Channel compensation method and device, electronic equipment and storage medium

Abstract

The application discloses a channel compensation method and device, electronic equipment and storage medium, and relates to the technical field of signal processing. The method comprises the following steps: acquiring a receiving end pilot signal and a transmitting end pilot signal; using an initial channel estimation algorithm to perform signal estimation on the receiving end pilot signal and the transmitting end pilot signal, and obtaining an initial channel compensation matrix; based on a pre-trained channel compensation matrix update model, correcting the initial channel compensation matrix to obtain a target channel compensation matrix; and using the target signal compensation matrix to perform channel compensation on the receiving end pilot signal to obtain an original pilot signal. The technical scheme of the embodiment of the application can reduce the physical layer deployment cost, improve the scalability of signal estimation, reduce the calculation cost, improve the signal compensation accuracy, significantly reduce the receiving end bit error rate, improve the communication rate, and meet the current communication system requirements for high speed, low latency and high reliability stability.

Description

Technical Field

[0001] This invention relates to the field of signal processing technology, and in particular to a channel compensation method, apparatus, electronic device, and storage medium. Background Technology

[0002] Existing channel estimation algorithms are mainly classified into reference signal-based channel estimation, blind estimation, and semi-blind estimation. Among them, reference signal-based channel estimation algorithms are the most widely used. Reference signal-based channel estimation algorithms calculate and process the received and transmitted signals to obtain the channel compensation matrix, thereby recovering the original signal.

[0003] However, the current wireless channel environment is becoming increasingly complex. Traditional channel estimation algorithms have insufficient scalability and high computational overhead, and can no longer meet the current communication system's requirements for high speed, low latency, and high reliability. Summary of the Invention

[0004] This invention provides a channel compensation method, apparatus, electronic device, and storage medium, which can reduce physical layer deployment costs, improve the scalability of signal estimation, reduce computational overhead, improve signal compensation accuracy, significantly reduce the bit error rate at the receiver, and increase the communication rate, thus meeting the current communication system's requirements for high speed, low latency, and high reliability.

[0005] According to one aspect of the present invention, a channel compensation method is provided, the method comprising:

[0006] Acquire the pilot signals from the receiving end and the pilot signals from the transmitting end;

[0007] An initial channel estimation algorithm is used to estimate the signal of the receiving pilot signal and the transmitting pilot signal to obtain the initial channel compensation matrix.

[0008] Based on the pre-trained channel compensation matrix update model, the initial channel compensation matrix is ​​corrected to obtain the target channel compensation matrix;

[0009] The target signal compensation matrix is ​​used to perform channel compensation on the receiving pilot signal to obtain the original pilot signal.

[0010] According to another aspect of the present invention, a channel compensation apparatus is provided, the apparatus comprising:

[0011] Pilot signal acquisition module, used to acquire the pilot signal from the receiving end and the pilot signal from the transmitting end;

[0012] The initial channel estimation module is used to perform signal estimation on the receiving pilot signal and the transmitting pilot signal using an initial channel estimation algorithm to obtain an initial channel compensation matrix.

[0013] The compensation matrix update module is used to correct the initial channel compensation matrix based on a pre-trained channel compensation matrix update model to obtain the target channel compensation matrix.

[0014] The channel compensation module is used to perform channel compensation on the receiving pilot signal using the target signal compensation matrix to obtain the original pilot signal.

[0015] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0016] At least one processor; and

[0017] A memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the channel compensation method according to any embodiment of the present invention.

[0019] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the channel compensation method according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program that, when executed by a processor, implements the channel compensation method according to any embodiment of the present invention.

[0021] The technical solution of this invention acquires the receiving pilot signal and the transmitting pilot signal, uses an initial channel estimation algorithm to estimate the signals of the receiving and transmitting pilot signals, obtains an initial channel compensation matrix, corrects the initial channel compensation matrix based on a pre-trained channel compensation matrix update model to obtain a target channel compensation matrix, and uses the target signal compensation matrix to perform channel compensation on the receiving pilot signal to obtain the original pilot signal. This approach leverages the powerful nonlinear representation capability of the pre-trained channel compensation matrix update model, has a fast computation speed, and can overcome the shortcomings of the initial channel compensation matrix. This invention can reduce physical layer deployment costs, improve the scalability of signal estimation, reduce computational overhead, improve signal compensation accuracy, significantly reduce receiver bit error rate, and increase communication speed. Furthermore, compared to directly performing model-based channel estimation on the transmitter and receiver signals, the technical solution of this invention adopts an initial channel estimation based on an initial channel estimation algorithm and a secondary channel estimation based on a channel compensation matrix update model. It can also take into account the advantages of the initial channel estimation algorithm in ensuring the robustness and anti-interference capability of the physical layer communication link, and can meet the current communication system's requirements for high speed, low latency, and high reliability.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of a channel compensation method provided according to Embodiment 1 of the present invention;

[0025] Figure 2 This is a flowchart of a channel compensation method provided according to Embodiment 2 of the present invention;

[0026] Figure 3 This is a flowchart of a channel compensation method provided according to Embodiment 2 of the present invention;

[0027] Figure 4A This is a schematic diagram of the transmitting pilot signal provided according to Embodiment 2 of the present invention;

[0028] Figure 4BThis is a schematic diagram of the receiving pilot signal provided according to Embodiment 2 of the present invention;

[0029] Figure 5 This is a schematic diagram of the structure of a channel compensation device according to Embodiment 3 of the present invention;

[0030] Figure 6 This is a schematic diagram of the structure of an electronic device that implements the channel compensation method of this invention. Detailed Implementation

[0031] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0033] Example 1

[0034] Figure 1 This is a flowchart illustrating a channel compensation method provided in Embodiment 1 of the present invention. This embodiment of the invention is applicable to situations involving channel compensation of pilot signals at the receiving end. The method can be executed by a channel compensation device, which can be implemented in hardware and / or software. This channel compensation device can be configured in an electronic device that carries channel estimation functionality.

