Apparatus and method for eliminating interference from receiver based on machine learning

Abstract

The present disclosure relates to an apparatus and a method for eliminating interference from a receiver based on machine learning, which may: estimate a channel; estimate covariance by using the estimated channel; estimate a shrinkage coefficient by using a first neural network; estimate a target matrix by using a second neural network; and correct the covariance by using the shrinkage coefficient and the target matrix.

Description

Device and method for eliminating interference in a receiver based on machine learning

[0001] The following embodiments relate to a technique for removing interference from a receiver based on machine learning.

[0002] MIMO (Multiple Input Multiple Output) systems are wireless communication systems that transmit data using multiple transmit and receive antennas. In these systems, the receiver plays a crucial role, and the Minimum Mean Square Error (MMSE) receiver is widely utilized.

[0003] The MMSE receiver operates by minimizing the mean square error of the received signal. This reduces interference between signals and improves the performance of the overall system.

[0004] However, signal interference can occur in MIMO systems due to various factors. One of these factors is multipath fading. Multipath fading occurs when a signal reaches the receiver through multiple paths, causing the signal strength to vary along each path. This can degrade the quality of the received signal and potentially degrade the performance of the entire MIMO system.

[0005] According to various embodiments of the present disclosure, a device and method for removing interference of a receiver based on machine learning are proposed.

[0006] A method for improving interference of a receiver according to one embodiment may include: an operation of estimating a channel; an operation of estimating covariance using the estimated channel; an operation of estimating a shrinkage coefficient using a first neural network; an operation of estimating a target matrix using a second neural network; and an operation of correcting the covariance using the shrinkage coefficient and the target matrix.

[0007] An interference improvement device of a receiver according to one embodiment may include a channel estimation unit for estimating a channel; a covariance estimation unit for estimating a covariance using the estimated channel; a first neural network for estimating a shrinkage coefficient; a second neural network for estimating a target matrix; and a covariance correction unit for correcting the covariance using the shrinkage coefficient and the target matrix.

[0008] FIG. 1 is a diagram illustrating a configuration of an electronic device for eliminating interference of a receiver according to an embodiment of the present invention.

[0009] FIG. 2 is a diagram illustrating a configuration of an electronic device that learns a neural network for interference removal according to one embodiment.

[0010] FIG. 3 is a diagram illustrating the configuration of a Shringe Gauge estimation unit according to one embodiment.

[0011] FIG. 4 is a diagram illustrating the configuration of a target matrix estimation unit according to one embodiment.

[0012] FIG. 5 is a flowchart illustrating a method for removing interference of a receiver in one embodiment.

[0013] FIG. 6 is a schematic diagram illustrating a configuration of an electronic device for eliminating interference of a receiver according to one embodiment.

[0014] FIG. 7 is a block diagram of an electronic device within a network environment according to one embodiment.

[0015] Hereinafter, embodiments are described in detail with reference to the attached drawings. However, the embodiments may be modified in various ways, and the scope of the patent application is not limited or restricted by these embodiments. It should be understood that all modifications, equivalents, or alternatives to the embodiments are included within the scope of the patent application.

[0016] The terms used in the examples are for illustrative purposes only and should not be construed as limiting. Singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, terms such as "comprise" or "have" are intended to indicate the presence of a feature, number, step, operation, component, part, or combination thereof described in the specification, but should be understood to not preclude the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0017] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art to which the embodiments pertain. Terms defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and shall not be interpreted in an idealized or overly formal sense unless explicitly defined herein.

[0018] In addition, when describing with reference to the attached drawings, identical components will be assigned the same reference numerals regardless of the drawing numbers, and redundant descriptions thereof will be omitted. When describing embodiments, if a detailed description of a related known technology is judged to unnecessarily obscure the gist of the embodiment, the detailed description will be omitted.

[0019] Additionally, terms such as first, second, A, B, (a), (b), etc. may be used to describe components of the embodiments. These terms are only intended to distinguish the components from other components, and the nature, order, or sequence of the components are not limited by the terms. When a component is described as being "connected," "coupled," or "connected" to another component, it should be understood that the component may be directly connected or connected to the other component, but another component may also be "connected," "coupled," or "connected" between each component.

[0020] Components included in one embodiment and components with common functions will be described using the same names in other embodiments. Unless otherwise stated, the descriptions given in one embodiment may also apply to other embodiments, and detailed descriptions will be omitted to the extent of overlap.

[0021] Hereinafter, a device and method for removing interference of a receiver based on machine learning according to an embodiment of the present invention will be described in detail with reference to the attached FIGS. 1 to 7.

[0022] FIG. 1 is a diagram illustrating a configuration of an electronic device for eliminating interference of a receiver according to an embodiment of the present invention.

[0023] Referring to FIG. 1, the electronic device (100) may be configured to include a channel estimation unit (110), a covariance estimation unit (120), a Shringe gauge estimation unit (130), a target matrix estimation unit (140), and a covariance correction unit (150).

[0024] The channel estimation unit (110) can estimate the channel of the received signal.

[0025] The channel estimation unit (110) can obtain channel information by estimating the channel of the received signal.

[0026] At this time, the channel information may include the number of receiving antennas, the number of samples, the number of transmission layers, the Doppler frequency, and the frequency selectivity. In addition, the channel information may further include at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured using a Demodulation Reference Signal (DMRS), noise per antenna measured using the DMRS, and interference per antenna measured using the DMRS.

