred to as a "transmitter," a "receiver," or a "transmitting and receiving unit." Furthermore, in the following description, transmission and reception performed via a wireless channel are used to mean that the wireless communication unit (210) performs the processing described above.

[0040] The backhaul communication unit (220) provides an interface for performing communication with other nodes within the network. That is, the backhaul communication unit (220) converts a bit string transmitted from the base station (110) to another node, such as another access node, another base station, an upper node, a core network, etc., into a physical signal, and converts a physical signal received from another node into a bit string.

[0041] The storage unit (230) stores data such as basic programs, application programs, and setting information for the operation of the base station (110). The storage unit (230) may include a memory. The storage unit (230) may be configured as a volatile memory, a non-volatile memory, or a combination of volatile and non-volatile memories. In addition, the storage unit (230) provides the stored data according to a request from the control unit (240). According to one embodiment, the storage unit (230) may store learning data for AI-based channel prediction and apply the stored learning data to a neural network structure of AI-based channel prediction.

[0042] The control unit (240) controls the overall operations of the base station (110). For example, the control unit (240) transmits and receives signals through the wireless communication unit (210) or the backhaul communication unit (220). In addition, the control unit (240) can record and read data in the storage unit (230). In addition, the control unit (240) can perform the functions of the protocol stack required by the communication standard. To this end, the control unit (240) can include at least one controller or processor. According to various embodiments, the control unit (240) or a combination of one or more control units can control the base station to perform operations according to various embodiments described below.

[0043] The configuration of the base station (110) illustrated in FIG. 2 is merely an example of a base station, and examples of base stations that perform various embodiments of the present disclosure are not limited to the configuration illustrated in FIG. 2. That is, some configurations may be added, deleted, or changed according to various embodiments.

[0044] Although the base station is described as a single entity in FIG. 2, the present disclosure is not limited thereto. The base station according to various embodiments of the present disclosure may be implemented to form an access network having not only an integrated deployment but also a distributed deployment. According to one embodiment, the base station may be divided into a central unit (CU) and a distributed unit (DU), and the CU may be implemented to perform upper layer functions (e.g., radio link control (RLC), packet data convergence protocol (PDCP), and radio resource control (RRC)) and the DU may be implemented to perform lower layer functions (e.g., medium access control (MAC), physical (PHY)). The DU of the base station may form beam coverage on a wireless channel.

[0045] FIG. 3 illustrates the configuration of a terminal in a wireless communication system according to various embodiments of the present disclosure. The configuration illustrated in FIG. 3 can be understood as the configuration of terminals (120, 130). Terms such as "...unit" and "...unit" used hereinafter refer to a unit that processes at least one function or operation, which can be implemented using hardware, software, or a combination of hardware and software.

[0046] Referring to FIG. 3, the terminal (120, 130) may include a communication unit (310), a storage unit (320), and a control unit (330).

[0047] The communication unit (310) performs functions for transmitting and receiving signals via a wireless channel. For example, the communication unit (310) performs a conversion function between a baseband signal and a bit stream according to the physical layer specifications of the system. For example, when transmitting data, the communication unit (310) generates complex symbols by encoding and modulating a transmission bit stream. In addition, when receiving data, the communication unit (310) restores a reception bit stream by demodulating and decoding the baseband signal. In addition, the communication unit (310) up-converts a baseband signal into an RF band signal and then transmits it through an antenna, and down-converts an RF band signal received through the antenna into a baseband signal. For example, the communication unit (310) may include a transmission filter, a reception filter, an amplifier, a mixer, an oscillator, a DAC, an ADC, etc.

[0048] In addition, the communication unit (310) may include a plurality of transmission and reception paths. Furthermore, the communication unit (310) may include an antenna unit. The communication unit (310) may include at least one antenna array composed of a plurality of antenna elements. In terms of hardware, the communication unit (310) may be composed of digital circuits and analog circuits (e.g., radio frequency integrated circuits (RFIC)). Here, the digital circuits and analog circuits may be implemented in a single package. In addition, the communication unit (310) may include a plurality of RF chains. The communication unit (310) may perform beamforming. The communication unit (310) may apply beamforming weights to a signal to be transmitted and received in order to impart directionality according to the settings of the control unit (330). According to one embodiment, the communication unit (310) may include an RF (radio frequency) block (or RF unit). The RF block may include first RF circuitry associated with the antenna and second RF circuitry associated with baseband processing. The first RF circuitry may be referred to as RF-A (antenna). The second RF circuitry may be referred to as RF-B (baseband).

[0049] In addition, the communication unit (310) can transmit and receive signals. For this purpose, the communication unit (310) can include at least one transceiver. The communication unit (310) can receive a downlink signal. The downlink signal can include a synchronization signal (SS), a reference signal (RS) (e.g., a cell-specific reference signal (CRS), a demodulation (DM)-RS), system information (e.g., MIB, SIB, remaining system information (RMSI), other system information (OSI)), a configuration message, control information, or downlink data. In addition, the communication unit (310) can transmit an uplink signal. The uplink signal may include a random access related signal (e.g., a random access preamble (RAP) (or Msg1 (message 1)), Msg3 (message 3)), a reference signal (e.g., a sounding reference signal (SRS), DM-RS), or a power headroom report (PHR).

[0050] Additionally, the communication unit (310) may include different communication modules to process signals of different frequency bands. Furthermore, the communication unit (310) may include multiple communication modules to support multiple different wireless access technologies. For example, different wireless access technologies may include Bluetooth low energy (BLE), wireless fidelity (Wi-Fi), WiGig (WiFi gigabyte), cellular networks (e.g., long term evolution (LTE), new radio (NR), etc.). In addition, different frequency bands may include upper-mid bands (e.g., 10 GHz, 13 GHz), super high frequency (SHF) (e.g., 2.5 GHz, 5 GHz) bands, millimeter wave (mm wave) (e.g., 38 GHz, 60 GHz, etc.) bands. In addition, the communication unit (310) may use the same type of wireless access technology on different frequency bands (e.g., unlicensed bands for licensed assisted access (LAA), citizens broadband radio service (CBRS) (e.g., 3.5 GHz)).

[0051] The communication unit (310) transmits and receives signals as described above. Accordingly, all or part of the communication unit (310) may be referred to as a "transmitter," a "receiver," or a "transmitting and receiving unit." Furthermore, in the following description, transmission and reception performed via a wireless channel are used to mean processing performed by the communication unit (310) as described above.

[0052] The storage unit (320) stores data such as basic programs, application programs, and setting information for the operation of the terminal (120). The storage unit (320) may be configured as volatile memory, non-volatile memory, or a combination of volatile and non-volatile memory. In addition, the storage unit (320) provides the stored data at the request of the control unit (330). In one embodiment, the storage unit (320) may store learning data for AI-based channel prediction according to AI parameters or channel prediction settings set by the base station.

[0053] The control unit (330) controls the overall operations of the terminals (120, 130). For example, the control unit (330) transmits and receives signals through the communication unit (310). In addition, the control unit (330) can record and read data in the storage unit (320). In addition, the control unit (330) can perform the functions of the protocol stack required by the communication standard. To this end, the control unit (330) may include at least one controller or processor. The control unit (330) may include at least one processor or microprocessor, or may be a part of a processor. In addition, a part of the communication unit (310) and the control unit (330) may be referred to as a CP (cellular processor). The control unit (330) may include various modules for performing communication. According to various embodiments, the control unit (330) or a combination of one or more control units may control the terminal to perform operations according to various embodiments described below.

[0054] The configuration of the terminals (120, 130) illustrated in FIG. 3 is merely an example of a terminal, and examples of terminals performing various embodiments of the present disclosure are not limited to the configuration illustrated in FIG. 3. That is, some configurations may be added, deleted, or changed according to various embodiments.