[0035] See Figure 1 The channel compensation method shown includes:

[0036] S110. Acquire the pilot signal from the receiving end and the pilot signal from the transmitting end.

[0037] Signal transmission involves both transmitting and receiving equipment. The transmitting pilot signal is the pilot signal emitted by the transmitting equipment. The receiving pilot signal is the pilot signal received by the receiving equipment. During signal transmission, interference is inevitable, causing distortion in the receiving pilot signal and making it difficult to guarantee the accuracy of signal transmission.

[0038] Specifically, a signal acquisition device can be used to acquire signals from both the transmitting and receiving devices, obtaining the receiving pilot signal and the transmitting pilot signal.

[0039] In an optional embodiment of the present invention, acquiring the receiver pilot signal and the transmitter pilot signal includes: acquiring the receiver signal and the transmitter signal; performing inverse Fourier transform on the receiver signal and the transmitter signal respectively to obtain a first receiver signal and a first transmitter signal; performing Heisenberg transform on the first receiver signal and the first transmitter signal respectively to obtain a second receiver signal and a second transmitter signal; performing Wigner transform on the second receiver signal and the second transmitter signal respectively to obtain a third receiver signal and a third transmitter signal; and performing sparse fast Fourier transform on the third receiver signal and the third transmitter signal respectively to obtain the receiver signal. The system includes pilot signals and transmitting pilot signals; wherein the transmitting pilot signals include a transmitting reference pilot signal, a transmitting first pilot signal, and a transmitting second pilot signal; the transmitting reference pilot signal is used for channel compensation assessment; after performing channel compensation on the receiving pilot signal using the target signal compensation matrix to obtain the original pilot signal, the system further includes: comparing the original reference pilot signal contained in the original pilot signal with the transmitting reference pilot signal; and processing the original first pilot signal and the original second pilot signal contained in the original pilot signal according to the pilot signal comparison result to obtain the data processing result of the original pilot signal.

[0040] The receiver signal can be the actual signal received by the receiver equipment. The transmitter signal can be the actual signal emitted by the transmitter equipment. Both the receiver and transmitter signals can be time-Doppler domain signals. The first receiver signal can be the inverse Fourier transform result of the receiver signal. The first transmitter signal can be the inverse Fourier transform result of the transmitter signal. Both the first receiver and first transmitter signals can be time-frequency domain signals. The second receiver signal can be the Heisenberg transform result of the first receiver signal. The second transmitter signal can be the Heisenberg transform result of the second transmitter signal. Both the second receiver and second transmitter signals can be time-domain signals. The third receiver signal can be the Wigner transform result of the second receiver signal. The third transmitter signal can be the Wigner transform result of the second transmitter signal. Both the third receiver and third transmitter signals can be time-frequency domain signals. The receiver pilot signal can be the sparse fast Fourier transform result of the third receiver signal. The transmitter pilot signal can be the sparse fast Fourier transform result of the third transmitter signal. Both the receiver and transmitter pilot signals can be time-Doppler domain signals. The transmitting reference pilot signal can be used to evaluate the signal compensation result of the receiving pilot signal. For example, the transmitting reference pilot signal can be the peak signal within the center range of the transmitting pilot signal. The transmitting first pilot signal and the transmitting second pilot signal can be used to characterize the information to be transmitted by the transmitting signal. The transmitting first pilot signal can be the peak signal in a range other than the center range of the transmitting pilot signal. The transmitting second pilot signal can be the valley signal in a range other than the center range of the transmitting pilot signal. The receiving reference pilot signal can be used to characterize the signal interference level of the receiving pilot signal. The receiving first pilot signal and the receiving second pilot signal can be used to characterize the information received by the receiving signal. The receiving first pilot signal can be the peak signal in a range other than the center range of the receiving pilot signal. The receiving second pilot signal can be the valley signal in a range other than the center range of the receiving pilot signal. The original reference pilot signal can be used to evaluate the signal compensation result of the receiving pilot signal. The original first pilot signal and the original second pilot signal can be used to characterize the transmitted information after signal compensation. The original first pilot signal can be a peak signal in a range other than the center range of the original pilot signal. The original second pilot signal can be a valley signal in a range other than the center range of the original pilot signal. The pilot signal comparison result is a comparison between the transmitting reference pilot signal and the original reference pilot signal. For example, the pilot signal comparison result may include a quantity comparison result and an amplitude comparison result between the transmitting reference pilot signal and the original reference pilot signal. The pilot signal comparison result can be used to characterize the signal compensation effect of the receiving pilot signal.This can be understood as follows: the closer the comparison results of the pilot signals, the better the signal compensation effect of the pilot signal at the receiving end; the greater the difference in the comparison results of the pilot signals, the worse the signal compensation effect of the pilot signal at the receiving end. The data processing result can be the signal processing result of the original first pilot signal and the original second pilot signal contained in the original pilot signal.