[0027] The covariance estimation unit (120) can estimate the covariance using the estimated channel.

[0028] The noise vector for calculating covariance regarding noise can be expressed as in <Mathematical Formula 1> below.

[0029]

[0030] Here, Is In the th DMRS symbol is the noise vector estimated from the subcarrier located at the th, Is The th subcarrier and is the received signal corresponding to the th DMRS symbol, is the estimated channel, is a transmission signal according to the MRS pattern.

[0031] The covariance estimation unit (120) uses the measured noise vector to estimate the noise covariance in a resource block group (RBG) representing the area of ​​a resource block (RB). It can be calculated as shown in <Mathematical Formula 2> below.

[0032]

[0033] Here, is the noise covariance, is the number of REs within the RBG used for calculation, Is And, Is In the th DMRS symbol is the noise vector estimated from the subcarrier located at the th position.

[0034] And, the covariance estimation unit (120) uses the estimated covariance to calculate the trace norm of the covariance and the Frobenius norm of the covariance as in <Mathematical Formula 3> below, and uses the ratio to express the shape of the covariance. can be calculated as in <Mathematical Formula 3> below.

[0035]

[0036] Here, is the trace norm of the covariance, is the Frobenius norm of the covariance.

[0037]

[0038] Here, is the ratio of the trace norm of the covariance to the Frobenius norm of the covariance.

[0039] And, the covariance estimation unit (120) can calculate the ratio of the maximum and minimum values ​​of the diagonal terms of the covariance as shown in <Mathematical Formula 5> below.

[0040]

[0041] Here, is the ratio of the maximum and minimum values ​​of the diagonal terms of the covariance, is the minimum value of the diagonal term of the covariance, is the maximum value of the diagonal term of the covariance.

[0042] The shrinkage estimation unit (130) can estimate the shrinkage coefficient using the first neural network.

[0043] FIG. 3 is a diagram illustrating the configuration of a Shringe Gauge estimation unit according to one embodiment.

[0044] Referring to FIG. 3, the Shringage estimation unit (130) can be configured to include a first neural network (310).

[0045] The first neural network (310) can estimate the Schringe gauge coefficient by receiving the ratio of the trace norm of covariance and the Frobenius norm of covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, the Doppler frequency, and the frequency selectivity.

[0046] When estimating the Shringage coefficient, the first neural network (310) can additionally receive the channel estimation result as input to estimate the Shringage coefficient.

[0047] At this time, the additional channel estimation result may include at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a DMRS (Demodulation Reference Signal), noise per antenna measured by a DMRS, and interference per antenna measured by a DMRS.

[0048] Returning to the description of Fig. 1, the target matrix estimation unit (140) can estimate the target matrix using the second neural network.

[0049] FIG. 4 is a diagram illustrating the configuration of a target matrix estimation unit according to one embodiment.

[0050] Referring to FIG. 4, the target matrix estimation unit (140) can be configured to include a second neural network (410) and a softmax (420).

[0051] The second neural network (410) can estimate multiple target matrix values ​​by receiving the ratio of the trace norm of covariance and the Frobenius norm of covariance, the ratio of the maximum and minimum values ​​of the diagonal terms of covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, and the Doppler frequency and frequency selectivity as inputs.

[0052] When estimating a target matrix, the second neural network (410) can additionally input channel estimation results and use them to estimate multiple target matrices.

[0053] At this time, the additional channel estimation result may include at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a DMRS (Demodulation Reference Signal), noise per antenna measured by a DMRS, and interference per antenna measured by a DMRS.

[0054] Softmax (420) can output probability values ​​of multiple target matrices by performing softmax on multiple target matrix values ​​output from the second neural network (410).

[0055] The target matrix estimation unit (140) can select the target matrix with the highest probability by using the probability values ​​of multiple target matrices output from the softmax (420).

[0056] At this time, the first neural network (310) of FIG. 3 and the second neural network (410) of FIG. 4 are generated through supervised machine learning and can be configured as a multi-layer perceptron (MLP).

[0057] Returning to the description of FIG. 1, the covariance correction unit (150) can correct the covariance using the Shringage coefficient and the target matrix.

[0058] The covariance correction unit (150)

[0059] The noise covariance estimated in the above-described <Mathematical Formula 2> can be corrected in the form of a linear shrinkage estimator as in <Mathematical Formula 6> below.

[0060]

[0061] Here, is the corrected covariance, is the estimated noise covariance, is the Shringe gauge coefficient, is the target matrix.

[0062] At this time, can be estimated through the Shringage estimation unit (130), can be estimated through the target matrix estimation unit (140).

[0063] FIG. 2 is a diagram illustrating a configuration of an electronic device that learns a neural network for interference removal according to one embodiment.

[0064] Referring to FIG. 2, the electronic device (200) may be configured to include a processor (210) and a memory (220).

[0065] The memory (220) stores an operating system, application programs, and storage data for controlling the overall operation of the electronic device (200). The memory (220) can store various instructions that can be executed by the processor (1620).

[0066] The processor (210) may be configured to include a first training unit (212) and a second training unit (214).

[0067] The processor (210) determines the optimal Shringage coefficient for learning through the first training unit (212) and the second training unit (214). and the optimal target matrix is generated in advance. At this time, the processor (210) can perform the role of a simulator.

