Training-based interference whitening method and device in multiple-input multiple-output receiver

Abstract

According to an embodiment, a method of an electronic device in a wireless communication system may comprise the operations of: calculating a noise covariance matrix on the basis of a received signal and an estimated channel; obtaining a sparse precision matrix from the noise covariance matrix by using an iterative algorithm or a neural network; deriving a plurality of indices for indicating non-zero elements in a matrix by applying a compressed sparse row (CSR) to the sparse precision matrix; and obtaining filtered channel information by performing pre-filtering on the basis of the plurality of indices and the estimated channel.

Description

Learning-based interference whitening method and device in a multiple input / output receiver

[0001] The present disclosure relates to an apparatus and a method thereof that support a learning-based interference whitening technique in a large-scale multiple input / output receiver.

[0002] In communication systems, the technology of using multiple antennas in transmitters or receivers to improve transmission quality can be referred to as multi-antenna technology. Improving transmission quality means increasing the maximum transmission speed per user, the overall capacity of a cell, or expanding cell coverage. The technology of using multiple antennas in transmitters and receivers can be referred to as Multiple Input Multiple Output (MIMO).

[0003] In wireless communication systems utilizing MIMO technology, multiple antennas can be used at both the transmitter and receiver. The channel capacity of wireless communication systems utilizing MIMO technology can be significantly improved compared to single-antenna technology. Both the base station and the terminal supporting MIMO use multiple antennas, and the channel capacity can be increased in proportion to the number of antennas used. For example, if the base station uses M antennas and the terminal uses N antennas, the average transmission capacity can increase by min(M, N).

[0004] The information described above may be provided as related art for the purpose of aiding understanding of the present disclosure. No claim or determination is made as to whether any of the foregoing may be applied as prior art related to the present disclosure.

[0005] The present disclosure relates to an invention concerning a receiving method for a device having multiple transmitting antennas, and proposes an algorithm and structure for improving the complexity and performance of interference whitening.

[0006] According to one embodiment, a method of an electronic device in a wireless communication system may include: calculating a noise covariance matrix based on a received signal and an estimated channel; obtaining a sparse precision matrix from the noise covariance matrix using an iterative algorithm or a neural network; applying a compressed sparse row (CSR) to the sparse precision matrix to derive a plurality of indices for indicating non-zero elements within the matrix; and performing pre-filtering based on the plurality of indices and the estimated channel to obtain filtered channel information.

[0007] According to one embodiment, an electronic device in a wireless communication system may include a transceiver; a memory; and at least one processor. The at least one processor may enable the electronic device to perform a plurality of operations by executing instructions stored in the memory. The plurality of operations may include: an operation of calculating a noise covariance matrix based on a received signal and an estimated channel; an operation of obtaining a sparse precision matrix from the noise covariance matrix using an iterative algorithm or a neural network; an operation of deriving a plurality of indices to indicate non-zero elements in the matrix by applying a compressed sparse row (CSR) to the sparse precision matrix; and an operation of obtaining filtered channel information by performing pre-filtering based on the plurality of indices and the estimated channel.

[0008] According to one embodiment, a storage medium storing at least one instruction readable by a computer may be implemented. When the at least one instruction is executed by at least one processor, the electronic device may perform a plurality of operations. The plurality of operations may include: an operation of calculating a noise covariance matrix based on a received signal and an estimated channel; an operation of obtaining a sparse precision matrix from the noise covariance matrix using an iterative algorithm or a neural network; an operation of deriving a plurality of indices to indicate non-zero elements in the matrix by applying a compressed sparse row (CSR) to the sparse precision matrix; and an operation of obtaining filtered channel information by performing pre-filtering based on the plurality of indices and the estimated channel.

[0009] A method and apparatus according to one embodiment of the present disclosure can reduce receiver complexity and improve reception performance by utilizing the inverse matrix of noise and interference covariance.

[0010] In relation to the description of the drawings, the same or similar reference numerals may be used for identical or similar components.

[0011] FIG. 1 is a block diagram of an electronic device in a network environment according to one embodiment of the present disclosure.

[0012] FIG. 2 is a diagram showing an example of a DL MU-MIMO environment in a wireless communication system according to one embodiment of the present disclosure.

[0013] FIG. 3 is a diagram showing an example of a DL SU-MIMO environment in a wireless communication system according to one embodiment of the present disclosure.

[0014] FIG. 4 is a drawing illustrating the structure of a wireless communication system according to one embodiment of the present disclosure.

[0015] FIG. 5 is a drawing illustrating the structure of a receiver according to one embodiment of the present disclosure.

[0016] FIG. 6 illustrates a process for designing a set of learnable parameters for unfolding according to one embodiment of the present disclosure.

[0017] FIG. 7 is a drawing for explaining an unfolding method according to one embodiment of the present disclosure.

[0018] FIG. 8 is a block diagram illustrating a learning-based interference whitening method according to one embodiment of the present disclosure.

[0019] FIG. 9 is a diagram illustrating a specific process of a learning-based interference whitening method according to one embodiment of the present disclosure.

[0020] FIG. 10 is a diagram illustrating an unrolling-based sparse precision matrix operation according to one embodiment of the present disclosure.

[0021] FIG. 11 is a diagram illustrating an Unfolded D-clime algorithm according to one embodiment of the present disclosure.

[0022] FIG. 12 is a diagram illustrating Compressed Sparse Row (CSR) and Sparse Precision Matrix Multiplication (SpMM) according to one embodiment of the present disclosure.

[0023] FIG. 13 shows a memory processing flowchart in an electronic device according to one embodiment of the present disclosure.

[0024] Hereinafter, embodiments of the present disclosure are described in detail with reference to the drawings so that those skilled in the art can easily practice them. However, the present disclosure may be embodied in various different forms and is not limited to the embodiments described herein. In relation to the description of the drawings, the same or similar reference numerals may be used for identical or similar components. Furthermore, in the drawings and related descriptions, descriptions of well-known functions and configurations may be omitted for clarity and brevity.

[0025] The present disclosure describes embodiments using terms used in some communication standards (e.g., 3GPP (3rd Generation Partnership Project)), but this is merely illustrative. Various embodiments of the present disclosure can be easily modified and applied to other communication systems.

[0026] FIG. 1 is a block diagram of an electronic device (101) in a network environment (100) according to one embodiment.

