Radar intermittent sampling forwarding interference reconstruction cancellation method based on slice matching

Abstract

The invention belongs to the technical field of radar anti-interference, and particularly relates to a radar intermittent sampling forwarding interference reconstruction cancellation method based on slice matching. The method specifically comprises the steps of interference slice outer edge estimation, interference slice number estimation, slice matching and interference slice reconstruction and cancellation, in the interference slice number estimation process, radar time domain echo signals are processed to obtain a binary time-frequency graph, and the binary time-frequency graph serves as a first time-frequency domain filter HOtsu to filter an original time-frequency graph M0; processing the first filtering result # imgabs0 # to obtain a binarized time-frequency graph, and filtering the original time-frequency graph M0 by taking the binarized time-frequency graph as a second time-frequency domain filter Hmean; and processing the second filtering result # imgabs1 # to obtain a binary time frequency graph # imgabs2 #, and estimating the number of interference slices according to the binary time frequency graph # imgabs3 #.

Description

Technical Field

[0001] The present invention belongs to the technical field of radar anti-interference, and in particular relates to a radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching. Background Art

[0002] As an essential component of modern electronic warfare, radar electronic countermeasures and electronic counter-countermeasures have long been a hot topic for researchers worldwide. The development of Digital Radio Frequency Memory (DRFM) technology has enriched the types of jamming. Interrupted Sampling Repeater Jamming (ISRJ) is a typical type of jamming generated by a DRFM jammer. After intercepting a radar signal, the jammer samples and stores a portion of it, then processes and forwards the stored signal slices. ISRJ is highly correlated with the target echo and, after pulse compression, gains gain, creating multiple false targets that are difficult to distinguish from real targets. This poses a significant challenge to the operation of linear frequency modulation radars.

[0003] In recent years, researchers have proposed a variety of methods to counter ISRJ, primarily categorized as filtering and reconstruction cancellation. Filtering methods exploit the differences in the characteristics of target and interference echoes in different domains to construct filters, which are then used to remove interference components. With the introduction of machine learning into radar anti-interference and its successful application in interference identification, interference suppression methods that utilize machine learning to construct adaptive filters have emerged. However, these methods are computationally complex and require a large number of training samples. Reconstruction cancellation methods primarily reconstruct the interference by estimating various interference parameters, and then use this reconstructed interference to cancel the interference components in the radar echo. Existing reconstruction cancellation methods rely on combining various interference parameter estimation methods to estimate parameters such as the number, width, number of retransmissions, and center frequency of interference slices. However, due to the limited accuracy of the parameter estimation methods, the accumulation of errors in the estimation of various interference parameters makes it difficult for these methods to accurately reconstruct the interference slices, resulting in a large amount of residual interference after cancellation. Summary of the Invention

[0004] To solve the above problems, the present invention proposes a radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching, which can achieve a robust improvement in the intermittent sampling and forwarding interference reconstruction cancellation performance.

[0005] The technical solutions for implementing the present invention are as follows:

[0006] A radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching, the specific process is as follows:

[0007] Interference slice outer edge estimation: Use the box difference filter h(k) to filter the radar time domain echo x(k), and estimate the interference slice outer edge based on the filtering result;

[0008] Interference slice number estimation: The radar time domain echo signal is processed to obtain a binary time-frequency graph, which is used as the first time-frequency domain filter H Otsu Filter the original time-frequency graph M0; filter the first filtering result The binary time-frequency graph is obtained by processing and used as the second time-frequency domain filter H mean Filter the original time-frequency graph M0; filter the second filtering result Processing to obtain binary time-frequency graph According to the binarized time-frequency graph Estimate the number of interfering slices; obtain the positions of overlapping edges of inner adjacent slices by averaging the outer edges;

[0009] Slice matching: Using the estimated start and end edge positions of the interference slice, each interference slice is extracted from the radar echo as a reference signal s ref and compare it with the radar transmission signal s t Perform matching to obtain the exact location of the interference slice;

[0010] Interference slice reconstruction and cancellation: Using the slice matching results, slices corresponding to the radar transmit signal are intercepted to reconstruct and cancel the interference slices.

