Electronic device and method for performing channel estimation, and storage medium

Abstract

This base station may comprise: at least one processor; and a memory for storing instructions. The instructions, when executed individually or collectively by the at least one processor, may instruct the base station to: identify weights, of a neural network, for estimating distance parameters and angle parameters for paths of channels by using channel information of the channels which are associated with receive antennas; obtain uplink signals including reference signals received by applying receive beamforming information generated using the weights to the receive antennas; identify distance parameters and angle parameters for paths of receive channels associated with the receive antennas on the basis of providing the reference signals to the neural network; and decode the uplink signals on the basis of channel information of the receive channels estimated using the distance parameters and the angle parameters for the paths of the receive channels.

Description

Electronic device, method, and storage medium for performing channel estimation

[0001] The following descriptions relate to electronic devices, methods, and storage media for performing channel estimation.

[0002] To improve signal transmission and reception performance, multiple antenna elements may be utilized. For example, the technologies utilized by the multiple antenna elements may include single-input multiple-output (SIMO) technology, multiple-input single-output (MISO) technology, and multiple-input multiple-output (MIMO) technology. The channel capacity of a wireless communication system utilizing the above technologies utilizing multiple antenna elements can be significantly improved compared to single-antenna technology.

[0003] As transmission capacity increases in wireless communication systems, functional splitting, which functionally separates base stations, is being implemented. Through functional splitting, base stations can be divided into distributed units (DUs) and radio units (RUs). A fronthaul interface is defined for communication between the DUs and RUs.

[0004] The above information may be provided as background art to aid in understanding the present disclosure. No claim or determination is made as to whether any of the above is applicable as prior art related to the present disclosure.

[0005] A base station may include at least one processor including a processing circuit. The base station may include a memory storing instructions and including one or more storage media. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify weights of a neural network for estimating distance parameters and angular parameters for paths of a channel associated with receive antennas using channel information of the channel. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to obtain uplink signals including reference signals received by applying receive beamforming information generated using the weights to the receive antennas. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify distance parameters and angular parameters for paths of a receive channel associated with the receive antennas based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to perform decoding of the uplink signals based on channel information of the receive channel estimated using the distance parameter and the angle parameter for the paths of the receive channel.

[0006] A lower network node may include an RF transceiver including receive antennas. The lower network node may include at least one processor including a processing circuit. The lower network node may include a memory storing instructions and including one or more storage media. The instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to obtain, from a higher network node, a neural network for estimating distance parameters and angle parameters for paths of a channel associated with the receive antennas using channel information of the channel. The instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to receive, from a terminal, uplink signals including reference signals by applying receive beamforming information generated using weights of the neural network to the receive antennas. The instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to identify distance parameters and angle parameters for paths of receive channels associated with the receive antennas between the lower network node and the terminal based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network. The instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to provide channel information for the receive channel estimated using the distance parameters and the angle parameters for the paths of the receive channel to the upper network node for decoding of the uplink signals.

[0007] A method performed by a base station may include identifying weights of a neural network for estimating distance parameters and angle parameters for paths of a channel using channel information associated with receive antennas. The method may include obtaining uplink signals including reference signals received by applying receive beamforming information generated using the weights to the receive antennas. The method may include identifying distance parameters and angle parameters for paths of a receive channel associated with the receive antennas based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network. The method may include performing decoding on the uplink signals based on channel information for the receive channel estimated using the distance parameters and the angle parameters for the paths of the receive channel.

[0008] A non-transitory computer-readable storage medium may store one or more programs comprising instructions that, when individually or collectively executed by at least one processor of a base station, cause the base station to identify weights of a neural network for estimating distance parameters and angle parameters for paths of a channel using channel information associated with receive antennas. The non-transitory computer-readable storage medium may store one or more programs comprising instructions that, when individually or collectively executed by the at least one processor, cause the base station to obtain uplink signals including received reference signals by applying receive beamforming information generated using the weights to the receive antennas. The non-transitory computer-readable storage medium may store one or more programs comprising instructions that, when individually or collectively executed by the at least one processor, cause the base station to identify distance parameters and angle parameters for paths of receive channels associated with the receive antennas based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network. The non-transitory computer-readable storage medium may store one or more programs comprising instructions that, when individually or collectively executed by the at least one processor, cause the base station to perform decoding of the uplink signals based on channel information for the receive channel estimated using the distance parameters and the angle parameters for the paths of the receive channel.

[0009] Figure 1 illustrates an example of a wireless communication system.

[0010] Figure 2 illustrates an example of an interface between an upper network node and a lower network node.

[0011] Figure 3a illustrates an example of a functional configuration of an electronic device.

[0012] Figure 3b shows an example of a resource structure in the time domain and frequency domain.

[0013] Figure 4 illustrates an example of a method for estimating a channel using beamforming information applied to receiving antennas of a base station.

[0014] Figure 5 illustrates an example of a neural network that estimates channel information using a received reference signal.

[0015] Figure 6a illustrates an example of an operational flow for how a base station performs channel estimation through a trained neural network.

[0016] Figure 6b shows an example of a neural network for estimating channel parameters.

[0017] Figure 6c illustrates an example of a method for adjusting channel parameters using a gradient descent algorithm.

[0018] FIG. 7a illustrates an example of a beam pattern of received beamforming information within a first wireless environment.

[0019] Figure 7b shows an example of a graph representing NMSE (normalized mean square error) according to SNR (signal to noise ratio) in a first wireless environment.

[0020] Figure 7c shows an example of a graph showing SE (spectral efficiency) according to SNR in a first wireless environment.

[0021] Figure 8a shows an example of a graph representing NMSE according to SNR in a second wireless environment.

[0022] Figure 8b shows an example of a graph representing SE according to SNR in a second wireless environment.

[0023] Figure 9 illustrates an example of an operational flow for a method of performing channel estimation using a received reference signal by applying received beamforming information generated using the weights of a learned neural network to receiving antennas.

[0024] 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.

[0025] 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.

[0026] In the following description, terms referring to signals (e.g., packet, message, signal, information, signaling), terms referring to resources (e.g., section, symbol, slot, subframe, radio frame, subcarrier, RE (resource element), RB (resource block), BWP (bandwidth part), band, spectrum), terms for operational states (e.g., step, operation, procedure), terms referring to data (e.g., packet, message, user stream, information, bit, symbol, codeword), terms referring to channels, terms referring to network entities (distributed unit (DU), radio unit (RU), central unit (CU), control plane (CU-CP), user plane (CU-UP), open radio access network (O-RAN) DU (O-DU), O-RAN RU (O-RU), Terms such as O-CU (O-RAN CU), O-CU-UP (O-RAN CU-CP), O-CU-CP (O-RAN CU-CP)), referring to components of the device, are examples for convenience of explanation. Therefore, the present disclosure is not limited to the terms described below, and other terms having equivalent technical meanings may be used. In addition, terms such as '... part', '... device', '... object', '... body', etc. used below may mean at least one shape structure or a unit that processes a function.

[0027] In addition, in the present disclosure, expressions such as "more than" or "less than" may be used to determine whether a specific condition is satisfied or fulfilled, but this is merely a description for expressing an example and does not exclude descriptions such as "more than" or "less than." A condition described as "more than" may be replaced with "more than," a condition described as "less than" may be replaced with "less than," and a condition described as "more than and less than" may be replaced with "more than and less than." In addition, hereinafter, "A" to "B" mean at least one of elements from A (including A) to B (including B). hereinafter, "C" and / or "D" mean at least one of "C" or "D," that is, including {"C", "D", "C" and "D"}.

[0028] Although this disclosure describes embodiments using terminology used in certain communication standards (e.g., 3rd Generation Partnership Project (3GPP)), this is merely an example for illustrative purposes. Embodiments of this disclosure can also be applied to other communication and broadcasting systems.

[0029] Figure 1 illustrates an example of a wireless communication system.

[0030] Referring to FIG. 1, FIG. 1 illustrates a base station (110) and a terminal (120) as some of the nodes utilizing a wireless channel in a wireless communication system. Although FIG. 1 illustrates only one base station, the wireless communication system may further include other base stations identical or similar to the base station (110).

[0031] The base station (110) is a network infrastructure that provides wireless access to the terminal (120). The base station (110) has coverage defined 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)', '5th generation node', 'next generation nodeB (gNB)', 'wireless point', 'transmission / reception point (TRP)', or other terms having equivalent technical meanings.

[0032] The terminal (120) is a device used by a user and communicates with the base station (110) via a wireless channel. The link from the base station (110) to the terminal (120) is referred to as a downlink (DL), and the link from the terminal (120) to the base station (110) is referred to as an uplink (UL). In addition, although not shown in FIG. 1, the terminal (120) and another terminal may communicate with each other via a wireless channel. In this case, the link between the terminal (120) and another terminal (device-to-device link, D2D) is referred to as a sidelink, and the sidelink may be used interchangeably with the PC5 interface. In some other embodiments, the terminal (120) may be operated without the involvement of a user. In one embodiment, the terminal (120) is a device that performs machine type communication (MTC) and may not be carried by the user. Additionally, according to one embodiment, the terminal (120) may be an NB (narrowband)-IoT (internet of things) device.

[0033] The terminal (120) may be referred to as a terminal, or other terms such as 'user equipment (UE),' 'customer premises equipment (CPE),' 'mobile station,' 'subscriber station,' 'remote terminal,' 'wireless terminal,' 'electronic device,' or 'user device,' or other terms having equivalent technical meanings.