[0068] The method of generating learning data is to generate a Shringe gauge coefficient ( ) and target matrix ( ) is given, and the MMSE (minimum mean squared error) result is based on this. ) and the actual transmission data signal generated by the processor (210) ) to minimize the MSE (mean squared error) of <Mathematical Formula 7> , Save. One of the candidates is decided, and the candidate's embodiment is , am.

[0069]

[0070] Here, In <Mathematical Formula 2> It refers to the number of data resources included in the RBG used in the operation. is the result of MMSE equalizer using target matrix F, which is the shrinkage coefficient estimated by neural network, is the signal from the transmitter that corresponds to the actual correct answer. This signal is obtained through simulation.

[0071] The first training unit (212) can generate a first neural network by training to minimize the difference between the learning Shringage coefficient included in the learning data and the Shringage index of the learning result using the loss function of the weighted mean square error.

[0072] The learning data used in the first training unit (212) may include the ratio of the trace norm of the learning covariance and the Frobenius norm of the learning covariance, the number of learning receiving antennas, the number of learning samples, the number of learning transmitting layers, the learning Doppler frequency, the learning frequency selectivity, and the learning Shringe gauge coefficient.

[0073] The learning data used in the first training unit (212) may further include at least one of a learning time offset, a learning frequency offset, a signal-to-noise ratio for each learning antenna, noise for each learning antenna, and interference for each learning antenna.

[0074] The first training unit (212) uses the weighted mean squared error as a loss function for training the first neural network (310), and can be expressed as in <Mathematical Formula 8> below.

[0075]

[0076] Here, is the optimal Shringe gauge coefficient derived from the simulator, is a weight In cases where the performance change due to the change is large, it can be given a high value. is the number of learning data during training, and is the number of test data during performance evaluation.

[0077] The second training unit (214) can generate a second neural network by training the target matrix with the highest probability among preset target matrices by using the optimal learning target matrix included in the learning data and the optimal target matrix according to the learning result as inputs to the loss function of weighted cross entropy.

[0078] The learning data used in the second training unit (214) may include the ratio of the trace norm of the learning covariance and the Frobenius norm of the learning covariance, the ratio of the maximum and minimum values ​​of the diagonal terms of the learning covariance, the number of learning receiving antennas, the number of learning samples, the number of learning transmitting layers, the learning Doppler frequency, the learning frequency selectivity, and the optimal learning target matrix.

[0079] 2 The learning data used in the training unit (214) may further include at least one of a learning time offset, a learning frequency offset, a signal-to-noise ratio for each learning antenna, noise for each learning antenna, and interference for each learning antenna.

[0080] The second training unit (214) uses weighted cross-entropy as a loss function for training the second neural network (410), and can be expressed as in <Mathematical Formula 9> below.

[0081]

[0082] Here, Is The th learning data is the probability of the th target matrix, is the weight, Is The th learning data In the training data as an indicator of the th matrix If the th target matrix is ​​optimal, am.

[0083] The second training unit (214) selects the target matrix with the highest probability when estimating the learned result.

[0084] Meanwhile, although FIGS. 1 and 2 are depicted as separate devices, they may be implemented as one device.

[0085] Hereinafter, the method according to the present disclosure configured as above will be described with reference to the drawings below.

[0086] FIG. 5 is a flowchart illustrating a method for removing interference of a receiver in one embodiment.

[0087] Referring to FIG. 5, in operation 510, an electronic device (e.g., the electronic device (100) of FIG. 1) according to an embodiment may estimate a channel. In operation 510, the electronic device may obtain channel information by estimating a channel of a received signal. At this time, the channel information may include a number of receiving antennas, a number of samples, a number of transmission layers, a Doppler frequency, and a frequency selectivity. In addition, the channel information may further include at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a DMRS (Demodulation Reference Signal), noise per antenna measured by the DMRS, and interference per antenna measured by the DMRS.

[0088] In operation 520, the electronic device can estimate the covariance using the estimated channel.

[0089] In operation 530, the electronic device can estimate the shrinkage coefficient using the first neural network (e.g., the first neural network (310) of FIG. 3).

[0090] In operation 530, the electronic device can estimate the Schringe gauge coefficient by inputting the ratio of the trace norm of covariance and the Frobenius norm of covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, the Doppler frequency, and the frequency selectivity into the first neural network.

[0091] In operation 530, the electronic device may further include at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a Demodulation Reference Signal (DMRS), noise per antenna measured by the DMRS, and interference per antenna measured by the DMRS as inputs to the first neural network.

[0092] In operation 540, the electronic device can estimate the target matrix using a second neural network (e.g., the second neural network (410) of FIG. 4).

[0093] In operation 540, the electronic device can estimate the target matrix by inputting the ratio of the trace norm of the covariance and the Frobenius norm of the covariance, the ratio of the maximum and minimum values ​​of the diagonal terms of the covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, and the Doppler frequency and frequency selectivity into the second neural network.

[0094] In operation 540, the electronic device may further include at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a Demodulation Reference Signal (DMRS), noise per antenna measured by the DMRS, and interference per antenna measured by the DMRS as inputs to the second neural network.

[0095] In the 550 motion, the electronic device can compensate for covariance using the Shringage coefficients and the target matrix.