[0027] Referring to FIG. 1, in a network environment (100), an electronic device (101) may communicate with an electronic device (102) through a first network (198) (e.g., a short-range wireless communication network) or with at least one of an electronic device (104) or a server (108) through a second network (199) (e.g., a long-range wireless communication network). According to one embodiment, the electronic device (101) may communicate with the electronic device (104) through a server (108). According to one embodiment, the electronic device (101) may include a processor (120), memory (130), input module (150), sound output module (155), display module (160), audio module (170), sensor module (176), interface (177), connection terminal (178), haptic module (179), camera module (180), power management module (188), battery (189), communication module (190), subscriber identification module (196), or antenna module (197). In some embodiments, at least one of these components (e.g., connection terminal (178)) may be omitted from the electronic device (101), or one or more other components may be added. In some embodiments, some of these components (e.g., sensor module (176), camera module (180), or antenna module (197)) may be integrated into a single component (e.g., display module (160)).

[0028] The processor (120) can control at least one other component (e.g., hardware or software component) of the electronic device (101) connected to the processor (120) by executing software (e.g., program (140)), for example, and can perform various data processing or operations. According to one embodiment, as at least part of the data processing or operations, the processor (120) can store commands or data received from other components (e.g., sensor module (176) or communication module (190)) in volatile memory (132), process the commands or data stored in volatile memory (132), and store the resulting data in non-volatile memory (134). According to one embodiment, the processor (120) may include a main processor (121) (e.g., central processing unit or application processor) or an auxiliary processor (123) that can operate independently or together with it (e.g., graphics processing unit, neural processing unit (NPU), image signal processor, sensor hub processor, or communication processor). For example, if the electronic device (101) includes a main processor (121) and an auxiliary processor (123), the auxiliary processor (123) may be configured to use lower power than the main processor (121) or to be specialized for a designated function. The auxiliary processor (123) may be implemented separately from the main processor (121) or as part thereof.

[0029] The auxiliary processor (123) may control at least some of the functions or states associated with at least one component of the electronic device (101) (e.g., display module (160), sensor module (176), or communication module (190)) on behalf of the main processor (121) while the main processor (121) is in an inactive (e.g., sleep) state, or together with the main processor (121) while the main processor (121) is in an active (e.g., application execution) state. According to one embodiment, the auxiliary processor (123) (e.g., image signal processor or communication processor) may be implemented as part of another functionally related component (e.g., camera module (180) or communication module (190)). According to one embodiment, the auxiliary processor (123) (e.g., neural network processing unit) may include a hardware structure specialized for processing an artificial intelligence model. The artificial intelligence model may be generated through machine learning. Such learning may be performed, for example, on the electronic device (101) itself where the artificial intelligence model is executed, or through a separate server (e.g., server (108)). The learning algorithm may include, for example, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning, but is not limited to the examples described above. The artificial intelligence model may include a plurality of artificial neural network layers.An artificial neural network may be a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), a restricted Boltzmann machine (RBM), a deep belief network (DBN), a bidirectional recurrent deep neural network (BRDNN), a deep Q-network, or a combination of two or more of the above, but is not limited to the examples described above. In addition to the hardware structure, the artificial intelligence model may include a software structure, either additionally or substantially.

[0030] The memory (130) can store various data used by at least one component of the electronic device (101) (e.g., processor (120) or sensor module (176)). The data may include, for example, input data or output data for software (e.g., program (140)) and related commands. The memory (130) may include volatile memory (132) or non-volatile memory (134).

[0031] The program (140) may be stored as software in memory (130) and may include, for example, an operating system (142), middleware (144), or an application (146).

[0032] The input module (150) can receive commands or data to be used for a component of the electronic device (101) (e.g., processor (120)) from outside the electronic device (101) (e.g., user). The input module (150) may include, for example, a microphone, a mouse, a keyboard, a key (e.g., a button), or a digital pen (e.g., a stylus pen).

[0033] The sound output module (155) can output a sound signal to the outside of the electronic device (101). The sound output module (155) may include, for example, a speaker or a receiver. The speaker may be used for general purposes, such as multimedia playback or recording playback. The receiver may be used to receive incoming calls. According to one embodiment, the receiver may be implemented separately from the speaker or as part thereof.

[0034] The display module (160) can visually provide information to an external (e.g., user) of the electronic device (101). The display module (160) may include, for example, a display, a holographic device, or a projector and a control circuit for controlling said device. According to one embodiment, the display module (160) may include a touch sensor configured to detect a touch, or a pressure sensor configured to measure the intensity of the force generated by said touch.

[0035] The audio module (170) can convert sound into an electrical signal or, conversely, convert an electrical signal into sound. According to one embodiment, the audio module (170) can acquire sound through the input module (150) or output sound through the sound output module (155) or an external electronic device (e.g., electronic device (102)) (e.g., speaker or headphones) connected directly or wirelessly to the electronic device (101).

[0036] The sensor module (176) can detect the operating state of the electronic device (101) (e.g., power or temperature) or the external environmental state (e.g., user state) and generate an electrical signal or data value corresponding to the detected state. According to one embodiment, the sensor module (176) may include, for example, a gesture sensor, a gyroscope sensor, a barometric pressure sensor, a magnetic sensor, an accelerometer sensor, a grip sensor, a proximity sensor, a color sensor, an IR (infrared) sensor, a biosensor, a temperature sensor, a humidity sensor, or an illuminance sensor.

[0037] The interface (177) may support one or more specified protocols that can be used for the electronic device (101) to be connected directly or wirelessly to an external electronic device (e.g., electronic device (102)). According to one embodiment, the interface (177) may include, for example, a high definition multimedia interface (HDMI), a universal serial bus (USB) interface, an SD card interface, or an audio interface.

[0038] The connection terminal (178) may include a connector through which the electronic device (101) can be physically connected to an external electronic device (e.g., electronic device (102)). According to one embodiment, the connection terminal (178) may include, for example, an HDMI connector, a USB connector, an SD card connector, or an audio connector (e.g., a headphone connector).

[0039] The haptic module (179) can convert an electrical signal into a mechanical stimulus (e.g., vibration or movement) or an electrical stimulus that the user can perceive through tactile or kinesthetic senses. According to one embodiment, the haptic module (179) may include, for example, a motor, a piezoelectric element, or an electric stimulation device.

[0040] The camera module (180) can capture still images and video. According to one embodiment, the camera module (180) may include one or more lenses, image sensors, image signal processors, or flashes.

[0041] The power management module (188) can manage the power supplied to the electronic device (101). According to one embodiment, the power management module (188) can be implemented, for example, as at least part of a power management integrated circuit (PMIC).

[0042] The battery (189) can supply power to at least one component of the electronic device (101). According to one embodiment, the battery (189) may include, for example, a non-rechargeable primary battery, a rechargeable secondary battery, or a fuel cell.