[0055] According to various embodiments of the present disclosure, the following describes an AI model included in a base station (110) or a terminal (120, 130). According to one embodiment, the base station (110) or the terminal (120, 130) may include an AI (artificial intelligence) model. According to one embodiment, an AI model learned based on a neural network may be operated through a control unit (240, 330) and a storage unit (230, 320) of the terminal (120, 130) or the base station (110). At this time, the control unit (240, 330) may be composed of one or more processors. The one or more processors may include the functions of a general-purpose processor such as a central processing unit (CPU), an application processor (AP), a digital signal processor (DSP), a graphics-only processor such as a graphics processing unit (GPU), a vision processing unit (VPU), or an artificial intelligence-only processor such as a neural processing unit (NPU). One or more processors can be controlled to process input data according to predefined operating rules or artificial intelligence models stored in the storage unit (230, 320). Alternatively, if one or more processors are dedicated artificial intelligence processors, the dedicated artificial intelligence processors may be designed with a hardware structure specialized for processing a specific artificial intelligence model. The dedicated artificial intelligence processors may not be included in the control unit (240, 330) but may be included as a separate component.

[0056] According to one embodiment, the predefined operation rules or artificial intelligence models are characterized by being created through learning. Here, being created through learning means that the basic artificial intelligence model is learned by a learning algorithm using a plurality of learning data, thereby creating a predefined operation rule or artificial intelligence model set to perform a desired characteristic (or purpose). This learning may be performed in the device itself on which the artificial intelligence according to the present disclosure is performed, or may be performed through a separate server and / or system. Examples of the learning algorithm include supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning, but are not limited to the examples described above. The control unit (240, 330) can learn an occurring event, a determined judgment, or collected or input information through the learning algorithm. The control unit (240, 330) can store the learning results in the storage unit (230, 320) (e.g., memory).

[0057] An artificial intelligence model (e.g., an AI model) may be composed of multiple neural network layers. Each of the multiple neural network layers has multiple weight values, and performs neural network operations through operations between the operation results of the previous layer and the multiple weights. The multiple weights of the multiple neural network layers may be optimized based on the learning results of the AI ​​model. For example, the multiple weights may be updated so that the loss value or cost value obtained from the AI ​​model during the learning process is reduced or minimized. The artificial neural network may include a deep neural network (DNN), and examples thereof include, but are not limited to, a convolutional neural network (CNN), a deep neural network (DNN), a recurrent neural network (RNN), a restricted Boltzmann machine (RBM), a deep belief network (DBN), a bidirectional recurrent deep neural network (BRDNN), a long short term memory (LSTM), or deep Q-Networks.

[0058] In one embodiment, the control unit (240, 330) may execute an algorithm for performing operations related to AI-based channel prediction (e.g., for convenience of explanation, interchangeably used with channel measurement, channel estimation, etc.). In one embodiment, an AI model learned to perform operations related to AI-based channel prediction may be configured in hardware, included in software, or configured through a combination of hardware and software in the control unit (240, 330). In other words, the control unit (240, 330) may include a control unit for AI-based channel prediction. The control unit for AI-based channel prediction may perform AI-based channel prediction, identification of prediction performance for the channel, determination of whether to report the identified result, and determination of whether to use AI-based channel calculation. In addition, according to various embodiments, the control unit (240, 330) may include an update unit. The update unit can obtain data updated by the base station alone or by a learning procedure between the terminal and the base station (e.g., data related to channel measurement between the terminal and the base station), and reconfigure the values ​​of parameters constituting the neural network (e.g., neural network structure, node layer-by-layer information, weight information between nodes) based on the data. Signaling for setting the parameters constituting the neural network is described in detail below according to various embodiments of the present disclosure. The AI-based channel prediction control unit and the update unit may be a set of instructions or codes stored in the storage unit (230, 320), and may be a command / code or a storage space that stores the command / code that is at least temporarily resided in the control unit (240, 330), or may be a part of the circuitry constituting the control unit (240, 330). According to various embodiments, the control unit (240, 330) can control the base station (110) or the terminal (120, 130) to perform operations according to various embodiments described below.

[0059] FIG. 4 illustrates an example of a wireless resource region in a wireless communication system according to various embodiments of the present disclosure. In various embodiments, the wireless resource region may include a structure in the time-frequency domain. In one embodiment, the wireless communication system may include an NR communication system.

[0060] Referring to Fig. 4, in the wireless resource domain, the horizontal axis represents the time domain, and the vertical axis represents the frequency domain. The length of a radio frame (404) is 10 ms. The radio frame (404) may be a time domain section composed of 10 subframes. The length of a subframe (403) is 1 ms. The unit of configuration in the time domain may be an OFDM (orthogonal frequency division multiplexing) and / or a DFT-s-OFDM (DFT (discrete Fourier transform)-spread-OFDM) symbol, and N symb OFDM and / or DFT-s-OFDM symbols (401) may be grouped to form one slot (402). In various embodiments, an OFDM symbol may include a symbol for transmitting and receiving a signal using an OFDM multiplexing scheme, and a DFT-s-OFDM symbol may include a symbol for transmitting and receiving a signal using a DFT-s-OFDM or SC-FDMA (single carrier frequency division multiple access) multiplexing scheme. The minimum transmission unit in the frequency domain is a subcarrier, and the carrier bandwidth constituting the resource grid is a total of N sc BWIt may be composed of subcarriers (205). In addition, in the present disclosure, an embodiment regarding downlink signal transmission and reception is described for convenience of explanation, but this can also be applied to an embodiment regarding uplink signal transmission and reception.

[0061] In some embodiments, the number of slots (402) constituting one subframe (403) and the length of the slots (402) may vary depending on the subcarrier spacing. This subcarrier spacing may be referred to as a numerology (μ). That is, the subcarrier spacing, the number of slots included in a subframe, the length of the slots, and the length of the subframe may be configured variably. For example, in an NR communication system, when the subcarrier spacing (SCS) is 15 kHz, one slot (402) constitutes one subframe (403), and the lengths of the slot (402) and the subframe (403) may each be 1 ms. In addition, for example, when the subcarrier spacing is 30 kHz, two slots may constitute one subframe (403). In this case, the length of the slot is 0.5 ms and the length of the subframe is 1 ms.

[0062] In some embodiments, the subcarrier spacing, the number of slots included in a subframe, the length of the slot, and the length of the subframe may be applied variably depending on the communication system. For example, in the case of an LTE system, the subcarrier spacing may be 15 kHz, two slots may constitute one subframe, and in this case, the length of the slot may be 0.5 ms and the length of the subframe may be 1 ms. As another example, in the case of an NR system, the subcarrier spacing (μ) may be one of 15 kHz, 30 kHz, 60 kHz, 120 kHz, and 240 kHz, and the number of slots included in one subframe depending on the subcarrier spacing (μ) may be 1, 2, 4, 8, and 16.

[0063] The basic unit of resources in the time-frequency domain may be a resource element (RE) (406), and the resource element (406) may be expressed by an OFDM symbol index and a subcarrier index. A resource block may include a plurality of resource elements. In an NR system, a resource block (RB) (or physical resource block (PRB)) (407) may be N in the frequency domain. SC RB can be defined as a series of consecutive subcarriers. The number of subcarriers N SC RB =12 can be. The frequency domain can include common resource blocks (CRBs). Physical resource blocks (PRBs) can be defined in the bandwidth part (BWP) of the frequency domain. The CRB and PRB numbers can be determined differently depending on the subcarrier spacing. In the LTE system, RBs are N in the time domain. symb N consecutive OFDM symbols in the frequency domain SC RB can be defined as a series of consecutive subcarriers.

[0064] In NR or LTE systems, scheduling information for downlink data or uplink data can be transmitted from a base station to a terminal via downlink control information (DCI). In various embodiments, DCI can be defined according to various formats, and each format can indicate whether the DCI includes scheduling information for uplink data (e.g., UL grant), scheduling information for downlink data (e.g., DL resource allocation), whether it is compact DCI with small control information size, whether it is fall-back DCI, whether spatial multiplexing using multiple antennas is applied, or whether it is DCI for power control. For example, NR DCI format 1_0 or NR DCI format 1_1 can include scheduling for downlink data. Also, for example, NR DCI format 0_0 or NR DCI format 0_1 ​​can include scheduling for uplink data.