[0041] Specifically, the system can acquire the transmitted signal from the transmitting device and the received signal from the receiving device. Inverse Fourier transforms can be performed on both the transmitted and received signals to obtain the first received signal and the first transmitted signal. Heisenberg transforms can be performed on both the first received and first transmitted signals to obtain the second received signal and the second transmitted signal. Wigner transforms can be performed on both the second received and second transmitted signals to obtain the third received signal and the third transmitted signal. Sparse Fast Fourier Transforms can be performed on both the third received and third transmitted signals to obtain the received pilot signal and the transmitted pilot signal. After using the target signal compensation matrix to perform channel compensation on the receiver pilot signal to obtain the original pilot signal, the original reference pilot signal contained in the original pilot signal can be compared with the transmitter reference pilot signal. If the comparison result shows that the transmitter reference pilot signal and the original reference pilot signal are close, the original first pilot signal and the original second pilot signal contained in the original pilot signal can be processed to obtain the data processing result of the original pilot signal. If the comparison result shows that the transmitter reference pilot signal and the original reference pilot signal are not close, the original first pilot signal and the original second pilot signal contained in the original pilot signal do not need to be processed.

[0042] This scheme pre-generates transmitting and receiving pilot signals by performing inverse Fourier transform, Heisenberg transform, Wigner transform, and sparse fast Fourier transform on the receiving and transmitting signals, respectively. This achieves the pre-division of reference and transmission information in the transmitting and receiving signals. By comparing the original reference pilot signal contained in the original pilot signal with the transmitting reference pilot signal, the signal compensation effect of the receiving pilot signal is evaluated. Based on the comparison results, the original first pilot signal and the original second pilot signal contained in the original pilot signal are processed to obtain the data processing result of the original pilot signal, further improving the data processing accuracy of the original pilot signal.

[0043] S120. Using the initial channel estimation algorithm, the pilot signals at the receiving end and the pilot signals at the transmitting end are estimated to obtain the initial channel compensation matrix.

[0044] The initial channel estimation algorithm can be used to perform preliminary channel estimation on both the receiver and transmitter pilot signals. For example, the initial channel estimation algorithm can be a reference signal-based channel estimation algorithm, such as LS (Least Squares) or MMSE (Minimum Mean Square Error). The initial channel compensation matrix can be the channel compensation matrix generated by the initial channel estimation algorithm.

[0045] Due to the limitations of the initial channel estimation algorithm, the initial channel compensation matrix suffers from insufficient scalability and high computational overhead, which can no longer meet the current stability requirements of communication systems for high speed, low latency, and high reliability. However, the initial channel estimation algorithm has the advantages of ensuring the robustness and anti-interference capability of the physical layer communication link. This scheme does not directly perform model-based channel estimation on the transmitter and receiver signals, but performs initial channel estimation based on the initial channel estimation algorithm and secondary channel estimation based on the channel compensation matrix update model. This can take into account the advantages of the initial channel estimation algorithm and further improve the shortcomings of the initial channel compensation matrix.

[0046] Specifically, an initial channel estimation algorithm can be used to calculate the initial channel compensation matrix by analyzing the pilot signals at the receiving end and the pilot signals at the transmitting end.

[0047] S130. Based on the pre-trained channel compensation matrix update model, the initial channel compensation matrix is ​​corrected to obtain the target channel compensation matrix.

[0048] The channel compensation matrix update model can be used to correct an initial channel compensation matrix. The input data of the channel compensation matrix update model is the initial channel compensation matrix, and the output result can be the target channel compensation matrix. For example, the channel compensation matrix update model can include a convolutional neural network model, a random forest, a reinforcement learning model, or a width learning model. The target channel compensation matrix can be the corrected result of the initial channel compensation matrix. Compared to the initial channel compensation matrix, the target channel compensation matrix provides better compensation. The target channel compensation matrix can be used to compensate for the pilot signal at the receiving end.

[0049] Specifically, the initial channel compensation matrix can be input into a pre-trained channel compensation matrix, and the initial channel compensation matrix can be corrected to obtain the target channel compensation matrix.

[0050] In an optional embodiment of the present invention, the initial channel compensation matrix is ​​modified based on a pre-trained channel compensation matrix update model to obtain a target channel compensation matrix, including: cleaning and normalizing the initial channel compensation matrix to update the initial channel compensation matrix; and modifying the updated initial channel compensation matrix based on the pre-trained channel compensation matrix update model to obtain the target channel compensation matrix.

[0051] Specifically, the initial channel compensation matrix can be cleaned, such as through noise filtering, missing value imputation, and / or detection based on the Z-score algorithm, before being updated. The updated initial channel compensation matrix can then be normalized, for example, by dividing each element by the maximum value of its elements, thus standardizing the data and controlling the data range to 0-1, before updating the initial channel compensation matrix again. This updated initial channel compensation matrix can then be input into the channel compensation matrix update model to output the target channel compensation matrix.