[0043] The communication module (190) can support the establishment of a direct (e.g., wired) communication channel or a wireless communication channel between an electronic device (101) and an external electronic device (e.g., electronic device (102), electronic device (104), or server (108)), and the performance of communication through the established communication channel. The communication module (190) may include one or more communication processors that operate independently of the processor (120) (e.g., application processor) and support direct (e.g., wired) communication or wireless communication. According to one embodiment, the communication module (190) may include a wireless communication module (192) (e.g., cellular communication module, short-range wireless communication module, or GNSS (global navigation satellite system) communication module) or a wired communication module (194) (e.g., LAN (local area network) communication module, or power line communication module). The corresponding communication module among these communication modules can communicate with an external electronic device (104) through a first network (198) (e.g., a short-range communication network such as Bluetooth, WiFi (wireless fidelity) direct, or IrDA (infrared data association)) or a second network (199) (e.g., a legacy cellular network, a 5G network, a next-generation communication network, the Internet, or a computer network (e.g., a LAN or WAN)). These various types of communication modules may be integrated into a single component (e.g., a single chip) or implemented as multiple separate components (e.g., multiple chips). The wireless communication module (192) can identify or authenticate the electronic device (101) within a communication network such as the first network (198) or the second network (199) using subscriber information (e.g., International Mobile Subscriber Identifier (IMSI)) stored in the subscriber identification module (196).

[0044] The wireless communication module (192) can support 5G networks and next-generation communication technologies following 4G networks, for example, new radio access technology. NR access technology can support high-speed transmission of high-capacity data (enhanced mobile broadband (eMBB)), minimization of terminal power and connection of multiple terminals (massive machine type communications (mMTC)), or high reliability and low latency (ultra-reliable and low-latency communications (URLLC)). The wireless communication module (192) can support a high-frequency band (e.g., mmWave band) to achieve a high data transmission rate, for example. The wireless communication module (192) can support various technologies for securing performance in the high-frequency band, such as beamforming, massive MIMO (multiple-input and multiple-output), full-dimensional MIMO (FD-MIMO), array antenna, analog beam-forming, or large-scale antenna. The wireless communication module (192) can support various requirements specified in the electronic device (101), external electronic device (e.g., electronic device (104)), or network system (e.g., second network (199)). According to one embodiment, the wireless communication module (192) can support a Peak data rate (e.g., 20 Gbps or more) for realizing eMBB, loss coverage (e.g., 164 dB or less) for realizing mMTC, or U-plane latency (e.g., downlink (DL) and uplink (UL) each 0.5 ms or less, or round trip 1 ms or less) for realizing URLLC.

[0045] An antenna module (197) can transmit a signal or power to or from an external source (e.g., an external electronic device). According to one embodiment, the antenna module (197) may include an antenna comprising a radiator made of a conductor or a conductive pattern formed on a substrate (e.g., a PCB). According to one embodiment, the antenna module (197) may include a plurality of antennas (e.g., an array antenna). In this case, at least one antenna suitable for a communication method used in a communication network, such as a first network (198) or a second network (199), may be selected from the plurality of antennas, for example, by a communication module (190). A signal or power may be transmitted or received between the communication module (190) and an external electronic device through the selected at least one antenna. According to some embodiments, in addition to the radiator, other components (e.g., a radio frequency integrated circuit (RFIC)) may be additionally formed as part of the antenna module (197).

[0046] According to various embodiments, the antenna module (197) may form a mmWave antenna module. According to one embodiment, the mmWave antenna module may include a printed circuit board, an RFIC disposed on or adjacent to a first surface (e.g., bottom surface) of the printed circuit board and capable of supporting a specified high frequency band (e.g., mmWave band), and a plurality of antennas (e.g., array antennas) disposed on or adjacent to a second surface (e.g., top surface or side surface) of the printed circuit board and capable of transmitting or receiving a signal of the specified high frequency band.

[0047] At least some of the above components can be connected to each other via a communication method between peripheral devices (e.g., bus, GPIO (general purpose input and output), SPI (serial peripheral interface), or MIPI (mobile industry processor interface)) and exchange signals (e.g., commands or data) with each other.

[0048] According to one embodiment, commands or data may be transmitted or received between the electronic device (101) and an external electronic device (104) through a server (108) connected to a second network (199). Each of the external electronic devices (102, or 104) may be the same or different type of device as the electronic device (101). According to one embodiment, all or part of the operations performed on the electronic device (101) may be performed on one or more of the external electronic devices (102, 104, or 108). For example, if the electronic device (101) needs to perform a function or service automatically or in response to a request from a user or another device, the electronic device (101) may request one or more external electronic devices to perform at least part of the function or service instead of performing the function or service itself or additionally. One or more external electronic devices that receive the above request may execute at least part of the requested function or service, or additional function or service related to the request, and transmit the result of the execution to the electronic device (101). The electronic device (101) may provide the result as is or additionally processed as at least part of the response to the request. For this purpose, for example, cloud computing, distributed computing, mobile edge computing (MEC), or client-server computing technology may be used. The electronic device (101) may provide ultra-low latency services using, for example, distributed computing or mobile edge computing. In another embodiment, the external electronic device (104) may include an Internet of Things (IoT) device. The server (108) may be an intelligent server using machine learning and / or neural networks. According to one embodiment, the external electronic device (104) or the server (108) may be included within a second network (199).The electronic device (101) can be applied to intelligent services (e.g., smart home, smart city, smart car, or healthcare) based on 5G communication technology and IoT-related technology.

[0049] FIG. 2 is a diagram showing an example of a MU (multi-user)-MIMO environment in a wireless communication system according to one embodiment of the present disclosure.

[0050] Referring to FIG. 2, a base station (210), and M r Each of the UEs (220, 230, 240) can be implemented as the electronic device (101) of FIG. 1.

[0051] In the present disclosure, a base station (210) is a network infrastructure that provides wireless access to UEs (220, 230, 240). The base station (210) has coverage defined based on the distance over which it can transmit signals. In addition to being a base station, the base station (210) may be referred to as an access point (AP), eNodeB (eNB), node, next generation nodeB (gNB), wireless point, transmission / reception point (TRP), electronic device, or other terms having an equivalent technical meaning.

[0052] In the present disclosure, each of the UEs (220, 230, 240) may be referred to as a terminal, customer premises equipment (CPE), mobile station, subscriber station, remote terminal, wireless terminal, electronic device, or user device, or any other term having an equivalent technical meaning.

