Communication signal interference elimination method and system

Abstract

The invention relates to the technical field of anti-interference, in particular to a communication signal interference elimination method and system, and the method comprises the following steps: obtaining communication signal time domain data, calculating a signal energy change rate, a signal abrupt change rate and a phase drift rate, marking and distinguishing a signal region according to a calculation result, and obtaining a communication signal non-stationary interval identification result. According to the method, the time domain data of the communication signal is subjected to fine-grained calculation, and the signal energy change rate, the signal abrupt change rate and the phase drift rate are extracted, so that accurate signal region distinguishing is realized, the regression modeling parameter range is adjusted, the influence of instantaneous data of an abrupt change region is reduced, and the stability of signal reconstruction is improved; according to the method, signal-to-noise ratio calculation and interference power fluctuation amplitude evaluation are combined, channel state parameters are dynamically adjusted, interference intensity evaluation is more accurate, the filtering step length, the weight coefficient and the gain coefficient are optimized based on error information, signal processing adapts to different types of interference, and integrity and stability of communication signals are enhanced.

Description

Technical Field

[0001] The present invention relates to the field of anti-interference technology, and in particular to a communication signal interference elimination method and system. Background Art

[0002] The field of anti-interference technology includes technical methods to suppress or eliminate interference signals in multiple scenarios such as wireless communications, satellite communications, radar systems and wired transmission. The core content of this technical field is to reduce the impact of environmental noise, electromagnetic interference, multipath effects and mutual interference on communication quality through signal processing. Anti-interference technology as a whole covers multiple aspects such as signal detection, interference identification, interference suppression and signal recovery. Signal detection is used to distinguish useful signals from interference signals. Interference identification uses spectrum analysis, time domain feature extraction and other methods to determine the type and characteristics of interference signals. Interference suppression uses adaptive filtering, coherent interference cancellation and other means to weaken or eliminate interference signals. Signal recovery includes error correction, channel equalization and other measures to ensure the integrity and reliability of useful signals.

[0003] The existing technology relies on fixed regression modeling parameters in the processing of mutation signals of communication signals, resulting in insufficient adaptability to mutation areas and affecting the signal separation effect. The channel state assessment uses a fixed channel model and lacks dynamic adjustment to interference power fluctuations, resulting in poor adaptability to complex channel environments. The filtering parameter optimization method is fixed and fails to adaptively adjust the weight and step size according to the error change. The filtering effect is unstable in different interference environments, affecting the recovery quality of communication signals. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to provide a communication signal interference elimination method and system.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: a communication signal interference elimination method, comprising the following steps: S1: Obtain the time domain data of the communication signal, calculate the signal energy change rate, signal mutation rate, and phase drift rate, distinguish the signal area according to the calculation result mark, and obtain the identification result of the non-stationary interval of the communication signal; S2: Based on the identification result of the non-stationary interval of the communication signal, call the mutation rate, phase drift rate, and energy change rate, adjust the range of regression modeling parameters, reduce the instantaneous data dependence in the mutation area, calculate the signal reconstruction error, adjust the regression modeling constraint weight and the input signal data weight, and obtain the communication signal segment prediction error value; S3: calling the communication signal segment prediction error value, calculating the communication channel signal-to-noise ratio and interference power fluctuation amplitude, obtaining the channel state parameters according to the weighted joint evaluation, and obtaining the communication channel interference intensity evaluation value; S4: Based on the communication channel interference intensity evaluation value, adjust the filter step size setting, call the signal error information to adjust the filter weight coefficient, the residual error threshold, and the gain coefficient, and obtain the communication signal filter optimization parameter set; S5: calling the communication signal filtering optimization parameter set, adjusting the filtering step size, weight coefficient, residual error threshold, and gain coefficient, calculating the signal energy change trend, adjusting the regression modeling parameter range, and obtaining the communication signal anti-interference reconstruction value.

[0006] The signal area includes a non-stationary area, a mutation area, and a signal interference area. The communication signal non-stationary interval identification result includes a signal energy change rate identification, a signal mutation rate identification, and a phase drift rate identification. The communication signal segmented prediction error value includes a mutation area error, a phase drift area error, and an energy change area error. The communication channel interference intensity evaluation value includes a signal-to-noise ratio evaluation value, an interference power fluctuation amplitude evaluation value, and a channel state parameter evaluation value. The communication signal filtering optimization parameter set includes a filtering step size setting value, a filtering weight coefficient, a residual error threshold, and a gain coefficient. The communication signal anti-interference reconstruction value includes a reconstructed signal energy value, a reconstructed signal phase value, and a reconstructed signal time domain eigenvalue.

[0007] As a further solution of the present invention, the step of obtaining the identification result of the non-stationary interval of the communication signal is specifically as follows: S101: Acquire time domain data of a communication signal, calculate the instantaneous energy of the signal at differentiated time points, and obtain the signal energy change rate by calculating the energy change rate of adjacent time points; S102: Based on the signal energy change rate, calculate the mutation degree of adjacent time points, define a mutation threshold, screen the time period in which the mutation rate exceeds the threshold, and obtain the signal mutation rate; S103: Based on the signal energy change rate and the signal mutation rate, calculate the change rate of the signal phase over time, using the formula: ;

[0008] The phase drift rate is obtained by signal phase difference operation, and the signal energy change rate and signal mutation rate are combined to classify the signal area and obtain the identification result of the non-stationary interval of the communication signal; in, represents the phase drift rate, Representative The phase value at a time point, Representative The phase value at a time point, Representative The signal energy at a time point, Representative The signal energy at a time point, Represents the total number of time points.

[0009] As a further solution of the present invention, the step of obtaining the communication signal segment prediction error value is specifically: S201: Based on the identification result of the non-stationary interval of the communication signal, the mutation rate, phase drift rate, and energy change rate are called to segment the time series of the communication signal, calculate the mutation rate and drift change trend in each segment, and screen the time window of the mutation area, set the data weight in the time window, reduce the instantaneous data dependence, and obtain the mutation area time window; S202: Based on the mutation region time window, adjust the regression modeling parameter range, dynamically adjust the input signal data weight according to the signal characteristics of the mutation region, set the weight distribution coefficient of the differentiation interval, calculate the signal reconstruction error, and use the formula: ; The signal reconstruction error is obtained by operation, and the regression modeling constraint weight is adjusted to obtain the modeling error constraint weight; in, represents the signal reconstruction error, Representative The signal value in a time window, Representative The weight value of the time window, Represents the total number of time windows, represents the sum of the signals in all time windows, Represents the sum of weights of all time windows; S203: calling the modeling error constraint weight, combining the characteristic data of the signal mutation area, adjusting the constraint conditions of the regression modeling, calculating the signal prediction error under the differentiation interval, and obtaining the communication signal segment prediction error value.