[0011] Optionally, the first filtering result of the present invention The binary time-frequency graph obtained by processing is: Calculate the first filtering result The mean η1 of the interference distribution area is used as the binarization threshold. Perform binarization processing to obtain a binary time-frequency graph.

[0012] Optionally, the second filtering result of the present invention The binary time-frequency graph obtained by processing is: Calculate the filtering result The mean value η2 of the interference distribution area is used as the binarization threshold. Perform binarization processing to obtain a binary time-frequency graph.

[0013] Optionally, the specific process of estimating the number of interference slices according to the present invention is as follows:

[0014] First, the time-frequency map M0 of the radar time-domain echo is calculated using STFT, and the time-frequency map M0 is binarized using Otsu. The binarized time-frequency map is used as the first time-frequency domain filter H Otsu, calculate the Hadamard product of the original time-frequency domain image M0 and the first time-frequency domain filter, complete the first time-frequency domain filtering, and the filtering result is recorded as

[0015] Secondly, calculate the first filtering result The mean η1 of the interference distribution area is used as the binarization threshold. Perform binarization processing, and the resulting binarized time-frequency graph is used as the second time-frequency domain filter H mean , calculate the original time-frequency map M0 and the second time-frequency domain filter H mean The Hadamard product of is used to complete the second time-frequency domain filtering. The filtering result is recorded as

[0016] Again, after time-frequency domain filtering, the filtering results are calculated The mean value η2 of the interference distribution area is used as the binarization threshold. Perform binarization processing to obtain the final binarized time-frequency graph

[0017] Finally, the binarized time-frequency graph After calculating the integral along the frequency dimension, a binarization operation is performed to estimate the number of interfering slices. The number of slices is combined with the position of the outer edge of the slice, and the position of the overlapping edges of the inner adjacent slices is obtained by averaging the outer edge, thereby obtaining the position of the starting and ending edges of the interfering slice.

[0018] Optionally, the box differential filter h(k) of the present invention is:

[0019]

[0020] Wherein, 2N+1 is the filter order, and the length M of the discrete input signal x(k) and the filter order 2N+1 usually satisfy M>>2N+1.

[0021] Optionally, the present invention uses a box differential filter h(k) to filter the radar time domain echo x(k) as follows:

[0022]

[0023] Where x(k) is the discrete input signal.

[0024] Optionally, the slice matching process of the present invention is: using the estimated positions of the start and end edges of the interference slice, each interference slice is extracted from the radar echo as a reference signal s ref , which corresponds to the impulse response function of the matched filter; the radar transmit signal s t Corresponding to the input signal of the matched filter, the following slice matching process is performed:

[0025] y=ifft[fft(s t )·fft(s ref ) * ]

[0026] Among them, ifft stands for inverse fast Fourier transform, and fft stands for fast Fourier transform.

[0027] Optionally, in the slice matching stage, the present invention uses an iterative optimization method to further optimize the matching results for the complete interference slice. The specific process is: within a certain position error range, the cancellation residual of the actual interference slice and the reconstructed interference slice is calculated for each position, and the position with the smallest cancellation residual is taken as the optimal interference slice position x opt ;

[0028]

[0029] Where slice0 is the actual interference slice in the radar echo, x0 is the interference slice position obtained by slice matching, slice(x) is the reconstructed interference slice, and n is the error range of the estimated position of the reconstructed slice.

[0030] Optionally, in the slice matching stage, the present invention, for incomplete interference slices: traverse the inner edge positions and calculate the maximum cumulative sum to obtain the correct inner edge position;

[0031]

[0032] Where I represents the position of the inner edge, y I (i) represents the matching result of the i-th slice.

[0033] Optionally, the specific process of interference slice reconstruction and cancellation described in the present invention is: using the results obtained by slice matching to intercept slices corresponding to the positions of the radar transmission signal as reconstructed interference slices; after performing pulse compression on the radar echo and the reconstructed interference slice respectively, the amplitude of the pulse compression result is normalized, and the normalized result is used to suppress the interference in the radar echo.