[0096] FIG. 6 is a schematic diagram illustrating a configuration of an electronic device for eliminating interference of a receiver according to one embodiment.

[0097] Referring to FIG. 6, the memory (620) stores an operating system, application programs, and storage data for controlling the overall operation of the electronic device (600). The memory (620) can store various instructions that can be executed by the processor (610).

[0098] The processor (610) can perform the operations of the channel estimation unit (110), the covariance estimation unit (120), the Shringe gauge estimation unit (130), the target matrix estimation unit (140), and the covariance correction unit (150) of the electronic device (100) of FIG. 1. That is, the processor (610) can include the configuration of the channel estimation unit (110), the covariance estimation unit (120), the Shringe gauge estimation unit (130), the target matrix estimation unit (140), and the covariance correction unit (150) of FIG. 1.

[0099] Additionally, the processor (610) can perform the operation of the processor (210) of the electronic device (200) of FIG. 2. That is, the processor (610) can include the configuration of the processor (210) of FIG. 2.

[0100] FIG. 7 is a block diagram of an electronic device within a network environment according to one embodiment.

[0101] Referring to FIG. 7, in a network environment (700), an electronic device (701) may communicate with an electronic device (702) via a first network (798) (e.g., a short-range wireless communication network), or may communicate with at least one of an electronic device (704) or a server (708) via a second network (799) (e.g., a long-range wireless communication network). According to one embodiment, the electronic device (701) may communicate with the electronic device (704) via the server (708). According to one embodiment, the electronic device (701) may include a processor (720), a memory (730), an input module (750), an audio output module (755), a display module (760), an audio module (770), a sensor module (776), an interface (777), a connection terminal (778), a haptic module (779), a camera module (780), a power management module (788), a battery (789), a communication module (790), a subscriber identification module (796), or an antenna module (797). In one embodiment, the electronic device (701) may have at least one of these components (e.g., the connection terminal (778)) omitted, or one or more other components added. In one embodiment, some of these components (e.g., the sensor module (776), the camera module (780), or the antenna module (797)) may be integrated into one component (e.g., the display module (760)).

[0102] The processor (720) may control at least one other component (e.g., hardware or software component) of the electronic device (701) connected to the processor (720) by executing, for example, software (e.g., program (740)), and may perform various data processing or operations. According to one embodiment, as at least a part of the data processing or operations, the processor (720) may store commands or data received from other components (e.g., sensor module (776) or communication module (790)) in volatile memory (732), process the commands or data stored in volatile memory (732), and store result data in non-volatile memory (734).

[0103] Meanwhile, the processor (720) can perform the operations of the channel estimation unit (110), covariance estimation unit (120), Shringage estimation unit (130), target matrix estimation unit (140), and covariance correction unit (150) of FIG. 1.

[0104] Additionally, the processor (720) can perform the operation of the processor (210) of FIG. 2. In addition, the processor (720) can perform the operation of the processor (610) of FIG. 6.

[0105] According to one embodiment, the processor (720) may include a main processor (721) (e.g., a central processing unit or an application processor) or an auxiliary processor (723) (e.g., a graphics processing unit, a neural processing unit (NPU), an image signal processor, a sensor hub processor, or a communication processor) that can operate independently or together with the main processor (721). For example, when the electronic device (701) includes the main processor (721) and the auxiliary processor (723), the auxiliary processor (723) may be configured to use less power than the main processor (721) or to be specialized for a given function. The auxiliary processor (723) may be implemented separately from the main processor (721) or as a part thereof.

[0106] The auxiliary processor (723) may control at least a portion of functions or states associated with at least one component (e.g., a display module (760), a sensor module (776), or a communication module (790)) of the electronic device (701), for example, on behalf of the main processor (721) while the main processor (721) is in an inactive (e.g., sleep) state, or together with the main processor (721) while the main processor (721) is in an active (e.g., application execution) state. In one embodiment, the auxiliary processor (723) (e.g., an image signal processor or a communication processor) may be implemented as a part of another functionally related component (e.g., a camera module (780) or a communication module (790)). In one embodiment, the auxiliary processor (723) (e.g., a neural network processing unit) may include a hardware structure specialized for processing artificial intelligence models. The artificial intelligence models may be generated through machine learning. This learning can be performed, for example, on the electronic device (701) itself where the artificial intelligence model is executed, or can be performed through a separate server (e.g., server (708)). The learning algorithm can 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 can include multiple artificial neural network layers.The artificial neural network may be one of 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, or alternatively to, a hardware structure, an artificial intelligence model may include a software structure.

[0107] The memory (730) can store various data used by at least one component (e.g., the processor (720) or the sensor module (776)) of the electronic device (701). The data can include, for example, software (e.g., the program (740)) and input data or output data for commands related thereto. The memory (730) can include a volatile memory (732) or a non-volatile memory (734).

[0108] Meanwhile, the memory (730) can perform the role of the memory (220) of FIG. 2 or the memory (620) of FIG. 6.

[0109] The program (740) may be stored as software in the memory (730) and may include, for example, an operating system (742), middleware (744), or an application (746).

[0110] The input module (750) can receive commands or data to be used in a component of the electronic device (701) (e.g., a processor (720)) from an external source (e.g., a user) of the electronic device (701). The input module (750) can include, for example, a microphone, a mouse, a keyboard, a key (e.g., a button), or a digital pen (e.g., a stylus pen).