[0010] As a further solution of the present invention, the step of obtaining the communication channel interference intensity evaluation value is specifically: S301: Based on the segmented prediction error value of the communication signal, calculate the signal-to-noise ratio of the communication channel, extract the ratio of the error value to the signal power, normalize the ratio, measure the signal quality on the same scale, remove the deviation data, and obtain the channel signal-to-noise ratio evaluation value; S302: calling the channel signal-to-noise ratio evaluation value, calculating the interference power fluctuation amplitude of the communication channel, extracting the maximum and minimum values ​​of the interference signal power, and calculating the difference, calling the average power of the communication signal, normalizing the average power fluctuation value, so as to measure the degree of fluctuation of the interference power relative to the signal power, and obtaining the interference power fluctuation ratio; S303: calling the channel signal-to-noise ratio evaluation value and the interference power fluctuation ratio to calculate the channel availability under the differentiated signal interference condition, using the formula: ; Obtain channel state parameters by operation, and calculate the correlation between the channel state parameters and the interference signal to obtain the communication channel interference intensity assessment value; in, represents the communication channel interference intensity assessment value, represents the average signal power, represents the variance of communication noise, represents the maximum interference signal power, Represents the minimum interference signal power, represents the weighting coefficient of the differentiated signal segment, represents the power of the kth signal segment, represents the interference power of the kth signal segment, Represents the total number of signal segments.

[0011] As a further solution of the present invention, the step of acquiring the communication signal filtering optimization parameter set is specifically: S401: Based on the communication channel interference intensity evaluation value, calculate the interference intensity change rate, identify the impact range, set the filter step adjustment value corresponding to the differentiated interference interval, compare the change of the step before and after the adjustment, and generate the filter step adjustment coefficient; S402: Call the filter step adjustment coefficient, adjust the filter weight coefficient according to the signal error information, and select the gain coefficient corresponding to the differentiated error range in combination with the residual error threshold, using the formula: ; Calculate the adjusted filter weight coefficient to obtain the optimized filter weight coefficient; in, represents the adjusted filter weight coefficient, Represents the filter weight coefficient before adjustment, Represents the current signal error value, represents the residual error threshold, represents the gain factor, Represents the filter step setting value at the previous moment, Represents the current filter step setting value, The representative constant is used to avoid the denominator approaching zero; S403: calling the optimized filter weight coefficient, screening the filter parameter set of the communication signal, matching multiple parameter combinations, calculating the filter optimization result, and obtaining the communication signal filter optimization parameter set.

[0012] As a further solution of the present invention, the step of obtaining the anti-interference reconstruction value of the communication signal is specifically as follows: S501: calling the communication signal filtering optimization parameter set, adjusting the filtering step size, weight coefficient, residual error threshold and gain coefficient, calculating the energy change rate of the signal, extracting the energy fluctuation trend of the continuous signal interval, comparing the change amplitude of adjacent intervals and screening the high change interval, and obtaining the signal energy change trend value; S502: Based on the signal energy change trend value, set the regression modeling parameter adjustment range, calculate the model error under the differentiated parameter combination, compare the error fluctuation and select the parameter combination with the smallest error, adjust the regression modeling strategy and perform data fitting to obtain the optimal regression modeling parameters; S503: Call the optimal regression modeling parameters, reconstruct the communication signal, and perform anti-interference adjustment in combination with the filter optimization parameters, and calculate the residual error value of the signal after anti-interference, using the formula: ; Calculate the anti-interference adjustment error value, and optimize and adjust the error to obtain the anti-interference reconstruction value of the communication signal; in, Represents the anti-interference reconstruction value of the communication signal, represents the original signal value, represents the signal value reconstructed by the regression model, represents the filter weight coefficient, Represents the dynamic change trend value of signal energy. Represents the energy adjustment threshold of the signal after anti-interference, represents the signal gain coefficient, Represents the total number of signal sampling points.