[0034] Beneficial effects:

[0035] First, the present invention's slice-matching-based radar intermittent sampling and forwarding interference reconstruction cancellation method differs from traditional reconstruction cancellation methods. This method employs an adaptive time-frequency domain filtering method to accurately estimate the number of interfering slices. Combined with the estimation of the outer edges of the interfering slices, the locations of the overlapping edges within consecutive forwarded slices can be more accurately determined.

[0036] Second, the present invention's radar intermittent sampling and forwarding interference reconstruction cancellation method, based on slice matching, differs from traditional reconstruction cancellation methods. This method employs a slice matching method to determine the position of each interference slice in the radar transmit signal. This method does not require the center frequency of the interference slice, a parameter that is difficult to accurately estimate and significantly impacts the final interference suppression performance. Simulation experiments demonstrate that the proposed method exhibits superior interference suppression performance compared to other methods in various scenarios.

[0037] Third, the proposed radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching differs from traditional reconstruction cancellation methods. Existing methods cannot handle the case of incomplete slices at the end of the radar echo, while the proposed method still has good interference suppression performance in this case.

[0038] Fourthly, the present invention can solve the problem that the performance of existing radar intermittent sampling and repeated forwarding interference reconstruction cancellation methods is easily affected by the accuracy of interference parameter estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0040] Figure 1 These are the schematic diagrams of three types of ISRJs. (a) is the schematic diagram of ISDRJ, (b) is the schematic diagram of ISPRJ, and (c) is the schematic diagram of ISPRJ.

[0041] Figure 2 Figures 1 and 2 show the time domain waveforms and time-frequency domain distributions of the radar transmission signal and three types of ISRJs. (a) shows the LFM time domain waveform, (b) shows the LFM time-frequency domain distribution, (c) shows the ISDRJ time domain waveform, (d) shows the ISDRJ time-frequency domain distribution, (e) shows the ISPRJ time domain waveform, (f) shows the ISPRJ time-frequency domain distribution, (g) shows the ISCRJ time domain waveform, and (h) shows the ISCRJ time-frequency domain distribution.

[0042] Figure 3 is a flow chart of the method proposed by the present invention;

[0043] Figure 4 is the DOB filtering result of ISPRJ;

[0044] Figure 5This is the adaptive time-frequency domain filtering process. (a) is the time-domain waveform of the input radar echo signal, (b) is the time-frequency domain distribution of the input radar echo signal, (c) is the first time-frequency domain filter, (d) is the first filtering result, (e) is the second time-frequency domain filter, and (f) is the second filtering result.

[0045] Figure 6 It is the result of binarized time-frequency distribution and frequency domain integration. Among them, (a) is the binarized time-frequency distribution, (b) is the frequency domain integration result.

[0046] Figure 7 is the slice matching process. Where (a) is the transmitted signal and the extracted interference slice, and (b) is the slice matching result;

[0047] Figure 8 This is a special case where the slice at the end of the PRT is incomplete. (a) is the result of evenly dividing the outer edge, and (b) is the result of reconstructing the interference using the division method of (a);

[0048] Figure 9 for Figure 8 Slice matching results of slices 1-4 in . Among them, (a) is the slice matching result of slice 1, (b) is the slice matching result of slice 2, (c) is the slice matching result of slice 3, and (d) is the slice matching result of slice 4;

[0049] Figure 10 is the edge estimation result after correction;

[0050] Figure 11 1 is an edge estimation result diagram and an interference reconstruction result diagram of the simulation data processed by the method of the present invention, wherein (a) is an edge estimation result diagram, and (b) is a time domain result of interference reconstruction and a pulse compression result diagram;

[0051] Figure 12 This is a graph showing the pulse compression normalization cancellation results of the simulation data processed by the method of the present invention;

[0052] Figure 13 Figures 2 and 3 are the pulse compression normalization cancellation results of different methods for processing simulation data, where (a) is the result of processing ISDRJ, (b) is the result of processing ISPRJ, and (c) is the result of processing ISCRJ.

[0053] Figure 14 The Monte Carlo simulation results of different methods for processing simulation data are shown in Figure 1, where (a) is the result of processing ISDRJ, (b) is the result of processing ISPRJ, and (c) is the result of processing ISCRJ.

