Impulse interference suppression method and related device

Abstract

The application provides a pulse interference suppression method and related equipment, and relates to the technical field of wireless communication, and the method comprises the following steps: detecting pulse interference points by performing pulse interference detection on a to-be-processed signal sequence based on a Myriad filter; and performing local smoothing repair on the pulse interference points based on an improved local weighted regression algorithm to obtain a smoothed signal sequence; the improved local weighted regression algorithm is an improved robust repair weight diagonal matrix based on a probability density function of a fitted Gaussian distribution; and the probability density function of the fitted Gaussian distribution is determined according to the amplitude of non-pulse interference points in the to-be-processed signal sequence. The application reduces the calculation complexity and improves the noise reduction efficiency through pulse interference detection and local smoothing repair; and the robust repair weight diagonal matrix is introduced based on the LOESS method, so that the noise reduction effect is improved.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to a pulse interference suppression method and related equipment. Background Technology

[0002] One of the significant challenges facing wireless communication systems is noise and interference during transmission, which can severely impact their performance. Impulse interference, in particular, with its long duration and high incidence, significantly degrades the signal quality at the receiver, leading to a substantial decrease in system performance. Therefore, applying impulse interference suppression techniques at both the transmitting and receiving ends is essential.

[0003] Among related technologies, the Robust Locally Regression Smoothing (RLRS) algorithm not only retains the good denoising effect of the Locally Weighted Regression (LOESS) algorithm on Gaussian background noise, but also improves the robustness to impulse interference by robustly repairing the weight matrix, and has a good denoising effect on both random impulse interference and background noise.

[0004] However, since the RLRS method requires recalculating the weighted least squares solution of the fitting parameters after each update of the robust repair coefficients of the fitting weights, its matrix inversion operation consumes a lot of computational resources and has very low noise reduction efficiency.

[0005] Therefore, it is necessary to provide a pulse interference suppression method that has good noise reduction effect and high noise reduction efficiency against random pulse interference. Summary of the Invention

[0006] This invention provides a pulse interference suppression method and related equipment to solve the problem that noise reduction efficiency and noise reduction effect cannot be satisfied at the same time in the prior art, and achieves good noise reduction effect and high noise reduction efficiency.

[0007] This invention provides a pulse interference suppression method, comprising:

[0008] Pulse interference detection is performed on the signal sequence to be processed based on the Myriad filter to determine the pulse interference points;

[0009] The improved local weighted regression algorithm is used to perform local smoothing repair on the pulse interference points to obtain a smooth signal sequence. The improved local weighted regression algorithm is based on the local weighted regression algorithm, and introduces a robust repair weight diagonal matrix constructed based on the probability density function of the fitted Gaussian distribution. The probability density function of the fitted Gaussian distribution is determined according to the amplitude of the non-pulse interference points in the signal sequence to be processed.

[0010] In some embodiments, the step of detecting impulse interference in the signal sequence to be processed based on the Myriad filter and determining the impulse interference point includes:

[0011] The signal sequence to be processed is input into the Myriad filter to obtain the loss parameter that minimizes the loss function of the Myriad filter, and the loss parameter is used as the impulse interference detection threshold.

[0012] Signal points in the signal sequence to be processed whose amplitude is greater than the pulse interference detection threshold are taken as pulse interference points.

[0013] In some embodiments, inputting the signal sequence to be processed into the Myriad filter to obtain the loss parameters that minimize the loss function of the Myriad filter includes:

[0014] The loss parameter is initialized to the amplitude of any signal point in the signal sequence to be processed, and the loss function value corresponding to the amplitude of each signal point is determined.

[0015] Sort the loss function values ​​corresponding to the amplitudes of all signal points in the signal sequence to be processed in ascending order, and determine the amplitudes of the two signal points corresponding to the first two loss function values.

[0016] The loss parameter that minimizes the loss function is obtained by searching between the amplitudes of the two signal points using the binary search method.

[0017] In some embodiments, the step of performing local smoothing repair on the impulse interference points based on the improved local weighted regression algorithm to obtain a smooth signal sequence includes:

[0018] Traverse the locations of all pulse interference points to obtain the smoothed signal sequence; at each pulse interference point location, execute:

[0019] The midpoint of the sliding window is taken as the location of the pulse interference point;

[0020] The signal points within the sliding window are weighted and regressed based on an improved local weighted regression algorithm to obtain smooth signal points; each signal point within the sliding window corresponds to one of the smooth signal points.

[0021] Replace the pulse interference points within the sliding window with the corresponding smooth signal points.

[0022] In some embodiments, the weighted regression of signal points within the sliding window based on the improved local weighted regression algorithm to obtain smooth signal points includes:

[0023] Determine the regression coefficient matrix;

[0024] The signal points within the sliding window are weighted and regressed based on the regression coefficient matrix to obtain smooth signal points.

[0025] In some embodiments, determining the regression coefficient matrix includes:

[0026] Based on the signal points within the sliding window, determine the position parameter matrix, the local data matrix, and the fitting weight diagonal matrix;

[0027] The regression coefficient matrix is ​​determined based on the location parameter matrix, the local data matrix, the fitting weight diagonal matrix, and the robust repair weight diagonal matrix.

[0028] The present invention also provides a pulse interference suppression device, comprising:

[0029] The detection module is used to detect pulse interference in the signal sequence to be processed based on the Myriad filter and to determine the pulse interference points.