[0013] A communication signal interference elimination system, the communication signal interference elimination system is used to perform the above communication signal interference elimination method, the system comprises: Signal area marking module: obtains the time domain data of the communication signal, calculates the signal energy change rate, signal mutation rate, and phase drift rate, determines the signal interval type based on the calculation results, and marks the interval where the mutation rate exceeds the set threshold, the phase drift rate is abnormal, or the energy change rate fluctuates as a non-stationary interval, and obtains the non-stationary interval identification result of the communication signal; Segmentation error calculation module: calling the identification result of the non-stationary interval of the communication signal, adjusting the range of regression modeling parameters according to the mutation rate, phase drift rate, and energy change rate, reducing the instantaneous data dependence in the mutation area, calculating the signal reconstruction error, and adjusting the regression modeling constraint weight and input signal data weight based on the error size to obtain the segmentation prediction error value of the communication signal; Channel state evaluation module: calling the communication signal segment prediction error value, calculating the communication channel signal-to-noise ratio and interference power fluctuation amplitude, based on weighted joint evaluation, screening parameters according to the weight of the influencing factor, calculating the channel state parameters, and obtaining the communication channel interference intensity evaluation value; Filter parameter optimization module: call the communication channel interference intensity evaluation value, adjust the filter step size setting according to the interference intensity, call the signal error information, dynamically adjust the filter weight coefficient, residual error threshold, and gain coefficient, calculate the filter performance index under the differentiated parameter configuration, screen the parameter combination with the best interference suppression effect, and obtain the communication signal filter optimization parameter set; Signal anti-interference reconstruction module: call the communication signal filtering optimization parameter set, adjust the filtering step size, weight coefficient, residual error threshold, gain coefficient, calculate the signal energy change trend, reconstruct the signal according to the adjustment result of the regression modeling parameter range, and obtain the communication signal anti-interference reconstruction value.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by performing fine-grained calculations on the time domain data of the communication signal, the signal energy change rate, signal mutation rate and phase drift rate are extracted to achieve accurate signal area distinction. The range of regression modeling parameters is adjusted to reduce the impact of instantaneous data in the mutation area and improve the stability of signal reconstruction. Combined with the signal-to-noise ratio calculation and the interference power fluctuation amplitude assessment, the channel state parameters are dynamically adjusted to make the interference intensity assessment more accurate. Based on the error information, the filter step size, weight coefficient and gain coefficient are optimized to adapt the signal processing to different types of interference, improve the anti-interference reconstruction capability, and enhance the integrity and stability of the communication signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram of the workflow of the present invention; Figure 2 A flowchart of the steps for obtaining the identification result of the non-stationary interval of the communication signal of the present invention; Figure 3 A flowchart of the steps for obtaining the segmented prediction error value of the communication signal of the present invention; Figure 4 A flowchart of the steps for obtaining the interference intensity evaluation value of the communication channel of the present invention; Figure 5 A flow chart of the steps for obtaining a communication signal filtering optimization parameter set according to the present invention; Figure 6 This is a flow chart of the steps for obtaining the anti-interference reconstruction value of the communication signal of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0018] Embodiment 1 See also Figure 1 The present invention provides a technical solution: a communication signal interference elimination method, comprising the following steps: S1: Obtain the time domain data of the communication signal, calculate the signal energy change rate, signal mutation rate, and phase drift rate, distinguish the signal area according to the calculation result mark, and obtain the identification result of the non-stationary interval of the communication signal; S2: Based on the identification results of the non-stationary interval of the communication signal, call the mutation rate, phase drift rate, and energy change rate, adjust the range of regression modeling parameters, reduce the instantaneous data dependence in the mutation area, calculate the signal reconstruction error, adjust the regression modeling constraint weight and input signal data weight, and obtain the communication signal segment prediction error value; S3: Call the communication signal segment prediction error value, calculate the communication channel signal-to-noise ratio and interference power fluctuation amplitude, obtain the channel state parameters based on the weighted joint evaluation, and obtain the communication channel interference intensity evaluation value; S4: Based on the communication channel interference intensity evaluation value, adjust the filter step size setting, call the signal error information to adjust the filter weight coefficient, residual error threshold, and gain coefficient, and obtain the communication signal filter optimization parameter set; S5: Call the communication signal filtering optimization parameter set, adjust the filtering step size, weight coefficient, residual error threshold, gain coefficient, calculate the signal energy change trend, adjust the regression modeling parameter range, and obtain the communication signal anti-interference reconstruction value.

[0019] The signal area includes a non-stationary area, a mutation area, and a signal interference area. The identification results of the non-stationary interval of the communication signal include a signal energy change rate identification, a signal mutation rate identification, and a phase drift rate identification. The segmented prediction error value of the communication signal includes a mutation area error, a phase drift area error, and an energy change area error. The communication channel interference intensity evaluation value includes a signal-to-noise ratio evaluation value, an interference power fluctuation amplitude evaluation value, and a channel state parameter evaluation value. The communication signal filtering optimization parameter set includes a filtering step size setting value, a filtering weight coefficient, a residual error threshold, and a gain coefficient. The communication signal anti-interference reconstruction value includes a reconstructed signal energy value, a reconstructed signal phase value, and a reconstructed signal time domain eigenvalue.

[0020] See also Figure 2 , the specific steps for obtaining the identification result of the non-stationary interval of the communication signal are: S101: Acquire time domain data of a communication signal, calculate the instantaneous energy of the signal at differentiated time points, and obtain the signal energy change rate by calculating the energy change rate of adjacent time points; First, it is necessary to obtain time domain data from the communication signal source. This data can be collected through wireless sensor devices or directly from a digital signal processor (DSP). The sampling frequency of the signal needs to be set according to the characteristics of the communication signal. For example, for cellular communication signals, a sampling rate of 10kHz to 100MHz can be selected to meet the signal analysis requirements. Then, the instantaneous energy of the signal at each time point is calculated. The instantaneous energy calculation formula is ,in For the The signal amplitude at a time point. For example, if the signal amplitude at a certain moment is , then the instantaneous energy , further, the energy change rate between adjacent time points is calculated as follows: ; Among them, if at time and The instantaneous energy values ​​calculated at and , then the energy change rate is , and so on, calculate the energy change rate of all adjacent time points in the complete data sequence and normalize it so that the data distribution is balanced in a standard interval, such as using minimum-maximum normalization ; If the maximum energy value , minimum , for a certain moment of energy value , the normalized value is ; Finally, the energy change rate sequence of the entire time domain signal is obtained, as shown in Table 1.

[0021] Table 1 Communication signal energy change rate table

[0022] As shown in Table 1, after the energy change rate is calculated, the data sequence is stored in the signal processing system, and the subsequent steps can be directly analyzed based on the data.

[0023] S102: Based on the signal energy change rate, calculate the mutation degree of adjacent time points, define a mutation threshold, screen the time period where the mutation rate exceeds the threshold, and obtain the signal mutation rate; First, calculate the mutation degree of the signal energy change rate. This value can be calculated by the local change rate of the signal. The specific calculation method is as follows: ; in, Indicates the time interval, usually a fixed sampling interval, such as 1ms. Based on the data in Table 1 in the previous part, calculate the mutation rate, for example: ; ; Next, set the mutation threshold , which can usually be calculated using the mean plus 3 times the standard deviation method: ; Set the mean , standard deviation , then the mutation threshold ; In this example, , so the time point is determined The mutation point and the mutation time period are recorded.

[0024] S103: Based on the signal energy change rate and the signal mutation rate, the change rate of the signal phase over time is calculated using the formula: ; The phase drift rate is obtained by signal phase difference operation, and the signal energy change rate and signal mutation rate are combined to classify the signal area and obtain the identification result of the non-stationary interval of the communication signal; in, represents the phase drift rate, Representative The phase value at a time point, Representative The phase value at a time point, Representative The signal energy at a time point, Representative The signal energy at a time point, Represents the total number of time points.