[0052] This scheme introduces data cleaning of the initial channel compensation matrix, realizes outlier handling of the initial channel compensation matrix, introduces data normalization of the initial channel compensation matrix, realizes data alignment of each element in the initial channel compensation matrix, and further improves the accuracy of the target channel compensation matrix obtained based on the secondary correction of the initial channel compensation matrix, thereby improving the accuracy of channel compensation, and thus improving the accuracy of data transmission between the transmitting and receiving devices.

[0053] S140. Using the target signal compensation matrix, channel compensation is performed on the pilot signal at the receiving end to obtain the original pilot signal.

[0054] The original pilot signal can be the channel compensation result of the receiver pilot signal. This can be understood as follows: during signal transmission, signal interference is inevitable, resulting in signal distortion in the received pilot signal. The original pilot signal can be the pilot signal obtained after recovering the distorted received pilot signal, thus achieving the recovery of the received pilot signal.

[0055] Specifically, the original pilot signal can be obtained by calculating the target signal pilot matrix and the receiving pilot signal.

[0056] In an optional embodiment of the present invention, a target signal compensation matrix is ​​used to perform channel compensation on the receiving pilot signal to obtain the original pilot signal, including: multiplying the target channel compensation matrix with the receiving pilot signal to obtain the original pilot signal, so as to realize the recovery of the receiving pilot signal.

[0057] Specifically, the target channel compensation matrix can be directly multiplied with the receiving pilot signal to obtain the original pilot signal, thereby realizing the recovery of the receiving pilot signal.

[0058] This scheme obtains the original pilot signal by multiplying the target channel compensation matrix with the receiver pilot signal, which simplifies the calculation process of the original pilot signal and improves the recovery efficiency of the original pilot signal.

[0059] The technical solution of this invention acquires the receiving pilot signal and the transmitting pilot signal, uses an initial channel estimation algorithm to estimate the signals of the receiving and transmitting pilot signals, obtains an initial channel compensation matrix, corrects the initial channel compensation matrix based on a pre-trained channel compensation matrix update model to obtain a target channel compensation matrix, and uses the target signal compensation matrix to perform channel compensation on the receiving pilot signal to obtain the original pilot signal. This approach leverages the powerful nonlinear representation capability of the pre-trained channel compensation matrix update model, has a fast computation speed, and can overcome the shortcomings of the initial channel compensation matrix. This invention can reduce physical layer deployment costs, improve the scalability of signal estimation, reduce computational overhead, improve signal compensation accuracy, significantly reduce receiver bit error rate, and increase communication speed. Furthermore, compared to directly performing model-based channel estimation on the transmitter and receiver signals, the technical solution of this invention adopts an initial channel estimation based on an initial channel estimation algorithm and a secondary channel estimation based on a channel compensation matrix update model. It can also take into account the advantages of the initial channel estimation algorithm in ensuring the robustness and anti-interference capability of the physical layer communication link, and can meet the current communication system's requirements for high speed, low latency, and high reliability.

[0060] Example 2

[0061] Figure 2 This is a flowchart of a channel compensation method provided in Embodiment 2 of the present invention. Based on the above embodiments, this embodiment further adds the following steps: "In an offline state, a training sample set is obtained; wherein, the training sample set includes actual channel compensation matrices and initial channel compensation matrix samples; the initial channel compensation matrix samples in the training sample set are input into the channel compensation matrix update model to obtain an estimated channel compensation matrix; based on the difference between the actual channel compensation matrix and the estimated channel compensation matrix, the parameters of the channel compensation matrix update model are adjusted." By pre-training the channel compensation matrix update model offline and using the channel compensation matrix update model to calculate the target channel compensation matrix online, the computational efficiency of the target channel compensation matrix is ​​further improved, which can further improve the data communication efficiency between the transmitting and receiving devices. It should be noted that parts not detailed in this embodiment can be found in the descriptions of other embodiments.

[0062] See Figure 2 The channel compensation method shown includes:

[0063] S210. Obtain the training sample set in offline mode.

[0064] The offline state indicates that the device is in an offline state. This can be understood as the device training the channel compensation matrix update model offline, calculating the target channel compensation matrix online, and performing channel compensation on the received pilot signal. The training sample set can be used to train the channel compensation matrix update model. The training sample set can include actual channel compensation matrices and initial channel compensation matrix samples. The actual channel compensation matrix can serve as the sample label for the channel compensation matrix update model. The initial channel compensation matrix samples can be the initial channel estimation results for the transmitted and received pilot signal samples using an initial channel estimation algorithm. The transmitted pilot signal samples can be the pilot signals of the transmitting device used as training samples. The received pilot signal samples can be the pilot signals of the receiving device used as training samples.

[0065] Specifically, pilot signal samples from both the transmitting and receiving ends can be acquired offline. An initial channel estimation algorithm can be used to estimate the signals from these samples, yielding an initial channel compensation matrix sample. A pre-calibrated actual channel compensation matrix can then be obtained. Both the actual and initial channel compensation matrix samples can be used as a training sample set.

[0066] S220. Input the initial channel compensation matrix samples from the training sample set into the channel compensation matrix update model to obtain the estimated channel compensation matrix.