[0065] While various embodiments of the present disclosure are described based on an LTE communication system or an NR communication system, the present disclosure is not limited thereto and can be applied to various wireless communication systems for transmitting DMRS for channel measurement. Furthermore, the present disclosure can also be applied to unlicensed bands as needed, in addition to licensed bands.

[0066] Hereinafter, in the present disclosure, higher layer signaling or higher signal may be a signal transmission method in which a base station transmits a signal to a terminal using a downlink data channel of a physical layer, or from a terminal to a base station using an uplink data channel of a physical layer. According to an embodiment, the higher layer signaling may include at least one of radio resource control (RRC) signaling, signaling according to an F1 interface between a central unit (CU) and a distributed unit (DU), or a signal transmission method transmitted through a media access control (MAC) control element (MAC CE). In addition, according to an embodiment, the higher layer signaling or higher signal may include system information commonly transmitted to a plurality of terminals, for example, a system information block (SIB).

[0067] In addition, in relation to channel prediction using a reference signal (RS), terms such as estimation, prediction, or estimation may be used interchangeably, and these should be distinguished not only by the terms themselves but also through the actual content of the operation.

[0068] According to various embodiments of the present disclosure, the contents of a demodulation reference signal (DMRS) transmitted and received for demodulation of an uplink (UL) or downlink (DL) signal are described below. The DMRS may be transmitted for demodulation of uplink or downlink transmission. In an NR communication system, uplink transmission may include physical uplink shared channel (PUSCH) transmission or physical uplink shared channel (PUCCH) transmission, and downlink transmission may include physical downlink shared channel (PDSCH) transmission or physical downlink control channel (PDCCH) transmission. PUSCH transmission may include transmission of data along an UL-SCH (shared channel), and PDSCH transmission may include transmission of data along a DL-SCH. PUCCH transmission may include transmission of uplink control information (UCI), and PDCCH may include transmission of downlink control information (DCI). Hereinafter, processes for transmission of downlink DMRS or channel prediction (or measurement) through PDSCH are disclosed according to various embodiments of the present disclosure, but these are only examples and are not limited thereto, and it goes without saying that the embodiments of the present disclosure can be applied in the same or similar manner to processes for transmission of uplink DMRS or channel prediction through PUSCH.

[0069] In one embodiment, the wireless channel experienced by the DMRS is assumed to be identical or similar to the wireless channel experienced by signals transmitted for uplink or downlink transmission. That is, the precoding applied to signals transmitted and received between the base station and the terminal can be applied equally to the DMRS. Similarly, the antenna ports utilized by the aforementioned signals can also be utilized by the DMRS.

[0070] According to one embodiment, when describing based on downlink DMRS, the terminal and the base station can know in advance the sequence of DMRS that the terminal will use for the downlink signal. The DMRS sequence can be defined in advance according to communication settings (e.g., whether transform precoding is used, cell identity, scrambling ID, DMRS configuration). In the downlink, the terminal can predict the channel by comparing the DMRS to be transmitted from the base station with the DMRS received. The terminal can demodulate the received downlink signal through channel prediction. Hereinafter, downlink transmission is described using PDSCH as an example, but the embodiments of the present disclosure can be applied to PDCCH in the same or similar manner. Similarly, in the case of uplink, the base station can predict the channel based on DMRS, and the embodiments of the present disclosure can be applied to PUSCH or PUCCH in the same or similar manner.

[0071] According to one embodiment, the symbol allocation of DMRS is described below. As described through FIG. 4, a slot may include 14 symbols. DMRS allocated within a slot are used for channel estimation and tuning / demodulation. Additional allocation of DMRS can improve channel estimation performance at a terminal or a base station. For example, if a PDSCH transmission includes two DMRS transmissions, the base station can estimate the channel in each of the two time intervals (e.g., symbols). Based on the estimated channel results, the base station can perform interpolation or extrapolation (hereinafter, for convenience, the description will be centered on interpolation) between each time interval. Here, interpolation can generally mean a process of finding the median value of data by fitting pieces of a function (typically a polynomial) appropriate to adjacent data points.

[0072] In one embodiment, as the number of DMRS symbols increases, the gap between channel estimation results decreases. For example, the reduced gap reduces the requirement for interpolation. In environments where the channel changes rapidly, such as high-speed movement scenarios, it is required to improve demodulation performance by allocating additional DMRS symbols. Meanwhile, as the number of DMRS symbols allocated within a slot increases, the area for allocating data symbols (DL-SCH data for PDSCH transmission or DCI for PDCCH transmission) relatively decreases. In other words, as the number of RSs increases, transmission overhead increases. Therefore, the base station is required to allocate to the terminal a number of DMRS symbols that is more appropriate for the channel environment.

[0073] Recently, discussions are actively underway regarding mid- to high-frequency bands (e.g., 10 GHz, 13 GHz) as frequencies for 6G wireless communication systems. In general, as the frequency band increases, a greater number of antennas can be integrated into the same area. Therefore, as described above, systems using mid- to high-frequency bands may require a greater number of antennas than those required in NR systems. In particular, next-generation MIMO systems, called extreme multiple-input and multiple-output (X-MIMO), are considering transmission through up to 256 ports and 64 layers.

[0074] As mentioned above, DMRS is required to be transmitted for every granted RB, and thus the overhead of DMRS can be determined according to the transmission period and the number of layers to be transmitted. In particular, in the X-MIMO system, since the number of ports can increase up to 8 times compared to the conventional technology, the overhead of DMRS can also increase significantly. Taking this into account, the DMRS period can be set to be long to reduce the DMRS overhead, but as a trade-off, the accuracy of channel estimation through DMRS can be significantly reduced, and also the throughput can be greatly reduced.

[0075] According to various embodiments of the present disclosure, an AI-based channel prediction method for reducing the overhead of DMRS in a 6G massive MIMO system environment is described. Channel prediction according to various embodiments may include at least one of frequency-domain channel prediction, time-domain channel prediction, or spatial-domain channel prediction.

[0076] More specifically, according to one embodiment, the base station and the terminal may perform pattern report or pattern configuration of DMRS to reduce overhead, depending on the entity that performed AI training.

[0077] Furthermore, as the importance of site / vendor-specific DMRS patterns grows, specific signaling strategies for these may be required. For example, each base station (or network (NW)) operator may implement base stations differently, such as antenna configuration and TxRU mapping. Furthermore, each location may have different propagation paths, path loss, scatter environments, and penetration models, requiring adaptive DMRS pattern configuration. This may imply the need for pairing between base stations and terminals for specific DMRS patterns.

[0078] After the pairing between the terminal and the base station as described above is completed, the base station can set the determined DMRS pattern to the terminal, and the terminal can perform channel estimation (including, for example, channel prediction in at least one of the frequency domain, the time domain, or the spatial domain) based on the determined DMRS pattern, thereby reducing DMRS overhead. This process can be applied similarly or identically to the uplink channel and uplink DMRS estimated by the base station. Through the above process, the DMRS overhead problem due to the increased number of antenna ports can be solved in the 5G massive MIMO system or the future 6G X-MIMO system.

[0079] FIG. 5 illustrates examples of resources for each demodulation reference signal (DMRS) port in a wireless communication system according to various embodiments of the present disclosure. More specifically, FIG. 5 illustrates a first resource grid (510) for a case with 12 DMRS layers and a second resource grid (520) for a case with 64 DMRS layers.

[0080] According to various embodiments of the present disclosure, as described above, wireless communication technologies have been developed in various ways to provide fast data transmission speeds, improved application ranges, and stable connections. In particular, with respect to the physical layer, MIMO technology, which increases the number of antennas for transmission and reception to obtain performance gains, may be one of them. For example, the 5G NR system can support a number of antenna ports of 32, but in the future 6G system, considering the commercialization of X-MIMO technology in a new frequency band (e.g., FR (frequency range)3: 10 GHz to 13 GHz), it is expected that up to 64 MIMO layer transmissions will be possible. Therefore, when applying the existing DMRS transmission / reception technique to the X-MIMO system, excessive DMRS overhead may occur, and thus the data transmission / reception efficiency may be significantly reduced.