[0025] formula: ; First, calculate the phase difference value , assuming that the phase sequence in a certain time period is , calculate the phase change: ; ; ; total ,but ; Next, calculate the normalized signal energy change: ; ; Finally, the phase drift rate The calculation result is ; This result indicates that the phase drift rate is relatively high, and combined with the signal mutation rate, the non-stationary interval can be further determined.

[0026] See also Figure 3 , the specific steps for obtaining the segmented prediction error value of the communication signal are: S201: Based on the identification result of the non-stationary interval of the communication signal, the mutation rate, phase drift rate, and energy change rate are called to segment the time series of the communication signal, calculate the mutation rate and drift change trend in each segment, and select the time window of the mutation area, set the data weight in the time window, reduce the instantaneous data dependence, and obtain the mutation area time window; First, extract the time series of the communication signal, and obtain the data of each time window in the order of the time axis. When the time series is segmented, calculate the mutation rate, phase drift rate and energy change rate in each time window. The mutation rate is calculated by counting the amplitude difference between the adjacent signal points before and after in a given time window, and calculating its mean. Set the mutation threshold. For example, when the mean of the amplitude difference is greater than 0.3 times the historical mean, it is considered that there is a mutation in the window; the calculation of the phase drift rate is based on the rate of change of the phase over time. By comparing the phase change values ​​of adjacent time windows, if the relative change exceeds the set threshold (such as 10 degrees / millisecond), it is judged that there is a significant drift; the energy change rate is calculated by comparing the signal energy of the current window with the signal energy change of the previous window. If the change rate exceeds the set value (such as greater than 20% of the energy mean of the previous window), it is considered that there is a significant change. After completing the above calculations, all time windows that meet the mutation judgment conditions are screened out, and the weights of the time windows are set. The weights are set in the mutation area to reduce the instantaneous data dependence. For example, lower weight values ​​are set for the two windows before and after the mutation window (such as the weight is reduced to 0.6), and the weight of the middle mutation window is set to 1. This ensures the stability of the time window in the mutation area, and finally obtains the time window in the mutation area.

[0027] S202: Based on the mutation region time window, adjust the regression modeling parameter range, dynamically adjust the input signal data weight according to the signal characteristics of the mutation region, set the weight distribution coefficient of the differentiation interval, calculate the signal reconstruction error, and use the formula: ; The signal reconstruction error is obtained by operation, and the regression modeling constraint weight is adjusted to obtain the modeling error constraint weight; in, represents the signal reconstruction error, Representative The signal value in a time window, Representative The weight value of the time window, Represents the total number of time windows, represents the sum of the signals in all time windows, Represents the sum of weights of all time windows; The specific approach is to set a higher weight for the time window within the mutation area, such as setting the weight of the mutation window to 1.2, and the weight of the front and back windows to 0.8, to ensure that the regression model can more accurately reflect the mutation trend. Next, calculate the signal reconstruction error, the calculation formula is as follows: ; in, Indicates The original signal value of the time window, Represents the weight value of the window, is the total number of all time windows, is the sum of the signals in all windows, is the sum of the weights of all windows. For example, suppose there are 5 time windows and their original signal values ​​are as follows: Table 2 Time window signal table

[0028] As shown in Table 2, according to the formula, the sum of the signals in all time windows is , the sum of the weights of all windows is , the weighted signal value is calculated as follows: ; ; Calculate the reconstruction error: ; ; After obtaining the regression modeling error, the regression modeling constraint weight is adjusted according to the error size. For example, when the error is greater than the set threshold (such as 1.0), the weight of the mutation window is adjusted to 1.5 to reduce the influence of the front and back windows, and finally the modeling error constraint weight is obtained.

[0029] S203: Call the modeling error constraint weight, combine the characteristic data of the signal mutation area, adjust the constraint conditions of the regression modeling, calculate the signal prediction error under the differentiation interval, and obtain the communication signal segment prediction error value.

[0030] The regression model involved in this scenario is a weighted linear regression model, whose core structure is as follows: the input features include the original signal value, phase drift rate, mutation rate, and energy change rate of each time window, and the weight term is pre-set for each time window. , used to strengthen the impact of the mutation window on modeling, the model form is a linear regression that minimizes the weighted error sum of squares: ; The output is the predicted value Sum signal reconstruction error , used to measure modeling accuracy and provide feedback to adjust weights. It does not involve nonlinear structures or neural networks, but only uses weight control and linear mapping to ensure more accurate modeling of mutation areas and stronger responses; in, Represents the parameter vector of the regression model, which contains the regression coefficients of each input feature in the signal prediction process. Indicates The input feature vector corresponding to a time window usually contains the current amplitude of the signal, the amplitude difference value with the previous window, the phase change rate, the energy change rate, and the mutation indication information, etc. Indicates The original signal value of the time window is used as the true target value of the model. Represents the model for The predicted output value of the signal value in the time window, Indicates The weighting coefficient of a time window is used to control the relative influence of the prediction error of the window in the objective function. Represents the total number of time windows, that is, the number of time windows included in the entire regression modeling process, which is used to control the complete interval of model training; The regression model relies on the weight system set in the mutation window identification process, where the weight of the center of the mutation window is usually set to 1.2, and the weight of the adjacent windows is set to 0.8 to increase the dominance of the mutation area on the modeling trend. If it exceeds the set threshold (for example, 1.0), the weight of the mutation window is further increased to 1.5, while the influence of windows in other regions is reduced, so as to achieve adaptive adjustment of regression modeling errors based on differentiated features, and ultimately achieve the goal of enhancing signal prediction accuracy in the mutation region.

[0031] First, the signal prediction error is calculated for the signal feature data in the differentiated interval. In the prediction stage, the historical data is weighted regressed using the adjusted weights, the model parameters are calculated using the least squares method, and the predicted value is calculated. The prediction error is calculated by comparing it with the actual value. For example, in the mutation window, the data of the past five time windows are set to predict the current window value. Assume that the signal values ​​of the past five windows are , the fitting model is , calculate the parameters by the least squares method , get the predicted value , calculate the prediction error: ; Assuming that after calculation , while the actual measured value is , then the prediction error If the prediction error exceeds the set threshold (such as 0.2), it is necessary to adjust the regression modeling parameters, such as increasing the historical weight of the mutation window (such as from 1.2 to 1.5), reducing the weight far from the mutation window, and finally obtaining the segmented prediction error value of the communication signal.