[0030] The repair module is used to perform local smoothing repair on the pulse interference points based on an improved local weighted regression algorithm to obtain a smooth signal sequence. The improved local weighted regression algorithm is based on the local weighted regression algorithm, and introduces a robust repair weight diagonal matrix constructed based on the probability density function of the fitted Gaussian distribution. The probability density function of the fitted Gaussian distribution is determined according to the amplitude of the non-pulse interference points in the signal sequence to be processed.

[0031] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the impulse interference suppression methods described above.

[0032] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the impulse interference suppression method as described above.

[0033] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements any of the impulse interference suppression methods described above.

[0034] The impulse interference suppression method and related equipment provided by this invention replace the repair process of adjusting the weight matrix multiple times in the RLRS method with impulse interference detection and local smoothing repair, which reduces the computational complexity and improves the noise reduction efficiency. Considering the amplitude distribution of signal and noise as well as the impact of impulse interference on the signal sequence to be processed, the amplitude distribution of signal points in the signal sequence to be processed after removing impulse interference is fitted to a Gaussian distribution, thereby calculating a robust repair weight diagonal matrix and improving the noise reduction effect. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0036] Figure 1 This is one of the flowcharts illustrating a pulse interference suppression method provided in an exemplary embodiment of the present invention;

[0037] Figure 2 This is a second schematic flowchart of a pulse interference suppression method provided in an exemplary embodiment of the present invention;

[0038] Figure 3 This is a schematic diagram of pulse interference point detection provided by the present invention;

[0039] Figure 4 This is the third flowchart illustrating the pulse interference suppression method provided in an exemplary embodiment of the present invention;

[0040] Figure 5 This is the fourth flowchart illustrating an exemplary embodiment of the pulse interference suppression method provided by the present invention;

[0041] Figure 6 This invention provides a schematic diagram of the pulse interference time-domain amplitude when the signal-to-interference ratio is -20dB.

[0042] Figure 7 A comparison chart of bit error rate performance of different methods when the signal-to-interference ratio is -20dB, provided by this invention;

[0043] Figure 8 This invention provides a schematic diagram of the pulse interference time-domain amplitude when the signal-to-interference ratio is -30dB.

[0044] Figure 9 A comparison chart of bit error rate performance of different methods when the signal-to-interference ratio is -30dB, provided by this invention;

[0045] Figure 10 A comparison of the bit error rate performance of the impulse interference suppression method provided by this invention under different impulse interference conditions;

[0046] Figure 11 This is a schematic diagram of the pulse interference suppression device provided by the present invention;

[0047] Figure 12 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. It should be noted that, unless otherwise specified, the embodiments and features of the embodiments of this invention can be combined with each other. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0050] It should be further noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0051] In this invention, "at least one" means one or more, and "more than one" means two or more. The terms "first," "second," "third," "fourth," etc. (if present) in this invention are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0052] In embodiments of the present invention, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in embodiments of the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0053] One of the significant challenges facing wireless communication systems is noise and interference during transmission, which can severely impact their performance. In longwave communication, the primary source of interference is atmospheric interference, mainly caused by lightning. In power line communication, there is man-made interference from ignition systems, heavy power lines, and current switches. These types of interference are typically characterized by high amplitude, wide bandwidth, and rapid onset, leading to signal quality degradation and reduced transmission capacity. Since this type of interference does not follow the characteristics of Gaussian noise, it requires pulse modeling for characterization; hence, it is also known as pulse interference.

[0054] The long duration and high incidence of impulse interference severely degrade signal quality at the receiver, leading to a significant reduction in the performance of the communication system. Therefore, it is essential to apply impulse interference suppression techniques at both the transmitting and receiving ends to suppress impulse interference.

[0055] Traditional impulse interference suppression methods can be divided into three categories: The first category consists of nonlinear preprocessing methods such as amplitude limiting and blanking. These methods are simple to implement in both hardware and software, but their performance is limited, and they are difficult to adjust in real-time according to changes in channel and noise characteristics. The second category consists of filtering methods such as robust local regression smoothing and median filtering. These methods evolved from low-pass linear filtering methods used to filter background noise, such as local weighted regression and mean filtering. They are effective at suppressing impulse interference that deviates from the normal data sequence, but they are prone to filtering distortion when the data changes rapidly. The third category consists of sparse reconstruction algorithms, such as compressed sensing. These algorithms utilize the sparse characteristics of impulse interference in the time domain to reconstruct and eliminate the impulse interference at the receiver. These methods can achieve good noise reduction, but they require a high degree of noise sparsity.

[0056] With the development of technology, the LOESS algorithm emerged. LOESS is a nonparametric method for local regression analysis. It can smooth noisy data samples by fitting a curve that conforms to the overall trend, thereby achieving denoising. In wireless communication systems, the LOESS algorithm can be applied to denoising Gaussian white noise. However, the presence of impulse interference causes smoothing distortion at the location of impulse interference, making it impossible to recover the original data.

[0057] To enhance the robustness of the LOESS algorithm against impulse interference, the RLRS algorithm was developed by introducing a robust repair weight matrix related to the smoothed residuals. The RLRS algorithm is a filtering method based on outlier detection and iterative weighted correction. By iteratively adjusting the weight matrix, it improves robustness to random impulse noise, exhibiting overall nonlinear filtering characteristics.