[0081] Referring to FIG. 5, the first resource grid (510) illustrates an example of antenna port mapping and DMRS resource allocation for three code-division multiplexing (CDM) groups when there are 12 DMRS layers. Based on the same technique, the second resource grid (520) illustrates an example of mapping to 24 antenna ports of three time division multiplexing (TDM) groups and an example of allocation of DMRS resources according to each layer when there are 64 DMRS layers. As illustrated in FIG. 5, if the currently used DMRS transmission technique and antenna port mapping technique are applied as is, it can be seen that the resource occupied by DMRS in downlink resources increases significantly due to an increase in the number of layers or the number of antenna ports in a high frequency band.

[0082] In one embodiment, when this is expressed numerically, a result as shown in Table 1 below can be derived.

[0083] DMRS PeriodicityEvery DL slotEvery 2DL slotsDMRS overheadin DL resources42.86%21.43%

[0084] That is, when transmitting data using a design such as the current DMRS transmission technique, up to 42.86% of the given downlink resources (e.g., 21.43% when transmitting every 2 slots) must be used for DMRS transmission, which inevitably leads to a significant reduction in resources for data transmission. Therefore, DMRS transmission in the X-MIMO system cannot be applied by directly extending the technique of general DMRS transmission, and a new type of DMRS pattern setting and a channel prediction method based on the same are required.

[0085] FIG. 6 illustrates a block diagram and signal flow for predicting a channel based on an artificial intelligence (AI) model according to various embodiments of the present disclosure.

[0086] According to various embodiments of the present disclosure, AI-based technologies are actively being researched to address issues such as the aforementioned DMRS overhead. In particular, such AI-based technologies are being utilized in various fields, including technologies that generate new data through AI's data-driven features (e.g., generative AI) and technologies that utilize the characteristics of given data to restore missing information.

[0087] Various embodiments of the present disclosure provide a method for reducing DMRS overhead through AI-based channel prediction (e.g., including channel prediction for at least one of the frequency domain (FD), time domain (TD), or spatial domain (SD)).

[0088] Furthermore, various embodiments provide a method for using site / vendor-specific AI optimized for each site or vendor, rather than using a common AI for all sites or vendors, to improve channel prediction performance while reducing overhead. For example, considering the characteristics of the site (e.g., propagation path, scattering environment, small-scale parameters, etc.) and the characteristics of the vendor (e.g., various antenna configurations, TxRU mapping, channel measurement techniques, etc.), an adaptive DMRS pattern may be set rather than a uniform DMRS pattern in terms of cell performance, and an advantageous DMRS pattern may vary depending on the environment and the provider. Therefore, when utilizing the above-described method, different optimized patterns may be derived for each base station, and a pairing procedure may be required between the base station and the terminal for channel prediction. Various embodiments of the present disclosure may include signaling for various procedures for channel prediction between the base station and the terminal.

[0089] Referring to FIG. 6, a procedure for channel prediction based on DMRS is illustrated through a block diagram (610) and a signal flow (620) between a terminal and a base station (e.g., a network).

[0090] In step (601), the base station may transmit a DMRS (or data including the DMRS) to the terminal. In one embodiment, the DMRS received by the terminal may occupy resources, such as the DMRS resource grid of block diagram (610). In one embodiment, when the base station measures a channel, an uplink DMRS may be transmitted from the terminal.

[0091] In step (602), the terminal (or base station) can measure the channel based on the received DMRS. For example, the terminal (or base station) can measure the quality of the channel corresponding to the DMRS by comparing the received DMRS with a predefined DMRS setting.

[0092] In step (603), the terminal (or base station) may measure a channel in the frequency domain by performing channel interpolation / extra-polation along the frequency axis. For example, the terminal may estimate a channel in the frequency domain using an AI model included in the terminal. According to one embodiment, the frequency domain channel predicted based on the AI ​​may be estimated along the frequency axis, such as the DMRS resource grid of block diagram (610). According to various embodiments, the process by which the terminal (or base station) predicts a channel based on the AI ​​model is described in detail in FIG. 7.

[0093] In step (604), the terminal (or base station) may measure a time-domain channel by performing channel interpolation / ex-interpolation along the time axis. For example, the terminal may estimate a time-domain channel using an AI model included in the terminal. According to one embodiment, the time-domain channel predicted based on the AI ​​may be further estimated along the time axis in addition to the channel predicted along the frequency axis, such as the DMRS resource grid of the block diagram (610). According to various embodiments, the process of the terminal (or base station) predicting a channel based on the AI ​​model is specifically described in FIG. 7. In addition, according to various embodiments of the present disclosure, although FIG. 6 illustrates a case where a channel in the frequency domain and a channel in the channel domain are sequentially predicted, this is merely an example, and it is to be understood that the channel in the time domain may be predicted first or they may be predicted simultaneously.

[0094] Below, through Fig. 7, the channel prediction process of the AI ​​model used by the terminal (or base station) to predict the channel and the AI ​​model generation / update process are described in detail.

[0095] FIG. 7 illustrates an example for training an AI model that performs channel prediction according to various embodiments of the present disclosure. Hereinafter, according to various embodiments, channel prediction may be described mainly focusing on channel prediction in the frequency domain, but is not limited thereto, and it is to be understood that the channel prediction of the present disclosure may include channel prediction in at least one of the frequency domain, the time domain, or the spatial domain. Referring to FIG. 7, a first algorithm (710) for estimating a channel based on DMRS reception and training (or learning) an AI model for this purpose, and a second algorithm (720) for generating a DMRS pattern for channel prediction are illustrated.

[0096] Referring to the first algorithm (710), an algorithm is illustrated for performing channel prediction based on a DMRS pattern set for a terminal (or base station) and training an AI model for the same. According to one embodiment, the AI ​​model is configured to predict a value of a measured DMRS channel ( ) as input values, and the estimated values ​​of the interpolated channels ( ) can be obtained as an output value. According to one embodiment, can mean the value of the channel corresponding to DMRS, may mean the value of a channel predicted based on an AI model. According to one embodiment, the first algorithm (710) may, as a loss function, calculate the ground-truth value of the channel ( )class One can train an AI model that generates ground truth channels (e.g., ground-truth channels) using the mean squared error (MSE) between them (e.g., using a procedure called backpropagation).

[0097] Referring to the second algorithm (720), an algorithm is illustrated for a case where a terminal (or base station) performs channel prediction and trains an AI model for this while generating a DMRS pattern. According to one embodiment, the AI ​​model is a ground-truth value ( ) as input values, and the estimated values ​​of the interpolated channels ( ) can be obtained as an output value. At this time, as an intermediate learnable parameter, the ground-truth value of the DMRS channel including the DMRS pattern ( ) can be obtained as an intermediate output value. According to one embodiment, the second algorithm (720) obtains, as a loss function, the ground-truth value of the channel ( )class An AI model that generates a ground truth channel (e.g., a ground truth channel) can be trained using the mean squared error (MSE) between the two (e.g., a procedure using backpropagation). According to various embodiments, is a trainable (or learnable) parameter according to the AI ​​model, and when learning is completed, a DMRS pattern that maximizes the channel estimation performance of the terminal (or base station) can be obtained. According to one embodiment, when learning is completed, the base station can use the algorithm of the front end of the second algorithm (720) as an AI model for generating a DMRS pattern, and the terminal (or base station) can use the algorithm of the FD channel prediction stage of the second algorithm (720) as an AI model for channel estimation.

[0098] According to various embodiments of the present disclosure, the procedure (or algorithm) for learning, training or channel prediction of the AI ​​model disclosed in FIG. 7 can be directly applied to the various processes for channel prediction or DMRS pattern generation of FIGS. 6 to 12, or can be indirectly and as a prerequisite to the above-described processes.

[0099] According to various embodiments of the present disclosure, as described above, a base station or a terminal may perform channel prediction based on a pattern generator or a channel predictor acquired through learning an AI model. At this time, the DMRS patterns between the base station and the terminal must be substantially aligned, and these DMRS patterns must be set and considered according to location / operator-specific characteristics or terminal capabilities. Specifically, among the candidate DMRS patterns to be used by the base station and the terminal, a pattern with good prediction performance must be selected, and a procedure for aligning the pattern between the base station and the terminal is required. The process for this is described below with reference to FIGS. 8A to 12.