[0034] The base station (110) and the terminal (120) can perform beamforming. The base station (110) and the terminal (120) can transmit and receive wireless signals in a relatively low frequency band (e.g., FR 1 (frequency range 1) of NR). In addition, the base station (110) and the terminal (120) can transmit and receive wireless signals in a relatively high frequency band (e.g., FR 2 (or, FR 2-1, FR 2-2, FR 2-3), FR 3 of NR), millimeter wave (mmWave) band (e.g., 28 GHz, 30 GHz, 38 GHz, 60 GHz)). To improve channel gain, the base station (110) and the terminal (120) can perform beamforming. Here, the beamforming can include transmission beamforming and reception beamforming. The base station (110) and the terminal (120) can impart directionality to the transmitted or received signal. To this end, the base station (110) and the terminal (120) can select serving beams through a beam search or beam management procedure. After the serving beams are selected, subsequent communication can be performed through resources that have a QCL relationship with the resource that transmitted the serving beams.

[0035] If large-scale characteristics of a channel carrying a symbol on a first antenna port can be inferred from a channel carrying a symbol on a second antenna port, the first antenna port and the second antenna port can be evaluated to have a QCL relationship. For example, the large-scale characteristics may include at least one of delay spread, Doppler spread, Doppler shift, average gain, average delay, and a spatial receiver parameter.

[0036] Although both the base station (110) and the terminal (120) are described as performing beamforming in FIG. 1, the embodiments of the present disclosure are not necessarily limited thereto. In some embodiments, the terminal may or may not perform beamforming. Furthermore, the base station may or may not perform beamforming. That is, either only one of the base station and the terminal may perform beamforming, or neither the base station nor the terminal may perform beamforming.

[0037] In the present disclosure, a beam refers to a spatial flow of a signal in a wireless channel, and is formed by one or more antennas (or antenna elements), and this forming process may be referred to as beamforming. Beamforming may include at least one of analog beamforming and digital beamforming (e.g., precoding). Reference signals transmitted based on beamforming may include, for example, a demodulation-reference signal (DM-RS), a channel state information-reference signal (CSI-RS), a synchronization signal / physical broadcast channel (SS / PBCH), and a sounding reference signal (SRS). In addition, as a configuration for each reference signal, an IE such as a CSI-RS resource or an SRS-resource may be used, and this configuration may include information associated with the beam. Information associated with a beam may mean whether the configuration (e.g., a CSI-RS resource) uses the same spatial domain filter as another configuration (e.g., another CSI-RS resource within the same CSI-RS resource set) or a different spatial domain filter, or whether it is quasi-co-located (QCL) with a reference signal, and if so, what type it is (e.g., QCL type A, B, C, D).

[0038] In the past, in communication systems with relatively large cell radius of base stations, each base station was installed to include the functions of a digital processing unit (or distributed unit (DU)) and a radio frequency (RF) processing unit (or radio unit (RU)). However, as higher frequency bands are used in 4G (4th generation) and / or subsequent communication systems (e.g., 5G) and the cell coverage of base stations becomes smaller, the number of base stations to cover a specific area has increased. The installation costs for operators to install base stations have also increased. In order to minimize the installation costs of base stations, a structure has been proposed in which the DU and RU of a base station are separated, one or more RUs are connected to one DU via a wired network, and one or more RUs are geographically distributed to cover a specific area. Hereinafter, the deployment structure and expansion examples of base stations according to various embodiments of the present disclosure are described through FIG. 2.

[0039] Figure 2 illustrates an example of an interface between an upper network node and a lower network node.

[0040] FIG. 2 illustrates an interface between an upper network node and a lower network node. The interface between the upper network node and the lower network node may include a fronthaul interface. Fronthaul refers to entities between a wireless LAN and a base station, unlike backhaul between a base station and a core network. FIG. 2 illustrates an example of a fronthaul structure between an upper network node (210) and one lower network node (220), but this is merely for convenience of explanation and the present disclosure is not limited thereto. In other words, embodiments of the present disclosure may also be applied to a fronthaul structure between one upper network node and multiple lower network nodes. For example, embodiments of the present disclosure may be applied to a fronthaul structure between one upper network node and two lower network nodes. Furthermore, embodiments of the present disclosure may also be applied to a fronthaul structure between one upper network node and three lower network nodes.

[0041] For example, an upper network node may include a digital unit / distributed unit (DU). The upper network node may be referred to as a DU. A lower network node may include a radio unit (RU) or a massive MIMO unit (MMU). The lower network node may be referred to as a RU or an MMU.

[0042] Referring to FIG. 2, a base station (110) may include an upper network node (210) and a lower network node (220). A fronthaul (215) between the upper network node (210) and the lower network node (220) may be operated via an Fx interface. For operation of the fronthaul (215), an interface such as an enhanced common public radio interface (eCPRI) or radio over ethernet (ROE) may be used, for example.

[0043] As communication technology develops, mobile data traffic increases, and accordingly, the bandwidth demand required in the fronthaul between the digital unit and the wireless unit has increased significantly. In a deployment such as a centralized / cloud radio access network (C-RAN), an upper network node (210) performs functions for packet data convergence protocol (PDCP), radio link control (RLC), media access control (MAC), and physical (PHY), and a lower network node (220) may be implemented to perform functions for the PHY layer in addition to the RF (radio frequency) function.

[0044] The upper network node (210) may be responsible for upper layer functions of a wireless network. For example, the upper network node (210) may perform functions of the MAC layer and a part of the PHY layer. Here, a part of the PHY layer refers to functions performed at a higher level among the functions of the PHY layer, and may include, for example, channel encoding (or channel decoding), scrambling (or descrambling), modulation (or demodulation), and layer mapping (or layer demapping). According to an embodiment, when the upper network node (210) complies with the O-RAN standard, it may be referred to as an O-DU (O-RAN DU) (or DU). The upper network node (210) may be replaced with a first network entity or DU for a base station (e.g., gNB) in embodiments of the present disclosure, as needed.

[0045] The lower network node (220) may be responsible for lower layer functions of the wireless network. For example, the lower network node (220) may perform a part of the PHY layer, an RF function. Here, a part of the PHY layer refers to functions of the PHY layer that are performed at a relatively lower level than the upper network node (210), and may include, for example, iFFT transformation (or FFT transformation), CP (cyclic prefix) insertion (CP removal), and digital beamforming. An example of such specific functional separation is described in detail in FIG. 4. The lower network node (220) may be referred to as an 'access unit (AU)', an 'access point (AP)', a 'transmission / reception point (TRP)', a 'remote radio head (RRH)', a 'radio unit (RU)', or other terms having an equivalent technical meaning thereto. In one embodiment, if a lower network node (220) complies with the O-RAN standard, it may be referred to as an O-RU (O-RAN RU) (or RU). The lower network node (220) may be replaced with a second network entity or RU for a base station (e.g., gNB) in embodiments of the present disclosure, as needed.

[0046] Although the above example describes that the upper network node (210) includes a DU and the lower network node (220) includes an RU, the embodiments of the present disclosure are not limited thereto. A base station according to the embodiments may be implemented in a distributed deployment according to a centralized unit (CU) configured to perform functions of upper layers of an access network (e.g., packet data convergence protocol (PDCP), radio resource control (RRC)) and a distributed unit (DU) configured to perform functions of lower layers. At this time, the distributed unit (DU) may include a digital unit (DU) and a radio unit (RU). Between a core (e.g., 5GC (5G core) or NGC (next generation core)) network and a radio network (RAN), the base station may be implemented in a structure in which CU, DU, and RU are arranged in that order. The interface between the CU and the distributed unit (DU) may be referred to as an F1 interface.

[0047] For example, a centralized unit (CU) may be connected to one or more DUs and may be responsible for functions at a higher layer than the DU. For example, the CU may be responsible for functions at the RRC (radio resource control) and PDCP (packet data convergence protocol) layers, while the DU and RU may be responsible for functions at lower layers. The DU may perform some functions (high PHY) of the RLC (radio link control), MAC (media access control), and PHY (physical) layers, and the RU may be responsible for the remaining functions (low PHY) of the PHY layer. In addition, for example, a digital unit (DU) may be included in a distributed unit (DU) depending on the implementation of a distributed deployment of a base station. Hereinafter, unless otherwise defined, the operations of DU and RU are described, but various embodiments of the present disclosure can be applied to both a base station deployment including a CU and a deployment in which the DU is directly connected to the core network (i.e., a base station in which the CU and DU are integrated into a single entity (e.g., an NG-RAN node)).

[0048] Figure 3a illustrates an example of a functional configuration of an electronic device.

[0049] The configuration of the electronic device (300) illustrated in FIG. 3A can be understood as a configuration of a base station (110), a terminal (120), an upper network node (210) (e.g., DU), or a lower network node (220) (e.g., RU or MMU). Terms such as "...unit" and "...unit" used hereinafter mean a unit that processes at least one function or operation, and this can be implemented by hardware, software, or a combination of hardware and software.

[0050] Referring to FIG. 3A, the electronic device (300) may include a transceiver (310), a memory (320), and a processor (330). However, the present disclosure is not limited thereto. For example, the electronic device (300) may not include at least some of the components illustrated in FIG. 3A, or may include further components not illustrated in FIG. 3A. For example, the electronic device (300) may not include the transceiver (310).

[0051] The transceiver (310) can perform functions for transmitting and receiving signals in a wired communication environment. The transceiver (310) can include a wired interface for controlling direct connections between devices via a transmission medium (e.g., copper wire, optical fiber). For example, the transceiver (310) can transmit electrical signals to other devices via copper wire, or perform conversion between electrical signals and optical signals.

[0052] The transceiver (310) may perform functions for transmitting and receiving signals in a wireless communication environment. For example, the transceiver (310) may perform a conversion function between baseband signals and bit streams according to the physical layer specifications of the system. For example, when transmitting data, the transceiver (310) encodes and modulates the transmitted bit stream to generate complex-valued symbols. Furthermore, when receiving data, the transceiver (310) demodulates and decodes the baseband signal to restore the received bit stream. Furthermore, the transceiver (310) may include multiple transmission and reception paths.

[0053] The transceiver (310) transmits and receives signals as described above. Accordingly, all or part of the transceiver (310) may be referred to as a "communication unit," a "transmitter," a "receiver," or a "transmitter-receiver unit." Furthermore, in the following description, transmission and reception performed via a wireless channel are used to mean that the transceiver (310) performs the processing described above.