[0058] The RLRS algorithm retains the good denoising effect of the LOESS algorithm on Gaussian background noise, and improves its robustness to impulse interference by robustly repairing the weight matrix, thus exhibiting good denoising effects on both random impulse interference and background noise. However, because the RLRS method needs to recalculate the weighted least squares solution of the fitting parameters after each update of the fitted weights and robust repair coefficients, its matrix inversion operation consumes a lot of computational resources, resulting in very low denoising efficiency.

[0059] Therefore, it is necessary to provide a pulse interference suppression method that has good noise reduction effect and high noise reduction efficiency against random pulse interference.

[0060] Figure 1 This is one of the flowcharts illustrating an exemplary embodiment of the pulse interference suppression method provided by the present invention, such as... Figure 1 As shown, this embodiment provides a pulse interference suppression method, which includes:

[0061] Step 110: Based on the Myriad filter, pulse interference detection is performed on the signal sequence to be processed to determine the pulse interference points.

[0062] Specifically, the signal sequence to be processed is the signal sequence received by the receiver that contains background noise and impulse interference. The signal sequence to be processed is denoted as Y = {y...} l |l=1,2,…L}, where y l L represents any signal point in the signal sequence to be processed, and L is the signal length of the signal sequence to be processed, i.e., the number of signal points contained in the signal sequence to be processed.

[0063] The Myriad filter is a robust nonlinear filter developed based on the mean and median filters. It is highly sensitive to outliers and can effectively suppress impulse interference. Utilizing the robustness of the Myriad filter to impulse interference, it is applied to impulse interference detection to identify impulse interference points in the signal sequence Y to be processed. The total number of impulse interference points is denoted as H.

[0064] Step 120: Local smoothing repair of impulse interference points is performed based on the improved local weighted regression algorithm to obtain a smooth signal sequence; the improved local weighted regression algorithm is based on the local weighted regression algorithm, and introduces a robust repair weight diagonal matrix constructed based on the probability density function of the fitted Gaussian distribution; the probability density function of the fitted Gaussian distribution is determined according to the amplitude of non-impulse interference points in the signal sequence to be processed.

[0065] Specifically, after identifying the pulse interference points in the signal sequence Y to be processed, the pulse interference points are removed from the signal sequence Y to obtain a signal sequence that does not contain pulse interference points. Since there are H pulse interference points and the signal length of the signal sequence Y to be processed is L, then the signal sequence... Represents signal sequence Any signal point in the middle.

[0066] Due to signal sequence The signal sequence contains no pulse interference points, therefore... The amplitude distribution of the LH signal points can be fitted using a Gaussian distribution. For the signal sequence... The mean and variance of the amplitudes of LH signal points are calculated to determine the parameters of the probability density function of the fitted Gaussian distribution, thus obtaining the probability density function of the fitted Gaussian distribution.

[0067] The expression for the probability density function is shown below:

[0068]

[0069] In the formula, f PDF () represents the probability density function, and σ represents the signal sequence. The variance corresponding to the amplitude of the signal point, μ represents the signal sequence. The mean value corresponding to the amplitude of the signal point, where x represents the input quantity of the probability density function.

[0070] The diagonal elements of the robust repair weight diagonal matrix are the robust repair weights δ corresponding to the signal points. n That is, the probability density function f that fits a Gaussian distribution. PDF The probability density value at that signal point.

[0071] Robust Repair Weight δ n The expression is as follows:

[0072]

[0073] In the formula, δ n Indicates that in |y n The probability density value at |y| n | represents the absolute value of the signal point's amplitude, and σ represents the signal sequence. The variance corresponding to the amplitude of the signal point, μ represents the signal sequence. The mean value corresponding to the amplitude of the signal point.

[0074] The improved LOESS algorithm introduces a robust repair weighted diagonal matrix based on the LOESS method. After identifying the impulse interference points, the improved LOESS algorithm is used to perform local smoothing repair on the impulse interference points to obtain a smooth signal sequence.

[0075] The impulse interference suppression method provided in this embodiment replaces the repair process of adjusting the weight matrix multiple times in the RLRS method with impulse interference detection and local smoothing repair, which reduces computational complexity and improves noise reduction efficiency. Considering the amplitude distribution of signal and noise as well as the impact of impulse interference on the signal sequence to be processed, the amplitude distribution of signal points in the signal sequence to be processed after removing impulse interference is fitted to a Gaussian distribution, thereby calculating a robust repair weight diagonal matrix and improving the noise reduction effect.

[0076] Please refer to Figure 2 , Figure 2 This is a second schematic flowchart of an exemplary embodiment of the present invention for pulse interference suppression. This embodiment further illustrates the foregoing embodiment, mainly describing the specific process of detecting pulse interference in the signal sequence to be processed based on a Myriad filter and determining the pulse interference points. Figure 2 As shown, the pulse interference suppression method provided in this embodiment includes:

[0077] Step 210: Input the signal sequence to be processed into the Myriad filter, obtain the loss parameter that minimizes the loss function of the Myriad filter, and use the loss parameter as the impulse interference detection threshold.

[0078] Specifically, the signal sequence to be processed is denoted as Y = {y l |l=1,2,…L}, where y l L represents any signal point in the signal sequence to be processed, and L is the signal length of the signal sequence to be processed.