[0100] FIGS. 8A and 8B illustrate an example for setting a DMRS pattern within a slot according to various embodiments of the present disclosure. The description of the setting or indication of a DMRS pattern described in FIGS. 8A and 8B according to various embodiments may be applied not only to setting a plurality of candidate DMRS patterns, but also to various signalings for transmitting a DMRS pattern according to the present disclosure, including signaling for indicating a DMRS pattern preferred by a terminal or setting a DMRS pattern set by a base station for channel prediction.

[0101] According to various embodiments of the present disclosure, to reduce DMRS overhead and determine an optimized DMRS pattern, a base station may configure various candidate DMRS patterns for a terminal. In one embodiment, the candidate DMRS patterns configured by the base station may include DMRS patterns generated based on the trained AI model described in FIG. 7. Here, the AI ​​model for generating the DMRS patterns may include an AI model deployed in the terminal or the base station.

[0102] According to one embodiment, the candidate DMRS patterns that the base station sets to the terminal may be set or indicated via RRC, MAC CE, or DCI. For example, when the candidate DMRS patterns are set via RRC signaling, the parameters in the existing RRC message may be modified or new RRC parameters may be included and set. According to one embodiment, the parameters in the existing RRC message may include parameters related to the configuration type (e.g., dmrs.DMRSConfigurationType={2, 3}), or the newly defined RRC parameters may include parameters related to RE position (e.g., dmrs.DMRSRePosition) or RB position (dmrs.DMRSRbPosition), etc. However, according to various embodiments of the present disclosure, the names of the parameters described below or the number of included parameters are only examples and are not limited thereto, and it is to be understood that parameters having substantially the same or similar contents may be modified and included.

[0103] In one embodiment, for the configuration type related parameter, when dmrs.DMRSConfigurationType=2, the configured DMRS pattern may include patterned DMRSs for each RB. Alternatively, when dmrs.DMRSConfigurationType=3, the configured DMRS pattern may include patterned DMRSs for selected RBs rather than for each RB. Furthermore, when dmrs.DMRSConfigurationType is indicated as 2 or 3, this may mean allocating a DMRS pattern according to various embodiments of the present disclosure (e.g., a location / operator specific DMRS pattern) rather than a conventional DMRS symbol allocation.

[0104] In one embodiment, when performing AI-based channel prediction, the base station may utilize location- and operator-specific DMRS patterns, requiring signaling to inform the terminal of the corresponding DMRS pattern. Unlike conventional DMRS allocation methods that allocate DMRS to each RB, the base station can adaptively allocate DMRS patterns to specific situations, thereby reducing wireless communication resource overhead.

[0105] According to one embodiment, the RE-related parameter (e.g., dmrs.DMRSRePosition) may indicate whether an RE within an RB contains a DMRS. Furthermore, according to one embodiment, the RB-related parameter (e.g., dmrs.DMRSRbPosition) may indicate whether an RB among multiple granted RBs contains a DMRS. According to various embodiments, the above-described parameters may be configured in the form of a bitmap, but are not limited thereto, and of course, may be configured as various pieces of information as long as they indicate substantially the same or similar content.

[0106] Referring to FIG. 8a, resource grids (810, 820, 830) illustrate examples of allocation of parameters and resources for indicating a DMRS pattern in units of one slot.

[0107] In one embodiment, the resource grid (810) may have a DMRS regularly allocated to each RB, with the configuration type indicated as 1. This may be substantially the same as a typical DMRS allocation.

[0108] In one embodiment, the resource grid (820) may have patterned DMRSs allocated to each RB, as indicated by a configuration type of 2. In one embodiment, the pattern information of the resource grid (820) may include RE-related parameters (e.g., dmrs.DMRSRePosition), and these RE-related parameters may indicate the position of the DMRS within the RB as a bitmap. For example, the DMRS pattern information of the resource grid (820) indicated by 'dmrs.DMRSRePosition=010010000100' may indicate a pattern in which the {2, 5, 10}th RE of each RB includes a DMRS.

[0109] In one embodiment, the resource grid (830) may have patterned DMRSs allocated only to selected RBs, as indicated by a configuration type of 3. In one embodiment, the pattern information of the resource grid (830) may include RE-related parameters (e.g., dmrs.DMRSRePosition) and RB-related parameters (e.g., dmrs.DMRSRbPosition). The RE-related parameters may indicate the position of a DMRS within an RB as a bitmap, and the RB-related parameters may indicate the position of an RB including a DMRS within the resource grid as a bitmap. For example, DMRS pattern information of a resource grid (830) indicated by 'dmrs.DMRSRbPosition=101' may indicate a pattern in which the {1, 3}th RB in the resource grid includes a DMRS, and DMRS pattern information of a resource grid (830) indicated by 'dmrs.DMRSRePosition=010010010100' may indicate a pattern in which the {2, 5, 8, 10}th RE of each RB including a DMRS includes a DMRS.

[0110] Referring to FIG. 8b, according to various embodiments of the present disclosure, even when DMRS is transmitted across multiple slots rather than a single slot, the same DMRS pattern information described above can be used. For example, when DMRS is allocated / transmitted across two slots, channel prediction performance can be improved by using different DMRS patterns in different slots.

[0111] According to one embodiment, the resource grid (840) may have patterned DMRSs allocated only to selected RBs, as indicated by a configuration type of 3. According to one embodiment, the pattern information of the resource grid (840) may include RE-related parameters (e.g., dmrs.DMRSRePosition) and RB-related parameters (e.g., dmrs.DMRSRbPosition). The RE-related parameters may indicate the position of a DMRS within an RB as a bitmap, and the RB-related parameters may indicate the position of an RB including a DMRS within the resource grid as a bitmap. At this time, in order to indicate a DMRS pattern allocated in units of two slots, the RE-related parameters or the RB-related parameters may include respective indication information (e.g., a bitmap) corresponding to each slot. For example, the DMRS pattern information of the resource grid (840) indicated by 'dmrs.DMRSRbPosition={101, 011}' may indicate a pattern in which the {1, 3}th RB of the first slot and the {2, 3}th RB of the second slot in the resource grid include DMRS, and the DMRS pattern information of the resource grid (840) indicated by 'dmrs.DMRSRePosition={010010010100, 100100101000}' may indicate a pattern in which the {2, 5, 8, 10}th RE of each RB including the DMRS of the first slot and the {1, 4, 7, 9}th RE of each RB including the DMRS of the second slot include DMRS.

[0112] However, this is only an example, and according to various embodiments, the DMRS pattern is not limited to two slot units and can be extended according to the design technique of channel prediction.

[0113] As described above, according to various embodiments of the present disclosure, a base station or a terminal can generate and instruct a DMRS pattern adaptively optimized for a local environment, base station performance, or operator characteristics (based on an AI model), thereby reducing the overhead of DMRS and utilizing efficient resources despite the increased number of antenna ports or layers due to the introduction of high-frequency bands.

[0114] Figure 9 illustrates a flow of operations for selecting a DMRS pattern based on channel measurements and an AI model according to various embodiments of the present disclosure. More specifically, Figure 9 illustrates a flow of operations for a terminal to determine an optimal DMRS pattern among multiple candidate DMRS patterns set by a base station to a terminal.

[0115] According to various embodiments of the present disclosure, a terminal selects optimal DMRS patterns (e.g., Top-k DMRS patterns (k)) from among a plurality of DMRS patterns (e.g., N DMRS patterns) transmitted by a base station (or network (NW)). <N))을 결정할 수 있고, 이를 보고할 수 있다. 이 때, 기지국은 복수 개의 DMRS 패턴들을 설정함과 함께 기지국에서 훈련시킨 AI 모델도 단말에게 전송할 수 있다. 다만, 기지국이 전송하는 AI 모델은 단말의 하드웨어를 고려하여 설계 또는 훈련시킨 것이 아닌 바, 단말의 채널 측정(channel estimation) 방법이나 단말의 주파수 오프셋 등에 따라 성능이 달라질 수 있다.