[0067] The estimated channel compensation matrix can be used to update the channel compensation matrix model based on the initial channel compensation matrix sample prediction obtained from the initial channel compensation matrix.

[0068] Specifically, the initial channel compensation matrix samples from the training sample set can be input into the channel compensation matrix update model to perform channel estimation on the initial channel compensation matrix samples, thereby obtaining the estimated channel compensation matrix.

[0069] S230. Based on the difference between the actual channel compensation matrix and the estimated channel compensation matrix, adjust the parameters of the channel compensation matrix update model.

[0070] Specifically, the loss value can be calculated based on the difference between the actual channel compensation matrix and the estimated channel compensation matrix, such as the mean squared error (MSE) or cross-entropy loss. The parameters of the channel compensation matrix update model can be adjusted with the goal of minimizing the loss value or achieving convergence. Once the channel compensation matrix update model is trained, it can be used to correct the initial channel compensation matrix.

[0071] S240: Acquire the receiving pilot signal and the transmitting pilot signal.

[0072] S250. Using the initial channel estimation algorithm, signal estimation is performed on the pilot signals at the receiving end and the pilot signals at the transmitting end to obtain the initial channel compensation matrix.

[0073] S260. Based on the pre-trained channel compensation matrix update model, the initial channel compensation matrix is ​​corrected to obtain the target channel compensation matrix.

[0074] S270. Using the target signal compensation matrix, channel compensation is performed on the pilot signal at the receiving end to obtain the original pilot signal.

[0075] The technical solution of this invention obtains a training sample set offline, inputs the initial channel compensation matrix samples from the training sample set into the channel compensation matrix update model to obtain an estimated channel compensation matrix, and adjusts the parameters of the channel compensation matrix update model based on the difference between the actual channel compensation matrix and the estimated channel compensation matrix. By pre-training the channel compensation matrix update model offline and using the channel compensation matrix update model to calculate the target channel compensation matrix online, the calculation efficiency of the target channel compensation matrix is ​​further improved, which can further improve the data communication efficiency between the transmitting and receiving devices.

[0076] In an optional embodiment of the present invention, the channel compensation matrix update model is a deep neural network model; after adjusting the parameters of the channel compensation matrix update model based on the difference between the actual channel compensation matrix and the estimated channel compensation matrix, the method further includes: obtaining a validation sample set; wherein the data type of the validation sample set is the same as that of the training sample set; using a K-fold cross-validation algorithm, the parameter-tuned channel compensation matrix is ​​validated based on the validation sample set to obtain at least one cross-validation parameter; comparing each cross-validation parameter, and determining the parameter of the channel compensation matrix update model corresponding to the minimum value of the cross-validation parameter as the optimal parameter of the channel compensation matrix update model.

[0077] Deep neural network models possess powerful nonlinear computational capabilities. A validation sample set can be used to update the parameters of the channel compensation matrix update model. The data type of the validation sample set is the same as that of the training sample set. This can be understood as the validation sample set also including samples of the actual channel compensation matrix and the initial channel compensation matrix. A preset sample ratio can exist between the training and validation sample sets. For example, the preset sample ratio can be 4:1. This allows for the training and validation of the channel compensation matrix update model. The cross-validation parameters can be the single-time validation parameters of the channel compensation matrix update model using a K-fold cross-validation algorithm. The cross-validation parameters can be used to characterize the performance of the channel compensation matrix update model corresponding to a single training iteration. The minimum value of the cross-validation parameters can be used to characterize the optimal parameters of the channel compensation matrix update model.

[0078] Specifically, after adjusting the parameters of the channel compensation matrix update model based on the difference between the actual and estimated channel compensation matrices, a validation sample set can be obtained in the same way as the training sample set. A K-fold cross-validation algorithm can be used to validate the parameters of the channel compensation matrix update model after each individual parameter tuning based on the validation sample set, obtaining the cross-validation parameters corresponding to each validation. The cross-validation parameters are then compared, and the parameters of the channel compensation matrix update model corresponding to the minimum cross-validation parameter value are determined as the optimal parameters of the channel compensation matrix update model.

[0079] This scheme improves the correction efficiency of the channel compensation matrix update model by verifying its parameters during the training process, thereby enhancing the channel compensation accuracy and data transmission accuracy based on the channel compensation matrix update model.