[0053] M in the DL (downlink) of a wireless communication system t A base station (210) using multiple antennas and M using at least one receiving antenna r A DL MU (multi-user)-MIMO environment including 220, 230, 240 UEs can be implemented. In addition, M in the UL (uplink) of the wireless communication system r A base station (210) using multiple antennas and M using at least one transmitting antenna r A UL MU-MIMO environment including 220, 230, 240 UEs can be implemented.

[0054] FIG. 3 is a diagram illustrating an example of a SU-MIMO environment in a wireless communication system according to an embodiment of the present disclosure.

[0055] Referring to FIG. 3, the base station (310) and the UE (350) can each be implemented as the electronic device (101) of FIG. 1. In the downlink of a wireless communication system, M t Base station (310) using multiple antennas and M r A DL SU (single user)-MIMO environment including a UE (350) using multiple receiving antennas can be implemented. Additionally, in the uplink of the wireless communication system, M t Base station (310) using multiple antennas and Mr A UL SU-MIMO environment including a UE (350) using multiple transmitting antennas can be implemented.

[0056] The embodiments of the present disclosure can be applied to MIMO environments in which the number of transmitting / receiving antennas of a base station and the number of transmitting / receiving antennas of a UE (or terminal) are implemented in various ways.

[0057] FIG. 4 is a drawing illustrating the structure of a wireless communication system according to one embodiment of the present disclosure.

[0058] FIG. 4 illustrates an example in which a UE (or terminal) moves and changes the connected base station in a wireless communication system to which embodiments of the present disclosure are applied. Base stations (420, 430) may be connected to some of the surrounding base stations, and base stations (420, 430) may be connected to a core network (CN) (440), such as an EPC (Evolved Packet Core), a 5GC (5G Core Network), or a 6G network.

[0059] The radio access technology of the base stations (420, 430) may be LTE, NR, Wi-Fi, 6G, etc., and is not limited to one example. For example, the base stations (420, 430) may be mobile communication base stations unrelated to the radio access technology.

[0060] The UE (410) can be connected to a base station to receive mobile communication services, and as the UE (410) moves, the connected base station may change, and the UE (410) can receive mobile communication services without interruption through a handover (HO; Handover, or handoff) procedure. In one example of FIG. 4, the UE (410) is connected to a base station (420), then disconnects from the base station (420) through a handover and connects to a new base station (430). The core network (440) can control multiple base stations to estimate the location of the UE (410).

[0061] The present disclosure describes embodiments using terms used in some communication standards (e.g., 3GPP (3rd Generation Partnership Project)), but this is merely illustrative. Various embodiments of the present disclosure can be easily modified and applied to other communication and broadcasting systems.

[0062] When transmitting an uplink data channel (physical uplink shared channel, PUSCH) signal in an interference environment, M within the receiving device (e.g., base station) rx N with antennas layer The received signal (y(k)) received at the k-th subcarrier in the layer can be expressed as Equation 1.

[0063] [Mathematical Formula 1]

[0064]

[0065] Here, H(k) corresponds to the k-th subcarrier It is a channel matrix of size, and x(k) corresponds to the k-th subcarrier It is a transmitted signal vector of size n(k) It is a noise and interference vector of magnitude. When interference from adjacent cells is present, the PUSCH reception performance of the user being served may deteriorate.

[0066] For example, in a minimum mean square error (MMSE) receiver within a receiving device (e.g., a base station), MMSE can be performed on the received signal (y(k)) as shown in Equation 2.

[0067] [Mathematical Formula 2]

[0068]

[0069] Here, is the k-th subcarrier This is the channel matrix estimated from a receiver of that size. is the covariance matrix of noise and interference. Is Since it is a matrix of size, the complexity is high to perform the inverse operation of this matrix every time in the RE (resource block) that calculates the MMSE weights, so a pre-filtering structure can be selected in which a whitening transformation as shown in Equation 3 below is performed first, and then the MMSE is performed.

[0070] [Mathematical Formula 3]

[0071]

[0072] Meanwhile, the receiving device (e.g., base station) may also apply a Pre-Whitening structure such as Equation 4.

[0073] [Mathematical Formula 4]

[0074]

[0075] Here, is. In this case, the inverse matrix of noise and interference variance (matrix inversion) We end up calculating, and the corresponding complexity is The inverse of the noise and interference covariance is not calculated for every RE, but rather performed only once for each RBG (resource block group), which is a group of REs with similar interference, so the overall complexity can be low. On the other hand The complexity of an operation like this is However, the complexity can be high because MMSE weights must be calculated for all REs that require computation.

[0076] In the case of Layer 1, when comparing the complexity per block for MMSE weight calculations, the whitening corresponding As the number of antennas increases, it can account for most of the complexity. Consequently, it is essential to reduce whitening complexity, and existing techniques can approximate some elements of the covariance to zero, as shown in the example below, regardless of the actual covariance shape of the interference, in order to lower complexity.

[0077] Mathematical Equation 5 below represents the process of performing an inverse matrix operation after approximating the noise and interference covariance by keeping only the diagonal elements and setting the rest to 0.

[0078] [Mathematical Formula 5]

[0079]

[0080] Using a diagonal approximation method, the covariance can be approximated in four forms based on machine learning, as shown in Equation 6, depending on the interference.

[0081] [Mathematical Formula 6]

[0082]

[0083] In addition, it is possible to approximate it using a block diagonal matrix, and the method of finding the inverse matrix by leaving the remainders as 0 to retain only the block diagonal form can be expressed, for example, by mathematical equation 7.

[0084] [Mathematical Formula 7]

[0085]

[0086] Additionally, a method for approximating noise and interference covariances with a precision matrix in the form of a banded matrix can be expressed, for example, by Equation 8.

[0087] [Mathematical Formula 8]

[0088]

[0089] In all three cases of mathematical equations 6 through 8, there may be a common disadvantage in that interference cancellation performance is significantly reduced by approximating specific elements to zero regardless of the form of interference.

[0090] Since approximation is performed regardless of the form of the interference covariance, a loss in interference removal performance may occur. To compensate for this, the present disclosure proposes a method of approximating elements that have little influence on interference performance to 0 and correcting the remaining non-zero elements to correct the error caused by the approximation. The present disclosure proposes a method of applying an iterative algorithm and an unrolling-based neural network. Furthermore, the present disclosure can improve memory usage and the complexity of whitening operations by applying the sparse matrix multiplication (SpMM) method to sparse matrices.

[0091] FIG. 5 is a drawing illustrating the structure of a receiver according to one embodiment of the present disclosure.