[0054] Figure 15This is the case where the interference slice at the end of the radar echo data is incomplete;

[0055] Figure 16 Figure 2 is a graph of the pulse compression normalization cancellation results of different methods for handling incomplete slices. DETAILED DESCRIPTION

[0056] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0057] It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments may be combined with each other; and, based on the embodiments in this disclosure, all other embodiments obtained by persons of ordinary skill in the art without creative work are within the scope of protection of this disclosure.

[0058] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on this disclosure, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement an apparatus and / or practice a method. In addition, other structures and / or functionalities other than one or more of the aspects described herein can be used to implement this apparatus and / or practice this method.

[0059] Use the Linear Frequency Modulation (LFM) signal as the radar transmission signal:

[0060]

[0061] Among them, A, T p , f0 represent the amplitude, pulse width and carrier frequency of the LFM signal respectively, k=B / T p is the frequency modulation slope of the signal, B is the bandwidth of the signal, Indicates that the time τ is the center and the width is T p rectangular wave.

[0062] This method conducts simulation analysis on three types of ISRJ, including Interrupted Sampling and Direct Repeater Jamming (ISDRJ), Interrupted Sampling and Periodic Repeater Jamming (ISPRJ), and Interrupted Sampling and Cyclic Repeater Jamming (ISCRJ).

[0063] The first type: Intermittent sampling direct forwarding interference (ISDRJ)

[0064] After partially sampling the intercepted radar signal, the jammer directly implements the forwarding process and then repeats this process until the pulse repetition time (PRT) ends. The formula is expressed as:

[0065]

[0066] Among them, N is the number of sampling times, n is the sampling sequence number, T j is the sampling duration, i.e. the slice width. The basic principle of ISDRJ is as follows Figure 1 As shown in (a).

[0067] The second type: Intermittent Sampling Repeated Forward Interference (ISPRJ)

[0068] After partially sampling the intercepted radar signal, the jammer forwards it multiple times continuously and repeats this process. The formula is:

[0069]

[0070] Among them, M is the number of consecutive forwarding times, and m is the forwarding sequence number. The basic principle of ISPRJ is as follows Figure 1 (b) shown.

[0071] The third type: Intermittent sampling cyclic forwarding interference (ISCRJ)

[0072] After the jammer samples part of the intercepted radar signal, it forwards all stored signal slices in reverse order and repeats this process. The formula is:

[0073]

[0074] Among them, m is the forwarding sequence number, a n T J is the starting time of the nth sampling, b mnis the delay corresponding to the mth forwarding of the signal slice obtained by the nth sampling. The basic principle of ISCRJ is as follows: Figure 1 (c) shown.

[0075] To better describe the interference, Figure 2 The time domain waveforms and time-frequency distributions of radar transmission signals and three types of ISRJ are given.

[0076] The embodiment of the present application proposes a radar intermittent sampling forwarding interference reconstruction cancellation method based on slice matching, such as Figure 3 As shown, the specific process is:

[0077] Step 1: Estimation of the outer edge of the interference slice;

[0078] The radar time domain echo x(k) is filtered using the box differential filter h(k), and the DOB filter with an order of 2N+1 is defined as:

[0079]

[0080] Then the process of DOB filtering of discrete input signal x(k), k=1,2,...,M can be expressed as:

[0081]

[0082] Among them, the length M of the discrete input signal x(k) and the DOB filter order 2N+1 usually need to satisfy M>>2N+1. Since this formula is actually a discrete convolution process, j The value range of is (-∞,+∞). The outer edge position of the interference slice is reflected in the filtering result. The DOB filter actually uses the amplitude discontinuity between the pulse and the noise floor to detect the pulse edge. Therefore, when the signal to be detected is a multiple-time continuous forwarding signal such as ISPRJ or ISCRJ, only the position of the outer edge can be extracted, and the position of the overlapping edge of the internal adjacent slices needs to be determined by other methods, such as Figure 4 shown.