[0080] Figure 3 This is a flowchart of a channel compensation method. Based on the above embodiments, Figure 3 This is a preferred embodiment of the present invention. The solution mainly consists of an offline training module and an online calculation module. For the offline training module, by using traditional channel estimation algorithms such as LS or MMSE, the initial channel compensation matrix can be obtained by calculating the pilot signals at the transmitting and receiving ends. The pilot structure is selected as a block structure. For example... Figure 4A As shown, the transmitting pilot signal may include a transmitting reference pilot signal, a transmitting first pilot signal, and a transmitting second pilot signal. The transmitting reference pilot signal may be located in the center of the data and a guard interval is set to prevent signal interference between the transmitting reference pilot signal, the transmitting first pilot signal, and the transmitting second pilot signal. Figure 4BAs shown, the receiver pilot signal may include a receiver reference pilot signal, a receiver first pilot signal, and a receiver second pilot signal. The receiver reference pilot signal may be located in the center of the data and a guard interval is set to prevent signal interference between the receiver reference pilot signal, the receiver first pilot signal, and the receiver second pilot signal. Figure 4A and Figure 4B As shown, the transmitting pilot signal experienced interference such as Doppler frequency shift, and the receiving pilot signal exhibited varying degrees of distortion, particularly noticeable in the receiving reference pilot signal. The actual channel compensation matrix and the estimated channel compensation matrix obtained through the initial channel estimation algorithm together form the original data sample for training the deep neural network. The sample label is the actual channel compensation matrix. Optionally, the original dataset can be generated through simulation based on the original data sample. The original dataset can be divided in a 4:1 ratio to obtain a training sample set and a validation sample set. Data preprocessing can be performed on the training sample set, such as data cleaning and data normalization. Data cleaning can filter out noise, impute missing values, and detect and process outliers based on the Z-score algorithm. Data normalization can limit the data size to the range of 0-1 by dividing each training sample by the maximum value of its elements to normalize the data. The training sample set can be used to train the deep neural network model. The parameters of the deep neural network model can be validated using a K-fold cross-validation algorithm. When the minimum RMSE (i.e., the minimum value of the cross-validation parameters) is obtained, the optimal solution for the parameters of the deep neural network model is obtained. The deep neural network model can be implemented using Keras (a high-level neural network API) from a third-party library. The model can be configured with three layers, 128, 64, and 16 neurons respectively, a learning rate of 0.001, and RULA (Rapid Upper Limb Assessment) as the parameter optimization algorithm. The deep neural network model iteratively updates its parameters. After training, the model structure and weights are saved as an HDF5 file. For the online calculation module, the initial channel compensation matrix obtained from the initial channel estimation algorithm in the actual scenario can be used as input to the deep neural network model to calculate the target channel compensation matrix. Multiplying this target channel compensation matrix sequentially with the received pilot signal recovers the original pilot signal, thus reducing the bit error rate.

[0081] This invention primarily addresses the problems of poor scalability, high computational overhead, and inability to meet the requirements of current wireless communication systems in traditional channel estimation algorithms. It leverages the advantages of deep neural network models, such as their powerful nonlinear representation capabilities and fast computation speed, and applies them to channel estimation and equalization techniques. This reduces physical layer deployment costs, improves signal compensation accuracy, significantly reduces receiver bit error rate, and increases communication speed.

[0082] Example 3

[0083] Figure 5 This is a schematic diagram of a channel compensation device provided in Embodiment 3 of the present invention. This embodiment of the invention is applicable to situations where channel compensation is performed on the pilot signal at the receiving end. The device can execute a channel compensation method and can be implemented in hardware and / or software. The device can be configured in an electronic device that carries a channel estimation function.

[0084] See Figure 5 The channel compensation device shown includes: a pilot signal acquisition module 510, an initial channel estimation module 520, a compensation matrix update module 530, and a channel compensation module 540. Specifically, the pilot signal acquisition module 510 acquires the receiving pilot signal and the transmitting pilot signal; the initial channel estimation module 520 uses an initial channel estimation algorithm to estimate the signal of the receiving pilot signal and the transmitting pilot signal to obtain an initial channel compensation matrix; the compensation matrix update module 530 corrects the initial channel compensation matrix based on a pre-trained channel compensation matrix update model to obtain a target channel compensation matrix; and the channel compensation module 540 uses the target signal compensation matrix to perform channel compensation on the receiving pilot signal to obtain the original pilot signal.

[0085] The technical solution of this invention acquires the receiving pilot signal and the transmitting pilot signal, uses an initial channel estimation algorithm to estimate the signal of the receiving pilot signal and the transmitting pilot signal to obtain an initial channel compensation matrix, and corrects the initial channel compensation matrix based on a pre-trained channel compensation matrix update model to obtain a target channel compensation matrix. The target channel compensation matrix is ​​then used to perform channel compensation on the receiving pilot signal to obtain the original pilot signal. This method utilizes the powerful nonlinear representation capability of the pre-trained channel compensation matrix update model, resulting in fast computation speed. It can overcome the shortcomings of the initial channel compensation matrix, reduce physical layer deployment costs, improve the scalability of signal estimation, reduce computational overhead, improve signal compensation accuracy, significantly reduce the receiving bit error rate, and increase communication speed. Furthermore, compared to directly performing model-based channel estimation on the transmitting and receiving signals, the technical solution of this invention uses an initial channel estimation based on the initial channel estimation algorithm and a secondary channel estimation based on the channel compensation matrix update model. This also takes into account the advantages of the initial channel estimation algorithm in ensuring the robustness and anti-interference capability of the physical layer communication link, and can meet the current communication system's requirements for high speed, low latency, and high reliability.

[0086] In an optional embodiment of the present invention, the apparatus further includes: a training sample set acquisition module, configured to acquire a training sample set in an offline state; wherein the training sample set includes an actual channel compensation matrix and initial channel compensation matrix samples; an estimated compensation matrix generation module, configured to input the initial channel compensation matrix samples from the training sample set into a channel compensation matrix update model to obtain an estimated channel compensation matrix; and a model parameter tuning module, configured to adjust the parameters of the channel compensation matrix update model based on the difference between the actual channel compensation matrix and the estimated channel compensation matrix.