[0054] Although not illustrated in FIG. 3A, the transceiver (310) may further include a backhaul transceiver for connection to the core network or other base stations. The backhaul transceiver provides an interface for communicating with other nodes within the network. That is, the backhaul transceiver converts a bit stream transmitted from the base station to other nodes, such as other access nodes, other base stations, upper nodes, the core network, etc., into a physical signal, and converts a physical signal received from other nodes into a bit stream.

[0055] The memory (320) stores data such as basic programs, application programs, and setting information for the operation of the electronic device (300). The memory (320) may be referred to as a storage unit. The memory (320) may be composed of volatile memory, nonvolatile memory, or a combination of volatile memory and nonvolatile memory. In addition, the memory (320) provides stored data upon request from the processor (330).

[0056] For example, the processor (330) may include various processing circuits and / or multiple processors. For example, the term "processor" as used herein, including in the claims, may include various processing circuits including at least one processor, one or more of which may be configured to individually and / or collectively perform the various functions described below in a distributed manner. As used herein, when "processor," "at least one processor," and "one or more processors" are described as being configured to perform various functions, these terms encompass, for example, and without limitation, situations where one processor performs some of the recited functions and other processor(s) perform other parts of the recited functions, and also situations where one processor may perform all of the recited functions. Additionally, the at least one processor may include a combination of processors that perform the various functions enumerated / disclosed, for example, in a distributed manner. At least one processor may execute program instructions to achieve or perform the various functions.

[0057] The processor (330) controls the overall operations of the electronic device (300). The processor (330) may be referred to as a control unit. For example, the processor (330) transmits and receives signals via the transceiver (310) (or via the backhaul communication unit). In addition, the processor (330) records and reads data from the memory (320). In addition, the processor (330) may perform functions of a protocol stack required by a communication standard. Although only the processor (330) is illustrated in FIG. 3A, the electronic device (300) may include two or more processors according to other implementation examples.

[0058] The configuration of the electronic device (300) illustrated in FIG. 3A is merely an example, and examples of electronic devices (300) that perform embodiments of the present disclosure are not limited to the configuration illustrated in FIG. 3A. In some embodiments, some configurations may be added, deleted, or changed.

[0059] For example, when the electronic device (300) is an RU, the electronic device (300) may further include a fronthaul transceiver. For example, the fronthaul transceiver may transmit and receive signals on a fronthaul interface. For example, the fronthaul transceiver may receive a management plane (M-plane) message. For example, the fronthaul transceiver may receive a management plane (S-plane) message. For example, the fronthaul transceiver may receive a control plane (C-plane) message. For example, the fronthaul transceiver may transmit a user plane (U-plane) message. For example, the fronthaul transceiver may receive a user plane message.

[0060] In addition, for example, when the electronic device (300) is an RU, the electronic device (300) may further include an RF (radio frequency) transceiver. For example, the RF transceiver may include antennas. The antenna performs functions for transmitting and receiving signals via a wireless channel. The antenna may include a radiator formed of a conductor or a conductive pattern formed on a substrate (e.g., an antenna PCB, an antenna substrate, the second substrate). The antenna may radiate an up-converted signal on a wireless channel or acquire a signal radiated by another device. Each antenna may be referred to as an antenna element or antenna element. For example, a plurality of antenna elements may be implemented as an antenna array (e.g., a sub-array) forming an array. For example, the antennas may be referred to as receiving antennas and / or transmitting antennas. As a non-limiting example, the distance between the antennas may correspond to a length of half the wavelength of a signal to be transmitted and received.

[0061] For example, the RF transceiver may include RF chains. For example, the RF chains may be referred to as RF paths. An RF path may be a unit of a path through which a signal received through an antenna or a signal radiated through an antenna passes. An RF chain may include a plurality of RF components. The RF components may include amplifiers, mixers, oscillators, DACs, ADCs, and the like.

[0062] Figure 3b shows an example of a resource structure in the time domain and frequency domain.

[0063] Figure 3b illustrates the basic structure of the time-frequency domain, which is a radio resource region in which data or control channels are transmitted in the downlink or uplink.

[0064] Referring to Figure 3b, the horizontal axis represents the time domain and the vertical axis represents the frequency domain. The minimum transmission unit in the time domain is an OFDM symbol, N symb OFDM symbols (352) are grouped to form one slot (356). The length of a subframe is defined as 1.0 ms, and the length of a radio frame (364) is defined as 10 ms. The minimum transmission unit in the frequency domain is a subcarrier, and the carrier bandwidth constituting the resource grid is N DL RB Dog (downlink) or N UL RB It consists of subcarriers (354) of the dog (uplink).

[0065] The basic unit of resources in the time-frequency domain is a resource element (RE) (362), which can be represented by an OFDM symbol index and a subcarrier index. A resource block may include multiple resource elements. In the LTE system, a resource block (RB) (or physical resource block (PRB)) is N in the time domain. symb N consecutive OFDM symbols and frequency domain SC RB is defined as a series of consecutive subcarriers. In an NR system, a resource block (RB) (358) is defined as N in the frequency domain. SC RB can be defined as a series of consecutive subcarriers (360). One RB (358) is N on the frequency axis. SC RB It contains REs (362). In general, the minimum transmission unit of data is RB and the number of subcarriers is N. SC RB=12. The frequency domain may include common resource blocks (CRBs). Physical resource blocks (PRBs) may be defined in the bandwidth part (BWP) of the frequency domain. The CRB and PRB numbers may be determined based on subcarrier spacing. The data rate may increase in proportion to the number of RBs scheduled to the terminal.

[0066] In the NR system, in the case of a frequency division duplex (FDD) system that operates the downlink and uplink by frequency division, the downlink transmission bandwidth and the uplink transmission bandwidth may be different. The channel bandwidth represents the radio frequency (RF) bandwidth corresponding to the system transmission bandwidth. [Table 1] shows part of the correspondence between the system transmission bandwidth, subcarrier spacing (SCS), and channel bandwidth defined in the NR system in a frequency band lower than x GHz (e.g., frequency range (FR) 1 (410 MHz to 7125 MHz)). And [Table 2] shows part of the correspondence between the transmission bandwidth, subcarrier spacing, and channel bandwidth defined in the NR system in a frequency band higher than y GHz (e.g., FR2 (24250 MHz - 52600 MHz) or FR2-2 (52600 MHz to 71000 MHz)). For example, an NR system with a 100 MHz channel bandwidth and a 30 kHz subcarrier spacing has a transmission bandwidth of 273 RBs. In [Table 1] and [Table 2], N / A may be a bandwidth-subcarrier combination not supported by the NR system.

[0067]

[0068]

[0069] The base station (110) (or the upper network node (210) and / or the lower network node (220) of FIG. 2) may support hybrid beamforming (or a hybrid beamforming system). The hybrid beamforming may include digital beamforming and analog beamforming. For example, the base station (110) may use the hybrid beamforming for transmitting and receiving signals. The base station (110) may perform communication with terminals within a communication environment (or channel environment, wireless communication environment, environment). For example, the base station (110) may perform communication with a terminal (120) within the communication environment through a channel. For example, the channel may represent signal change characteristics between the base station (110) and the terminal (120) within the communication environment. Since the terminals may transmit mutually orthogonal signals, channel estimation for each of the terminals may be performed independently. For example, the base station (110) can perform more effective communication with the terminal (120) by performing estimation for the channel. In one example, estimation (or channel estimation) for the channel can be performed using a reference signal.

[0070] The electronic device, method, and storage medium according to the present disclosure may utilize a neural network to perform estimation of the channel. For example, the neural network may be trained to estimate parameters for channel estimation using channel information (or a channel matrix). For example, the neural network may have weights (or a weight matrix) that can improve signal gain for a specific direction (or path) within a specific communication environment based on the training. For example, the neural network may be referred to as a model, an artificial intelligence model, an artificial neural network, a module, an engine, machine learning, an algorithm, or a function.

[0071] The electronic device, method, and storage medium according to the present disclosure can generate a beamformer for receiving a reference signal for channel estimation using the weights of the learned neural network. For example, the beamformer can be referred to as reception beamforming information, a reception matrix, a beamforming matrix, a combiner, or a reception beamformer. In the above example, a beamformer used to receive a reference signal is described, but the present disclosure is not limited thereto. As a non-limiting example, the beamformer can be referred to as transmission beamforming information, a transmission matrix, a precoder, a precoding matrix, or a transmission beamformer. For example, the electronic device, method, and storage medium according to the present disclosure can receive a reference signal by applying the beamformer generated using the weights to reception antennas, and use the received reference signal as an input of the neural network. The electronic device, method, and storage medium according to the present disclosure can obtain parameters for channel estimation (hereinafter, channel parameters) based on providing a received reference signal to the neural network by applying the beamformer generated using the weights to the receiving antennas.

[0072] As described above, the electronic device, method, and storage medium according to the present disclosure can perform more accurate channel estimation for a specific communication environment by performing channel estimation using the channel parameters. For example, the present disclosure can improve the accuracy of channel estimation by using a beamformer optimized for the specific communication environment and a neural network (or channel estimation function) for channel estimation. Accordingly, the signal gain (or beam gain, reception gain) within the specific communication environment can be improved. The electronic device, method, and storage medium according to the present disclosure can effectively perform communication with a terminal within the specific communication environment. Hereinafter, FIG. 4 illustrates a method by which a base station performs estimation of a channel between a terminal and a base station within a communication environment.

[0073] Figure 4 illustrates an example of a method for estimating a channel using beamforming information applied to receiving antennas of a base station.