[0083] Step 2: Estimation of the number of interference slices;

[0084] The time-frequency distribution of radar echo is calculated by STFT, and the input signal (such as Figure 5 (a)) of the STFT time-frequency diagram (as shown in Figure 5 (b) is binarized, and the resulting binarized time-frequency image is used as the first time-frequency domain filter (as shown in Figure 5 (c)), calculate the Hadamard product of the original time-frequency graph and the time-frequency domain filter, and complete the first time-frequency domain filtering (as shown in Figure 5 (d)

[0085]

[0086] Among them, M0 represents the time-frequency diagram of the input signal, that is, the time-frequency domain matrix, H Otsu represents the matrix obtained by binarizing M0 using the Otsu method, i.e., the first time-frequency domain filter. "⊙" represents the Hadamard product operation. The filtering result obtained in this step preserves the approximate distribution of the interference slices, ensuring that the next step is not affected by noise and target echoes.

[0087] Then, the mean value η1 of the interference distribution area in the first time-frequency domain filtering result image is calculated as the binarization threshold. Perform binarization processing and obtain the binarized time-frequency graph H mean As the second time-frequency domain filter (such as Figure 5 (e)), calculate the original time-frequency map M0 and the second time-frequency domain filter H mean The Hadamard product of , completes the second time-frequency domain filtering (such as Figure 5 (f)), expressed as:

[0088]

[0089] Among them, |·| represents the modulo operation, N represents The number of points in the interference distribution area, (m,n) represents the coordinates of each point. Filtering results The distribution area of interfering slices is further reduced, which can prevent the slices after binarization from being incomplete or inseparable.

[0090] After time-frequency domain filtering, calculate The mean value η2 of the interference distribution area is used as the binarization threshold. Perform binarization processing to obtain the final binarized time-frequency graph

[0091]

[0092] The results are as follows Figure 6 As shown in (a), the binarized time-frequency graph After integrating along the frequency dimension, binarization is performed to estimate the number of interfering slices. Combining the number of slices with the location of the outer edges of the slices, the locations of the overlapping edges of adjacent inner slices can be obtained by averaging the outer edges.

[0093] Step 3: Slice matching;

[0094] Using the estimated positions of the interference slice start and end edges, each interference slice is extracted from the radar echo as a reference signal s ref, which corresponds to the impulse response function of the matched filter; the radar transmit signal s t Corresponding to the input signal of the matched filter, the following slice matching process is performed:

[0095] y=ifft[fft(s t )·fft(s ref ) * ]

[0096] There are two cases of interference slices in radar echoes. Generally, the interference slices are relatively complete, but the interference slices at the end of the PRT may be incomplete and require additional processing.

[0097] 1. Matching method of complete interference slice

[0098] like Figure 7 As shown in (a), the signal in the lower sub-graph is the reference signal, the signal in the upper sub-graph is the radar transmission signal, and the red dotted line indicates the actual position of the interference slice. Figure 7 (b) The peak value in the slice matching result is used to obtain the position of the interference slice in the transmitted signal. On this basis, the reconstructed interference slice can be extracted from the transmitted signal. Repeat the above steps until all interference slices are reconstructed. Considering that a lower JNR or a smaller slice width may affect the performance of slice matching, that is, cause a certain position error, an iterative optimization method can be used to further optimize the matching results. The specific operation is: within a certain position error range, the cancellation residual of the actual interference slice and the reconstructed interference slice is calculated for each position, and the position with the smallest cancellation residual is taken as the optimal interference slice position x opt The iterative optimization process is expressed as:

[0099]

[0100] Where slice0 is the actual interference slice in the radar echo, x0 is the interference slice position obtained by slice matching, slice(x) is the reconstructed interference slice, and n is the error range of the estimated position of the reconstructed slice.