[0087] In an optional embodiment of the present invention, the channel compensation matrix update model is a deep neural network model; the device further includes: a verification sample set acquisition module, used to acquire a verification sample set after adjusting the parameters of the channel compensation matrix update model based on the difference between the actual channel compensation matrix and the estimated channel compensation matrix; wherein the data type of the verification sample set is the same as that of the training sample set; a model parameter verification module, used to verify the parameter-tuned channel compensation matrix based on the verification sample set using a K-fold cross-validation algorithm to obtain at least one cross-validation parameter; and a model training completion module, used to compare each of the cross-validation parameters and determine the parameter of the channel compensation matrix update model corresponding to the minimum value of the cross-validation parameter as the optimal parameter of the channel compensation matrix update model.

[0088] In an optional embodiment of the present invention, the pilot signal acquisition module 510 includes: acquiring a receiver signal and a transmitter signal; a first signal conversion unit, configured to perform inverse Fourier transform on the receiver signal and the transmitter signal respectively to obtain a first receiver signal and a first transmitter signal; a second signal conversion unit, configured to perform Heisenberg transform on the first receiver signal and the first transmitter signal respectively to obtain a second receiver signal and a second transmitter signal; a third signal conversion unit, configured to perform Wigner transform on the second receiver signal and the second transmitter signal respectively to obtain a third receiver signal and a third transmitter signal; and a fourth signal conversion unit, configured to perform sparse fast transform on the third receiver signal and the third transmitter signal respectively. Fourier transform is performed to obtain the receiver pilot signal and the transmitter pilot signal; wherein, the transmitter pilot signal includes a transmitter reference pilot signal, a transmitter first pilot signal, and a transmitter second pilot signal; the transmitter reference pilot signal is used for channel compensation evaluation; the device further includes: a pilot signal comparison module, used to compare the original reference pilot signal contained in the original pilot signal with the transmitter reference pilot signal after channel compensation is performed on the receiver pilot signal using the target signal compensation matrix to obtain the original pilot signal; a data processing module, used to process the original first pilot signal and the original second pilot signal contained in the original pilot signal according to the pilot signal comparison result to obtain the data processing result of the original pilot signal.

[0089] In an optional embodiment of the present invention, the channel compensation module 540 includes: a channel compensation unit, used to multiply the target channel compensation matrix with the receiving pilot signal to obtain the original pilot signal, so as to realize the recovery of the receiving pilot signal.

[0090] In an optional embodiment of the present invention, the compensation matrix update module 530 includes: a data preprocessing unit, used to perform data cleaning and data normalization on the initial channel compensation matrix to update the initial channel compensation matrix; and a compensation matrix update unit, used to correct the updated initial channel compensation matrix based on a pre-trained channel compensation matrix update model to obtain a target channel compensation matrix.

[0091] The channel compensation device provided in the embodiments of the present invention can execute the channel compensation method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.

[0092] In the technical solutions of this invention, the acquisition, storage, and application of receiver pilot signals, transmitter pilot signals, training sample sets, actual channel compensation matrices, initial channel compensation matrix samples, receiver signals, and transmitter signals all comply with relevant laws and regulations and do not violate public order and good morals.

[0093] Example 4

[0094] Figure 6 A schematic diagram of an electronic device 600 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0095] like Figure 6 As shown, the electronic device 600 includes at least one processor 601 and a memory, such as a read-only memory (ROM) 602 or a random access memory (RAM) 603, communicatively connected to the at least one processor 601. The memory stores computer programs executable by the at least one processor. The processor 601 can perform various appropriate actions and processes based on the computer program stored in the ROM 602 or loaded into the RAM 603 from storage unit 608. The RAM 603 may also store various programs and data required for the operation of the electronic device 600. The processor 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0096] Multiple components in electronic device 600 are connected to I / O interface 605, including: input unit 606, such as keyboard, mouse, etc.; output unit 607, such as various types of displays, speakers, etc.; storage unit 608, such as disk, optical disk, etc.; and communication unit 609, such as network card, modem, wireless transceiver, etc. Communication unit 609 allows electronic device 600 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0097] Processor 601 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 601 performs the various methods and processes described above, such as channel compensation methods.

[0098] In some embodiments, the channel compensation method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 608. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by processor 601, one or more steps of the channel compensation method described above may be performed. Alternatively, in other embodiments, processor 601 may be configured to perform the channel compensation method by any other suitable means (e.g., by means of firmware).

[0099] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0100] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0101] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0102] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0103] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0104] A computing system can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system. It addresses the shortcomings of traditional physical hosts and VPS (Virtual Private Server) services, such as high management difficulty and weak business scalability.