[0032] See also Figure 4 , the steps for obtaining the communication channel interference intensity evaluation value are specifically as follows: S301: Calculate the signal-to-noise ratio of the communication channel based on the segmented prediction error value of the communication signal, extract the ratio of the error value to the signal power, normalize the ratio, measure the signal quality at the same scale, remove the deviation data, and obtain the channel signal-to-noise ratio evaluation value; First, the received signal is segmented and the duration of each segment is set to 0.1 seconds to ensure that the channel characteristics remain relatively stable in a short period of time. The data set of each signal segment includes the signal power value and the noise interference value, which are respectively used and It means that for each segment signal, its prediction error is calculated, that is, the difference between the theoretical received signal power and the actual received signal power is calculated. The error value is calculated using the mean square error method, that is: ; in, is the theoretical ideal signal power, is the received signal power, is the number of sampling points in the signal segment. For example, assuming , then the calculated error value Reflect the distortion of the signal segment, and then calculate the signal-to-noise ratio (SNR) of the communication channel using the formula: ; In order to improve the stability of the calculation, the ratio of the error value to the signal power is calculated, that is: ; In order to ensure the consistency of the calculation scale, the normalization method is used to process the ratio value, that is: ; in, and are the minimum and maximum error ratios of all signal segments, respectively. For example, assuming , , error ratio of a certain signal segment is 0.05, then the calculated normalized ratio is: ; After normalization, the signal quality is judged and all signal segments are normalized according to the set quality threshold. To filter, for example, if you set , then all The signal segment is retained, otherwise it is removed, and the channel signal-to-noise ratio evaluation value is finally calculated. .

[0033] S302: calling the channel signal-to-noise ratio evaluation value, calculating the interference power fluctuation amplitude of the communication channel, extracting the maximum and minimum values ​​of the interference signal power, and calculating the difference, calling the average power of the communication signal, normalizing the average power fluctuation value, so as to measure the fluctuation degree of the interference power relative to the signal power, and obtaining the interference power fluctuation ratio; First, extract the maximum power of the interference signal and minimum power , set the sampling window to 1 second, record multiple interference power values ​​per second, and calculate the maximum and minimum values ​​in the window. For example, at a certain moment, the interference power data is: ; but mW, mW, calculate the interference power fluctuation amplitude: ; Next, call the average power of the communication signal , calculate the normalized interference power fluctuation value: ; Assumptions mW, then: ; The relationship between interference power fluctuation and signal power is measured based on this ratio, and finally the interference power fluctuation value is obtained. .

[0034] S303: Call the channel signal-to-noise ratio evaluation value and the interference power fluctuation ratio to calculate the channel availability under the differentiated signal interference condition, using the formula: ; Obtain channel state parameters by operation, and calculate the correlation between the channel state parameters and the interference signal to obtain the communication channel interference intensity assessment value; in, represents the communication channel interference intensity assessment value, represents the average signal power, represents the variance of communication noise, represents the maximum interference signal power, Represents the minimum interference signal power, represents the weighting coefficient of the differentiated signal segment, represents the power of the kth signal segment, represents the interference power of the kth signal segment, Represents the total number of signal segments.

[0035] First, calculate the channel interference strength evaluation value , according to the formula: ; in, is the weighting coefficient of the differentiated signal segment, set to 0.8, Representative The power of the signal segment, Represents the interference power of the signal segment, set , the data of each signal segment are as follows: Table 3 Signal segment power and interference power data table

[0036] Table 3 lists the signal power and interference power data of the five signal segments, and calculates the difference between the power of each signal segment and the interference power: ; Perform a weighted sum: ; Assumptions mW, mW, substitute into the formula: ; The result shows that the interference strength evaluation value of the communication channel is 252.92, which is used to measure the correlation between the channel state parameters and the interference signal.

[0037] See also Figure 5 , the specific steps for obtaining the communication signal filtering optimization parameter set are: S401: Based on the communication channel interference intensity evaluation value, the interference intensity change rate is calculated, and the impact range is identified. At the same time, the filter step adjustment value corresponding to the differentiated interference interval is set, and the change amount of the step before and after the adjustment is compared to generate the filter step adjustment coefficient; First, it is necessary to collect the interference signal in the communication channel in real time, set a certain time window, such as 100ms as a sampling period, and obtain the signal strength data of the communication channel in each period. and interference intensity data , calculate the average strength of the interference signal in this time window and variance , calculate the interference intensity change rate through the data of multiple consecutive time windows To determine the increase or decrease trend of interference intensity, for example, in the past 5 time windows, if The interference intensity change rate is -80dBm, -78dBm, -75dBm, -72dBm, and -68dBm respectively. , when the value is greater than 2dBm / cycle, it can be determined that the interference is significantly increased; then identify the impact range of the interference, that is, evaluate the change in interference intensity of different communication nodes, select multiple base station signal nodes, such as three base stations A, B, and C, whose interference intensity change rates are 2.1dBm / cycle, 1.5dBm / cycle, and 2.6dBm / cycle respectively, then the change rates of A and C are both greater than 2dBm / cycle, and should be classified into the high interference area, while B is in the medium interference area, and then set the filter step adjustment value corresponding to the differentiated interference interval, for example, the filter step increment is set to 0.02 in the high interference area, the filter step increment is set to 0.01 in the medium interference area, and no adjustment is made in the low interference area, and the change in the step before and after the adjustment is compared. For example, the original step in the high interference area is 0.05, and the adjusted step becomes 0.05+0.02=0.07, and finally the filter step adjustment coefficient is generated, and the adjustment coefficient is calculated by the step change rate , if the step length before adjustment is 0.05, the adjusted step size is 0.07, then .

