A method, system, storage medium, and electronic device for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering.
By using a matched filtering-based method, utilizing prior orbital information, feature subspace projection, and robust principal component analysis, the mutual interference problem of spaceborne synthetic aperture radar was solved, achieving high-precision interference suppression and image quality improvement.
Patent Information
- Application Number
- CN202411701363.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-26
- Publication Date
- 2025-11-14
- Estimated Expiration
- 2044-11-26
AI Technical Summary
Spaceborne synthetic aperture radars are subject to mutual radio frequency interference during missions, resulting in a decrease in image quality. Existing interference suppression methods are insufficient in terms of high accuracy and effectiveness.
A matched filtering-based method is adopted to estimate the mutual interference parameters by using prior track information and time-frequency relationship. Combined with maximum eigenvalue sequence detection, feature subspace projection and robust principal component analysis, the interference signal is separated and suppressed.
It effectively suppresses mutual interference of spaceborne synthetic aperture radar, improves image quality, protects useful signals, and enhances system survivability.
Smart Images
Figure CN119535366B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of signal processing technology, and in particular to a method, system, storage medium, and electronic device for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering. Background Technology
[0002] Currently, Synthetic Aperture Radar (SAR) is an active remote sensing device. SAR systems can be carried by satellites and utilize their unique synthetic aperture technology to achieve high-resolution imaging of the Earth's surface. Compared to traditional optical remote sensing, spaceborne SAR has advantages such as all-weather, all-time operation, and the ability to penetrate clouds and fog, and has been widely used in Earth science research, disaster monitoring, and environmental monitoring. However, with the rapid development of global communication and the increasing scarcity of wireless spectrum, many electromagnetic devices share the same frequency band. Consequently, spaceborne SAR inevitably receives radio frequency interference from the electromagnetic environment during missions. In recent years, with the launch of more and more SAR satellites, spaceborne SAR satellites in the same band may receive mutual radio frequency interference from other SAR satellites at the orbital transition points. For example, the C-band European Sentinel-1 experienced prolonged mutual interference with China's Gaofen-3 and Canada's Radarsat-2 during its mission. Mutual radio frequency interference, due to its long duration, large bandwidth, and unstable pulse energy, severely degrades the quality of SAR images, greatly affecting subsequent product applications and posing a significant challenge to interference suppression.
[0003] Interference suppression is an effective means of removing the effects of interference and thus obtaining high-quality SAR image data. In order to reduce the negative impact of interference on SAR systems, research teams at home and abroad have conducted extensive research on the problem of interference suppression in SAR data and proposed a variety of classic interference suppression methods. These methods can be mainly divided into three categories: parametric, semi-parametric and non-parametric methods.
[0004] The core idea of parameterization methods is to establish an accurate mathematical model of the interference signal, thereby separating the interference signal from SAR echo data and achieving suppression. Typical parameterization methods include maximum likelihood estimation, least squares estimation, and parameter maximum likelihood and minimum mean square estimation algorithms. Researchers such as Zhou Feng proposed regularized time-frequency filtering and second-order multi-synchronization compression transform to suppress broadband interference. Yang Huizhang et al. proposed a two-dimensional spectral filtering method using a parametric linear frequency modulated signal model, which can suppress direct wave interference between SAR satellites. Yang et al. also proposed a block subspace filtering method to suppress interference in single-look complex SAR images, which can suppress various common radio frequency interferences in spaceborne SAR.
[0005] Semi-parametric methods suppress interference by transforming the complex signal separation problem into a hyperparameter optimization problem. For example, Nguyen et al. employed a sparsity-driven optimization strategy to separate interference signals from SAR data; Liu Hongqing et al. used a sparse representation reconstruction algorithm in the range-time-frequency domain to achieve interference suppression, providing new ideas for the design of subsequent related algorithms. Huang Yan et al. proposed a series of "low-rank + sparse" models, which showed good suppression effects. The core of current methods of this kind is to construct optimization models for different types of interference based on the low-rank and sparse characteristics of SAR echoes containing interference, so as to achieve the separation of interference from SAR signals.
[0006] Non-parametric methods primarily suppress interference by utilizing the energy characteristics of the interfering signal. Currently, classic non-parametric methods include notch filtering, subband cancellation, and subspace projection. Notch filtering suppresses interference by setting the interfering component to zero within a specific domain; subband cancellation mainly operates on the focused image data, using the power spectrum to cancel interference; subspace projection can be used in both the time and frequency domains, performing well in handling both narrowband and wideband interference and effectively protecting the useful signal. However, when the interference energy is too strong, it can over-map the subspace, resulting in abnormal responses.