[0105] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0106] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A channel compensation method, characterized in that, The method includes: Acquiring receiver pilot signals and transmitter pilot signals includes: acquiring receiver signals and transmitter signals; performing inverse Fourier transforms on the receiver signals and transmitter signals respectively to obtain a first receiver signal and a first transmitter signal; performing Heisenberg transforms on the first receiver signals and the first transmitter signals respectively to obtain a second receiver signal and a second transmitter signal; performing Wigner transforms on the second receiver signals and the second transmitter signals respectively to obtain a third receiver signal and a third transmitter signal; performing sparse fast Fourier transforms on the third receiver signals and the third transmitter signals respectively to obtain receiver pilot signals and transmitter pilot signals; wherein, the transmitter pilot signals include a transmitter reference pilot signal, a transmitter first pilot signal, and a transmitter second pilot signal; the transmitter reference pilot signal is used for channel compensation assessment. An initial channel estimation algorithm is used to estimate the signal of the receiving pilot signal and the transmitting pilot signal to obtain the initial channel compensation matrix. The initial channel compensation matrix is ​​cleaned and normalized to update it. The data cleaning includes noise filtering, missing value imputation, and outlier handling based on the Z-score algorithm. The data normalization includes dividing each element in the initial channel compensation matrix by the maximum value of the matrix elements to control the data range between 0 and 1. Based on the pre-trained channel compensation matrix update model, the updated initial channel compensation matrix is ​​corrected to obtain the target channel compensation matrix; The target channel compensation matrix is ​​multiplied by the receiving pilot signal to obtain the original pilot signal, thereby realizing the recovery of the receiving pilot signal; The original reference pilot signal contained in the original pilot signal is compared with the transmitting end reference pilot signal; Based on the comparison results of the pilot signals, the original first pilot signal and the original second pilot signal contained in the original pilot signal are processed to obtain the data processing result of the original pilot signal.

2. The method according to claim 1, characterized in that, Also includes: In offline mode, a training sample set is acquired; wherein, the training sample set includes actual channel compensation matrix and initial channel compensation matrix samples; The initial channel compensation matrix samples in the training sample set are input into the channel compensation matrix update model to obtain the estimated channel compensation matrix; Based on the difference between the actual channel compensation matrix and the estimated channel compensation matrix, the parameters of the channel compensation matrix update model are adjusted.

3. The method according to claim 2, characterized in that, The channel compensation matrix update model is a deep neural network model; After adjusting the parameters of the channel compensation matrix update model based on the difference between the actual channel compensation matrix and the estimated channel compensation matrix, the method further includes: Obtain a verification sample set; wherein the data type of the verification sample set is the same as that of the training sample set; The K-fold cross-validation algorithm is used to verify the parameter-tuned channel compensation matrix based on the validation sample set, and at least one cross-validation parameter is obtained. The cross-validation parameters are compared, and the parameters of the channel compensation matrix update model corresponding to the minimum value of the cross-validation parameters are determined as the optimal parameters of the channel compensation matrix update model.

4. A channel compensation device, characterized in that, The device includes: Pilot signal acquisition module, used to acquire the pilot signal from the receiving end and the pilot signal from the transmitting end; The pilot signal acquisition module includes: a first signal conversion unit, a second signal conversion unit, a third signal conversion unit, and a fourth signal conversion unit; The first signal conversion unit is used to acquire the receiving end signal and the transmitting end signal; and to perform inverse Fourier transform on the receiving end signal and the transmitting end signal respectively to obtain the first receiving end signal and the first transmitting end signal. The second signal conversion unit is used to perform Heisenberg transformation on the first receiving end signal and the first transmitting end signal respectively to obtain the second receiving end signal and the second transmitting end signal; The third signal conversion unit is used to perform Wigner transformation on the second receiving end signal and the second transmitting end signal respectively to obtain the third receiving end signal and the third transmitting end signal; The fourth signal conversion unit is used to perform sparse fast Fourier transform on the third receiving end signal and the third transmitting end signal respectively to obtain the receiving end pilot signal and the transmitting end pilot signal; wherein, the transmitting end pilot signal includes a transmitting end reference pilot signal, a transmitting end first pilot signal and a transmitting end second pilot signal; the transmitting end reference pilot signal is used for channel compensation evaluation; The initial channel estimation module is used to perform signal estimation on the receiving pilot signal and the transmitting pilot signal using an initial channel estimation algorithm to obtain an initial channel compensation matrix. The compensation matrix update module includes: a data preprocessing unit and a compensation matrix update unit; The data preprocessing unit is used to perform data cleaning and data normalization on the initial channel compensation matrix and update the initial channel compensation matrix; wherein, the data cleaning includes noise filtering, missing value imputation and outlier handling based on the Z-score algorithm, and the data normalization includes dividing each element in the initial channel compensation matrix by the maximum value of the matrix elements to control the data range between 0 and 1. The compensation matrix update unit is used to correct the updated initial channel compensation matrix based on a pre-trained channel compensation matrix update model to obtain the target channel compensation matrix. The channel compensation module includes: a channel compensation unit; The channel compensation unit is used to multiply the target channel compensation matrix with the receiving pilot signal to obtain the original pilot signal, so as to realize the recovery of the receiving pilot signal; The pilot signal comparison module is used to compare the original reference pilot signal contained in the original pilot signal with the transmitter reference pilot signal; The data processing module is used to process the original first pilot signal and the original second pilot signal contained in the original pilot signal according to the pilot signal comparison result, so as to obtain the data processing result of the original pilot signal.

5. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the channel compensation method according to any one of claims 1-3.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the channel compensation method according to any one of claims 1-3.

7. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the channel compensation method according to any one of claims 1-3.

Images