[0092] Referring to FIG. 5, the receiver may include a channel estimation unit (510), a frequency domain equalizer (520), a de-mapper (530), and a decoder (540).

[0093] The received signal (received I / Q) (y) received by the receiver can be expressed by Equation 9.

[0094] [Mathematical Formula 9]

[0095]

[0096] Here, H is a channel matrix representing the channel that the received signal (received I / Q) (y) undergoes, x is a vector for the transmitted signal sent by the transmitter, n is a vector representing noise, and I can be a vector representing interference.

[0097] The channel estimation unit (510) can estimate the channel using the received signal (received I / Q) (y) and transmit information about the estimated channel (Estimated Channel) to the frequency domain equalizer (520). The frequency domain equalizer (520) can obtain an Equalized I / Q using the received signal (received I / Q) (y) and information about the estimated channel (Estimated Channel). The frequency domain equalizer (520) transmits the Equalized I / Q to the de-mapping unit (530), and the de-mapping unit (530) can calculate the LLR (log-likelihood ratio) using the Equalized I / Q. The de-mapping unit (530) transmits the calculated LLR to the decoder (540), and the decoder (540) can derive uncoded bits using the LLR.

[0098] FIG. 6 illustrates a process for designing a set of learnable parameters for unfolding according to one embodiment of the present disclosure, and FIG. 7 illustrates an unfolding method according to one embodiment of the present disclosure.

[0099] A method and apparatus according to one embodiment of the present disclosure may approximate elements within a matrix that have little effect on interference performance to zero and correct the remaining non-zero elements within the matrix to correct the error caused by the approximation. A method and apparatus according to one embodiment of the present disclosure may apply an iterative algorithm (IA) and an unfolding and / or unrolling-based neural network to correct the non-zero elements within the matrix.

[0100] Referring to FIGS. 5 and 6, an electronic device including the receiver of FIG. 5 can identify an optimization problem (P) (610) and perform an iterative algorithm (IA) to solve the optimization problem (P) (620). An electronic device including the receiver of FIG. 5 can unfold and / or unroll the iterative algorithm (IA) (630). An electronic device including the receiver of FIG. 5 can design and / or select a trainable parameter set through unfolding and / or unrolling the iterative algorithm (IA) (640).

[0101] Referring to FIGS. 5 to 7, an electronic device including the receiver of FIG. 5 can perform the i-th iteration algorithm (IA) using a pre-fixed parameter and a (-1)-th output value to solve an optimization problem (P) (710).

[0102] The electronic device including the receiver of FIG. 5 can perform unfolding for the iterative algorithm (IA) (720). The electronic device including the receiver of FIG. 5 has learnable parameters Using this, it is possible to perform iteration algorithms (IA), such as the first iteration (730) … the Nth iteration (740).

[0103] FIG. 8 is a block diagram illustrating a learning-based interference whitening method according to one embodiment of the present disclosure.

[0104] Referring to FIG. 8, an electronic device performing a learning-based interference whitening method may include a channel estimation unit (810), a noise covariance matrix calculation unit (820), an AI / Non-AI sparse precision matrix estimation unit (830), a compressed sparse row (CSR) application unit (840), and a sparse matrix multiplication (SpMM) application pre-filtering unit (850).

[0105] The channel estimation unit (810) estimates a channel representing the channel environment that the signal experiences, and the noise covariance matrix calculation unit (820) can calculate (or obtain) a dense noise covariance matrix based on the received signal and the estimated channel.

[0106] The AI / Non-AI based precision matrix estimator (830) can estimate (or calculate) a sparse precision matrix from a dense noise covariance matrix using an iterative algorithm (IA) or a neural network.

[0107] The CSR application unit (840) can apply CSR (compressed sparse row) to a sparse precision matrix to derive at least one row index (or row pointer), at least one column index, and at least one value index (or non-zero value) of a non-zero element in the matrix and store them in memory within the electronic device.

[0108] The SpMM application pre-filtering unit (850) can perform pre-filtering by applying SpMM to at least one row index (or row pointer), at least one column index, and at least one value index (or non-zero value) of a non-zero element.

[0109] The noise covariance matrix calculation unit (820) calculates the noise covariance using the received signal corresponding to the k-th subcarrier and the rs-th RS (reference symbol). , estimated channel , transmission signal according to DMRS (DeModulation Reference Signal) pattern Uses the estimated noise and interference vectors at the RE located at the k-th position in the rs-th DMRS symbol. It can be expressed as mathematical formula 10.

[0110] [Mathematical Formula 10]

[0111]

[0112] Using the noise and interference vectors estimated in Equation 10, the noise and interference covariance can be estimated as in Equation 11.

[0113] [Mathematical Formula 11]

[0114]

[0115] Here, the RBG (resource block group) is a set composed of resource blocks that share the same interference characteristics. Generally, the estimated S has dense characteristics. When interference is present, off-diagonal elements exist densely due to the interference; even in the absence of interference, the noise and interference matrices generally exhibit dense characteristics due to statistical errors resulting from a finite number of samples or errors arising from channel estimation.

[0116] The AI / Non-AI based precision matrix estimation unit (830) is in the SpMM applied pre-filtering unit (850). To reduce the complexity, instead of performing inverse matrix operations, a sparse precision matrix P can be obtained ( The AI / Non-AI based precision matrix estimator (830) may use a non-AI based Graphical Lasso / D Lasso or a similar non-AI based sparse precision matrix estimation method and a neural network.

[0117] The AI / Non-AI based precision matrix estimation unit (830) may apply the graphical least absolute shrinkage and selection operator (Lasso) as an example of the first method, the non-AI method. The optimization problem of the Graphical Lasso can be expressed by Equation 12.

[0118] [Mathematical Formula 12]

[0119]

[0120] The optimization problem of Equation 12 above can be solved using the block-wise coordinate descent (BCD) method or the alternative direction method of multipliers (ADMM) method. In addition, other non-AI-based sparse precision matrix estimation methods may be used. Since the first method has high computational complexity, the second method can use a neural network to take the noise and interference covariance S estimated like a neural network as input and output a sparse precision matrix P.

[0121] FIG. 9 is a diagram illustrating a specific process of a learning-based interference whitening method according to one embodiment of the present disclosure.

[0122] Referring to FIGS. 8 and 9, an electronic device for performing a learning-based interference whitening method may include a noise covariance estimation unit (910), an unfolded application unit (Unfolded D-Clime) (920), a matrix format conversion unit (930), and a pre-filtering unit (940) using SpMM.

[0123] The noise covariance estimation unit (910) can estimate the noise covariance (S) based on the estimated channel information (H) and the received signal (y).