[0007] In summary, to obtain high-quality SAR images, it is of great significance to carry out high-precision suppression research on mutual interference of SAR scattered waves. Effective mutual interference suppression methods are crucial to improving the survivability of SAR systems. Summary of the Invention
[0008] The purpose of this invention is to provide a method, system, storage medium, and electronic device for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering.
[0009] The technical solution adopted in this invention is as follows:
[0010] A method for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering includes the following steps:
[0011] Step S101: Estimate the relevant parameters of mutual interference based on the prior orbit information and time-frequency relationship, and perform range-direction pulse compression on the interference signal in the original echo data;
[0012] Step S102: Use the maximum eigenvalue sequence detection method to detect whether there is interference in the range pulse;
[0013] Step S103: Based on the interference detection results in step S102, calculate the relative signal-to-interference ratio of the interference pulse;
[0014] Step S104: Classify the interference pulses by relative signal-to-interference ratio (SINR) to obtain two parts of data: high SINR and low SINR, namely, interference signals under high SINR conditions and interference signals under low SINR conditions.
[0015] Step S105: Reconstruct the interference signal under high signal-to-interference-plus-noise ratio conditions using the feature subspace projection method;
[0016] Step S106: Use robust principal component analysis to extract interference signals under low signal-to-interference-plus-noise ratio conditions;
[0017] Step S107: Based on the interference signals reconstructed and extracted in steps S105 and S106, filter them out from the compressed synthetic aperture radar echo data. Finally, perform inverse range compression processing on the data after interference filtering to restore the interference-free normal echo.
[0018] Step S101 is as follows:
[0019] Mutual interference signals are very similar to SAR signals, with D m B is the coefficient of the constant term. m The bandwidth represents mutual interference, and f represents the frequency. c and f m The carrier frequencies of the SAR satellite and the interference source satellite are respectively, K m For interference frequency modulation, t m The time delay from transmission to reception of the interference; j is the imaginary unit; the subscripts are used to distinguish between the interference and the target signal and have no actual meaning. The subscript c represents the target signal, and the subscript m represents the interference signal. The spectrum of the mutual interference signal can be expressed as:
[0020]
[0021] The design of frequency-domain matched filters to address mutual interference can eliminate the quadratic phase in the model, expressed as:
[0022]
[0023] Carrier frequency f is estimated based on orbital data of the onboard SAR and the interfering satellite at the time of interference. m Relevant parameter B m and K m Then, the estimation is made by combining the time-frequency relationship of strong interference in the echo; generally speaking, the carrier frequencies f of the two SAR satellites are... c and f m The magnitudes are similar or consistent; range compression of the interfering SAR signal is performed using the estimated parameters; ignoring the π / 4 phase term, the range compression of the mutual interference signal can be expressed as:
[0024]
[0025] The SAR echo time-domain model after frequency-domain matched filtering of the interference is as follows:
[0026]
[0027] Where D c t c K c and B c These represent the amplitude, time delay, frequency modulation (FM), and bandwidth of the normal SAR echo signal, respectively. The subscript 'c' represents the target signal, and the subscript 'm' represents the interference signal. Modeling after mutual interference range compression processing shows that the target SAR signal was not successfully compressed like the mutual interference signal due to mismatched filter parameters. The range compression effect of mutual interference is related to the carrier frequency offset of the two SAR satellites. Generally, the carrier frequency f that satisfies the conditions for mutual interference formation... c and f m If they are similar or identical, and if the two are the same or their minor differences are ignored, the mutual interference distance compression effect is the best.
[0028] In step S102, the maximum eigenvalue sequence detection method is used to detect whether there is interference in the range pulse, specifically:
[0029] The maximum eigenvalue sequence detection method is used here to detect whether there is mutual interference in the pulse: the maximum eigenvalue sequence is obtained by matrix eigenvalue decomposition, and then a threshold is set to obtain the final detection result of whether there is mutual interference; based on the signal after mutual interference distance compression, N a and N r Let i represent the number of sampling points in the azimuth and range directions of the SAR signal, respectively, assuming i = 1, 2, ..., N. a and z = 1, 2, ..., N r , with S i (γ z Let ) represent the z-th sampling point of the i-th distance-direction pulse signal. Then, the Hankel matrix corresponding to this pulse can be represented as:
[0030]
[0031] Where M = N r -N+1, from which the covariance matrix C corresponding to this pulse can be constructed. m =H m H m HN is the dimension of the covariance matrix, typically set to 16 to 128 in practice, depending on the situation. The constructed covariance matrix is then subjected to eigenvalue decomposition, and the largest eigenvalue is obtained. This process is repeated pulse by pulse to obtain a sequence composed of the largest eigenvalues of all pulses, represented as...