[0038] S402: Call the filter step adjustment coefficient, adjust the filter weight coefficient according to the signal error information, and select the gain coefficient corresponding to the differentiated error range in combination with the residual error threshold, using the formula: ; Calculate the adjusted filter weight coefficient to obtain the optimized filter weight coefficient; in, represents the adjusted filter weight coefficient, Represents the filter weight coefficient before adjustment, represents the current signal error value, represents the residual error threshold, represents the gain factor, Represents the filter step setting value at the previous moment, Represents the current filter step setting value, The representative constant is used to avoid the denominator approaching zero; First, obtain the signal error information , that is, calculate the current signal value The expected signal value If the current signal value is -60dBm and the expected signal value is -58dBm, then , and then set the residual error threshold , if the threshold is set to 1.5dBm, then This indicates that the current error is large and the filter weight needs to be adjusted. Then select the gain coefficient corresponding to the differential error range. , assuming that the gain factor is set to 0.8 when the error is between 1 and 2 dBm, and to 1.2 when the error exceeds 2 dBm, then the current error of -2 dBm makes the gain factor , and then calculate the filter weight coefficient using the following formula: ; Set the basic filter weight The filter step sizes before and after are and ,constant To prevent the denominator from approaching zero, calculate: ; ; ; ; ; Since this value is too low and unreasonable, the gain coefficient needs to be adjusted or Threshold To ensure that the weight value is within a reasonable range.

[0039] S403: Calling the optimized filter weight coefficient, screening the filter parameter set of the communication signal, matching multiple parameter combinations, calculating the filter optimization result, and obtaining the communication signal filter optimization parameter set.

[0040] Set multiple filter parameter combinations, such as filter order , filter cutoff frequency , filter gain Etc., select different parameter values ​​for combination matching, as shown in the following table: Table 4 Filter parameter combination table

[0041] As shown in Table 4, in high interference areas, higher order and higher gain filter parameters can be selected, such as combination C, while in low interference areas, lower order combination A can be selected, and then the filter optimization results are calculated, such as comparing the signal quality indicators before and after filtering. (Signal-to-noise ratio), assuming the original signal-to-noise ratio dB, after applying combination C, it is increased to dB, it indicates that the optimization effect of the combination is better. Finally, the communication signal filtering optimization parameter set is obtained and applied to the system.

[0042] See also Figure 6 , the specific steps for obtaining the anti-interference reconstruction value of the communication signal are: S501: Call the communication signal filtering optimization parameter set, adjust the filtering step size, weight coefficient, residual error threshold and gain coefficient, calculate the energy change rate of the signal, extract the energy fluctuation trend of the continuous signal interval, compare the change amplitude of adjacent intervals and filter the high change interval, and obtain the signal energy change trend value; First, the initial filtering step size is set based on the historical signal data. The initial value setting can refer to the mean value of the signal change rate and be adjusted in combination with empirical parameters. For example, the average signal change rate is , then the initial step size is set to to Then the filter weight coefficient is calculated. The weight coefficient can be calculated by weighted summing the changes of the first five data points to reflect the signal trend. If the energy change of the current data point increases by more than the previous data point, , a larger weight is given, such as setting the range to to , then the current point can be set with weight ; Next, set the residual error threshold and calculate the threshold range based on the signal noise root mean square (RMS). If the historical noise RMS value is , the threshold can be set at to At the same time, the setting of the gain coefficient can be calculated based on the signal energy mean, such as the current signal energy mean is , then the gain factor can be set to to ; When calculating the energy change rate of a signal, it is necessary to calculate the energy difference between consecutive signal points and perform statistical analysis after normalization. For example, the signal energy at a certain moment is , the previous moment is , then the rate of change is ; Then, the energy fluctuation trend of the continuous signal interval is extracted, the energy mean curve is calculated by moving the window, and the mean changes of adjacent windows are compared. For example, if the window size is set to 10 sampling points and the mean of the previous interval is , the mean of the latter interval is , then the change range is ; When filtering high change intervals, you can set a change threshold, such as selecting an energy change greater than The interval, as mentioned above, varies is selected, and finally the signal energy change trend value is obtained.

[0043] Table 5 Signal filtering optimization parameter setting table

[0044] As shown in Table 5, each parameter setting is obtained by calculating based on the signal characteristics to ensure the rationality of signal processing.

[0045] S502: Based on the signal energy change trend value, set the regression modeling parameter adjustment range, calculate the model error under the differentiated parameter combination, compare the error fluctuation and select the parameter combination with the smallest error, adjust the regression modeling strategy and perform data fitting to obtain the optimal regression modeling parameters; It is necessary to analyze the distribution of trend values ​​first, such as setting the adjustment range based on the historical change amplitude of the signal. Percentile value, such as the 90% percentile of the trend value is , the adjustment range can be set to to ; When calculating the model error under differentiated parameter combinations, it is necessary to perform error calculations for multiple parameter combinations. For example, if the parameter combination is set to , calculation error , represents the actual measurement data, represents the value obtained through prediction, Indicates the error value, assuming that a certain combination corresponds to the error value , another combined error value , then the parameter combination with the smallest screening error ; When comparing error fluctuations, it is necessary to calculate the standard deviation of the error of each parameter combination. For example, if the error of a set of data is , calculate the standard deviation as , if it is lower than the set threshold , then the fluctuation is considered small; subsequently, when adjusting the regression modeling strategy, the model is adjusted according to the parameter combination with the smallest error, such as the aforementioned combination with the smallest error is , then the parameter is used for regression modeling; finally, data fitting is performed and the fitting error is calculated, such as the true signal value , fitting signal value , error calculation , and finally obtain the optimal regression modeling parameters.

[0046] S503: Call the optimal regression modeling parameters, reconstruct the communication signal, and perform anti-interference adjustment in combination with the filter optimization parameters, and calculate the residual error value of the signal after anti-interference, using the formula: ; Calculate the anti-interference adjustment error value, and optimize and adjust the error to obtain the anti-interference reconstruction value of the communication signal; in, Represents the anti-interference reconstruction value of the communication signal, represents the original signal value, represents the signal value reconstructed by the regression model, represents the filter weight coefficient, Represents the dynamic change trend value of signal energy. Represents the energy adjustment threshold of the signal after anti-interference, represents the signal gain coefficient, Represents the total number of signal sampling points.

[0047] First, the reconstructed signal is calculated based on the optimal parameters, such as the aforementioned optimal parameters , calculate the reconstruction value , for a signal point ,calculate ; Combined with the filter optimization parameters for anti-interference adjustment, the residual error value of the signal after anti-interference is calculated using the formula: ; Among them, ,set up , , calculate the sum of squared errors: ; set up , , ; Calculate the denominator: ; ; ; Add the signal gain term ,have to: Denominator sum = 0.014 + 0.03 + 0.032 + 3.2 = 3.276; Finally, the anti-interference reconstruction value of the communication signal is calculated: .