[0111] The audio output module (755) can output audio signals to the outside of the electronic device (701). The audio output module (755) can include, for example, a speaker or a receiver. The speaker can be used for general purposes, such as multimedia playback or recording playback. The receiver can be used to receive incoming calls. According to one embodiment, the receiver can be implemented separately from the speaker or as part of the speaker.

[0112] The display module (760) can visually provide information to an external party (e.g., a user) of the electronic device (701). The display module (760) may include, for example, a display, a holographic device, or a projector, and a control circuit for controlling the device. According to one embodiment, the display module (760) may include a touch sensor configured to detect a touch, or a pressure sensor configured to measure the intensity of a force generated by the touch. In this case, the display module (760) may perform the role of the touch screen (70) of FIG. 1.

[0113] The audio module (770) can convert sound into an electrical signal, or vice versa, convert an electrical signal into sound. According to one embodiment, the audio module (770) can acquire sound through the input module (750), output sound through the sound output module (755), or an external electronic device (e.g., electronic device (702)) (e.g., speaker or headphone) directly or wirelessly connected to the electronic device (701).

[0114] The sensor module (776) can detect the operating status (e.g., power or temperature) of the electronic device (701) or the external environmental status (e.g., user status) and generate an electrical signal or data value corresponding to the detected status. According to one embodiment, the sensor module (776) can include, for example, a gesture sensor, a gyro sensor, a barometric pressure sensor, a magnetic sensor, an acceleration sensor, a grip sensor, a proximity sensor, a color sensor, an IR (infrared) sensor, a biometric sensor, a temperature sensor, a humidity sensor, or an illuminance sensor.

[0115] The interface (777) may support one or more designated protocols that may be used to directly or wirelessly connect the electronic device (701) with an external electronic device (e.g., the electronic device (702)). In one embodiment, the interface (777) may include, for example, a high definition multimedia interface (HDMI), a universal serial bus (USB) interface, an SD card interface, or an audio interface.

[0116] The connection terminal (778) may include a connector through which the electronic device (701) may be physically connected to an external electronic device (e.g., the electronic device (702)). According to one embodiment, the connection terminal (778) may include, for example, an HDMI connector, a USB connector, an SD card connector, or an audio connector (e.g., a headphone connector).

[0117] The haptic module (779) can convert electrical signals into mechanical stimuli (e.g., vibration or movement) or electrical stimuli that a user can perceive through tactile or kinesthetic sensations. According to one embodiment, the haptic module (779) can include, for example, a motor, a piezoelectric element, or an electrical stimulation device.

[0118] The camera module (780) can capture still images and videos. According to one embodiment, the camera module (780) may include one or more lenses, image sensors, image signal processors, or flashes.

[0119] The power management module (788) can manage the power supplied to the electronic device (701). According to one embodiment, the power management module (788) can be implemented as, for example, at least a part of a power management integrated circuit (PMIC).

[0120] A battery (789) may power at least one component of the electronic device (701). In one embodiment, the battery (789) may include, for example, a non-rechargeable primary battery, a rechargeable secondary battery, or a fuel cell.

[0121] The communication module (790) may support the establishment of a direct (e.g., wired) communication channel or a wireless communication channel between the electronic device (701) and an external electronic device (e.g., electronic device (702), electronic device (704), or server (708)), and the performance of communication through the established communication channel. The communication module (790) may operate independently from the processor (720) (e.g., application processor) and may include one or more communication processors that support direct (e.g., wired) communication or wireless communication. According to one embodiment, the communication module (790) may include a wireless communication module (792) (e.g., a cellular communication module, a short-range wireless communication module, or a global navigation satellite system (GNSS) communication module) or a wired communication module (794) (e.g., a local area network (LAN) communication module, or a power line communication module). Among these communication modules, a corresponding communication module can communicate with an external electronic device (704) via a first network (798) (e.g., a short-range communication network such as Bluetooth, wireless fidelity (WiFi) direct, or infrared data association (IrDA)) or a second network (799) (e.g., a long-range communication network such as 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 can 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 (792) can verify or authenticate the electronic device (701) within a communication network such as the first network (798) or the second network (799) by using subscriber information (e.g., an international mobile subscriber identity (IMSI)) stored in the subscriber identification module (796). The communication module (790) can perform the role of the communication unit (120) of FIG. 1.

[0122] The wireless communication module (792) can support 5G networks and next-generation communication technologies following the 4G network, such as NR access technology (new radio access technology). The NR access technology can support high-speed transmission of high-capacity data (eMBB (enhanced mobile broadband)), minimization of terminal power and connection of multiple terminals (mMTC (massive machine type communications)), or high reliability and low latency communications (URLLC (ultra-reliable and low-latency communications)). The wireless communication module (792) can support, for example, a high-frequency band (e.g., mmWave band) to achieve a high data transmission rate. The wireless communication module (792) may support various technologies for securing performance in a high-frequency band, such as beamforming, massive multiple-input and multiple-output (MIMO), full dimensional MIMO (FD-MIMO), array antenna, analog beam-forming, or large scale antenna. The wireless communication module (792) may support various requirements specified in the electronic device (701), an external electronic device (e.g., the electronic device (704)), or a network system (e.g., the second network (799)). According to one embodiment, the wireless communication module (792) may support a peak data rate (e.g., 20 Gbps or more) for eMBB realization, a loss coverage (e.g., 164 dB or less) for mMTC realization, or a U-plane latency (e.g., 0.5 ms or less for downlink (DL) and uplink (UL), or 1 ms or less for round trip) for URLLC realization.