[0032]
[0033] In E max In the sequence, the eigenvalues corresponding to pulses without mutual interference are relatively small and fluctuate gently, while the maximum eigenvalues corresponding to pulses with mutual interference exhibit more anomalous behavior. For all range-direction pulses, we classify them by presence or absence of mutual interference, and use a robust K-means clustering algorithm to perform binary segmentation on the sequence of maximum eigenvalues to obtain the final detection results.
[0034]
[0035] Where σ represents the threshold for adaptive clustering segmentation, a value of 1 indicates that the pulse contains mutual interference, and 0 indicates that it does not contain mutual interference.
[0036] In step S103, based on the interference detection results of step S102, the relative signal-to-interference ratio (SIR) of the interference pulse is calculated, specifically as follows:
[0037] Based on the correlation of SAR signals and combined with interference detection results, the average power of all pulses without mutual interference is calculated. This can be used as a benchmark for further judging the signal-to-interference-plus-noise ratio (SNR) of pulses containing interference. Assume the number of pulses without mutual interference is N. m The number of pulses containing mutual interference can be expressed as N. a -N m The average power of signals without mutual interference is calculated as follows:
[0038]
[0039] Where ε represents the sequence number of the pulse containing mutual interference; P s This can be used as a basis for judging the signal-to-interference-plus-noise ratio (SIR), because when background noise is ignored, P s The power of the k-th pulse, which contains both mutual interference and the useful signal, can be expressed as:
[0040]
[0041] With P s Using the interference-free signal power as a reference, the relative signal-to-interference ratio of the k-th pulse can be calculated as follows:
[0042]
[0043] Perform the above relative signal-to-interference ratio calculation for each pulse with mutual interference one by one.
[0044] In step S105, use the eigen-subspace projection method to reconstruct the interference signal under high signal-to-interference-plus-noise ratio conditions, specifically:
[0045] When the R(k) value is marked as 1, that is, all pulse data corresponding to high signal-to-interference-plus-noise ratio, use the subspace projection method to perform mutual interference reconstruction one by one; the subspace projection method performs eigenvalue decomposition on the covariance matrix constructed by the pulses, and then projects the interference and the target signal into different subspaces, which can be expressed as
[0046] C m =H m H m H =UΛU H
[0047] where Λ = diag(λ1, λ2,..., λ N ), λ corresponds to the eigenvalue sequence and is sorted from large to small; U = [μ1, μ2,..., μ N is the eigenvector corresponding to each eigenvalue; according to the maximum eigenvalue sequence in the interference detection step, first eliminate the eigenvalues containing interference pulses, then fit the eigenvalues without interference to obtain a dynamic threshold, and mark the a (1 < a < N) eigenvalues higher than the threshold as the eigenvalues corresponding to interference, and the eigenvectors are类推; expressed as
[0048] Λ a =diag(λ1, λ2,..., λ a )
[0049] U a =[μ1, μ2,..., μ a
[0050] The interference subspace is formed by expanding the eigenvectors, expressed as span(U a ); perform interference subspace projection based on the Hankel matrix H m :
[0051] S a =U a U a H H m
[0052] The interference subspace S a contains all the information of the mutual interference under high signal-to-interference-plus-noise ratio conditions; rearrange it to obtain a new vector, which is the interference reconstruction result under high signal-to-interference-plus-noise ratio.
[0053] In step S106, robust principal component analysis is used to extract interference signals under low signal-to-interference-plus-noise ratio (SINNR) conditions, specifically as follows:
[0054] Based on the sparsity of the echo signals, the problem of mutual interference extraction can be transformed into a semi-parametric optimization problem: Under low signal-to-interference-plus-noise ratio (SNR), the amplitudes of the SAR target signal and the mutual interference signal differ significantly. Therefore, the overall SAR signal can be modeled as a low-rank smooth background, while the mutual interference, after range compression, can be modeled as a sparse target with energy anomalies. Thus, sparsity constraints are applied to the mutual interference to solve the semi-parametric optimization problem. For the observed echo signal matrix S, neglecting background noise, the above target optimization problem can be expressed as:
[0055]
[0056] Where λ represents the hyperparameter, and ||·||0 is the L0 norm; because the target signal S r It possesses linear correlation and can be projected into a lower-dimensional subspace at low signal-to-interference-plus-noise ratios, represented as a low-rank matrix. However, finding the rank of the matrix is non-convex, making it difficult to solve using correspondence optimization. This leads to its convex approximation optimization, i.e., the nuclear norm ‖·‖. * The above optimization problem can be updated to:
[0057]
[0058] Where ||·||1 is the L1 norm, and minimizing it is beneficial for sparse constraints on mutual interference signals. By balancing the low-rank term and the sparse term with the weight parameter λ, when the optimization model meets the convergence condition or reaches the maximum number of iterations, it is possible to effectively extract strong energy mutual interference in the case of low signal-to-interference-plus-noise ratio of echo data.