[0116] Accordingly, a process may be required to select k DMRS patterns (or AI models) with good performance from the terminal's perspective among the DMRS patterns set by the base station. Here, the number of DMRS patterns, N or k, is a value directly set by the user (e.g., a hyperparameter). The base station can directly inform the terminal of this value through signaling, or the terminal can arbitrarily determine k and report it to the base station.

[0117] Referring to FIG. 9, a specific process in which a terminal determines k DMRS patterns is illustrated.

[0118] In step (905), the terminal may perform channel measurement for a plurality of candidate DMRS patterns. In one embodiment, the candidate DMRS patterns may include DMRS patterns generated by the base station and set for the terminal, or may include DMRS patterns generated by an AI model included in the terminal. In one embodiment, the terminal may receive a downlink signal (e.g., y_dmrs,1 to y_dmrs,N) each including a plurality (e.g., N) of DMRS patterns, and measure a channel for the same.

[0119] In steps (915-1 to 915-N), the terminal may perform channel prediction using an AI model corresponding to each of a plurality of DMRS patterns. For example, the terminal may perform channel prediction using a channel value for each DMRS pattern (e.g., Inland ) to perform channel prediction based on AI model to estimate the interpolated channel value (e.g., Inland ) can be obtained. Here, the process of the terminal predicting the channel based on the AI ​​model may include a process of sequentially or simultaneously performing channel estimation for the frequency domain or channel estimation for the time domain. According to one embodiment, when pattern generalization is applied, the AI ​​model used by the terminal may be smaller than the number of multiple DMRS patterns. According to one embodiment, the AI ​​model used by the terminal may include an AI model trained by the terminal or an AI model trained by the base station and transmitted to the terminal.

[0120] In step (925), the terminal may determine one or more DMRS patterns having superior performance metrics based on a plurality of estimated channel values ​​(e.g., N or a smaller number in the case of pattern generalization). For example, the terminal may perform channel equalization, channel detection, and channel decoding for compensating for channel loss on the estimated channel values, and may identify a quality metric for each signal accordingly. In one embodiment, the terminal may measure a decoding signal to interference plus noise ratio (SINR) for each signal (or channel) through channel decoding, and determine a DMRS pattern for one or more signals having an SINR above a specific threshold value. In one embodiment, the terminal may identify a bit error rate (BER) or block error rate (BLER) value for each signal, and determine a DMRS pattern for one or more signals having superior quality by comparing the BER or BLER value with a specific threshold value. According to one embodiment, the threshold value that the terminal compares with the measured value of the signal to determine one or more DMRS patterns may be preset or may be received through separate signaling from the base station.

[0121] In one embodiment, the terminal may then report one or more determined DMRS patterns (e.g., the selected Top-k DMRS patterns) to the base station. Through the above-described process, the base station can identify DMRS patterns optimized for or available to the terminal.

[0122] FIG. 10 illustrates an example of performing channel prediction in the frequency domain and time domain based on an AI model according to various embodiments of the present disclosure. More specifically, FIG. 10 illustrates an algorithm for a terminal or a base station to predict a channel in the frequency domain or a channel in the time domain for a channel estimated based on a DMRS pattern. According to one embodiment, the AI ​​model or channel prediction algorithm used in FIG. 10 may be similar to the algorithm of the AI ​​model for channel prediction described in FIG. 7.

[0123] According to various embodiments of the present disclosure, with reference to FIG. 10, channel prediction performed by a terminal or a base station may be performed according to a third algorithm (1010) that first predicts a channel for a frequency domain based on a signal including a DMRS pattern and then sequentially predicts a channel for a time domain, or a fourth algorithm (1020) that simultaneously performs channel prediction for a frequency domain and channel prediction for a time domain based on a signal including a DMRS pattern. According to one embodiment, before performing the channel estimation of FIG. 10, the terminal or base station may first perform measurement for a channel on which a DMRS is received.

[0124] Referring to the third algorithm (1010), the terminal or base station may perform frequency-domain channel estimation (e.g., channel interpolation) based on the measured channel value and the AI ​​model for the signal including the DMRS pattern. Thereafter, the terminal or base station may perform time-domain channel estimation (e.g., channel interpolation) for each estimated frequency-domain channel value. In one embodiment, since the third algorithm (1010) estimates AI model-based channels separately in the frequency domain and the time domain, two AI models may be required. Alternatively, in one embodiment, the terminal or base station may perform non-AI-based time-domain channel estimation (e.g., legacy technique channel interpolation) for the estimated frequency-domain channel value.

[0125] With reference to the fourth algorithm (1020), a terminal or base station can simultaneously perform channel estimation (e.g., channel interpolation) in the frequency domain and the time domain based on the measured channel value and the AI ​​model for a signal including a DMRS pattern. According to one embodiment, the channel prediction of the fourth algorithm (1020) can be performed using a single AI model. Accordingly, the output value of the AI ​​model through the fourth algorithm (1020) can include a result value of channel prediction that is channel interpolated in both the time domain and the frequency domain. According to one embodiment, the fourth algorithm (1020) may require a higher computational load and a larger AI model size than the third algorithm (1010). Therefore, a terminal or base station including an AI model capable of performing a higher computational load can utilize the fourth algorithm (1020) to perform more accurate channel estimation.

[0126] According to various embodiments of the present disclosure, the following describes an operational flow for a terminal or a base station to align DMRS patterns and perform channel prediction based on the embodiments described in FIGS. 6 to 10 described above. However, not all of the embodiments or operations are considered essential components, and it is obvious that embodiments of the present disclosure may include at least one of all, some, or a combination of some of the disclosed components.

[0127] According to various embodiments, a base station (e.g., a network (NW)) may train an AI model and perform operations for channel prediction accordingly, and a terminal may train an AI model and perform operations for channel prediction accordingly. For convenience, these may be referred to as base station model-based channel prediction and terminal model-based channel prediction, respectively.

[0128] FIG. 11 illustrates a signal flow for measuring a channel based on an AI model of a network according to various embodiments of the present disclosure. More specifically, referring to FIG. 11, a signal flow for base station model-based channel prediction is illustrated. According to an embodiment, as described in FIG. 10, channel prediction may mean channel prediction for at least one of a frequency domain, a time domain, or a spatial domain. According to an embodiment, FIG. 11 describes a process for predicting a channel based on a downlink DMRS signal, but is not limited thereto, and of course, it can be similarly or identically applied to a process in which a base station predicts a channel based on an uplink DMRS signal.

[0129] In step (1105), the terminal may report capability information to the base station. According to various embodiments, the capability information reported by the terminal may be used for selecting an AI model of the base station in the case of base station model-based channel prediction.

[0130] In one embodiment, the capability information transmitted by the terminal to the base station may include an AI-running capability report. In one embodiment, the terminal may transmit the capability information to request a trained AI model from the base station. Accordingly, the base station may select an appropriate trained AI model for the terminal to use.

[0131] According to one embodiment, the capability information transmitted by the terminal may include at least one of whether the terminal supports AI (e.g., non-AI or AI-indicating instruction information), AI computing-related capabilities (e.g., AI operating power consumption, floating point operations per second (FLOPS), input / output (I / O) memory bandwidth, etc.), or AI model-related capabilities (NPU memory, model storage, quantization type, etc.), depending on the implementation of the terminal.

[0132] In one embodiment, the terminal may report at least one of the aforementioned capability information to the base station via separate signaling, or, if each of the aforementioned information is categorized, may report an index for each category. In one embodiment, the categories of AI model-related information may be predefined or include information previously set by the base station to the terminal.

[0133] Category indexquantization typeFLOPSAI memory (MB)Category 0INT810 6 100Category 1INT810 7 Category 2FLOAT1610 8 200Category 3FLOAT1610 9

[0134] According to one embodiment, the terminal may transmit to the base station the capability information including an index for some of the categories as shown in Table 2. However, the above-described Table 2 is merely an example and is not limited thereto, and the capability information transmitted by the terminal to the base station may include more diverse information when it is a capability related to an AI model required for channel prediction, and the category determined by the terminal may include a combination of capability information related to various AI models.