[0079] The expression for the loss function of the Myriad filter is as follows:

[0080]

[0081] In the formula, D sum Let β represent the loss function, β represent the loss parameter, and K represent the scaling parameter, which is the maximum value of the signal amplitude in the signal sequence Y to be processed, i.e., K = max|Y|, |y| i | represents the amplitude of any signal point in the signal sequence Y to be processed, and L is the signal length of the signal sequence Y to be processed.

[0082] The signal sequence Y to be processed is input into the Myriad filter, that is, the signal points in the signal sequence Y are substituted into the loss function to determine the loss parameter β. * loss parameter β * It is the loss parameter that minimizes the loss function. Let the loss parameter β... * The threshold τ is used for pulse interference detection.

[0083] Loss parameter β * The expression is as follows:

[0084]

[0085] In the formula, β * Let β represent the loss parameter that minimizes the loss function, K represent the scaling parameter, and |y| represent the loss parameter. i | represents the amplitude of any signal point in the signal sequence Y to be processed, and L is the signal length of the signal sequence Y to be processed.

[0086] Step 220: Select signal points in the signal sequence to be processed whose amplitude is greater than the pulse interference detection threshold as pulse interference points.

[0087] Specifically, the amplitude of the signal point in the signal sequence Y to be processed is compared with the pulse interference detection threshold τ to determine whether the amplitude of the signal point in the signal sequence Y to be processed is greater than the pulse interference detection threshold τ.

[0088] Figure 3 This is a schematic diagram of pulse interference point detection provided by the present invention, as shown below. Figure 3 As shown, if the amplitude of a signal point in the signal sequence Y to be processed is less than or equal to the impulse interference detection threshold τ, the signal point is considered a non-impulse interference point. If the amplitude of a signal point in the signal sequence Y to be processed is greater than the impulse interference detection threshold τ, the signal point is considered an impulse interference point, and the location of the impulse interference point is marked.

[0089] Let the set of locations of the pulse interference points be denoted as U = {u h |h=1,2,…H}, where U represents the set of locations of all pulse interference points, u h This indicates the location of the h-th pulse interference point, and H represents the total number of pulse interference points.

[0090] The impulse interference suppression method provided in this embodiment improves the probability of correct detection and keeps the probability of false alarms at a low level by using the loss parameter that minimizes the loss function as the impulse interference detection threshold and using signal points in the signal sequence to be processed with amplitudes greater than the impulse interference detection threshold as impulse interference points.

[0091] Please refer to Figure 4 , Figure 4 This is the third flowchart illustrating an exemplary embodiment of the pulse interference suppression method provided by the present invention. This embodiment further explains the foregoing embodiment, mainly illustrating: inputting the signal sequence to be processed into the Myriad filter to obtain the loss parameter that minimizes the loss function of the Myriad filter. For example... Figure 4 As shown, the pulse interference suppression method provided in this embodiment includes:

[0092] Step 410: Initialize the loss parameter to the amplitude of any signal point in the signal sequence to be processed, and determine the loss function value corresponding to the amplitude of each signal point.

[0093] Step 420: Sort the loss function values ​​corresponding to the amplitudes of all signal points in the signal sequence to be processed in ascending order, and determine the amplitudes of the two signal points corresponding to the first two loss function values.

[0094] Step 430: Search between the amplitudes of the two signal points using the bisection method to obtain the loss parameter that minimizes the loss function.

[0095] Specifically, the loss parameter is initialized to the amplitude of any signal point in the signal sequence to be processed, i.e., the loss parameter β = |y l |,y l For any ∈Y, l=1,2,…,L, the loss function becomes:

[0096]

[0097] In the formula, D sum Let β represent the loss function, K represent the scaling parameter, and σk represent the maximum amplitude of the signal in the signal sequence Y to be processed. i | represents the amplitude of any signal point in the signal sequence Y to be processed, |y l | represents the amplitude of any signal point in the signal sequence Y to be processed, and L is the signal length of the signal sequence Y to be processed.

[0098] Calculate the amplitude |y| at each signal point l |Corresponding loss function value D sum The loss function value D corresponding to the amplitude of all signal points in the signal sequence Y to be processed is... sum Sort the signals in ascending order, determine the amplitudes of the two signal points corresponding to the first two loss function values, and denot these two signal point amplitudes as β. m1 and β m2 In β m1 and β m2 The loss function value D is obtained by using a binary search method. sum Minimum β * .

[0099] The impulse interference suppression method provided in this invention first locks the range of the loss parameter to the amplitude of the signal points in the signal sequence to be processed, then narrows the range of the loss parameter to the amplitude between two signal points by using the loss function value corresponding to the amplitude of each signal point, and finally searches for the loss parameter that minimizes the loss function between the amplitudes of the two signal points using a bisection method. In this way, the range of the loss parameter is gradually narrowed, thereby achieving accurate acquisition of the loss parameter β*.

[0100] Please refer to Figure 5 , Figure 5 This is the fourth flowchart illustrating an exemplary embodiment of the pulse interference suppression method provided by the present invention. This embodiment further explains the foregoing embodiments, mainly illustrating: local smoothing repair of pulse interference points based on an improved local weighted regression algorithm to obtain a smooth signal sequence. For example... Figure 5 As shown, the pulse interference suppression method provided in this embodiment includes:

[0101] Step 510: Use the location of the pulse interference point as the midpoint of the sliding window.

[0102] Specifically, the sliding window contains N signal points, and the midpoint of the sliding window is the location u of the pulse interference point. h Optionally, N is an odd number so that the number of signal points on both sides of the pulse interference point is the same.