[0101] 2. Matching method for incomplete interference slices

[0102] like Figure 8 As shown in (a), the blue, yellow, and green parts at the end of the PRT are interference slices, of which the green part is incomplete. The parts marked with "2", "3", and "4" are slices obtained by evenly dividing the outer edge. It can be seen that in this case, simply dividing the outer edge by the number of slices will obviously cause edge estimation errors, thereby leading to incorrect interference reconstruction results, such as Figure 8 As shown in (b), for slice 2, since only a part of the slice is left, its corresponding peak is lower than that of slice 1, as shown in Figure 9 (a) and Figure 9 (b) shows that slices 3 and 4 contain components from other slices. This difference is reflected in the fact that after these slices are processed by the aforementioned slice matching method, two peaks with different positions will appear. The size of the peak is related to the width of the two slice components. Therefore, either of the two peaks must be lower than the peak corresponding to slice 1, as shown in Figure 9 (c) and Figure 9 As shown in (d). Take the maximum value of the matching results of slices 2, 3, and 4 and add them together. When the inner edge is correctly divided, the cumulative sum reaches the maximum value. Considering that the outer edge estimate of the continuously forwarded slices is accurate and the number of slices within the outer edge has also been estimated, we only need to traverse the inner edge position and calculate the maximum cumulative sum to obtain the correct inner edge position. This process can be expressed as:

[0103]

[0104] Where I represents the position of the inner edge, y I (i) represents the matching result of the i-th slice.

[0105] In this case, the slice width obtained by the average division method is always smaller than the actual slice width. Therefore, we can use the inner edge position obtained by the average division as the starting point and gradually move the inner edge backward to traverse. The corrected edge estimation result is as follows Figure 10 shown.

[0106] Step 4: Interference slice reconstruction and cancellation;

[0107] Based on the estimated information above, the interference can be reconstructed by intercepting the transmitted signal. Using the slice matching results, slices corresponding to the radar transmitted signal are intercepted and used as reconstructed interference slices. Because the energy of the interference in the radar echo is unknown, after all interference slices are reconstructed, pulse compression amplitude normalization is used to cancel it out. Specifically, after pulse compression is performed on the radar echo and the reconstructed interference slices, the amplitude of the pulse compression results is normalized. The interference in the radar echo is suppressed using the following method:

[0108] X out =X in -X rec

[0109] Among them, X in is the normalized result of radar echo pulse compression, X rec is the normalized result of pulse compression for reconstructing the interference slice.

[0110] The simulation uses Matlab software to build the ISRJ model. A Ku-band LFM radar is used. The received radar echo includes one target and one ISPRJ echo. The simulation parameters are shown in Table 1.

[0111] Table 1 Simulation parameter settings

[0112]

[0113]

[0114] At this time, since the actual slice width obtained at the last sampling is less than the sampling time, two slice widths and two interference types (ISPRJ and ISDRJ) are formed simultaneously in the entire radar echo, as shown in Figure 2. Figure 11 (a) The results of edge estimation, interference slice reconstruction and pulse compression normalization cancellation are shown as follows: Figure 11 and Figure 12 As shown (for comparison, Figure 11 (b) The original echo in the upper subfigure has been pre-normalized. It can be seen that the proposed method can accurately estimate all edges of consecutive forwarded slices, whether they are outer edges or overlapping edges of adjacent slices. The slice matching results also accurately indicate the location of the interfering slice in the transmitted signal. Therefore, the reconstructed interfering slice achieves good cancellation of the interference components in the original radar echo.

[0115] On this basis, the parameter settings in Table 1 remain unchanged, and the interference types are set to ISDRJ, ISPRJ, and ISCRJ respectively. The reconstruction cancellation method based on accurate edge detection (hereinafter referred to as AED) and the reconstruction cancellation method based on bidirectional constant false alarm rate detector (hereinafter referred to as Bi-CFAR) are compared. The pulse compression results processed by the three methods are shown in Figure 2. Figure 13 As shown in the figure, (a) shows the result of ISDRJ processing, (b) shows the result of ISPRJ processing, and (c) shows the result of ISCRJ processing. It can be seen that Bi-CFAR processing leaves the most residual interference components, and the AED's interference suppression effect is improved by approximately 5 to 10 dB compared to Bi-CFAR. In comparison, the proposed method has the best interference suppression performance, almost completely eliminating the interference components in the original echo.