[0074] The terminal (120) of FIG. 4 may be an example of the terminal (120) of FIG. 1. The base station (110) of FIG. 4 may be an example of the base station (110) of FIG. 1. In FIG. 4, an example (400) of a communication environment (407) related to the base station (110) and the terminal (120) is illustrated, but the present disclosure is not limited thereto. For example, the base station (110) may be implemented as a plurality of network nodes through functional separation. For example, the terminal (120) may provide a signal to a lower network node among the plurality of network nodes (e.g., a lower network node (220) of FIG. 2), and an upper network node among the plurality of network nodes (e.g., an upper network node (210) of FIG. 2) may obtain the provided signal.

[0075] FIG. 4 illustrates an example (400) in which a base station (110) communicates with a terminal (120) within a communication environment (407). In example (400), for convenience of explanation, an example in which the base station (110) communicates with one terminal (120) is illustrated; however, the present disclosure is not limited thereto. For example, the base station (110) may communicate with multiple terminals.

[0076] For example, the terminal (120) may transmit the transmission reference signal (420) to the base station (110) via the antenna. For example, the transmission reference signal (420) may be referred to as a signal transmitted (or, radiated, output) via the antenna of the terminal (120). In the example (400) of FIG. 4, for convenience of explanation, the terminal (120) is described as transmitting the transmission reference signal (420) using a single antenna, but the present disclosure is not limited thereto. For example, the terminal (120) may transmit the transmission reference signal (420) on a resource allocated by the base station (110). For example, the allocated resource may include a frequency resource having M subcarriers.

[0077] For example, the base station (110) can receive a reference signal from the terminal (120). The 'reference signal' mentioned in the present disclosure can be understood to include one or more reference signals. The reference signal can be referred to as a transmission reference signal or a reception reference signal. For example, the base station (110) can receive (or acquire) a reference signal from the terminal (120) using the reception antennas (410). In the example (400), for convenience of explanation, the reception antennas (410) include N antennas and can be implemented as a uniform linear array (ULA). However, the present disclosure is not limited thereto.

[0078] For example, the receiving antennas (410) may include a first receiving antenna (410-1), a second receiving antenna (410-2) to an Nth receiving antenna (410-N). For example, the receiving antennas (410) may include phase shifters (413) for analog beamforming and N RF The RF chains (or RF paths) (415) may be connected to a processor (430). For example, the processor (430) may include at least a portion of the processor (330) of FIG. 3A. For example, the processor (430) may include at least a portion of the transceiver (310) of FIG. 3A.

[0079] For example, the reference signal that the base station (110) receives from the terminal (120) through the receiving antennas (410) may represent a signal that has been changed by the channel between the base station (110) and the terminal (120) from the transmission reference signal (420). For example, the reference signal received through the receiving antennas (410) may be referred to as the reception reference signal (440). For example, the reception reference signal (440) that has been changed by the channel from the transmission reference signal (420) may be referred to as in the following mathematical equation.

[0080]

[0081]

[0082] Referring to mathematical expression 1, a reception reference signal (440) can be received within a specific time resource (or slot p). When the reception reference signal (440) is received during a time interval (e.g., P) including a plurality of slots, an extended reception reference signal (440) for the time resource can be referenced as in mathematical expression below.

[0083]

[0084]

[0085] Referring to mathematical expression 2, a reception reference signal (440) can be received within a time interval (P). When the reception reference signal (440) is received on a frequency interval (e.g., M subcarriers) including multiple frequency resources, an extended reception reference signal (440) for the frequency resources can be referenced as in the mathematical expression below.

[0086]

[0087]

[0088]

[0089] As illustrated in example (400), the transmission reference signal (420) may be reflected by external factors (409) (e.g., buildings, trees) in the communication environment (407) between the base station (110) and the terminal (120). The communication environment (407) between the base station (110) and the terminal (120) may be referred to as a far-field. For example, the transmission reference signal (420) may be transmitted to the base station (110) via the first path (421) (or the third path (423)) that is reflected by the external factor (409). Additionally, for example, the transmission reference signal (420) may be transmitted to the base station (110) via the second path (422) without being affected by the external factor (409). In the present disclosure, a path (e.g., a first path (421), a second path (422), a third path (423)) may represent a virtual path along which a signal is provided. For example, the path may be defined by a distance along which the signal travels (or a distance along the path) and an angle at which the signal is received (or a reception angle). For example, the path may be related to a space (or area) through which the signal is transmitted. In this case, the space may represent a location of an electronic device (e.g., a terminal (120)) transmitting (or providing) the signal.

[0090] The transmission reference signal (420) provided through the first path (421) or the third path (423) may have a relatively high path loss. Accordingly, the transmission reference signal (420) provided through the first path (421) or the third path (423) may not be transmitted to the base station (110). As in the above example, a channel having the characteristic that a signal provided from the terminal (120) to the base station (110) is provided through some of the multiple paths (e.g., the second path (422)) may be referred to as a sparse channel. For example, the channel may include or be related to multiple paths. The channel for the mth subcarrier may be referred to by the following mathematical equation.

[0091]

[0092]

[0093]

[0094]

[0095]

[0096] Referring to the above, in the case of a communication environment (407) as illustrated in example (400), an orthogonal matching pursuit (OMP) may be used to estimate (or restore) a channel having the characteristics of a sparse channel. For example, the OMP may include a simultaneous orthogonal matching pursuit (SOMP). For example, the OMP may be an algorithm used to solve a sparse recovery problem. The sparse recovery problem may represent a problem of estimating the array response and path gain for each path of the channel using a reception reference signal (440), when the array response is composed of a linear combination as illustrated in Equation 5 and candidates for possible array responses are known.

[0097]

[0098] In one example, a neural network may be used to estimate a channel. For example, channel estimation based on the neural network may utilize an equidirectional beamformer, as exemplified in the SOMP. A specific example of a neural network that estimates channel information using a reception reference signal (440) as input may be referenced in FIG. 5.

[0099] Figure 5 illustrates an example of a neural network that estimates channel information using a received reference signal.

[0100] Fig. 5 illustrates an example (500) of a neural network (510) that estimates channel information using a received reference signal as input. In the example (500) of Fig. 5, for convenience of explanation, it is assumed that the reference signal is received over a frequency range (e.g., M subcarriers).

[0101]

[0102]

[0103]

[0104]

[0105] As described above, when a beamformer capable of increasing signal gain for a path with good channel characteristics among the paths of a channel in a specific communication environment is used, a reference signal can be received with a high signal-to-noise ratio (SNR), and the accuracy of channel estimation can be increased. In other words, when a reference signal is received using an equidirectional beamformer, as in the example (500) of FIG. 5, the channel estimation accuracy may be lowered, unlike when a beamformer with directionality is used. In the following FIGS. 6A to 6C, an example of a method for performing channel estimation using a beamformer with directionality, which learns a neural network and generates weights of the learned neural network, according to the present disclosure, is described.

[0106] Figure 6a illustrates an example of an operational flow for how a base station performs channel estimation through a trained neural network.

[0107] At least some of the above methods of FIG. 6A may be performed by the base station (110) of FIG. 1. For example, the base station (110) may be an example of the electronic device (300) of FIG. 3A. However, the present disclosure is not limited thereto. For example, the base station (110) may be implemented as a plurality of network nodes through functional separation. For example, the terminal (120) may provide a signal to a lower network node among the plurality of network nodes (e.g., the lower network node (220) of FIG. 2), and an upper network node among the plurality of network nodes (e.g., the upper network node (210) of FIG. 2) may obtain the provided signal. For example, at least some of the above methods may be controlled by the processor (330) of the electronic device (300). In the following embodiments, each operation may be performed sequentially, but is not necessarily performed sequentially. For example, the order of each operation may be changed, and at least two operations may be performed in parallel.

[0108] In operation (600), the base station (110) may train a neural network to estimate channel parameters using channel information about channels for the receiving antennas. For example, the receiving antennas may be included in the base station (110). In one example, when the base station (110) is implemented with multiple network nodes, the receiving antennas may be included in a lower network node (e.g., lower network node (220) of FIG. 2).

[0109] For example, the channels for the receiving antennas may be formed in a communication environment (e.g., communication environment (407) of FIG. 4) in which the base station (110) is installed (or positioned). For example, the channels for the receiving antennas may be related to the space of the base station (110) (or the lower network node (220)). For example, each of the channels may include one or more paths. For example, a path of a channel may represent a path along which a signal is provided. The path of the channel may be referred to as a receiving path (or a transmitting path).

[0110] For example, the channel information for the channels may include channel information (H) (or channel matrix) of the channels through which reference signals received via the receiving antennas of the base station (110) are provided. The channel information (H) (or channel matrix) of the channels may be related to the communication environment in which the receiving antennas are used. The channel information for the channels may be data that has been previously acquired (or stored). For example, the channel information for the channels may be referenced as samples for training the neural network. In other words, the channel information for the channels may indicate signal change characteristics of a channel that are already known. For example, the channel information for the channels may include channel information that is estimated (or collected, acquired, or identified) before training of the neural network is performed.

[0111]

[0112] For example, the base station (110) may include the neural network. By utilizing the neural network, the base station (110) can simultaneously optimize (or jointly optimize) the neural network (or a function for channel estimation of the neural network) together with a beamformer. The learning of the neural network may be performed at the base station (110). For a specific example of the neural network according to the present disclosure, reference may be made to FIG. 6B.

[0113] Figure 6b shows an example of a neural network for estimating channel parameters.

[0114] FIG. 6B illustrates an example (631) of a method for training a neural network (655) according to the present disclosure. For example, the neural network (655) may have weights (or weight values, weight information, weight metrics) and biases (or bias values, bias information). As a non-limiting example, each of the weights and biases may be implemented as a tuple.

[0115]

[0116]

[0117]

[0118]

[0119]

[0120]

[0121]

[0122]

[0123]

[0124]

[0125]

[0126]

[0127]

[0128]

[0129]

[0130]

[0131]

[0132] According to example (631), the neural network (655) (or function) that has completed learning can be exemplified as the following mathematical formula.