[0048] The calculation shows that the error of the signal after anti-interference is controlled within a reasonable range and meets the optimization goal.

[0049] A communication signal interference elimination system, the communication signal interference elimination system is used to perform the above communication signal interference elimination method, the system comprises: Signal area marking module: obtains the time domain data of the communication signal, calculates the signal energy change rate, signal mutation rate, and phase drift rate, determines the signal interval type based on the calculation results, and marks the interval where the mutation rate exceeds the set threshold, the phase drift rate is abnormal, or the energy change rate fluctuates as a non-stationary interval, and obtains the non-stationary interval identification result of the communication signal; Segment error calculation module: call the identification result of the non-stationary interval of the communication signal, adjust the range of regression modeling parameters according to the mutation rate, phase drift rate, and energy change rate, reduce the instantaneous data dependence in the mutation area, calculate the signal reconstruction error, and adjust the regression modeling constraint weight and input signal data weight based on the error size to obtain the segment prediction error value of the communication signal; Channel state assessment module: call the communication signal segment prediction error value, calculate the communication channel signal-to-noise ratio and interference power fluctuation amplitude, select parameters according to the weighted joint assessment, calculate the channel state parameters according to the weight of the influencing factor, and obtain the communication channel interference intensity assessment value; Filter parameter optimization module: call the communication channel interference intensity evaluation value, adjust the filter step size setting according to the interference intensity, call the signal error information, dynamically adjust the filter weight coefficient, residual error threshold, and gain coefficient, calculate the filter performance index under differentiated parameter configuration, screen the parameter combination with the best interference suppression effect, and obtain the communication signal filter optimization parameter set; Signal anti-interference reconstruction module: call the communication signal filtering optimization parameter set, adjust the filtering step size, weight coefficient, residual error threshold, gain coefficient, calculate the signal energy change trend, reconstruct the signal based on the adjustment results of the regression modeling parameter range, and obtain the communication signal anti-interference reconstruction value.

[0050] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for eliminating communication signal interference, characterized in that: The following steps are involved: S1: Obtain the time domain data of the communication signal, calculate the signal energy change rate, signal mutation rate, and phase drift rate, distinguish the signal area according to the calculation result mark, and obtain the identification result of the non-stationary interval of the communication signal; S2: Based on the identification result of the non-stationary interval of the communication signal, call the mutation rate, phase drift rate, and energy change rate, adjust the range of regression modeling parameters, reduce the instantaneous data dependence in the mutation area, calculate the signal reconstruction error, adjust the regression modeling constraint weight and the input signal data weight, and obtain the communication signal segment prediction error value; S3: calling the communication signal segment prediction error value, calculating the communication channel signal-to-noise ratio and interference power fluctuation amplitude, obtaining the channel state parameters according to the weighted joint evaluation, and obtaining the communication channel interference intensity evaluation value; S4: Based on the communication channel interference intensity evaluation value, adjust the filter step size setting, call the signal error information to adjust the filter weight coefficient, the residual error threshold, and the gain coefficient, and obtain the communication signal filter optimization parameter set; S5: calling the communication signal filtering optimization parameter set, adjusting the filtering step size, weight coefficient, residual error threshold, and gain coefficient, calculating the signal energy change trend, adjusting the regression modeling parameter range, and obtaining the communication signal anti-interference reconstruction value.

2. The communication signal interference elimination method according to claim 1, characterized in that: The signal area includes a non-stationary area, a mutation area, and a signal interference area. The communication signal non-stationary interval identification result includes a signal energy change rate identification, a signal mutation rate identification, and a phase drift rate identification. The communication signal segmented prediction error value includes a mutation area error, a phase drift area error, and an energy change area error. The communication channel interference intensity evaluation value includes a signal-to-noise ratio evaluation value, an interference power fluctuation amplitude evaluation value, and a channel state parameter evaluation value. The communication signal filtering optimization parameter set includes a filtering step size setting value, a filtering weight coefficient, a residual error threshold, and a gain coefficient. The communication signal anti-interference reconstruction value includes a reconstructed signal energy value, a reconstructed signal phase value, and a reconstructed signal time domain eigenvalue.

3. The communication signal interference elimination method according to claim 2, characterized in that: The steps for obtaining the identification result of the non-stationary interval of the communication signal are specifically as follows: S101: Acquire time domain data of a communication signal, calculate the instantaneous energy of the signal at differentiated time points, and obtain the signal energy change rate by calculating the energy change rate of adjacent time points; S102: Based on the signal energy change rate, calculate the mutation degree of adjacent time points, define a mutation threshold, screen the time period in which the mutation rate exceeds the threshold, and obtain the signal mutation rate; S103: Based on the signal energy change rate and the signal mutation rate, calculate the change rate of the signal phase over time, using the formula: ； The phase drift rate is obtained by signal phase difference operation, and the signal energy change rate and signal mutation rate are combined to classify the signal area and obtain the identification result of the non-stationary interval of the communication signal; in, represents the phase drift rate, Representative The phase value at a time point, Representative The phase value at a time point, Representative The signal energy at a time point, Representative The signal energy at a time point, Represents the total number of time points.

4. The communication signal interference elimination method according to claim 3, characterized in that: The step of obtaining the communication signal segment prediction error value is specifically as follows: S201: Based on the identification result of the non-stationary interval of the communication signal, the mutation rate, phase drift rate, and energy change rate are called to segment the time series of the communication signal, calculate the mutation rate and drift change trend in each segment, and screen the time window of the mutation area, set the data weight in the time window, reduce the instantaneous data dependence, and obtain the mutation area time window; S202: Based on the mutation region time window, adjust the regression modeling parameter range, dynamically adjust the input signal data weight according to the signal characteristics of the mutation region, set the weight distribution coefficient of the differentiation interval, calculate the signal reconstruction error, and use the formula: ； The signal reconstruction error is obtained by operation, and the regression modeling constraint weight is adjusted to obtain the modeling error constraint weight; in, represents the signal reconstruction error, Representative The signal value in a time window, Representative The weight value of the time window, Represents the total number of time windows, represents the sum of the signals in all time windows, Represents the sum of weights of all time windows; S203: calling the modeling error constraint weight, combining the characteristic data of the signal mutation area, adjusting the constraint conditions of the regression modeling, calculating the signal prediction error under the differentiation interval, and obtaining the communication signal segment prediction error value.