[0123] The antenna module (797) can transmit or receive signals or power to or from an external device (e.g., an external electronic device). According to one embodiment, the antenna module (797) may include an antenna including a radiator formed of a conductor or a conductive pattern formed on a substrate (e.g., a PCB). According to one embodiment, the antenna module (797) 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 the first network (798) or the second network (799), may be selected from the plurality of antennas, for example, by the communication module (790). A signal or power may be transmitted or received between the communication module (790) and an external electronic device via the selected at least one antenna. According to one embodiment, in addition to the radiator, another component (e.g., a radio frequency integrated circuit (RFIC)) may be additionally formed as a part of the antenna module (797).

[0124] In one embodiment, the antenna module (797) may form a mmWave antenna module. In one embodiment, the mmWave antenna module may include a printed circuit board, an RFIC disposed on or adjacent a first side (e.g., a bottom side) of the printed circuit board and capable of supporting a designated high-frequency band (e.g., a mmWave band), and a plurality of antennas (e.g., an array antenna) disposed on or adjacent a second side (e.g., a top side or a side side) of the printed circuit board and capable of transmitting or receiving signals in the designated high-frequency band.

[0125] At least some of the above components can be interconnected and exchange signals (e.g., commands or data) with each other via a communication method between peripheral devices (e.g., a bus, GPIO (general purpose input and output), SPI (serial peripheral interface), or MIPI (mobile industry processor interface)).

[0126] According to one embodiment, commands or data may be transmitted or received between the electronic device (701) and an external electronic device (704) via a server (708) connected to a second network (799). Each of the external electronic devices (702, or 1304) may be the same or a different type of device as the electronic device (701). According to one embodiment, all or part of the operations executed in the electronic device (701) may be executed in one or more of the external electronic devices (702, 1304, or 1308). For example, when the electronic device (701) is to perform a certain function or service automatically or in response to a request from a user or another device, the electronic device (701) may, instead of or in addition to executing the function or service itself, request one or more external electronic devices to perform the function or at least a part of the service. One or more external electronic devices that receive the request may execute at least a portion of the requested function or service, or an additional function or service related to the request, and transmit the result of the execution to the electronic device (701). The electronic device (701) may process the result as is or additionally and provide it as at least a portion of a response to the request. For this purpose, cloud computing, distributed computing, mobile edge computing (MEC), or client-server computing technology may be used, for example. The electronic device (701) may provide an ultra-low latency service by using distributed computing or mobile edge computing, for example. In another embodiment, the external electronic device (704) may include an Internet of Things (IoT) device. The server (708) may be an intelligent server utilizing machine learning and / or a neural network. According to one embodiment, the external electronic device (704) or the server (708) may be included in the second network (799).The electronic device (701) 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.

[0127] Electronic devices according to embodiments of the present disclosure may take various forms. Electronic devices may include, for example, portable communication devices (e.g., smartphones), computer devices, portable multimedia devices, portable medical devices, cameras, wearable devices, or home appliances. Electronic devices according to embodiments of the present disclosure are not limited to the aforementioned devices.

[0128] The embodiments of the present disclosure and the terminology used herein are not intended to limit the technical features described in the present disclosure to specific embodiments, but should be understood to include various modifications, equivalents, or substitutes of the 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 the items, unless the context clearly indicates otherwise. In the present disclosure, each of the phrases "A or B", "at least one of A and B", "at least one of A or B", "A, B, or C", "at least one of A, B, and C", and "at least one of A, B, or C" can include any one of the items listed together in the corresponding phrase among the phrases, or all possible combinations thereof. Terms such as "first," "second," or "first" or "second" may be used merely to distinguish one component from another, and do not limit the components in any other respect (e.g., importance or order). When a component (e.g., a first component) is referred to as "coupled" or "connected" to another component (e.g., a second component), with or without the terms "functionally" or "communicatively," it means that the component can be connected to the other component directly (e.g., wired), wirelessly, or through a third component.

[0129] The term "module" used in one embodiment of the present disclosure 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 an integral component, or a minimum unit or part of such a component 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).

[0130] An embodiment of the present disclosure may be implemented as software (e.g., a program (740)) including one or more instructions stored in a storage medium (e.g., an internal memory (736) or an external memory (738)) readable by a machine (e.g., an electronic device (701)). For example, a processor (e.g., a processor (720)) of the machine (e.g., an electronic device (701)) may call at least one instruction among the one or more instructions stored from the storage medium and execute it. This enables the machine to operate 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 executable by an interpreter. The machine-readable storage medium may be provided in the form of a non-transitory storage medium. Here, 'non-transitory' simply means that the storage medium is a tangible device and does not contain signals (e.g., electromagnetic waves), and the term does not distinguish between cases where data is stored semi-permanently or temporarily on the storage medium.

[0131] According to one embodiment, a method according to one embodiment of the present disclosure may be provided as included in a computer program product. The computer program product may be traded as a product between a seller and a buyer. 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 generated in a machine-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or a relay server.