[0135] In step (1115), the base station may configure multiple candidate DMRS patterns for the terminal. In one embodiment, the base station may consider the terminal's capability information to configure the DMRS patterns to be applied to the terminal. In one embodiment, for base station model-based channel prediction, the base station may transmit to the terminal an AI model trained by the base station along with the configuration of candidate DMRS patterns.

[0136] In one embodiment, the multiple candidate DMRS patterns set by the base station to the terminal may be applied based on the DMRS patterns specifically described in FIGS. 8A and 8B . For example, the base station may indicate to the terminal the DMRS pattern indices corresponding to each DMRS pattern indication information determined based on the contents described in FIGS. 8A and 8B .

[0137] In step (1125), the terminal may determine one or more DMRS patterns and report them to the base station. According to one embodiment, step (1125) of determining one or more DMRS patterns among multiple candidate DMRS patterns set by the base station is described in detail in FIG. 9. According to one embodiment, the terminal may report the determined one or more DMRS patterns to the base station based on the DMRS pattern indication method described in FIGS. 8a and 8b.

[0138] In step (1135), the base station may set a DMRS pattern to the terminal. According to one embodiment, the base station may set at least one DMRS pattern to the terminal based on one or more DMRS patterns determined by the terminal. According to one embodiment, the at least one DMRS pattern set by the base station to the terminal may be set or indicated via RRC, MAC CE, or DCI. According to one embodiment, the DMRS patterns set by the base station to the terminal may be applied based on the DMRS patterns specifically described in FIGS. 8A and 8B . For example, the base station may indicate to the terminal DMRS pattern indices corresponding to each DMRS pattern indication information determined based on the contents described in FIGS. 8A and 8B .

[0139] According to one embodiment, the base station can set the index of DMRS patterns to the terminal in various ways. For example, the base station can set the index for one DMRS pattern (e.g., {0}), can set the change of DMRS pattern periodically (e.g., 0 -> 1 -> 2 -> 0 -> 1 ->), can set odd DMRS transmission and even DMRS transmission to use different DMRS patterns (e.g., 0 -> 1 -> 0 -> 1 -> 0), or can set the DMRS pattern to be used in a repetition form (e.g., 0 -> 0 -> 1 -> 1 -> 0). Alternatively, according to one embodiment, the base station can set the DMRS pattern to be changed when a specific event occurs. For example, the occurrence of a specific event may include changing the DMRS pattern when a performance metric deteriorates while performing AI / ML-related LCM (life-cycle-management), changing the DMRS pattern according to a decrease in throughput when monitoring cell throughput as a performance metric, or changing the DMRS pattern for each specific location range within the cell (e.g., zone 1, zone 2, ...). However, this is merely an example and is not limited thereto, and the method of a base station setting DMRS patterns to a terminal may include various methods of setting in consideration of FIGS. 8A and 8B.

[0140] According to various embodiments of the present disclosure, through step (1135), the terminal can check the DMRS pattern set by the base station. If multiple DMRS patterns are set, the terminal can use all of the estimated multiple DMRS channels as input values ​​of the AI ​​model and improve the performance of channel prediction and interpolation by utilizing domain-specific correlation. For example, in case of frequency-domain channel interpolation, the terminal can utilize frequency-domain correlation, and in case of time-domain channel interpolation, the terminal can utilize time-domain correlation.

[0141] In step (1145), the base station may transmit a PDSCH to the terminal. According to one embodiment, the base station may transmit a PDSCH including a DMRS to the terminal based on a configured DMRS pattern.

[0142] In step (1155), the terminal may measure the channel based on the received PDSCH. In one embodiment, the terminal may perform channel measurement on the DMRS included in the PDSCH. In one embodiment, the DMRS-based channel measurement performed by the terminal may include channel measurement based on legacy techniques such as least squares, linear minimum mean square error (LMMSE), or channel measurement based on an AI model.

[0143] In step (1165), the terminal can predict a channel based on measured channel values ​​and an AI model. According to one embodiment, the process of predicting a channel using an AI model algorithm based on measured channel values ​​is described in detail in FIG. 10.

[0144] According to various embodiments of the present disclosure, FIG. 11 illustrates base station model-based channel prediction. Although not illustrated, in a series of processes, the base station can train an AI model for channel prediction in advance and generate or determine an optimized DMRS pattern based on the AI ​​model. Accordingly, the trained AI model can be transmitted to the terminal, and the terminal can use the AI ​​model to predict the channel. According to one embodiment, the process of training the AI ​​model and generating the DMRS pattern described above is described in detail in FIGS. 6 and 7.

[0145] FIG. 12 illustrates a signal flow for measuring a channel based on an AI model of a terminal according to various embodiments of the present disclosure. More specifically, referring to FIG. 12, a signal flow for terminal model-based channel prediction is illustrated. According to an embodiment, as described in FIG. 10, channel prediction may mean channel prediction for at least one of a frequency domain, a time domain, or a spatial domain. According to an embodiment, FIG. 12 describes a process for predicting a channel based on a downlink DMRS signal, but is not limited thereto, and can be similarly or identically applied to a process in which a base station predicts a channel based on an uplink DMRS signal.

[0146] Although not illustrated in FIG. 12, prior to step (1205), the terminal may report capability information to the base station. According to various embodiments, the capability information reported by the terminal may be used to report DMRS patterns based on an AI model trained by the terminal in the case of terminal model-based channel prediction. According to one embodiment, the process of transmitting the capability information of the terminal may include a process similar to or identical to the process specifically described in step (1105) of FIG. 11.

[0147] Although not illustrated in FIG. 12, prior to step (1205), the base station may configure multiple candidate DMRS patterns for the terminal. According to various embodiments, the terminal may train an AI model using the multiple candidate DMRS patterns configured by the base station, or select an optimal DMRS among them. According to one embodiment, the process of transmitting the terminal's capability information may include a process similar to or identical to the process specifically described in step (1115) of FIG. 11.

[0148] Although not illustrated in FIG. 12, according to various embodiments of the present disclosure, FIG. 12 describes terminal model-based channel prediction, wherein, in a series of processes, the terminal can train an AI model for channel prediction in advance and generate or determine an optimized DMRS pattern accordingly. Accordingly, the terminal can report the trained AI model and the DMRS patterns generated / determined accordingly to the base station. According to one embodiment, the process of training the AI ​​model and generating the DMRS pattern described above is specifically described in FIG. 6 and FIG. 7.

[0149] In step (1205), the base station may configure the terminal to report candidate DMRS patterns. For example, the base station may instruct the terminal to report candidate DMRS patterns. This process is because the terminal performs AI model training, and therefore the base station needs to report the DMRS patterns trained by the terminal. Specifically, since each terminal may have different methods for channel measurement / balancing (estimation / equalization), the AI ​​models (e.g., DMRS patterns) learned in the environment where each terminal performs AI model training may be different. Furthermore, since the optimized AI configuration according to the hardware implementation of each terminal may also differ, to compensate for this, the terminal may report candidate DMRS patterns to the base station.

[0150] According to one embodiment, the candidate DMRS pattern report set by the base station may include at least one of information about the period of the DMRS pattern report or information about the number of DMRS patterns to be reported.

[0151] In one embodiment, the DMRS pattern reporting cycle may include the time at which the base station instructs the terminal to report the DMRS pattern, such as when initial access or event-driven reporting is required. In one embodiment, the need for event-driven reporting may include cases where the available DMRS patterns change depending on the AI ​​operating status of the terminal (e.g., battery, memory, etc.), or cases where performance of the DMRS pattern used by the terminal deteriorates, requiring a new DMRS pattern.

[0152] According to one embodiment, the number of DMRS patterns to be reported may include the number of all DMRS patterns trained by the terminal or some of the DMRS patterns among all of the DMRS patterns, taking into account AI operation capabilities (e.g., monitoring capabilities, model management capabilities, etc.).