[0124] The unfolding application unit (920) is a block corresponding to the AI / Non-AI based precision matrix estimation unit (830) of FIG. 8, and can perform multiple Unfolded Layer operations using noise covariance (S). The unfolding application unit (920) performs N Unfolded Layer operations (0 to (N-1)), and the result of the (N-1)th Unfolded Layer operation is Only the result is used for pre-filtering, and the remaining operation results can be intermediate values ​​for Unfolded Layer operations.

[0125] The unfolding application part (920) uses noise covariance (S) It can derive. is the estimation result of the precision matrix for the i-th layer, and represents diagonal loading, can be parameters derived using noise covariance (S).

[0126] The unfolding application part (920) is Unfolded Layer operation using (Unfolded Layer 0 th Performs ) and as the result of the operation It can obtain. Afterwards, the unfolding application part (920) Unfolded Layer operation using (Unfolded Layer 1 st Performs ) and as the result of the operation It can obtain. Afterwards, the unfolding application part (920) Using (N-1)th Unfolded Layer operation(Unfolded Layer (N-1) th Performs ) and as the result of the operation You can obtain . At this time, Only the parts are actually used for pre-filtering, and the rest may be intermediate results for Unfolded Layer operations.

[0127] The matrix format conversion unit (930) is the result of the operation of the unfolding application unit (920). Using this, you can obtain the row index r[k], column index c[k], and value index v[k] for the k-th subcarrier.

[0128] The pre-filtering unit (940) using SpMM uses r[k], c[k], v[k] and channel information (H) to pre-filter a filtered channel. You can obtain ( ).

[0129] FIG. 10 is a diagram illustrating an unrolling-based sparse precision matrix operation according to one embodiment of the present disclosure.

[0130] In FIG. 10, the electronic device can reduce non-iteration using an unrolling-based method and configure operation blocks for a sparse precision matrix.

[0131] The parameter set in the existing iterative algorithm It varies by layer and can be used in the current layer by being calculated based on the output of the previous layer and the results of the current layer.

[0132] Referring to FIG. 10, the electronic device can perform a plurality of Unfolded Layer operations using the noise covariance (S). The electronic device performs Unfolded Layer operations (Unfolded Layer 0) using the noise covariance (S). th )( Perform ) and as the result of the operation, Can obtain. Unfolded Layer operation (Unfolded Layer 0 th The result of ) is a parameter set It is passed to the neural network to update, and Unfolded Layer operations (Unfolded Layer 0 th The parameter set transmitted from the above neural network at ) can be used. Parameter set It can be passed to the next neural network.

[0133] The electronic device noise covariance (S), Unfolded Layer operation using (Unfolded Layer 1 st )( Perform ) and as the result of the operation, Can obtain. Unfolded Layer operation (Unfolded Layer 1 st The result of ) is a parameter set It is passed to the neural network for updating, and Unfolded Layer operations (Unfolded Layer 1 st The parameter set transmitted from the above neural network at ) This can be used. Parameter set It can be passed to the next neural network.

[0134] The electronic device noise covariance (S), Unfolded Layer operation using (Unfolded Layer (n-1) th )( Can perform ). Unfolded Layer operation (Unfolded Layer 1 st The result of ) is a parameter set It is passed to the neural network for updating, and Unfolded Layer operation(Unfolded Layer (n-1) th The parameter set transmitted from the above neural network This can be used.

[0135] According to one embodiment, the loss function for learning is a filtered channel ( The MSE (mean square error) value of ) can be expressed by the following mathematical formula 13.

[0136] [Mathematical Formula 13]

[0137]

[0138] Here, is the actual noise and interference covariance generated by the simulator. During training, the ground truth channel H(k) can use the actual channel generated by the simulator.

[0139] According to one embodiment, the loss function for learning can be expressed as Equation 14, which minimizes the MSE of the MMSE weights calculated using the inverse matrix of the MMSE weights and the ideal noise and interference covariance when using an MMSE receiver.

[0140] [Mathematical Formula 14]

[0141]

[0142] Here As the parameter is set larger, the sparsity of P increases, but If is too large, interference rejection performance may be degraded. Similarly, for the MMSE weights corresponding to the ground truth data, the channel, noise, and interference covariance generated by simulation can be used. Additionally, by looking at the normalized gain value corresponding to the performance of the equalizer, the negative form of the normalized gain can be used as the loss function as shown in Equation 15.

[0143] [Mathematical Formula 15]

[0144]

[0145] When used for whitening prior to ML detection, Equation 16 can be used as the loss function to minimize the cross entropy of LLR.

[0146] [Mathematical Formula 16]

[0147]

[0148] Here, is the sigmoid result of LLR, which is the log value of the ratio of the output after the input passes through the neural network.

[0149] Rows, values, and columns corresponding to non-zero elements can be stored in a sparse precision matrix P. To store rows, values, and columns corresponding to non-zero elements, the CSR (compressed sparse row) rule can be followed.

[0150] Subsequently, during pre-filtering, a filtered channel can be obtained by performing SpMM (sparse matrix multiplication) as shown in Equation 17.

[0151] [Mathematical Formula 17]

[0152]

[0153] Here, is the element of the filtered channel corresponding to the i-th row and j-th column. is a value in which non-zero elements are stored sequentially based on CSR, of the sparse precision matrix P It is the same as. is the i-th stored row pointer and the k-th stored column index according to the CSR format.

[0154] FIG. 11 is a diagram illustrating an Unfolded D-clime algorithm according to one embodiment of the present disclosure.

[0155] Referring to FIG. 11, an electronic device performing the Unfolded D-clime algorithm uses the noise covariance (S) as in Equation 18 It can derive. is the estimation result of the precision matrix for the i-th layer, and represents diagonal loading, can be parameters derived using noise covariance (S).

[0156] [Mathematical Formula 18]

[0157]

[0158] The electronic device performing the Unfolded D-clime algorithm Unfolded Layer operation using (Unfolded Layer 0 th Performs ) and as the result of the operation It can obtain. Subsequently, the electronic device Unfolded Layer operation using (Unfolded Layer 1 st Performs ) and as the result of the operation You can obtain. For example, Unfolded Layer operation (Unfolded Layer 1 st ) can be performed through the method described in Table 1.

[0159] [Table 1]

[0160]

[0161] Subsequently, the electronic device performing the Unfolded D-clime algorithm Using (N-1)th Unfolded Layer operation(Unfolded Layer (N-1) th Performs ) and as the result of the operation You can obtain.

[0162] According to one embodiment, a trainable parameter is It can be implemented with my parameters.