5. The communication signal interference elimination method according to claim 4, characterized in that: The steps for obtaining the communication channel interference intensity evaluation value are specifically as follows: S301: Based on the segmented prediction error value of the communication signal, calculate the signal-to-noise ratio of the communication channel, extract the ratio of the error value to the signal power, normalize the ratio, measure the signal quality on the same scale, remove the deviation data, and obtain the channel signal-to-noise ratio evaluation value; S302: calling the channel signal-to-noise ratio evaluation value, calculating the interference power fluctuation amplitude of the communication channel, extracting the maximum and minimum values ​​of the interference signal power, and calculating the difference, calling the average power of the communication signal, normalizing the average power fluctuation value, so as to measure the degree of fluctuation of the interference power relative to the signal power, and obtaining the interference power fluctuation ratio; S303: calling the channel signal-to-noise ratio evaluation value and the interference power fluctuation ratio to calculate the channel availability under the differentiated signal interference condition, using the formula: ； Obtain channel state parameters by operation, and calculate the correlation between the channel state parameters and the interference signal to obtain the communication channel interference intensity assessment value; in, represents the communication channel interference intensity assessment value, represents the average signal power, represents the variance of communication noise, represents the maximum interference signal power, Represents the minimum interference signal power, represents the weighting coefficient of the differentiated signal segment, represents the power of the kth signal segment, represents the interference power of the kth signal segment, Represents the total number of signal segments.

6. The communication signal interference elimination method according to claim 5, characterized in that: The steps of obtaining the communication signal filtering optimization parameter set are specifically as follows: S401: Based on the communication channel interference intensity evaluation value, calculate the interference intensity change rate, identify the impact range, set the filter step adjustment value corresponding to the differentiated interference interval, compare the change of the step before and after the adjustment, and generate the filter step adjustment coefficient; S402: Call the filter step adjustment coefficient, adjust the filter weight coefficient according to the signal error information, and select the gain coefficient corresponding to the differentiated error range in combination with the residual error threshold, using the formula: ； Calculate the adjusted filter weight coefficient to obtain the optimized filter weight coefficient; in, represents the adjusted filter weight coefficient, Represents the filter weight coefficient before adjustment, Represents the current signal error value, represents the residual error threshold, represents the gain factor, Represents the filter step setting value at the previous moment, Represents the current filter step setting value, The representative constant is used to avoid the denominator approaching zero; S403: calling the optimized filter weight coefficient, screening the filter parameter set of the communication signal, matching multiple parameter combinations, calculating the filter optimization result, and obtaining the communication signal filter optimization parameter set.

7. The communication signal interference elimination method according to claim 6, characterized in that: The steps for obtaining the communication signal anti-interference reconstruction value are specifically as follows: S501: calling the communication signal filtering optimization parameter set, adjusting the filtering step size, weight coefficient, residual error threshold and gain coefficient, calculating the energy change rate of the signal, extracting the energy fluctuation trend of the continuous signal interval, comparing the change amplitude of adjacent intervals and screening the high change interval, and obtaining the signal energy change trend value; S502: Based on the signal energy change trend value, set the regression modeling parameter adjustment range, calculate the model error under the differentiated parameter combination, compare the error fluctuation and select the parameter combination with the smallest error, adjust the regression modeling strategy and perform data fitting to obtain the optimal regression modeling parameters; S503: Call the optimal regression modeling parameters, reconstruct the communication signal, and perform anti-interference adjustment in combination with the filter optimization parameters, and calculate the residual error value of the signal after anti-interference, using the formula: Calculate the anti-interference adjustment error value, and optimize and adjust the error to obtain the anti-interference reconstruction value of the communication signal; in, Represents the anti-interference reconstruction value of the communication signal, represents the original signal value, represents the signal value reconstructed by the regression model, represents the filter weight coefficient, Represents the dynamic change trend value of signal energy. Represents the energy adjustment threshold of the signal after anti-interference, represents the signal gain coefficient, Represents the total number of signal sampling points.

8. A communication signal interference elimination system, characterized in that: According to the communication signal interference elimination method according to any one of claims 1 to 7, the system comprises: Signal area marking module: obtains the time domain data of the communication signal, calculates the signal energy change rate, signal mutation rate, and phase drift rate, determines the signal interval type based on the calculation results, and marks the interval where the mutation rate exceeds the set threshold, the phase drift rate is abnormal, or the energy change rate fluctuates as a non-stationary interval, and obtains the non-stationary interval identification result of the communication signal; Segmentation error calculation module: calling the identification result of the non-stationary interval of the communication signal, adjusting the range of regression modeling parameters according to the mutation rate, phase drift rate, and energy change rate, reducing the instantaneous data dependence in the mutation area, calculating the signal reconstruction error, and adjusting the regression modeling constraint weight and input signal data weight based on the error size to obtain the segmentation prediction error value of the communication signal; Channel state evaluation module: calling the communication signal segment prediction error value, calculating the communication channel signal-to-noise ratio and interference power fluctuation amplitude, based on weighted joint evaluation, screening parameters according to the weight of the influencing factor, calculating the channel state parameters, and obtaining the communication channel interference intensity evaluation value; Filter parameter optimization module: call the communication channel interference intensity evaluation value, adjust the filter step size setting according to the interference intensity, call the signal error information, dynamically adjust the filter weight coefficient, residual error threshold, and gain coefficient, calculate the filter performance index under the differentiated parameter configuration, screen the parameter combination with the best interference suppression effect, and obtain the communication signal filter optimization parameter set; Signal anti-interference reconstruction module: call the communication signal filtering optimization parameter set, adjust the filtering step size, weight coefficient, residual error threshold, gain coefficient, calculate the signal energy change trend, reconstruct the signal according to the adjustment result of the regression modeling parameter range, and obtain the communication signal anti-interference reconstruction value.

Images