[0133]

[0134]

[0135] Referring back to FIG. 6A, in operation (605), the base station (110) may identify channel parameters based on providing the received reference signal to the trained neural network. For example, the base station (110) may generate reception beamforming information using the weights of the trained neural network (e.g., the neural network (655) of FIG. 6B). For example, the following mathematical equation may be used as a reference for a method of generating the reception beamforming information.

[0136]

[0137]

[0138] Referring to mathematical expression 13, the reception beamforming information can be generated as a sum between the first sub-matrix and the second sub-matrix of the optimized weight of the neural network (655).

[0139]

[0140]

[0141]

[0142]

[0143]

[0144]

[0145]

[0146]

[0147]

[0148]

[0149]

[0150]

[0151] Referring to Equations 17 and 12, the function for estimating the neural network (655) may be substantially the same as the function for learning the neural network (655). For example, the function for estimating the neural network (655) may not include (or may exclude, omit) a preprocessing operation (or an additional layer for preprocessing) for learning the neural network (655).

[0152]

[0153]

[0154]

[0155]

[0156] Figure 6c illustrates an example of a method for adjusting channel parameters using a gradient descent algorithm.

[0157]

[0158]

[0159]

[0160]

[0161]

[0162]

[0163]

[0164]

[0165]

[0166]

[0167]

[0168]

[0169]

[0170]

[0171]

[0172]

[0173]

[0174]

[0175]

[0176]

[0177]

[0178]

[0179]

[0180]

[0181]

[0182]

[0183]

[0184]

[0185]

[0186]

[0187]

[0188]

[0189]

[0190]

[0191] To calculate Equations 23 and 24, the results of Equations 30 and 32 can be used.

[0192]

[0193] Referring to operation (610) of FIG. 6A and FIG. 6C, the base station (110) can perform adjustment (or refinement) of channel parameters. When performing adjustment (or refinement) of channel parameters, the base station (110) can perform more fine adjustment of channel parameters by utilizing distance parameters as well as angle parameters, thereby increasing the accuracy of channel estimation. In FIG. 6A, operation (610) is illustrated as being performed after operation (605), but the present disclosure is not limited thereto. For example, the adjustment (or refinement) of the channel parameters may be omitted.

[0194] Referring back to FIG. 6A, in operation (615), the base station (110) may generate channel information using channel parameters. For example, the channel parameters for generating channel information may include channel parameters identified according to operation (605) or channel parameters adjusted according to operation (610).

[0195] For example, the base station (110) can generate estimated channel information by applying the channel parameter to mathematical expression 4. Generating the channel information can be referred to as performing channel estimation.

[0196] In operation (620), the base station (110) may perform decoding on uplink signals based on channel information. For example, the base station (110) may perform decoding on the uplink signals including the reference signal based on the generated channel information. In the above example, the channel information is described as being used to perform decoding on the received (or acquired) uplink signals, but the present disclosure is not limited thereto. For example, when channel reciprocity is satisfied, the base station (110) may use the channel information to provide (or transmit) downlink signals to a terminal (e.g., terminal (120) of FIG. 1) that transmitted the reference signal.

[0197]

[0198]

[0199]

[0200]

[0201]

[0202]

[0203]

[0204]

[0205]

[0206] Referring to the above, the base station (110) may utilize a neural network (e.g., the neural network (655) of FIG. 6B) trained by further considering noise information based on Equations 33 to 36. For example, the base station (110) may estimate channel parameters (or channel information) based on the trained neural network by using a received reference signal.

[0207] In FIGS. 6A to 6C , the neural network (e.g., the neural network (655) of FIG. 6B ) is illustrated as being included in the base station (110), but the present disclosure is not limited thereto. For example, if the base station (110) is implemented with multiple network nodes, the neural network may be included in an upper network node (210) or a lower network node (220).

[0208] In the above example, the learning of the neural network can be performed in the upper network node (210). For example, the upper network node (210) can collect (or store, acquire) information for performing learning of the neural network (e.g., channel information (640) and channel parameters (660) of FIG. 6B), and perform learning of the neural network using the collected information.

[0209] In the above example, the estimation of the channel based on the neural network may be performed in the upper network node (210) and / or the lower network node (220). For example, the lower network node (220) may acquire the neural network trained in the upper network node (210) and estimate the channel parameter (or channel information) using the reference signal received using the acquired neural network. Alternatively, for example, the upper network node (210) may acquire the reference signal received through the lower network node (220) and estimate the channel parameter (or channel information) based on the neural network.

[0210] In FIGS. 6A to 6C, reception beamforming information applicable to the base station (110) is illustrated, but the present disclosure is not limited thereto. For example, when the base station (110) is implemented with a plurality of network nodes, the reception beamforming information may be applied to reception antennas of the lower network node (220) (e.g., reception antennas (410) of FIG. 4) (or, phase shifter (413) of FIG. 4). The upper network node (210) may obtain a reference signal received by the lower network node (220).

[0211] Hereinafter, FIGS. 7A to 8B illustrate improved performance compared to various channel estimation techniques by using a channel estimation technique according to the present disclosure. For example, the channel estimation technique according to the present disclosure may include a first estimation technique that utilizes estimated channel parameters based on providing a received reference signal to a neural network (e.g., the neural network (655) of FIG. 6B). Furthermore, for example, the channel estimation technique according to the present disclosure may include a second estimation technique that utilizes adjusted channel parameters generated by further adjusting the channel parameters estimated through the first estimation technique. Furthermore, for example, the channel estimation technique according to the present disclosure may include a third estimation technique that utilizes channel information directly estimated based on providing a received reference signal to a neural network (e.g., the neural network (655) of FIG. 6B).

[0212] In contrast, the various channel estimation techniques may include a fourth estimation technique that performs channel estimation from a reference signal based on the neural network (510) of FIG. 5. In addition, for example, the various channel estimation techniques may include a fifth estimation technique that performs LS (linear square) on the reference signal. In addition, for example, the various channel estimation techniques may include a sixth estimation technique based on SOMP using angular parameters. In addition, for example, the various channel estimation techniques may include a seventh estimation technique based on SIGW (simultaneous iterative gridless weighted) using adjusted values ​​of the OMP parameters, which are initial values.

[0213] In addition, the channel estimation technique according to the present disclosure can be compared with a computational method that utilizes theory (or mathematical formula) using the acquired information to verify the accuracy of the channel estimation. For example, the computational method may include a first computational method that utilizes LS according to the proposed beamformer (e.g., reception beamforming information using the weights of the neural network (655)). For example, the computational method may include a second computational method that utilizes LS (least square) according to an equidirectional beamformer.

[0214]

[0215] In the following FIGS. 7A to 7C, a case where the communication environment (e.g., the communication environment (409) of FIG. 4) is a first wireless environment is described, and in FIGS. 8A and 8B, a case where the communication environment is a second wireless environment is described. For example, the first wireless environment may represent an environment (or an uneven distribution environment) in which the gain of a transmitted signal is uneven depending on the path (or distance and / or angle). For example, the second wireless environment may represent an environment (or an even distribution environment) in which the gain of a transmitted signal is equal regardless of the path (or distance and / or angle).

[0216] In the following Figures 7a to 8b, for convenience of explanation, the base station (110) includes receiving antennas according to ULA, the number (N) of the receiving antennas is 32, and the number (N) of RF chains connected to the receiving antennas RF ) is 8. In addition, in the following FIGS. 7a to 8b, it is assumed that the number (M) of subcarriers, which are frequency resources allocated for transmission (or reception) of a reference signal, is 8, the number (P) of slots, which are allocated time resources, is 4, and the number (L) of paths through (or provided with) the reference signal is 2. In addition, in the following FIGS. 7a to 8b, the network type of the neural network (e.g., the neural network (655) of FIG. 6b) included in the base station (110) is a multi-layer perceptron (MLP), the number of hidden layers (e.g., K) of the neural network is 3, and the number (Q) of samples used for learning the neural network is 10. 5 , and it is assumed that the number of times the neural network has been trained (or epochs) is 500. The above examples are merely examples for the convenience of explanation, and the present disclosure is not limited to the base station (110) and neural network according to the above examples.

[0217] FIG. 7a illustrates an example of a beam pattern of received beamforming information within a first wireless environment.

[0218] FIG. 7A illustrates an example of a graph (700) representing beam gain according to angle and distance in the first wireless environment having an uneven distribution. For example, the horizontal axis of the graph (700) may represent an angle (e.g., -1 to 1), and the vertical axis of the graph (700) may represent a distance (unit: meters). An angle of -1 on the horizontal axis may be replaced with -180°, and an angle of 1 may be replaced with 180°.

[0219] For example, the signal may be transmitted in a first area (711) or a second area (712). For example, the first area (711) may represent a space defined by a distance (e.g., about 7 to 9.4 m) from a location where the signal is actually transmitted and an angle (e.g., a first range (701)) at which the signal is received. For example, the second area (712) may represent a space defined by a distance (e.g., about 18 to 20.4 m) from a location where the signal is actually transmitted and an angle (e.g., a second range (702)) at which the signal is received. For example, a neural network according to the present disclosure (e.g., a neural network (655) of FIG. 6B) may be trained using data (e.g., channel information or channel parameters) about signals transmitted in the first area (711) and the second area (722).

[0220] Graph (700) represents the beam pattern of a beamformer when generating a beamformer using the weights of a neural network according to the present disclosure. A brighter graph (700) may indicate a higher beam gain (e.g., 1). Conversely, a darker graph (700) may indicate a lower beam gain (e.g., 0).

[0221] Referring to the graph (700), when the angle at which a signal (e.g., a reference signal) is received has a value within a first range (701) or a second range (702), the beam gain may be relatively high. For example, the first range (701) may be defined as -0.3 to -0.1. For example, the second range (702) may be defined as 0.5 to 0.7. For example, the first range (701) and the second range (702) may exhibit a uniform distribution. Referring to the above, the beam gain may be independent of the distance and dependent on the angle.