[0059] In step S107, the interference signals extracted in steps S105 and S106 are filtered out from the spaceborne synthetic aperture radar echo data. Finally, the data undergoes inverse range compression processing to restore the interference-free normal echo. Specifically:
[0060] By combining the reconstruction and extraction results of mutual interference under different signal-to-interference-to-noise ratio (SNR) data, the signals are arranged according to their original pulse order, with zeros filled in for pulses without mutual interference, to obtain the final extracted mutual interference echo. The extracted mutual interference is then extracted from the interference-compressed echo signal S. MF Elimination, i.e., suppression of mutual interference, is represented as...
[0061]
[0062] Where M E This represents the extracted mutual interference signal matrix. To obtain the final signal matrix after removing mutual interference, additional processing is needed to recover the target signal affected by matched filtering. This is because range compression of mutual interference introduces unnecessary phase into the useful SAR signal. The matched filter design requires eliminating this phase. express The distance spectrum is then represented as follows:
[0063]
[0064] Where M F * (f) represents the matched filter M F (f) complex conjugate, then with respect to S r (f) Performing an inverse Fourier transform will complete the data recovery and yield the final SAR signal after mutual interference suppression.
[0065] A spaceborne synthetic aperture radar scattering wave mutual interference suppression system based on matched filtering includes:
[0066] An interference signal pulse compression unit is configured to estimate the relevant parameters of the interference and perform range-directed pulse compression on the interference signal in the echo data.
[0067] The interference detection unit is configured to use the maximum eigenvalue sequence detection method to detect whether there is interference in all range pulses;
[0068] The relative signal-to-interference ratio (SINR) calculation unit is configured to calculate the relative SINR of all interfering pulses based on power comparison of the detection results based on interference.
[0069] The interference signal-to-interference-plus-noise ratio (SINR) classification unit is configured to classify interference pulses based on the obtained relative SINR, and obtain two parts of data: high SINR and low SINR.
[0070] The interference extraction unit is configured to reconstruct and extract interference signals under high signal-to-interference-plus-noise ratio (SNR) and low SNR conditions using the feature subspace projection method and robust principal component analysis method, respectively.
[0071] The interference suppression and data recovery unit is configured to filter out the extracted interference signals from the echo data and then perform mutual interference inverse range compression processing on the data to restore the normal echo without interference.
[0072] A storage medium storing multiple programs, said programs being loaded and executed by a processor, wherein a method for suppressing mutual interference of scattering waves from spaceborne synthetic aperture radar based on matched filtering is provided.
[0073] An electronic device includes a storage medium and a processor; the processor is adapted to execute various programs; a memory is used to store multiple programs; characterized in that, when the memory executes the programs on the processor, it implements the aforementioned method for suppressing mutual interference of spaceborne synthetic aperture radar (SAR) scattering waves based on matched filtering. The method for suppressing mutual interference of spaceborne SAR scattering waves based on matched filtering provided in this application first estimates the relevant parameters of mutual interference based on prior orbit information and time-frequency relationships, and performs range pulse compression on the interference signals in the echo data; then, it uses the maximum eigenvalue sequence detection method to detect whether the range pulses contain interference; based on the correlation of the echo signals and combined with the interference detection results, it calculates the relative signal-to-interference ratio (SNR) of the interference pulses; it classifies the interference pulses by the relative SNR, obtaining two parts of data: high SNR and low SNR; it reconstructs and extracts the interference signals under high SNR and low SNR conditions using the eigenspace projection method and robust principal component analysis method, respectively; finally, it filters out the interference signals from the echo data and then performs mutual interference inverse range compression processing on the data to restore the interference-free normal echo. This invention enables the effective utilization of the compression gain of mutual interference, thereby more effectively extracting mutual interference signals from synthetic aperture radar echoes and achieving high-precision interference suppression while better protecting useful signals. Attached Figure Description
[0074] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0075] Figure 1 This is a schematic diagram of the process of the present invention;
[0076] Figure 2 This is a characteristic diagram of the scattered wave mutual interference signal described in this invention;
[0077] Figure 3 This is a signal bandwidth distribution diagram of Gaofen-3 and Sentinel-1 as described in this invention;
[0078] Figure 4 This is a schematic diagram of the process of the present invention;
[0079] Figure 5 This is a comparison diagram showing the effects of the invention before and after actual use. Detailed Implementation
[0080] The present application will now be described in detail with reference to the accompanying drawings and embodiments. Various examples are provided by way of explanation and not by way of limitation. In fact, those skilled in the art will recognize that modifications and variations can be made to the present application without departing from the scope or spirit thereof. For example, a feature shown or described as part of one embodiment may be used in another embodiment to produce yet another embodiment. Therefore, it is desirable that the present application encompass such modifications and variations that fall within the scope of the appended claims and their equivalents.