[0103] Let the signal sequence within the sliding window be denoted as... The position corresponding to the signal point within the sliding window is denoted as

[0104] Step 520: Perform weighted regression on the signal points within the sliding window based on the improved LOESS algorithm to obtain smooth signal points.

[0105] Specifically, a weighted regression is performed on the signal points within the sliding window based on the improved LOESS algorithm. This involves adding different regression coefficients to the signal points within the sliding window to reduce the amplitude of impulse interference points and increase the amplitude of non-impulse interference points, thereby obtaining smooth signal points and achieving smooth repair of the signal points within the sliding window. Each signal point within the sliding window corresponds to a smoothed signal point.

[0106] Step 530: Replace the pulse interference points in the sliding window with the corresponding smooth signal points.

[0107] Specifically, after obtaining the smoothed signal points, the pulse interference points within the sliding window are replaced with the corresponding smoothed signal points, that is, the signal sequence within the sliding window is... midpoint y (N+1) / 2 Replace with a sequence Y consisting of smooth signal points GLOESS The midpoint.

[0108] Step 540: Determine whether to traverse the locations of all pulse interference points.

[0109] Specifically, it determines whether the positions of all pulse interference points have been traversed, that is, whether the index h of the current pulse interference point is equal to H. If the index h of the current pulse interference point is not equal to H, it is determined that the positions of all pulse interference points have not been traversed, and step 550 is executed; if the index h of the current pulse interference point is equal to H, it is determined that the positions of all pulse interference points have been traversed, and step 560 is executed.

[0110] Step 550: Slide the sliding window to the location of the next pulse interference point.

[0111] Specifically, after the sliding window slides to the location of the next pulse interference point, steps 510 to 540 are repeated.

[0112] Step 560: Obtain the smoothed signal sequence.

[0113] Specifically, after performing local smoothing repair on all pulse interference points, that is, after replacing the signal points corresponding to the locations of all pulse interference points, the replaced signal sequence is taken as the smoothed signal sequence, and the smoothed signal sequence is denoted as...

[0114] The impulse interference suppression method provided in this embodiment traverses the locations of all impulse interference points and performs the following at each impulse interference point location: weighted regression is performed on the signal points within the sliding window based on an improved local weighted regression algorithm to obtain smooth signal points; the impulse interference points within the sliding window are replaced with the corresponding smooth signal points, thereby transforming the signal sequence to be processed into a smooth signal sequence, and further achieving good impulse interference suppression.

[0115] Optionally, this embodiment further illustrates the foregoing embodiments, mainly describing the specific process of obtaining smooth signal points by performing weighted regression on signal points within a sliding window based on the improved LOESS algorithm. The impulse interference suppression method provided in this embodiment includes:

[0116] Determine the regression coefficient matrix;

[0117] Weighted regression is performed on the signal points within the sliding window based on the regression coefficient matrix to obtain smooth signal points.

[0118] Specifically, the regression coefficient α is determined based on the weighted least squares method. GLOESS After obtaining the regression coefficient matrix α GLOESS Then, multiply the position parameter matrix X by α. GLOESS That is, based on the regression coefficient matrix α GLOESS Weighted regression is performed on the signal points within the sliding window to obtain an N x 1 matrix. This N x 1 matrix is ​​then used as the locally smoothed sequence Y. GLOESS ∈R N×1 That is, the locally smoothed sequence YGLOESS It contains N smooth signal points.

[0119] Local smoothing sequence Y GLOESS The expression is as follows:

[0120] Y GLOESS =Xα GLPESS

[0121] In the formula, Y GLPESS Let X represent a locally smoothed sequence, and α represent the position parameter matrix. GLOESS This represents the regression coefficient matrix.

[0122] The impulse interference suppression method provided in this embodiment achieves weighted regression of signal points within the sliding window through a regression coefficient matrix, thereby further realizing good impulse interference suppression.

[0123] Optionally, this embodiment is a further explanation of the foregoing embodiments, mainly illustrating the specific process of determining the regression coefficient matrix. The impulse interference suppression method provided in this embodiment includes:

[0124] Based on the signal points within the sliding window, determine the position parameter matrix, local data matrix, and fitting weight diagonal matrix;

[0125] The regression coefficient matrix is ​​determined based on the location parameter matrix, the local data matrix, the fitting weight diagonal matrix, and the robust repair weight diagonal matrix.

[0126] Smooth signal points are obtained based on the regression coefficient matrix and the location parameter matrix.

[0127] Specifically, the position parameter matrix X is an N-row, M+1-column matrix, i.e., X∈R N×(M+1) M is the polynomial order, and its value is chosen based on a trade-off between simulation error performance and computational complexity. The element in the nth row and mth column of the position parameter matrix X represents the position x of the nth signal point within the sliding window. n The m-1th power, that is

[0128] The local data matrix Y′ is an N x 1 matrix, i.e., Y′∈R. N×1 In the local data matrix Y′, each row element represents the corresponding signal point within the sliding window; that is, the nth row element represents the nth signal point y. n .

[0129] The fitting weights diagonal matrix W is an N x N matrix, i.e., W ∈ R. N×N The diagonal elements of the fitting weights diagonal matrix W are the fitting weights ω corresponding to the position of the nth signal point within the sliding window. n .