[0116] To compare the performance stability of the three methods, the input JNR range was set to 10-30dB with a step size of 1dB, and 200 Monte Carlo simulations were performed. The Signal-to-Interference Ratio Improvement Factor (SJRIF) was used as the evaluation metric:

[0117]

[0118] SJRIF=SJR after -SJR before

[0119] Among them, P s 、P j are the target and interference powers after pulse compression, respectively. s Calculated by the pulse compression amplitude at the actual target position, P j Calculated by the maximum pulse compression amplitude within the false target formation range. before 、SJR after They represent the SJR of the pulse compression results before and after interference suppression. Figure 14 As shown. Among them, (a) is the result diagram of processing ISDRJ, (b) is the result diagram of processing ISPRJ, and (c) is the result diagram of processing ISCRJ. It can be seen that in the AED method, due to the insufficient accuracy of its parameter estimation method, the improvement of the output signal-to-interference ratio is limited. Specifically, the output of the edge estimation method TDDC used by the AED method depends on the estimated slice width of the interference waveform. In the case of continuous forwarding slices, the estimated interference slice width is actually the sum of the widths of multiple continuous forwarding slices. Therefore, the periodic pulses in the TDDC edge estimation result are difficult to reflect the true slice edge position. In addition, this method will also produce certain errors when estimating the slice center frequency, which greatly affects the interference suppression performance. In Bi-CFAR, the bidirectional CFAR method used for edge estimation becomes unstable as the input JNR decreases. When the input JNR is greater than 20dB, as the input JNR increases, the interference cancellation residual of AED and Bi-CFAR continues to increase, resulting in a very slow improvement in SJRIF. Compared with AED and Bi-CFAR, the proposed method not only significantly improves SJRIF but also becomes more stable with the change of input JNR.

[0120] Figure 15 The example shows an incomplete jammer slice at the end of the radar echo data. The simulation parameters for this case remain the same as in Table 1, except that the jammer sampling duration and target delay are modified to 2 μs and 8.8 μs, respectively. Figure 16 Figure 2 shows the pulse compression normalization cancellation results for different methods in this situation. It can be seen that due to the lack of an edge estimation strategy for incomplete slices, the interference residual after AED and Bi-CFAR processing increases significantly. The proposed method uses slice matching to not only reconstruct the interfering slices but also correct the slice edges. Compared with the Bi-CFAR method, which also averages the overlapping edges of consecutively forwarded slices, the proposed method still maintains stable and excellent interference suppression performance.

[0121] To avoid the loss of interference suppression performance caused by the accumulation of multiple parameter estimation errors, this paper proposes a radar intermittent sampling forwarding interference reconstruction and cancellation method based on slice matching. First, DOB filtering is performed to accurately estimate the outer edges of the slices. Then, the number of slices is estimated using STFT and time-frequency filtering. Combined with the outer edge estimation results, the overlapping edges of consecutive forwarded slices are located. Finally, interfering slices are extracted from the radar echoes for slice matching. Interference reconstruction and cancellation are achieved after pulse compression amplitude normalization. Simulation results demonstrate the superior performance of this method.

[0122] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A radar intermittent sampling forwarding interference reconstruction cancellation method based on slice matching, characterized in that: The specific process is: Step 1: Estimation of the outer edge of the interference slice: Filter the radar time domain echo x(k) using a box difference filter h(k); Step 2: Estimation of the number of interference slices: Process the filtered time domain echo signal to obtain a binary time-frequency graph, and use it as the first time-frequency domain filter H Otsu Filtering the original time-frequency map M0; The first filtering result The binary time-frequency graph is obtained by processing and used as the second time-frequency domain filter H mean Filter the original time-frequency map M0; The second filtering result Processing to obtain binary time-frequency graph According to the binarized time-frequency graph Estimate the number of interfering slices; Obtain the positions of the overlapping edges of the inner adjacent slices by averaging the outer edges; Step 3: Slice matching: Using the estimated start and end edge positions of the interference slice, each interference slice is extracted from the radar echo as a reference signal s ref and compare it with the radar transmission signal s t Perform matching to obtain the exact location of the interference slice; Step 4: Interference slice reconstruction and cancellation: Using the slice matching results, slices corresponding to the radar transmit signal are intercepted to reconstruct and cancel the interference slices.