[0222] As described above, when using a beamformer generated using the weights of a neural network according to the present disclosure, the accuracy of channel estimation for a path (or a channel including a path) having a specific angular range (e.g., a first range (701) and a second range (702)) can be relatively increased compared to when using an equidirectional beamformer. In other words, the beamformer according to the present disclosure can have a value effectively adapted to a first wireless environment.

[0223] Figure 7b shows an example of a graph representing NMSE (normalized mean square error) according to SNR (signal to noise ratio) in a first wireless environment.

[0224] FIG. 7b illustrates a graph (720) representing NMSE according to signal-to-noise ratio (SNR) to identify the accuracy of channel estimation. For example, the horizontal axis of the graph (720) may represent SNR (unit: decibel ([dB])), and the vertical axis of the graph (720) may represent NMSE.

[0225] The graph (720) includes a first line (721) representing the first estimation technique, a second line (723) representing the second estimation technique, a third line (725) representing the third estimation technique, a fourth line (727) representing the fourth estimation technique, a fifth line (729) representing the fifth estimation technique, a sixth line (731) representing the sixth estimation technique, a seventh line (733) representing the seventh estimation technique, an eighth line (735) representing the first operation method, and a ninth line (737) representing the second operation method.

[0226] Referring to graph (720), referring to the eighth line (735) representing the first operation method using the beamformer proposed in the present disclosure and the ninth line (737) representing the second operation method using the equidirectional beamformer, the NMSE values ​​according to the SNR of the eighth line (735) may have lower values ​​than the NMSE values ​​according to the SNR of the ninth line (737). In theory, using the beamformer according to the present disclosure may result in higher accuracy of channel estimation than using the equidirectional beamformer.

[0227] Also, referring to the graph (720), the first line (721), the second line (723), and the third line (725) according to the present disclosure may have NMSE values ​​according to a relatively low SNR, compared to the fourth line (727) representing a channel estimation technique using the neural network (510) of FIG. 5. In other words, the channel estimation technique using the neural network (e.g., the neural network (655) of FIG. 6B) according to the present disclosure may have higher accuracy of channel estimation than the existing channel estimation technique. In particular, the accuracy of channel estimation in the case of using the second estimation technique using the adjusted channel parameter, as shown in the second line (723), may have an accuracy similar to that of the ninth line (737), which represents a case where the channel is theoretically estimated while knowing the angle and distance at which the actual signal is transmitted.

[0228] In addition, referring to the graph (720), the first line (721), the second line (723), and the third line (725) according to the present disclosure may have NMSE values ​​according to relatively lower SNRs compared to the fifth line (729), the sixth line (731), and the seventh line (733) representing other channel estimation techniques that do not use a neural network. In other words, the channel estimation technique using a neural network according to the present disclosure may have higher channel estimation accuracy than the existing channel estimation technique in the first wireless environment.

[0229] Figure 7c shows an example of a graph showing SE (spectral efficiency) according to SNR in a first wireless environment.

[0230] Figure 7c illustrates a graph (740) representing SE (spectral efficiency) according to SNR (signal to noise ratio) when transmitting data through an estimated channel. For example, the horizontal axis of the graph (740) may represent SNR (unit: decibel), and the vertical axis of the graph (740) may represent SE (unit: bps (bit per second)).

[0231] The graph (740) includes a first line (741) representing the first estimation technique, a second line (743) representing the second estimation technique, a third line (745) representing the third estimation technique, a fourth line (747) representing the fifth estimation technique, a fifth line (749) representing the sixth estimation technique, and a sixth line (751) representing the seventh estimation technique.

[0232] Referring to the graph (740), the first line (741), the second line (743), and the third line (745) according to the present disclosure may have SE values ​​according to a relatively high SNR compared to the fourth line (747), the fifth line (749), and the sixth line (751) representing other channel estimation techniques that do not use a neural network. In other words, the channel estimation technique using a neural network according to the present disclosure may have a higher data transmission rate than the existing channel estimation techniques. More specifically, the lower the SNR, the greater the difference in values ​​between the first line (741), the second line (743), and the third line (745) and the fourth line (747), the fifth line (749), and the sixth line (751). That is, the lower the SNR, the greater the difference in data transmission rate.

[0233] Figure 8a shows an example of a graph representing NMSE according to SNR in a second wireless environment.

[0234] FIG. 8A illustrates a graph (800) representing NMSE according to signal-to-noise ratio (SNR) for identifying the accuracy of channel estimation in the second wireless environment having a uniform distribution. For example, the horizontal axis of the graph (800) may represent SNR (unit: decibel ([dB])), and the vertical axis of the graph (800) may represent NMSE.

[0235] The graph (800) includes a first line (801) representing the first estimation technique, a second line (803) representing the second estimation technique, a third line (805) representing the third estimation technique, a fourth line (807) representing the fourth estimation technique, a fifth line (809) representing the fifth estimation technique, a sixth line (811) representing the sixth estimation technique, and a seventh line (813) representing the seventh estimation technique.

[0236] Referring to graph (800), compared to graph (720) of FIG. 7B, the fourth line (807), fifth line (809), sixth line (811), and seventh line (813) according to existing channel estimation techniques may have NMSE values ​​according to substantially the same SNR. In other words, existing channel estimation techniques may not cause the NMSE value according to SNR to change even if the wireless communication environment changes.

[0237] Referring to graph (800), the second line (803) and the third line (805) according to the present disclosure may have lower NMSE values ​​as the SNR increases, compared to the fourth line (807) representing a general channel estimation technique using a neural network. In other words, the channel estimation technique using a neural network according to the present disclosure may have higher channel estimation accuracy than the existing channel estimation technique as the SNR increases in the second wireless environment.

[0238] Figure 8b shows an example of a graph representing SE according to SNR in a second wireless environment.

[0239] Figure 8b illustrates a graph (820) representing SE (spectral efficiency) according to SNR (signal to noise ratio) when transmitting data through an estimated channel. For example, the horizontal axis of the graph (820) may represent SNR (unit: decibel), and the vertical axis of the graph (820) may represent SE (unit: bps (bit per second)).

[0240] The graph (820) includes a first line (821) representing the first estimation technique, a second line (823) representing the second estimation technique, and a third line (825) representing the third estimation technique.

[0241] Referring to graph (820), the first line (821), the second line (823), and the third line (825) according to the present disclosure may have higher SE values ​​as the SNR increases, similar to graph (740) of FIG. 7C. For example, the value of the third line (825) may have a higher SE value than the SE values ​​of the first line (821) and the second line (823). Accordingly, in the second wireless environment, the third estimation technique that directly estimates channel information using a neural network may be used to secure a higher data transmission rate.

[0242] Figure 9 illustrates an example of an operational flow for a method of performing channel estimation using a received reference signal by applying received beamforming information generated using the weights of a learned neural network to receiving antennas.

[0243] At least some of the methods of FIG. 9 may be performed by the base station (110) of FIG. 1. For example, the base station (110) may be an example of the electronic device (300) of FIG. 3A. However, the present disclosure is not limited thereto. For example, the base station (110) may be implemented as a plurality of network nodes through functional separation. For example, the terminal (120) may provide a signal to a lower network node among the plurality of network nodes (e.g., the lower network node (220) of FIG. 2), and an upper network node among the plurality of network nodes (e.g., the upper network node (210) of FIG. 2) may obtain the provided signal. For example, at least some of the methods may be controlled by the processor (330) of the electronic device (300). In the following embodiments, each operation may be performed sequentially, but is not necessarily performed sequentially. For example, the order of each operation may be changed, and at least two operations may be performed in parallel.

[0244] In operation (910), the base station (110) may identify weights of a neural network for estimating distance parameters and angle parameters for paths of a channel using channel information associated with the receiving antennas. For example, the receiving antennas may be included in the base station (110). In one example, when the base station (110) is implemented with multiple network nodes, the receiving antennas may be included in a lower network node (e.g., lower network node (220) of FIG. 2).

[0245] For example, the channel associated with the receiving antennas may be formed in a communication environment (e.g., communication environment (407) of FIG. 4) in which the base station (110) is installed (or positioned). For example, the channels for the receiving antennas may be related to the space of the base station (110) (or the lower network node (220)). For example, the channel may include one or more paths. For example, the path of the channel may represent the path along which a signal is provided. The path of the channel may be referred to as a receiving path (or a transmitting path).

[0246] For example, the neural network may be in a learned state. For example, the neural network may be in a learned state using channel information about the channels for the receiving antennas, as described in operation (600) of FIG. 6A and example (631) of FIG. 6B. For specific details related thereto, reference may be made to FIGS. 6A through 6C.

[0247] For example, the weights of the neural network may represent the weights of the learned neural network. For example, the weights of the neural network may include weights adjusted based on mathematical expression 7.

[0248] In operation (920), the base station (110) can obtain uplink signals including received reference signals by applying the received beamforming information generated using the weights to the receiving antennas.

[0249] For example, the base station (110) can generate the reception beamforming information using the weights of the learned neural network. For example, the method for generating the reception beamforming information may refer to the contents related to mathematical expression 13. For example, the reception beamforming information may be generated as the sum between the first sub-matrix and the second sub-matrix of the weights. For example, the first sub-matrix may include diagonal elements. For example, the second sub-matrix may include off-diagonal elements.

[0250] For example, the base station (110) can obtain the uplink signals including the reference signals by applying the generated reception beamforming information to the reception antennas. For example, the base station (110) can receive the uplink signals by applying the reception beamforming information to the reception antennas of the base station (110) (e.g., the reception antennas (410) of FIG. 4) (or the phase shifter (413) of FIG. 4).

[0251] In operation (930), the base station (110) may identify distance parameters and angle parameters for paths of receive channels associated with the receive antennas based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network. For example, the base station (110) may estimate distance parameters and angle parameters for paths of the receive channel associated with the receive antennas by using the received reference signals as inputs to the neural network. For example, the receive channel may include a channel through which the reference signals are provided. For example, the receive channel may represent a channel between the base station (110) and a terminal (e.g., terminal (120) of FIG. 1) that provided the reference signals.