[0163] This can be an operation that squares each matrix element using the Frobneius norm and adds them together. soft(A,B) can be an operation that performs the operation on matrices AB and then changes any negative matrix elements to 0. clip(A,b) can be an operation that changes any elements of matrix A that are greater than b to b. This can be a matrix element-wise multiplication operation.

[0164] FIG. 12 is a diagram illustrating Compressed Sparse Row (CSR) and Sparse Precision Matrix Multiplication (SpMM) according to one embodiment of the present disclosure.

[0165] In Fig. 12, values ​​corresponding to non-zero elements in the sparse precision matrix (P) are stored as value (v[k]), and RowPtr (r[k]) for the row and colmunIndex (c[k]) for the column can be stored according to the CSR (compressed sparse row) rule.

[0166] According to one embodiment, v[k], r[k], and c[k] can be determined based on Equation 19.

[0167] [Mathematical Formula 19]

[0168]

[0169] According to one embodiment, Sparse Precision Matrix Multiplication (SpMM) can be computed based on Table 2.

[0170] [Table 2]

[0171]

[0172] During pre-filtering, a filtered channel can be obtained by performing sparse matrix multiplication (SpMM) as shown in Equation 20.

[0173] [Mathematical Formula 20]

[0174]

[0175] Here, is the element of the filtered channel corresponding to the i-th row and j-th column. is a value in which non-zero elements are stored sequentially based on CSR, of the sparse precision matrix P It is the same as. is the i-th stored row pointer and the k-th stored column index according to the CSR format.

[0176] FIG. 13 shows a memory processing flowchart in an electronic device according to one embodiment of the present disclosure.

[0177] Referring to FIG. 13, the electronic device stores the noise covariance (S) in memory in response to a first command (calculate noise covariance matrix) instructing to calculate the noise covariance matrix. It can be stored in memory.

[0178] The electronic device loads the noise covariance (S) from memory in response to a second command (calculate sparse precision matrix) instructing to calculate the sparse precision matrix, and overwrites the memory with the sparse precision matrix (P). It can be stored in memory. Afterwards, the memory can be cleared.

[0179] The electronic device loads the noise covariance (S) from memory in response to a third command instructing the CSR to obtain row index, column index, and value, and stores the obtained row index, column index, and value in memory, thereby storing integer index elements and complex value elements in memory.

[0180] According to one embodiment, a method of an electronic device in a wireless communication system may include: calculating a noise covariance matrix based on a received signal and an estimated channel; obtaining a sparse precision matrix from the noise covariance matrix using an iterative algorithm or a neural network; applying a compressed sparse row (CSR) to the sparse precision matrix to derive a plurality of indices for indicating non-zero elements within the matrix; and performing pre-filtering based on the plurality of indices and the estimated channel to obtain filtered channel information.

[0181] According to one embodiment, the plurality of indices may include first indices representing the values ​​of the non-zero elements in the sparse precision matrix, second indices representing the row of each of the first indices in the sparse precision matrix, and third indices representing the column of each of the first indices in the sparse precision matrix.

[0182] According to one embodiment, the method of the electronic device may further include an operation of performing a sparse matrix multiplication (SpMM) operation when performing the pre-filtering.

[0183] According to one embodiment, the method of the electronic device may further include an operation of performing a plurality of Unfolded Layer operations using the noise covariance matrix to obtain the sparse precision matrix.

[0184] According to one embodiment, the method of the electronic device may further include: an operation of storing the noise covariance matrix in response to a first command (calculate noise covariance matrix) instructing to calculate the noise covariance matrix; and an operation of loading the noise covariance matrix and overwriting the sparse precision matrix in response to a second command (calculate sparse precision matrix) instructing to calculate the sparse precision matrix. According to one embodiment, the method of the electronic device may further include an operation of loading the noise covariance matrix and storing the plurality of indices in response to a third command instructing to obtain the plurality of indices by the CSR.

[0185] According to one embodiment, the method of the electronic device may further include the operation of applying an unfolding or unrolling-based neural network to the iterative algorithm; and the operation of designing a trainable parameter set.

[0186] According to one embodiment, an electronic device in a wireless communication system may include a transceiver; a memory; and at least one processor. The at least one processor may enable the electronic device to perform a plurality of operations by executing instructions stored in the memory. The plurality of operations may include: an operation of calculating a noise covariance matrix based on a received signal and an estimated channel; an operation of obtaining a sparse precision matrix from the noise covariance matrix using an iterative algorithm or a neural network; an operation of deriving a plurality of indices to indicate non-zero elements in the matrix by applying a compressed sparse row (CSR) to the sparse precision matrix; and an operation of obtaining filtered channel information by performing pre-filtering based on the plurality of indices and the estimated channel.

[0187] According to one embodiment, a storage medium storing at least one instruction readable by a computer may be implemented. When the at least one instruction is executed by at least one processor, the electronic device may perform a plurality of operations. The plurality of operations may include: an operation of calculating a noise covariance matrix based on a received signal and an estimated channel; an operation of obtaining a sparse precision matrix from the noise covariance matrix using an iterative algorithm or a neural network; an operation of deriving a plurality of indices to indicate non-zero elements in the matrix by applying a compressed sparse row (CSR) to the sparse precision matrix; and an operation of obtaining filtered channel information by performing pre-filtering based on the plurality of indices and the estimated channel.

[0188] The present disclosure proposes a method to reduce receiver complexity and improve performance by obtaining a sparse precision matrix of noise and interference, which is the inverse of the noise and interference covariance, and using it for pre-filtering / whitening. By using the unrolling method proposed in the present disclosure, it operates with sparsely low complexity when interference is low, which can reduce the overall receiver complexity according to the SpMM method.

[0189] The embodiments of this document and the terms used therein are not intended to limit the technical features described in this document to specific embodiments, and should be understood to include various modifications, equivalents, or substitutions of said embodiments. In connection with the description of the drawings, similar reference numerals may be used for similar or related components. The singular form of a noun corresponding to an item may include one or more of said items unless the relevant context clearly indicates otherwise. In this document, phrases such as "A or B," "at least one of A and B," "at least one of A or B," "A, B or C," "at least one of A, B and C," and "at least one of A, B, or C" each may include any one of the items listed together in the corresponding phrase, or all possible combinations thereof. Terms such as "first," "second," or "first" or "second" may be used simply to distinguish said components from other said components and do not limit said components in any other aspect (e.g., importance or order). Where any (e.g., 1st) component is referred to as “coupled” or “connected” to another (e.g., 2nd) component, with or without the terms “functionally” or “communicationly,” it means that said any component may be connected to said other component directly (e.g., via a wire), wirelessly, or through a third component.