[0132] According to one embodiment, each component (e.g., a module or a program) of the above-described components may include one or more entities, and some of the entities may be separated and placed in other components. According to one embodiment, one or more 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., a module or a program) 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 one embodiment, 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.

[0133] According to one embodiment, a method for improving interference of a receiver may include: an operation of estimating a channel; an operation of estimating covariance using the estimated channel; an operation of estimating a shrinkage coefficient using a first neural network (310); an operation of estimating a target matrix using a second neural network (410); and an operation of correcting the covariance using the shrinkage coefficient and the target matrix.

[0134] According to one embodiment, the operation of estimating the Shringe gauge coefficient may estimate the Shringe gauge coefficient by inputting the ratio of the trace norm of the covariance and the Frobenius norm of the covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, the Doppler frequency, and the frequency selectivity into the first neural network (310).

[0135] According to one embodiment, the operation of estimating the Shringage coefficient may further include at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a Demodulation Reference Signal (DMRS), noise per antenna measured by the DMRS, and interference per antenna measured by the DMRS as inputs of the first neural network (310).

[0136] According to one embodiment, the operation of estimating the target matrix may estimate the target matrix by inputting the ratio of the trace norm of the covariance and the Frobenius norm of the covariance, the ratio of the maximum and minimum values ​​of the diagonal terms of the covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, and the Doppler frequency and frequency selectivity into the second neural network (410).

[0137] According to one embodiment, the operation of estimating the target matrix may further include at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a DMRS (Demodulation Reference Signal), noise per antenna measured by the DMRS, and interference per antenna measured by the DMRS as inputs of the second neural network (410).

[0138] According to one embodiment, the first neural network (310) and the second neural network (410) can be generated through supervised machine learning.

[0139] According to one embodiment, the first neural network (310) and the second neural network (410) may be configured as a multi-layer perceptron (MLP).

[0140] According to one embodiment, the first neural network (310) may further be generated by training the first neural network (310) so that the difference between the learning Shringage coefficient included in the learning data and the Shringage index of the learning result is minimized using a loss function of the weighted mean square error.

[0141] According to one embodiment, the learning data may include a ratio of a trace norm of a learning covariance and a Frobenius norm of the learning covariance, a number of learning receiving antennas, a number of learning samples, a number of learning transmitting layers, a learning Doppler frequency, a learning frequency selectivity, and a learning Shringe gauge coefficient.

[0142] According to one embodiment, the learning data may further include at least one of a learning time offset, a learning frequency offset, a learning antenna-specific signal-to-noise ratio, a learning antenna-specific noise, and a learning antenna-specific interference.

[0143] According to one embodiment, the second neural network (410) may further include an operation of generating the second neural network by training the target matrix with the highest probability among preset target matrices by using the optimal learning target matrix included in the learning data and the optimal target matrix according to the learning result as inputs of the loss function of the weighted cross entropy.

[0144] According to one embodiment, the training data may include a ratio of a trace norm of a learning covariance and a Frobenius norm of the learning covariance, a ratio of a maximum value and a minimum value of a diagonal term of the learning covariance, a number of learning receiving antennas, a number of learning samples, a number of learning transmitting layers, a learning Doppler frequency, a learning frequency selectivity, and the optimal learning target matrix.

[0145] According to one embodiment, the learning data may further include at least one of a learning time offset, a learning frequency offset, a learning antenna-specific signal-to-noise ratio, a learning antenna-specific noise, and a learning antenna-specific interference.

[0146] According to one embodiment, an interference improvement device of a receiver may include a channel estimation unit (110) for estimating a channel; a covariance estimation unit (120) for estimating a covariance using the estimated channel; a first neural network (310) for estimating a shrinkage coefficient; a second neural network (410) for estimating a target matrix; and a covariance correction unit (150) for correcting the covariance using the shrinkage coefficient and the target matrix.

[0147] According to one embodiment, the first neural network (310) can estimate the Schringe gauge coefficient by inputting the ratio of the trace norm of the covariance and the Frobenius norm of the covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, the Doppler frequency, and the frequency selectivity into the first neural network.

[0148] According to one embodiment, when estimating the Shringage coefficient, the first neural network (310) may further include at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a DMRS (Demodulation Reference Signal), noise per antenna measured by the DMRS, and interference per antenna measured by the DMRS as inputs of the first neural network.

[0149] According to one embodiment, the second neural network (410) can estimate the target matrix by inputting the ratio of the trace norm of the covariance and the Frobenius norm of the covariance, the ratio of the maximum and minimum values ​​of the diagonal terms of the covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, and the Doppler frequency and frequency selectivity into the second neural network.

[0150] According to one embodiment, when estimating the target matrix, the second neural network (410) may further include at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a DMRS (Demodulation Reference Signal), noise per antenna measured by the DMRS, and interference per antenna measured by the DMRS as inputs of the second neural network.

[0151] According to one embodiment, the first neural network (310) and the second neural network (410) may be generated through supervised machine learning and configured as a multi-layer perceptron (MLP).

[0152] The method according to the embodiment may be implemented in the form of program commands that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may store program commands, data files, data structures, etc., singly or in combination. The program commands recorded on the medium may be those specially designed and configured for the embodiment or may be those known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specially configured to store and execute program commands, such as ROMs, RAMs, and flash memories. Examples of the program commands include not only machine language codes generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter, etc. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiment, and vice versa.