[0252] For example, the base station (110) can estimate channel parameters based on the neural network using the received reference signals. For specific details related to this, reference may be made to operation (605) of FIG. 6A and example (632) of FIG. 6B.

[0253] Although not illustrated in FIG. 9, the base station (110) may adjust the identified channel parameters. For example, the base station (110) may adjust the estimated channel parameters based on a neural network. For specific details related thereto, reference may be made to operation (610) of FIG. 6A and FIG. 6C.

[0254] In operation (940), the base station (110) can perform decoding on the uplink signals based on channel information for the reception channel estimated using the distance parameter and the angle parameter for the paths of the reception channel.

[0255] For example, the base station (110) may generate channel information using channel parameters. For example, the channel parameters for generating channel information may include channel parameters identified according to operation (930). However, the present disclosure is not limited thereto. For example, the channel parameters for generating channel information may include channel parameters adjusted according to FIG. 6c.

[0256] For example, the base station (110) can generate estimated channel information by applying the channel parameter to mathematical expression 4. Generating the channel information can be referred to as performing channel estimation.

[0257] For example, the base station (110) can perform decoding on the uplink signals based on the estimated channel information. For example, the base station (110) can perform decoding on the uplink signals including the reference signal based on the generated (or estimated) channel information. In the above example, the channel information is described as being used to perform decoding on the received (or acquired) uplink signals, but the present disclosure is not limited thereto. For example, when channel reciprocity is satisfied, the base station (110) can use the channel information to provide (or transmit) downlink signals to a terminal (e.g., terminal 120 of FIG. 1) that transmitted the reference signal.

[0258] Additionally, although not described in FIG. 9, the neural network may be trained by further utilizing noise information. For specific details regarding this, reference may be made to the descriptions related to Equations 33 to 36. In other words, the base station (110) may utilize a neural network trained by further considering noise information based on Equations 33 to 36. For example, the base station (110) may estimate channel parameters (or channel information) based on the trained neural network by utilizing a received reference signal.

[0259] In one example, the base station (110) may directly estimate channel information using reference signals. In other words, the base station (110) may directly estimate channel information rather than estimating channel parameters using reference signals.

[0260] An electronic device, a method, and a storage medium according to the present disclosure can learn a neural network for estimating channel parameters using channel information, which is data about a specific wireless environment. For example, the electronic device, the method, and the storage medium according to the present disclosure can generate a beamformer having directionality using the weights of the learned neural network. The electronic device, the method, and the storage medium according to the present disclosure can use the beamformer that increases the signal gain for a path with good channel characteristics among the paths of the channels in a specific communication environment. The electronic device, the method, and the storage medium according to the present disclosure can receive a reference signal with a high signal-to-noise ratio (SNR), and the accuracy of channel estimation can be increased. In other words, the electronic device, the method, and the storage medium according to the present disclosure can increase the accuracy of channel estimation by receiving a signal using a directional beamformer, compared to a case where a reference signal is received using an equidirectional beamformer. In addition, the electronic device, method, and storage medium according to the present disclosure can improve the accuracy of channel estimation by performing adjustments to channel parameters output through a neural network using not only angle parameters but also distance parameters.

[0261] 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.

[0262] As described above, a base station may include at least one processor including a processing circuit. The base station may include a memory storing instructions and including one or more storage media. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify weights of a neural network for estimating distance parameters and angle parameters for paths of a channel using channel information associated with receive antennas. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to obtain uplink signals including received reference signals by applying receive beamforming information generated using the weights to the receive antennas. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify distance parameters and angle parameters for paths of a receive channel associated with the receive antennas based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to perform decoding of the uplink signals based on channel information of the receive channel estimated using the distance parameters and the angle parameters for the paths of the receive channel.

[0263] According to one embodiment, the reception beamforming information may be used to increase the reception gain of a path through which the reference signals are provided among the paths of the reception channel associated with the reception antennas.

[0264] According to one embodiment, the weights may be determined such that an average of difference values ​​for channels associated with the receiving antennas, including the channel associated with the receiving antennas, has a minimum value. Each of the difference values ​​may include an error between an estimated distance parameter and an estimated angle parameter for paths of each of the channels, and a distance parameter and an angle parameter for the paths of each of the channels calculated from channel information of each of the channels.

[0265] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify, based on the neural network, the estimated distance parameter and the estimated angle parameter for the paths of the channel, using the channel information of the channel, the initial weight of the neural network, and the initial bias of the neural network, before identifying the weights, to train the neural network. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify, before identifying the weights, a difference value between the estimated distance parameter and the estimated angle parameter for the paths of the channel, and the distance parameter and the angle parameter for the paths of the channel calculated from the channel information of the channel, to train the neural network. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to apply the weights and biases to the neural network, the weights and biases being determined based on the difference values, including the difference values, to train the neural network before identifying the weights.

[0266] According to one embodiment, each of the initial weight and the initial bias can be set to an arbitrary value.

[0267] In one embodiment, the neural network may include a plurality of layers. For the input layer among the plurality of layers, the product between channel information and the weights of the neural network and the sum of the bias of the neural network may be used as input. For the output layer among the plurality of layers, the estimated distance parameter and the estimated angle parameter may be output.

[0268] According to one embodiment, for the input layer among the plurality of layers, the product between noise information and the weights of the neural network may be further used as input.

[0269] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify a first sub-matrix and a second sub-matrix of the symmetric matrix from the weights, the weights comprising a symmetric matrix having a size according to the number of the receiving antennas and the number of receiving chains connected to the receiving antennas. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to generate the receiving beamforming information comprising a matrix formed by the sum of the first sub-matrix and the second sub-matrix.

[0270] According to one embodiment, the first sub-matrix may be included in a diagonal block matrix of the symmetric matrix. The second sub-matrix may be included in a non-diagonal block matrix of the symmetric matrix. The size of each of the first sub-matrix and the second sub-matrix may be 1 / 4 of the size of the symmetric matrix.

[0271] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify an adjusted angular parameter for the paths of the receive channel by performing a gradient descent algorithm for the angular parameter for the paths of the receive channel a specified number of iterations, using the distance parameter for the paths of the receive channel. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify an adjusted distance parameter for the paths of the receive channel by performing a gradient descent algorithm for the distance parameter for the paths of the receive channel a specified number of iterations, using the adjusted angular parameter for the paths of the receive channel.

[0272] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify path gains for the paths of the receive channel using the adjusted distance parameters for the paths of the receive channel and the adjusted angle parameters for the paths of the receive channel. The instructions, when individually or collectively executed by the at least one processor, may cause the base station to identify channel information for the receive channel using the adjusted distance parameters for the paths of the receive channel, the adjusted angle parameters for the paths of the receive channel, and the path gains for the paths of the receive channel.

[0273] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the base station to generate the channel information for the receive channel based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network.

[0274] According to one embodiment, the reference signals received by applying the reception beamforming information to the reception antennas may be defined based on the reception beamforming information and the estimated channel information for the reception channel. The reference signals received by applying the reception beamforming information to the reception antennas may be used as inputs of an input layer among a plurality of layers of the neural network to estimate the distance parameter and the angle parameter for the paths of the reception channel.

[0275] As described above, a lower network node may include an RF transceiver including receive antennas. The lower network node may include at least one processor including a processing circuit. The lower network node may include a memory storing instructions and including one or more storage media. The instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to obtain, from a higher network node, a neural network for estimating distance parameters and angle parameters for paths of a channel associated with the receive antennas using channel information of the channel. The instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to receive, from a terminal, uplink signals including reference signals by applying receive beamforming information generated using weights of the neural network to the receive antennas. The instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to identify distance parameters and angle parameters for paths of receive channels associated with the receive antennas between the lower network node and the terminal based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network. The instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to provide channel information for the receive channel estimated using the distance parameters and the angle parameters for the paths of the receive channel to the upper network node for decoding of the uplink signals.

[0276] According to one embodiment, the reception beamforming information may be used to increase the reception gain of a path through which the reference signals are provided among the paths of the reception channel associated with the reception antennas.

[0277] According to one embodiment, the weights may be determined such that an average of difference values ​​for channels associated with the receiving antennas, including the channel associated with the receiving antennas, has a minimum value. Each of the difference values ​​may include an error between an estimated distance parameter and an estimated angle parameter for paths of each of the channels, and a distance parameter and an angle parameter for the paths of each of the channels calculated from channel information of each of the channels.

[0278] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to identify a first submatrix and a second submatrix of the symmetric matrix from the weights, the weights being a symmetric matrix having a size according to the number of the receiving antennas and the number of receiving chains of the RF transceiver connected to the receiving antennas. The instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to generate the receiving beamforming information, the receiving beamforming information being a matrix formed by the sum of the first submatrix and the second submatrix.

[0279] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to identify an adjusted angular parameter for the paths of the receive channel by performing a gradient descent algorithm for the angular parameter for the paths of the receive channel a specified number of iterations using the distance parameter for the paths of the receive channel. The instructions, when individually or collectively executed by the at least one processor, may cause the lower network node to identify an adjusted distance parameter for the paths of the receive channel by performing a gradient descent algorithm for the distance parameter for the paths of the receive channel a specified number of iterations using the adjusted angular parameter for the paths of the receive channel.

[0280] The method performed by the base station as described above may include an operation of identifying weights of a neural network for estimating distance parameters and angle parameters for paths of a channel using channel information associated with receive antennas. The method may include an operation of obtaining uplink signals including reference signals received by applying receive beamforming information generated using the weights to the receive antennas. The method may include an operation of identifying distance parameters and angle parameters for paths of a receive channel associated with the receive antennas based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network. The method may include an operation of performing decoding on the uplink signals based on channel information for the receive channel estimated using the distance parameters and the angle parameters for the paths of the receive channel.