[0130] Fitting weights ωn The expression is as follows:

[0131]

[0132] In the formula, ω n x represents the fitted weight corresponding to the position of the nth signal point. s This indicates the filtering position, specifically the location u of the pulse interference point within the sliding window. h d(x) s ) represents x s The farthest distance to other signal points within the sliding window, i.e., d(x) s )=max{|x s -x i |},,x i This indicates the location of all signal points within the sliding window, excluding pulse interference points.

[0133] The regression coefficient matrix is ​​obtained based on the location parameter matrix X, the local data matrix Y′, the fitting weight diagonal matrix W, and the robust repair weight diagonal matrix Δ. The expression for the regression coefficient matrix is ​​shown below:

[0134] α GLOESS =(X T WΔX) -1 X T WΔY′

[0135] In the formula, α GLOESS Let R represent the regression coefficient matrix, X represent the location parameter matrix, W represent the fitting weight diagonal matrix, and Δ represent the robust repair weight diagonal matrix, where Δ∈R N×N The superscript T indicates the transpose operation, and Y′ represents the local data matrix.

[0136] The impulse interference suppression method provided in this invention determines the position parameter matrix, local data matrix, and fitting weight diagonal matrix based on the signal points within the sliding window; and adds a robust repair weight diagonal matrix to the position parameter matrix, local data matrix, and fitting weight diagonal matrix to determine the regression coefficient matrix, thereby achieving accurate determination of the regression coefficient matrix and further facilitating subsequent local smoothing repair.

[0137] The technical effects of the present invention will be illustrated below through a data comparison of an example.

[0138] The simulation uses a Minimum Shift Keying (MSK) signal as the input signal, which is generated using orthogonal modulation. The impulse interference signal is modeled using a Bernoulli-Gaussian model, and the signal received by the receiver is the sum of the MSK signal, the impulse interference signal, and Gaussian white noise. The impulse interference suppression method provided in this invention processes the received signal, and then calculates the bit error rate after MSK demodulation.

[0139] Figure 6 This is a schematic diagram of the pulse interference time-domain amplitude when the signal-to-interference ratio is -20dB, provided by the present invention. Figure 7 This diagram compares the bit error rate (BER) performance of different methods at a signal-to-interference ratio (SINR) of -20 dB, as provided by this invention. The BER performance of the proposed method (Myriad-GLOESS) is compared with other methods (RLRS, clipping algorithm, and no-impact interference) under impulse interference conditions of p = 0.05 and SINR = -20 dB. The impulse interference diagram used in the simulation is shown below. Figure 6 As shown. Where p represents the probability of impulse interference occurring, SINR represents the power ratio of the useful signal to the impulse interference, i.e., the signal-to-interference ratio (SINR), and E... b / N0 represents the power ratio of the useful signal to the Gaussian white noise, i.e., the signal-to-noise ratio.

[0140] like Figure 7 As shown, the bit error rate performance of the method proposed in this invention is superior to that of the RLRS algorithm and the traditional clipping algorithm. At a signal-to-noise ratio of 14 dB, the bit error rate of the method proposed in this invention can reach 10%. -7 When the bit error rate is 10 -5 At that time, the signal-to-noise ratio required by the method proposed in this invention differs from that in the absence of interference by only 1 dB.

[0141] Figure 8 This is a schematic diagram of the pulse interference time-domain amplitude when the signal-to-interference ratio is -30dB, provided by the present invention. Figure 9 This is a comparison chart of the bit error rate (BER) performance of different methods when the signal-to-interference ratio (SINR) is -30 dB, as provided by this invention. The BER performance of the proposed method (Myriad-GLOESS) is compared with other comparative methods under impulse interference conditions of p = 0.05 and SINR = -30 dB. The impulse interference diagram used in the simulation is shown below. Figure 8 As shown.

[0142] like Figure 9 As shown, when the bit error rate is 10... -5 At that time, the signal-to-noise ratio required by the method proposed in this invention differs from that in the absence of interference by only 0.5 dB. Since the RLRS algorithm improves its robustness to impulse interference through an iterative process, its performance is significantly affected by the number of iterations. Insufficient iterations can lead to a bit error plateau, causing system performance degradation, such as... Figure 9As shown. The method proposed in this invention introduces a fitted Gaussian robust repair matrix, eliminating the need for iterative repair, thus making the proposed method more robust to impulse interference.

[0143] Figure 10 This chart compares the bit error rate performance of the Myriad-GLOESS impulse interference suppression method provided by this invention under different impulse interference conditions. Figure 10 As can be seen, the method provided by this invention has better bit error rate performance when the probability of impulse interference (p) and the signal-to-interference ratio (SINR) are both lower. When the probability of impulse interference is low, fewer signal samples are affected by the interference. Furthermore, when the SINR is even lower, the difference in amplitude between the impulse interference and the signal is greater, making it easier to detect impulse interference in the first step of impulse noise detection. When p = 0.01 and SINR = -30 dB, the bit error rate curve of the method proposed in this invention asymptotically coincides with the bit error rate curve under interference-free conditions, indicating that the proposed method effectively suppresses impulse interference.

[0144] The simulation results above demonstrate that the bit error rate performance of the proposed method is significantly improved compared to the RLRS algorithm and the amplitude limiting algorithm, and it is more robust to impulse interference.