[0153] Software may include a computer program, code, instructions, or a combination of one or more of these, and may configure a processing device to perform a desired operation or, independently or collectively, command the processing device. The software and / or data may be stored on any type of machine, component, physical device, virtual equipment, computer storage medium, or device, for interpretation by the processing device or for providing instructions or data to the processing device. The software may also be distributed over networked computer systems, and stored or executed in a distributed manner. The software and data may be stored on one or more computer-readable recording media.

[0154] Although the embodiments described above have been described with limited drawings, those skilled in the art will appreciate that various technical modifications and variations can be applied based on the above. For example, appropriate results can still be achieved even if the described techniques are performed in a different order than described, and / or components of the described systems, structures, devices, circuits, etc. are combined or combined in a different manner than described, or are replaced or substituted with other components or equivalents.

[0155] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims described below.

Claims

1. The action of estimating the channel; The operation of estimating covariance using the estimated channel; An operation to estimate the shrinkage coefficient using the first neural network; An operation of estimating a target matrix using a second neural network; and An operation of correcting the covariance using the above Shringage coefficient and the above target matrix. A method for improving interference of a receiver including:

2. In paragraph 1, The operation of estimating the above Shringe Gauge coefficient is as follows: The ratio of the trace norm of the above covariance and the Frobenius norm of the above covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, the Doppler frequency and the frequency selectivity. Estimating the Shringage coefficient by inputting it into the first neural network Method for improving interference in receivers.

3. In any one of paragraphs 1 and 2, The operation of estimating the above Shringe Gauge coefficient is as follows: The input of the first neural network further includes at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a DMRS (Demodulation Reference Signal), noise per antenna measured by the DMRS, and interference per antenna measured by the DMRS. Method for improving interference in receivers.

4. In any one of paragraphs 1 to 3, The operation of estimating the above target matrix is as follows: The target matrix is estimated by inputting the ratio of the trace norm of the covariance and the Frobenius norm of the covariance, the ratio of the maximum and minimum values of the diagonal terms of the covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, and the Doppler frequency and frequency selectivity into the second neural network. Method for improving interference in receivers.

5. In any one of paragraphs 1 to 4, The operation of estimating the above target matrix is as follows: The input of the second neural network further includes at least one of a time offset, a frequency offset, a signal-to-noise ratio per antenna measured by a DMRS (Demodulation Reference Signal), noise per antenna measured by the DMRS, and interference per antenna measured by the DMRS. Method for improving interference in receivers.

6. In any one of paragraphs 1 to 5, The above first neural network and the above second neural network, Generated through supervised machine learning Method for improving interference in receivers.

7. In any one of paragraphs 1 to 6, The above first neural network and the above second neural network, It consists of a multi-layer perceptron (MLP). Method for improving interference in receivers.

8. In any one of paragraphs 1 to 7, An operation of generating the first neural network by training it so that the difference between the learning Shringage coefficient included in the learning data and the Shringage index of the learning result is minimized using the loss function of the weighted mean square error. A method for improving interference of a receiver including more.

9. In any one of paragraphs 1 to 8, The above learning data is, The ratio of the trace norm of the learning covariance and the Frobenius norm of the learning covariance, the number of learning receiving antennas, the number of learning samples, the number of learning transmitting layers, the learning Doppler frequency, the learning frequency selectivity, and the learning Shringe gauge coefficients are included. Method for improving interference in receivers.

10. In any one of paragraphs 1 to 9, The above learning data is, At least one of a learning time offset, a learning frequency offset, a learning antenna-specific signal-to-noise ratio, a learning antenna-specific noise, and a learning antenna-specific interference is further included. Method for improving interference in receivers.

11. In any one of paragraphs 1 to 10, An operation of generating the second neural network by training the target matrix with the highest probability among preset target matrices by using the optimal learning target matrix included in the learning data and the optimal target matrix according to the learning result as inputs to the loss function of weighted cross entropy. A method for improving interference of a receiver including more.

12. In any one of paragraphs 1 to 11, The above learning data is, The ratio of the trace norm of the learning covariance and the Frobenius norm of the learning covariance, the ratio of the maximum and minimum values of the diagonal terms of the learning covariance, the number of learning receiving antennas, the number of learning samples, the number of learning transmitting layers, the learning Doppler frequency, the learning frequency selectivity, and the optimal learning target matrix. Method for improving interference in receivers.

13. In any one of paragraphs 1 to 12, The above learning data is, At least one of a learning time offset, a learning frequency offset, a learning antenna-specific signal-to-noise ratio, a learning antenna-specific noise, and a learning antenna-specific interference is further included. Method for improving interference in receivers.

14. Channel estimation unit that estimates the channel; A covariance estimation unit that estimates covariance using the estimated channel; A first neural network that estimates shrinkage coefficients; a second neural network that estimates the target matrix; and A covariance correction unit that corrects the covariance using the above Shringage coefficient and the target matrix. An interference improvement device for a receiver including a .

15. In paragraph 14, The above first neural network, The ratio of the trace norm of the covariance and the Frobenius norm of the covariance, the number of receiving antennas, the number of samples, the number of transmitting layers, the Doppler frequency, and the frequency selectivity are input to the first neural network to estimate the Schringe gauge coefficient. Interference improvement device of the receiver.

Images