[0190] The term “module” as used in the embodiments of this document may include a unit implemented in hardware, software, or firmware, and may be used interchangeably with terms such as logic, logic block, component, or circuit, for example. A module may be a component formed integrally, or a minimum unit of said component or a part thereof that performs one or more functions. For example, according to one embodiment, a module may be implemented in the form of an application-specific integrated circuit (ASIC).

[0191] One embodiment of the present document may be implemented as software (e.g., program (140)) comprising one or more instructions stored in a storage medium (e.g., internal memory (136) or external memory (138)) readable by a machine (e.g., electronic device (101)). For example, a processor (e.g., processor (120)) of the machine (e.g., electronic device (101)) may call at least one of the one or more instructions stored in the storage medium and execute it. This enables the machine to be operated to perform at least one function according to the at least one called instruction. The one or more instructions may include code generated by a compiler or code that can be executed by an interpreter. The storage medium readable by the machine may be provided in the form of a non-transitory storage medium. Here, 'non-temporary' simply means that the storage medium is a tangible device and does not contain a signal (e.g., electromagnetic waves), and the term does not distinguish between cases where data is stored semi-permanently and cases where it is stored temporarily.

[0192] According to one embodiment, the method according to the embodiments disclosed herein may be provided by being included in a computer program product. The computer program product may be traded between a seller and a buyer as a product. The computer program product may be distributed in the form of a device-readable storage medium (e.g., compact disc read-only memory (CD-ROM)), or distributed online (e.g., download or upload) through an application store (e.g., Play Store™) or directly between two user devices (e.g., smartphones). In the case of online distribution, at least a portion of the computer program product may be temporarily stored or temporarily created on a device-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or a relay server.

[0193] According to one embodiment, each component (e.g., module or program) of the components described above may include a singular or multiple entities, and some of the multiple entities may be separated and placed in other components. According to one embodiment, 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. Generally or additionally, multiple components (e.g., module or program) may be integrated into a single component. In this case, the integrated component may perform one or more functions of each of the multiple components in the same or similar manner as those performed by the corresponding component among the multiple components prior to integration. According to one embodiment, operations performed by the module, program, or other components 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.

Claims

1. In a method of an electronic device in a wireless communication system, An operation to calculate a noise covariance matrix based on a received signal and an estimated channel; An operation to obtain a sparse precision matrix from the noise covariance matrix using an iterative algorithm or a neural network; The operation of deriving a plurality of indices for indicating non-zero elements within the matrix by applying a compressed sparse row (CSR) to the above sparse precision matrix; and A method comprising the operation of obtaining filtered channel information by performing pre-filtering based on the plurality of indices and the estimated channel.

2. In paragraph 1, the plurality of indices are, A method comprising first indices representing the values ​​of the non-zero elements within the sparse precision matrix, second indices representing the rows of each of the first indices within the sparse precision matrix, and third indices representing the columns of each of the first indices within the sparse precision matrix.

3. In Paragraph 1, A method further comprising an operation to perform SpMM (sparse matrix multiplication) operations when performing the above pre-filtering.

4. In Paragraph 1, A method further comprising the operation of performing a plurality of Unfolded Layer operations using the noise covariance matrix to obtain the above sparse precision matrix.

5. In Paragraph 1, An operation of storing the noise covariance matrix in response to a first command (calculate noise covariance matrix) instructing to calculate the noise covariance matrix; and A method further comprising the operation of loading the noise covariance matrix and overwriting the sparse precision matrix in response to a second command (calculate sparse precision matrix) instructing to calculate the sparse precision matrix.

6. In Paragraph 5, A method further comprising the operation of loading the noise covariance matrix and storing the plurality of indices in response to a third command instructing to acquire the plurality of indices by the above CSR.

7. In Paragraph 1, An operation of applying an unfolding or unrolling-based neural network to the above iterative algorithm; and A method that further includes an operation to design a trainable parameter set.

8. In an electronic device in a wireless communication system, transceiver; Memory; and It includes at least one processor, The above at least one processor enables the electronic device to perform a plurality of operations by executing instructions stored in the memory, and The above plurality of operations are, An operation to calculate a noise covariance matrix based on a received signal and an estimated channel; An operation to obtain a sparse precision matrix from the noise covariance matrix using an iterative algorithm or a neural network; The operation of deriving a plurality of indices for indicating non-zero elements within the matrix by applying a compressed sparse row (CSR) to the above sparse precision matrix; and A device comprising the operation of obtaining filtered channel information by performing pre-filtering based on the plurality of indices and the estimated channel.

9. In paragraph 8, the plurality of indices are, An apparatus comprising first indices representing values ​​of non-zero elements within the sparse precision matrix, second indices representing rows of each of the first indices within the sparse precision matrix, and third indices representing columns of each of the first indices within the sparse precision matrix.

10. In paragraph 8, the plurality of operations are, A device further comprising an operation to perform SpMM (sparse matrix multiplication) operations when performing the above pre-filtering.

11. In paragraph 8, the plurality of operations are, A device further comprising the operation of performing a plurality of Unfolded Layer operations using the noise covariance matrix to obtain the above sparse precision matrix.

12. In paragraph 8, the plurality of operations are, An operation of storing the noise covariance matrix in response to a first command (calculate noise covariance matrix) instructing to calculate the noise covariance matrix; and A device further comprising the operation of loading the noise covariance matrix and overwriting the sparse precision matrix in response to a second command (calculate sparse precision matrix) instructing to calculate the sparse precision matrix.

13. In Clause 12, the above plurality of operations are, A device further comprising the operation of loading the noise covariance matrix and storing the plurality of indices in response to a third command instructing to acquire the plurality of indices by the above CSR.

14. In paragraph 8, the plurality of operations are, An operation of applying an unfolding or unrolling-based neural network to the above iterative algorithm; and A device that further includes an operation for designing a trainable parameter set.

15. In a storage medium storing at least one instruction readable by a computer, The above at least one instruction causes an electronic device to perform a plurality of operations when executed by at least one processor, and The above plurality of operations are, An operation to calculate a noise covariance matrix based on a received signal and an estimated channel; An operation to obtain a sparse precision matrix from the noise covariance matrix using an iterative algorithm or a neural network; The operation of deriving a plurality of indices for indicating non-zero elements within the matrix by applying a compressed sparse row (CSR) to the above sparse precision matrix; and A storage medium comprising an operation of obtaining filtered channel information by performing pre-filtering based on the plurality of indices and the estimated channel.

Images