[0145] From a time complexity perspective, the RLRS algorithm requires regression smoothing at each signal sample and improves robustness to impulse interference through iterative calculations, which consumes a significant amount of time. The method proposed in this invention first uses the Myriad algorithm, which is sensitive to impulse interference, to detect the location of the interference. Smoothing repair is performed only at the location of the interference, greatly reducing the computational load. Furthermore, by introducing a new robust repair weight diagonal matrix, robustness to impulse interference is improved without requiring iterative calculations, thus significantly saving computation time and drastically reducing computational complexity.

[0146] Both algorithms are based on the least squares method. The RLRS algorithm has L*I least squares operations, where L is the signal sample length and I is the number of iterations. The method proposed in this invention has H least squares operations, where H is the number of impulse interferences detected by the Myriad module. Since general impulse interferences have time-domain sparsity, H << L. Therefore, the method provided in this invention effectively reduces time complexity and improves noise reduction efficiency.

[0147] The pulse interference suppression device provided by the present invention is described below. The pulse interference suppression device described below can be referred to in correspondence with the pulse interference suppression method described above.

[0148] Figure 11 This is a schematic diagram of the pulse interference suppression device provided by the present invention, as shown below. Figure 11 As shown, the pulse interference suppression device provided by the present invention includes: a detection module 1110 and a repair module 1120. Wherein:

[0149] Detection module 1110 is used to detect pulse interference in the signal sequence to be processed based on Myriad filter and determine the pulse interference point;

[0150] Repair module 1120 is used to perform local smoothing repair on the pulse interference points based on an improved local weighted regression algorithm to obtain a smooth signal sequence; the improved local weighted regression algorithm is based on the local weighted regression algorithm, and introduces a robust repair weight diagonal matrix constructed based on the probability density function of the fitted Gaussian distribution; the probability density function of the fitted Gaussian distribution is determined according to the amplitude of the non-pulse interference points in the signal sequence to be processed.

[0151] In some embodiments, the detection module 1110 includes: a first acquisition submodule and a second acquisition submodule; wherein:

[0152] The first acquisition submodule is used to input the signal sequence to be processed into the Myriad filter, acquire the loss parameter that minimizes the loss function of the Myriad filter, and use the loss parameter as the impulse interference detection threshold.

[0153] The second acquisition submodule is used to identify signal points in the signal sequence to be processed whose amplitude is greater than the pulse interference detection threshold as the pulse interference points.

[0154] In some embodiments, the first acquisition submodule includes: a first determining unit, a second determining unit, and an acquisition unit; wherein:

[0155] The first determining unit is used to initialize the loss parameter to the amplitude of any signal point in the signal sequence to be processed, and to determine the loss function value corresponding to the amplitude of each signal point.

[0156] The second determining unit is used to sort the loss function values ​​corresponding to the amplitudes of all signal points in the signal sequence to be processed from smallest to largest, and determine the amplitudes of the two signal points corresponding to the first two loss function values.

[0157] The acquisition unit is used to search between the amplitudes of the two signal points based on the bisection method to obtain the loss parameter that minimizes the loss function.

[0158] In some embodiments, the repair module 1120 includes a third acquisition submodule; the third acquisition submodule is used to traverse the locations of all pulse interference points to acquire the smoothed signal sequence; the third acquisition submodule includes: a sliding unit, a weighting unit, and a replacement unit; wherein:

[0159] A sliding unit is used to make the midpoint of the sliding window the position of the pulse interference point.

[0160] A weighting unit is used to perform weighted regression on the signal points within the sliding window based on an improved local weighted regression algorithm to obtain smooth signal points; each signal point within the sliding window corresponds to one of the smooth signal points;

[0161] The replacement unit is used to replace the pulse interference points in the sliding window with the corresponding smooth signal points.

[0162] In some embodiments, the weighting unit includes: a determining subunit and an obtaining subunit; wherein:

[0163] Determine the sub-units to determine the regression coefficient matrix;

[0164] The sub-unit is used to perform weighted regression on the signal points within the sliding window based on the regression coefficient matrix to obtain smooth signal points.

[0165] In some embodiments, determining the subunit is specifically used for:

[0166] Based on the signal points within the sliding window, determine the position parameter matrix, the local data matrix, and the fitting weight diagonal matrix;

[0167] The regression coefficient matrix is ​​determined based on the location parameter matrix, the local data matrix, the fitting weight diagonal matrix, and the robust repair weight diagonal matrix.

[0168] It should be noted that the pulse interference suppression device provided by the present invention can implement all the method steps implemented in the above method embodiments and can achieve the same technical effect. Here, the parts that are the same as those in the method embodiments and the beneficial effects will not be described in detail.

[0169] Figure 12 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 12As shown, the electronic device may include a processor 1210, a communication interface 1220, a memory 1230, and a communication bus 1240. The processor 1210, communication interface 1220, and memory 1230 communicate with each other via the communication bus 1240. The processor 1210 can call logic instructions in the memory 1230 to execute a pulse interference suppression method. This method includes: detecting pulse interference in the signal sequence to be processed based on a Myriad filter to determine pulse interference points; performing local smoothing repair on the pulse interference points based on an improved local weighted regression algorithm to obtain a smooth signal sequence; the improved local weighted regression algorithm is based on a local weighted regression algorithm, introducing a robust repair weight diagonal matrix constructed based on a probability density function of a fitted Gaussian distribution; the probability density function of the fitted Gaussian distribution is determined according to the amplitude of non-pulse interference points in the signal sequence to be processed.