[0153] In step (1215), the terminal may report a candidate DMRS pattern to the base station. In one embodiment, the terminal may report to the base station one or more DMRS patterns determined based on the candidate DMRS pattern configuration and an AI model trained by the terminal. In one embodiment, the DMRS pattern determined by the AI ​​model transmitted by the terminal may be applied based on the DMRS patterns and corresponding instruction methods specifically described in FIGS. 8A and 8B .

[0154] In step (1225), the base station may transmit the reported DMRS to the terminal. In one embodiment, the base station may transmit a PDSCH including the DMRS based on the candidate DMRS pattern reported to the terminal. In one embodiment, a series of processes according to step (1225) may be performed to pair DMRS patterns between the base station and the terminal. For example, the base station may, through the following process, determine which DMRS pattern trained by the terminal is more suitable for the base station's location / operator environment.

[0155] In step (1235), the terminal may determine one or more DMRS patterns and report them to the base station. According to one embodiment, step (1235) of determining one or more DMRS patterns for the DMRS patterns included in the PDSCH transmitted by the base station may include a process similar to or identical to the process specifically described in FIG. 9 and specifically described in step (1125) of FIG. 11. According to one embodiment, the terminal may report the determined one or more DMRS patterns to the base station based on the DMRS pattern indication method described in FIGS. 8A and 8B.

[0156] In step (1245), the base station may set a DMRS pattern for the terminal. According to one embodiment, the base station may set at least one DMRS pattern for the terminal based on one or more DMRS patterns determined by the terminal. According to one embodiment, the at least one DMRS pattern set by the base station for the terminal may be set or indicated via RRC, MAC CE, or DCI. According to one embodiment, the DMRS patterns set by the base station for the terminal may be applied based on the DMRS patterns specifically described in FIGS. 8A and 8B . According to one embodiment, step (1245) may include a process similar to or identical to the process specifically described in step (1135) of FIG. 11 .

[0157] In step (1255), the base station may transmit a PDSCH to the terminal. In one embodiment, the base station may transmit a PDSCH including a DMRS to the terminal based on a configured DMRS pattern. In one embodiment, step (1255) may include a process similar to or identical to the process specifically described in step (1145) of FIG. 11 .

[0158] In step (1265), the terminal may measure the channel based on the received PDSCH. In one embodiment, the terminal may perform channel measurement for the DMRS included in the PDSCH. In one embodiment, step (1265) may include a process similar to or identical to the process specifically described in step (1155) of FIG. 11 .

[0159] In step (1175), the terminal may predict a channel based on the measured channel value and the AI ​​model. According to one embodiment, the process of predicting a channel using an AI model algorithm based on the measured channel value is described in detail in FIG. 10 . According to one embodiment, step (1275) may include a process similar to or identical to the process described in detail in step (1165) of FIG. 11 .

[0160] 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.

[0161] When 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 methods according to embodiments described in the claims or specification of the present disclosure.

[0162] 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-ROM (CD-ROM), 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.

[0163] Additionally, the 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. Such a storage device may be connected to a device implementing an embodiment of the present disclosure via an external port. Additionally, a separate storage device on the communication network may be connected to a device implementing an embodiment of the present disclosure.

[0164] In the specific embodiments of the present disclosure described above, components included in the disclosure are expressed singularly or plurally, depending on the specific embodiment 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.

[0165] While the detailed description of this disclosure has described specific embodiments, it should be understood that various modifications are possible without departing from the scope of this disclosure. Therefore, the scope of this disclosure should not be limited to the described embodiments, but should be defined not only by the scope of the claims described below, but also by equivalents thereof.

Claims

1. In a wireless communication system, a terminal (user equipment, UE) is Transceiver; and Including a controller coupled with the above transceiver, The above controller, Receive first information from a base station that sets a set of candidate DMRS (demodulation reference signal) patterns, Based on the AI ​​(artificial intelligence) model, a first channel prediction is performed for the set of candidate DMRS patterns, Based on the results of the first channel prediction, one or more DMRS patterns are identified, Transmitting information about one or more DMRS patterns to the base station, and A terminal configured to receive, from the base station, second information for setting at least one DMRS pattern based on the one or more DMRS patterns.

2. In claim 1, if the AI ​​model is an AI model trained from the base station, the controller, Transmitting to the base station, the terminal's capability information regarding AI-based channel prediction, and A terminal further configured to receive information about the AI ​​model from the base station.

3. In claim 1, the controller, From the base station, a PDSCH (physical downlink shared channel) is received based on at least one DMRS pattern, and A terminal further configured to perform second channel prediction based on the AI ​​model and the PDSCH.

4. In claim 1, The above first information is a terminal including at least one of information indicating whether DMRS is included per RB (resource block) for each of one or two slots or information indicating the location of DMRS within the RB.

5. In claim 1, if the AI ​​model is an AI model trained from the terminal, the controller, To the base station, based on the AI ​​model, transmit information about DMRS patterns that the terminal can use among the set of candidate DMRS patterns, and Further configured to receive a PDSCH including a plurality of DMRS patterns based on DMRS patterns that can be used by the terminal from the base station, The above first channel prediction is performed for the plurality of DMRS patterns based on the AI ​​model.

6. In a wireless communication system, a base station is Transceiver; and Including a controller coupled with the above transceiver, The above controller, Transmitting first information for setting a set of candidate DMRS (demodulation reference signal) patterns to a terminal (user equipment, UE), Receive information about one or more DMRS patterns from the terminal, and configured to transmit to the terminal, second information setting at least one DMRS pattern based on the one or more DMRS patterns, A base station, wherein said one or more DMRS patterns are identified based on a result of a first channel prediction for a set of candidate DMRS patterns, based on an AI model of said terminal.

7. In claim 6, if the AI ​​model is an AI model trained from the base station, the controller, Receive from the terminal, the terminal's capability information regarding AI-based channel prediction, and A base station further configured to transmit information about the AI ​​model to the terminal.

8. In claim 6, the controller, A base station further configured to transmit a PDSCH (physical downlink shared channel) to the terminal based on the at least one DMRS pattern.

9. In claim 6, A base station, wherein the first information includes at least one of information indicating whether a DMRS is included in each RB (resource block) for one or two slots or information indicating a DMRS location within an RB.

10. In claim 6, if the AI ​​model is an AI model trained from the terminal, the controller, From the terminal, based on the AI ​​model, information about DMRS patterns that the terminal can use among the set of candidate DMRS patterns is received, and Further configured to transmit, to the terminal, a PDSCH including a plurality of DMRS patterns based on DMRS patterns that the terminal can use, The base station wherein the first channel prediction is performed for the plurality of DMRS patterns based on the AI ​​model.

11. In a wireless communication system, a method performed by a user equipment (UE) is as follows: A step of receiving first information setting a set of candidate DMRS (demodulation reference signal) patterns from a base station; A step of performing a first channel prediction for a set of candidate DMRS patterns based on an AI (artificial intelligence) model; A step of identifying one or more DMRS patterns based on the result of the first channel prediction; a step of transmitting information about one or more DMRS patterns to the base station; and A method comprising the step of receiving second information setting at least one DMRS pattern based on the one or more DMRS patterns from the base station.

12. In claim 11, if the AI ​​model is an AI model trained from the base station, the method, A step of transmitting, to the base station, capability information of the terminal regarding AI-based channel prediction; and A method further comprising the step of receiving information about the AI ​​model from the base station.

13. In claim 11, the method comprises: A step of receiving a PDSCH (physical downlink shared channel) based on at least one DMRS pattern from the base station; and A method further comprising the step of performing second channel prediction based on the AI ​​model and the PDSCH.

14. In claim 11, A method wherein the first information includes at least one of information indicating whether a DMRS is included in each RB (resource block) for each of one or two slots or information indicating a DMRS location within an RB.

15. In claim 11, if the AI ​​model is an AI model trained from the terminal, the method, A step of transmitting, to the base station, information about DMRS patterns that the terminal can use among the set of candidate DMRS patterns based on the AI ​​model; and Further comprising a step of receiving a PDSCH including a plurality of DMRS patterns based on DMRS patterns that can be used by the terminal from the base station, A method in which the first channel prediction is performed for the plurality of DMRS patterns based on the AI ​​model.

Images