2. The radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching according to claim 1 is characterized in that: The first filtering result The binary time-frequency graph obtained by processing is: Calculate the first filtering result The mean η1 of the interference distribution area is used as the binarization threshold. Perform binarization processing to obtain a binary time-frequency graph.

3. The radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching according to claim 1 or 2 is characterized in that: The second filtering result The binary time-frequency graph obtained by processing is: Calculate the filtering result The mean value η2 of the interference distribution area is used as the binarization threshold. Perform binarization processing to obtain a binary time-frequency graph.

4. The radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching according to claim 3 is characterized in that: The specific process of step S2 is: First, the time-frequency map M0 of the radar time-domain echo is calculated using STFT, and the time-frequency map M0 is binarized using Otsu. The binarized time-frequency map is used as the first time-frequency domain filter H Otsu , calculate the Hadamard product of the original time-frequency domain image M0 and the first time-frequency domain filter, complete the first time-frequency domain filtering, and the filtering result is recorded as Secondly, calculate the first filtering result The mean η1 of the interference distribution area is used as the binarization threshold. Perform binarization processing, and the resulting binarized time-frequency graph is used as the second time-frequency domain filter H mean , calculate the original time-frequency map M0 and the second time-frequency domain filter H mean The Hadamard product of is used to complete the second time-frequency domain filtering. The filtering result is recorded as Again, after time-frequency domain filtering, the filtering results are calculated The mean value η2 of the interference distribution area is used as the binarization threshold. Perform binarization processing to obtain the final binarized time-frequency graph Finally, the binarized time-frequency graph After calculating the integral along the frequency dimension, a binarization operation is performed to estimate the number of interfering slices. The number of slices is combined with the position of the outer edge of the slice, and the position of the overlapping edges of the inner adjacent slices is obtained by averaging the outer edge, thereby obtaining the position of the starting and ending edges of the interfering slice.

5. The radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching according to claim 1 is characterized in that: The box differential filter h(k) is: Wherein, 2N+1 is the filter order, and the length M of the discrete input signal x(k) and the filter order 2N+1 usually satisfy M>>2N+1.

6. The radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching according to claim 5 is characterized in that: The filtering of the radar's time domain echo x(k) using the box differential filter h(k) is: y(k)=x(k)*h(k) Where x(k) is the discrete input signal.

7. The radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching according to claim 1 is characterized in that: The slice matching process is as follows: using the estimated positions of the interference slice start and end edges, each interference slice is extracted from the radar echo as a reference signal s ref , which corresponds to the impulse response function of the matched filter; the radar transmit signal s t Corresponding to the input signal of the matched filter, the following slice matching process is performed: y=ifft[fft(s t )·fft(s ref ) * ] Among them, ifft stands for inverse fast Fourier transform, and fft stands for fast Fourier transform.

8. The radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching according to claim 7 is characterized in that: In the slice matching stage, the iterative optimization method is used to further optimize the matching results for the complete interference slice. The specific process is: within a certain position error range, the cancellation residual of the actual interference slice and the reconstructed interference slice is calculated for each position, and the position with the smallest cancellation residual is taken as the optimal interference slice position x opt ; Where slice0 is the actual interference slice in the radar echo, x0 is the interference slice position obtained by slice matching, slice(x) is the reconstructed interference slice, and n is the error range of the estimated position of the reconstructed slice.

9. The radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching according to claim 7 is characterized in that: In the slice matching stage, for incomplete interference slices: traverse the inner edge positions and calculate the maximum cumulative sum to obtain the correct inner edge position; Where I represents the position of the inner edge, y I (i) represents the matching result of the i-th slice.

10. The radar intermittent sampling and forwarding interference reconstruction cancellation method based on slice matching according to claim 1 is characterized in that: The specific process of interference slice reconstruction and cancellation is as follows: slices corresponding to the positions of the radar transmission signal are intercepted using the results of slice matching as reconstructed interference slices; after pulse compression is performed on the radar echo and the reconstructed interference slices respectively, the amplitude of the pulse compression result is normalized, and the normalized result is used to suppress the interference in the radar echo.

Images