[0170] Furthermore, the logical instructions in the aforementioned memory 1230 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0171] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the impulse interference suppression method provided by the above methods. The method includes: detecting impulse interference in the signal sequence to be processed based on a Myriad filter to determine impulse interference points; performing local smoothing repair on the impulse interference points based on an improved local weighted regression algorithm to obtain a smooth signal sequence; the improved local weighted regression algorithm is based on the local weighted regression algorithm by introducing a robust repair weight diagonal matrix constructed based on a probability density function of a fitted Gaussian distribution; the probability density function of the fitted Gaussian distribution is determined according to the amplitude of non-impulse interference points in the signal sequence to be processed.

[0172] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the impulse interference suppression method provided by the above methods. The method includes: detecting impulse interference in a signal sequence to be processed based on a Myriad filter to determine impulse interference points; performing local smoothing repair on the impulse interference points based on an improved local weighted regression algorithm to obtain a smooth signal sequence; the improved local weighted regression algorithm is based on a local weighted regression algorithm, introducing a robust repair weight diagonal matrix constructed based on a probability density function of a fitted Gaussian distribution; the probability density function of the fitted Gaussian distribution is determined according to the amplitude of non-impulse interference points in the signal sequence to be processed.

[0173] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0174] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0175] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for suppressing impulse interference, characterized in that, include: Pulse interference detection is performed on the signal sequence to be processed based on the Myriad filter to determine the pulse interference points; The improved local weighted regression algorithm is used to perform local smoothing repair on the pulse interference points to obtain a smooth signal sequence. The improved local weighted regression algorithm is based on the local weighted regression algorithm, and introduces a robust repair weight diagonal matrix constructed based on the probability density function of the fitted Gaussian distribution. The probability density function of the fitted Gaussian distribution is determined according to the amplitude of the non-pulse interference points in the signal sequence to be processed. The improved local weighted regression algorithm is used to perform local smoothing repair on the pulse interference points to obtain a smooth signal sequence, including: Traverse the locations of all pulse interference points to obtain the smoothed signal sequence; at each pulse interference point location, execute: The midpoint of the sliding window is taken as the location of the pulse interference point; The signal points within the sliding window are weighted and regressed based on an improved local weighted regression algorithm to obtain smooth signal points; each signal point within the sliding window corresponds to one of the smooth signal points. Replace the pulse interference points within the sliding window with the corresponding smooth signal points.

2. The pulse interference suppression method according to claim 1, characterized in that, The step of detecting pulse interference in the signal sequence to be processed based on the Myriad filter and determining the pulse interference points includes: The signal sequence to be processed is input into the Myriad filter to obtain the loss parameter that minimizes the loss function of the Myriad filter, and the loss parameter is used as the impulse interference detection threshold. Signal points in the signal sequence to be processed whose amplitude is greater than the pulse interference detection threshold are taken as pulse interference points.

3. The pulse interference suppression method according to claim 2, characterized in that, The signal sequence to be processed is input into the Myriad filter to obtain the loss parameters that minimize the loss function of the Myriad filter, including: The loss parameter is initialized to the amplitude of any signal point in the signal sequence to be processed, and the loss function value corresponding to the amplitude of each signal point is determined. Sort the loss function values ​​corresponding to the amplitudes of all signal points in the signal sequence to be processed in ascending order, and determine the amplitudes of the two signal points corresponding to the first two loss function values. The loss parameter that minimizes the loss function is obtained by searching between the amplitudes of the two signal points using the binary search method.

4. The pulse interference suppression method according to claim 1, characterized in that, The improved local weighted regression algorithm performs weighted regression on the signal points within the sliding window to obtain smooth signal points, including: Determine the regression coefficient matrix; The signal points within the sliding window are weighted and regressed based on the regression coefficient matrix to obtain smooth signal points.

5. The pulse interference suppression method according to claim 4, characterized in that, The determination of the regression coefficient matrix includes: Based on the signal points within the sliding window, determine the position parameter matrix, the local data matrix, and the fitting weight diagonal matrix; The regression coefficient matrix is ​​determined based on the location parameter matrix, the local data matrix, the fitting weight diagonal matrix, and the robust repair weight diagonal matrix.

6. A pulse interference suppression device, characterized in that, include: The detection module is used to detect pulse interference in the signal sequence to be processed based on the Myriad filter and to determine the pulse interference points. The repair module is used to perform local smoothing repair on the pulse interference points based on an improved local weighted regression algorithm to obtain a smooth signal sequence. The improved local weighted regression algorithm is based on the local weighted regression algorithm, and introduces a robust repair weight diagonal matrix constructed based on the probability density function of the fitted Gaussian distribution. The probability density function of the fitted Gaussian distribution is determined according to the amplitude of the non-pulse interference points in the signal sequence to be processed. The improved local weighted regression algorithm is used to perform local smoothing repair on the pulse interference points to obtain a smooth signal sequence, including: Traverse the locations of all pulse interference points to obtain the smoothed signal sequence; at each pulse interference point location, execute: The midpoint of the sliding window is taken as the location of the pulse interference point; The signal points within the sliding window are weighted and regressed based on an improved local weighted regression algorithm to obtain smooth signal points; each signal point within the sliding window corresponds to one of the smooth signal points. Replace the pulse interference points within the sliding window with the corresponding smooth signal points.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the pulse interference suppression method as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the pulse interference suppression method as described in any one of claims 1 to 5.

9. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the pulse interference suppression method as described in any one of claims 1 to 5.

Images