[0281] The non-transitory computer-readable storage medium as described above may store one or more programs including instructions that, when individually or collectively executed by at least one processor of a base station, cause the base station to identify weights of a neural network for estimating distance parameters and angle parameters for paths of a channel using channel information associated with receive antennas. The non-transitory computer-readable storage medium may store one or more programs including instructions that, when individually or collectively executed by the at least one processor, cause the base station to obtain uplink signals including received reference signals by applying receive beamforming information generated using the weights to the receive antennas. The non-transitory computer-readable storage medium may store one or more programs comprising instructions that, when individually or collectively executed by the at least one processor, cause the base station to identify distance parameters and angle parameters for paths of receive channels associated with the receive antennas based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network. The non-transitory computer-readable storage medium may store one or more programs comprising instructions that, when individually or collectively executed by the at least one processor, cause the base station to perform decoding of the uplink signals based on channel information for the receive channel estimated using the distance parameters and the angle parameters for the paths of the receive channel.

[0282] A method performed by a lower network node including receive antennas as described above may include obtaining, from an upper network node, a neural network for estimating distance parameters and angle parameters for paths of a channel associated with the receive antennas using channel information of the channel. The method may include receiving, from a terminal, uplink signals including reference signals by applying receive beamforming information generated using weights of the neural network to the receive antennas. The method may include identifying distance parameters and angle parameters for paths of a receive channel associated with the receive antennas between the network node and the terminal based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network. The method may include providing, to the upper network node, channel information for the receive channel estimated using the distance parameters and the angle parameters for the paths of the receive channel for decoding the uplink signals.

[0283] The non-transitory computer-readable storage medium as described above may store one or more programs including instructions that, when individually or collectively executed by at least one processor of a lower-level network node including an RF transceiver including receive antennas, cause the lower-level network node to obtain, from an upper-level network node, a neural network for estimating distance parameters and angle parameters for paths of a channel using channel information of the channel associated with the receive antennas. The non-transitory computer-readable storage medium may store one or more programs including instructions that, when individually or collectively executed by the at least one processor, cause the lower-level network node to receive, from a terminal, uplink signals including reference signals by applying receive beamforming information generated using weights of the neural network to the receive antennas. The non-transitory computer-readable storage medium may store one or more programs including instructions that, when individually or collectively executed by the at least one processor, cause the lower network node to identify distance parameters and angular parameters for paths of receive channels associated with the receive antennas between the lower network node and the terminal based on providing the reference signals received by applying the receive beamforming information to the receive antennas to the neural network.The non-transitory computer-readable storage medium, when executed individually or collectively by the at least one processor, may cause the lower network node to provide, to the upper network node, channel information for the reception channel estimated using the distance parameter and the angle parameter for the paths of the reception channel for decoding the uplink signals.

[0284] 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.

[0285] 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 to be executed by one or more processors in 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. The one or more programs may be provided as a computer program product. The computer program product may be traded between a seller and a buyer as a commodity. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., compact disc read only memory (CD-ROM)), or may be distributed online (e.g., downloaded or uploaded) via an application store (e.g., Play Store™) or directly between two user devices (e.g., smart phones). In the case of online distribution, at least a portion of the computer program product may be temporarily stored or temporarily created in a device-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or an intermediary server.

[0286] 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.

[0287] 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.

[0288] 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.

[0289] According to embodiments, one or more of the components or operations of the aforementioned components may be omitted, or one or more other components or operations may be added. Alternatively or additionally, a plurality of components (e.g., modules or programs) may be integrated into a single component. In such a case, the integrated component may perform one or more functions of each of the plurality of components identically or similarly to those performed by the corresponding component among the plurality of components prior to the integration. According to embodiments, the operations performed by a module, program, or other component may be executed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be executed in a different order, omitted, or one or more other operations may be added.

[0290] Meanwhile, although the detailed description of the present disclosure has described specific embodiments, it is obvious that various modifications are possible within the scope of the present disclosure.

Claims

1. At the base station, At least one processor comprising a processing circuit; and A memory for storing instructions and including one or more storage media, The above instructions, when individually or collectively executed by the at least one processor, cause the base station to: Identifying weights of a neural network for estimating distance parameters and angle parameters for paths of the channel using channel information associated with the receiving antennas; By applying the received beamforming information generated using the above weights to the receiving antennas, uplink signals including the received reference signals are obtained; Identifying distance parameters and angle parameters for paths of reception channels associated with the reception antennas based on providing the reference signals received by applying the reception beamforming information to the reception antennas to the neural network; and Causing decoding of the uplink signals based on channel information of the receiving channel estimated using the distance parameter and the angle parameter for the paths of the receiving channel. Base station.

2. In claim 1, The above receiving beamforming information is used to increase the receiving gain of a path through which the reference signals are provided among the paths of the receiving channel associated with the receiving antennas. Base station.

3. In claim 1, The weights are determined such that the average value of the difference values ​​for the channels associated with the receiving antennas, including the channels associated with the receiving antennas, has a minimum value, and Each of the above difference values ​​includes an error between the estimated distance parameter and the estimated angle parameter for the paths of each of the channels and the distance parameter and the angle parameter for the paths of each of the channels calculated from the channel information of each of the channels. Base station.

4. In claim 3, The above instructions, when individually or collectively executed by the at least one processor, cause the base station to: Before identifying the above weights, to train the neural network: Using the channel information of the channel, the initial weight of the neural network, and the initial bias of the neural network, identifying the estimated distance parameter and the estimated angle parameter for the paths of the channel based on the neural network; Identifying the difference values ​​between the estimated distance parameters and the estimated angle parameters for the paths of the channel and the distance parameters and the angle parameters for the paths of the channel calculated from the channel information of the channel; and Causing the neural network to apply the weights and biases, which are determined based on the difference values ​​including the difference values, Base station.

5. In claim 4, Each of the above initial weights and the above initial biases are set to arbitrary values. Base station.

6. In claim 4, The above neural network includes multiple layers, For the input layer among the plurality of layers, the product between the channel information and the weight of the neural network and the sum of the bias of the neural network are used as input, and Among the above multiple layers, the output layer outputs the estimated distance parameter and the estimated angle parameter. Base station.

7. In claim 6, Among the plurality of layers, for the input layer, the product between noise information and the weight of the neural network is further used as input. Base station.

8. In claim 1, The above instructions, when individually or collectively executed by the at least one processor, cause the base station to: Identifying a first submatrix and a second submatrix of the symmetric matrix from the weights, which are composed of a symmetric matrix having a size according to the number of the receiving antennas and the number of receiving chains connected to the receiving antennas; and Causing the reception beamforming information to be generated as a matrix formed by the sum of the first sub-matrix and the second sub-matrix, Base station.

9. In claim 8, The above first submatrix is ​​included in the diagonal block matrix of the above symmetric matrix, The second submatrix is ​​included in the off-diagonal block matrix of the symmetric matrix, and The size of each of the first submatrix and the second submatrix is ​​1 / 4 of the size of the symmetric matrix, Base station.

10. In claim 1, The above instructions, when individually or collectively executed by the at least one processor, cause the base station to: Identifying an adjusted angle parameter for the paths of the receiving channel by repeatedly performing a gradient descent algorithm for the angle parameter for the paths of the receiving channel a specified number of times using the distance parameter for the paths of the receiving channel; and By repeatedly performing a gradient descent algorithm for the distance parameter for the paths of the receiving channel a specified number of times using the adjusted angle parameter for the paths of the receiving channel, thereby causing the adjusted distance parameter for the paths of the receiving channel to be identified, Base station.

11. In claim 10, The above instructions, when individually or collectively executed by the at least one processor, cause the base station to: Identifying path gains for the paths of the receiving channel using the adjusted distance parameters for the paths of the receiving channel and the adjusted angle parameters for the paths of the receiving channel; and By using the adjusted distance parameter for the paths of the receiving channel, the adjusted angle parameter for the paths of the receiving channel, and the path gain for the paths of the receiving channel, causing the channel information to be identified in the receiving channel, Base station.

12. In claim 1, The above instructions, when individually or collectively executed by the at least one processor, cause the base station to: By applying the received beamforming information to the receiving antennas, and providing the reference signals received to the neural network, the channel information for the receiving channel is generated. Base station.

13. In claim 1, The reference signals received by applying the above reception beamforming information to the above reception antennas are defined based on the reception beamforming information and the estimated channel information for the reception channel, and The reference signals received by applying the reception beamforming information to the reception antennas are used as inputs of an input layer among a plurality of layers of the neural network to estimate the distance parameter and the angle parameter for the paths of the reception channel. Base station.

14. In a method performed by a base station, An operation of identifying weights of a neural network for estimating distance parameters and angle parameters for paths of a channel using channel information associated with receiving antennas; An operation of obtaining uplink signals including received reference signals by applying received beamforming information generated using the above weights to the above receiving antennas; An operation of identifying distance parameters and angle parameters for paths of reception channels associated with the reception antennas based on providing the reference signals received by applying the reception beamforming information to the reception antennas to the neural network; and An operation of performing decoding on the uplink signals based on channel information for the reception channel estimated using the distance parameter and the angle parameter for the paths of the reception channel, method.

15. In a non-transitory computer-readable storage medium, when individually or collectively executed by at least one processor of a base station, the base station: Identifying weights of a neural network for estimating distance parameters and angle parameters for paths of the channel using channel information associated with the receiving antennas; By applying the received beamforming information generated using the above weights to the receiving antennas, uplink signals including the received reference signals are obtained; Identifying distance parameters and angle parameters for paths of reception channels associated with the reception antennas based on providing the reference signals received by applying the reception beamforming information to the reception antennas to the neural network; and Storing one or more programs including instructions that cause decoding of the uplink signals based on channel information for the receiving channel estimated using the distance parameter and the angle parameter for the paths of the receiving channel. Non-transitory computer-readable storage medium.

Images