[0081] Exemplary methods
[0082] like Figure 1 As shown, the method for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering includes:
[0083] Step S101: Estimate the relevant parameters of the interference and perform range pulse compression on the interference signal in the echo data;
[0084] Specifically, based on prior orbital information and time-frequency relationships, the relevant parameters of mutual interference are estimated, and range-direction pulse compression is performed on the interference signals in the echo data.
[0085] Distance compression of the mutual interference signal in the echo can concentrate the energy of each pulse into a shorter time, thereby increasing the peak energy of the mutual interference signal on the time axis. This makes the mutual interference, which was originally distributed throughout the entire pulse width, more concentrated, thus improving the ability to distinguish mutual interference and is very beneficial for the high-precision extraction of subsequent mutual interference.
[0086] Mutual interference signals are very similar to SAR signals, with D m B is the coefficient of the constant term. m The bandwidth represents mutual interference, and f represents the frequency. c and f m The carrier frequencies of the SAR satellite and the interference source satellite are respectively, K m For interference frequency modulation, t m The interference signal's latency from transmission to reception is considered. The spectrum of the mutual interference signal can be represented as:
[0087]
[0088] The design of frequency-domain matched filters to address mutual interference can eliminate the quadratic phase in the model, expressed as:
[0089]
[0090] Carrier frequency f is estimated based on orbital data of the onboard SAR and the interfering satellite at the time of interference. m Relevant parameter B m and K mThen, the estimation is performed by combining the time-frequency relationship of strong interference in the echo. Generally speaking, the carrier frequency f of the two SAR satellites... c and f m The magnitudes are similar or consistent. Range compression of the interfering SAR signal is performed using the estimated parameters. Ignoring the π / 4 phase term, the range compression of the interfering signal can be expressed as:
[0091]
[0092] The SAR echo time-domain model after frequency-domain matched filtering of the interference is as follows:
[0093]
[0094] Where D c t c K c and B c These represent the amplitude, time delay, frequency modulation (FM), and bandwidth of the normal SAR echo signal, respectively. Modeling after mutual interference range compression processing shows that the target SAR signal was not successfully compressed like the mutual interference signal due to filter parameter mismatch. The range compression effect of mutual interference is related to the carrier frequency offset of the two SAR satellites. Generally speaking, the carrier frequency f that satisfies the mutual interference formation conditions... c and f m If they are similar or identical, and if the two are the same or their minor differences are ignored, the mutual interference distance compression effect is the best.
[0095] like Figure 2 The echo and image characteristics of the scattered wave mutual interference signal shown are characterized by ultra-wideband, long duration, and unstable energy, which makes it difficult for some traditional methods to effectively detect the interference. Furthermore, traditional methods lose a significant amount of useful information, and traditional suppression methods also lose a large amount of useful information. The spaceborne synthetic aperture radar scattered wave mutual interference suppression method based on matched filtering can overcome this problem. The interference source is Gaofen-3, and the SAR satellite is Sentinel-1.
[0096] Step S102: Use the maximum eigenvalue sequence detection method to detect whether there is interference in the range pulse;
[0097] Specifically, the maximum eigenvalue sequence detection method is used here to detect whether there is mutual interference in pulses. The maximum eigenvalue sequence method benefits from the fact that the energy of pulses containing mutual interference is usually inconsistent with that of normal pulses, resulting in high-precision detection results. Furthermore, it is not limited by bandwidth and has high robustness. With the addition of mutual interference distance compression, the accuracy of the maximum eigenvalue sequence method can also be improved to some extent.
[0098] The sequence of maximum eigenvalues is obtained through matrix eigenvalue decomposition, and then a threshold is set to obtain the final detection result for the presence or absence of mutual interference. Based on the signal after mutual interference distance compression, N... a and N r Let i represent the number of sampling points in the azimuth and range directions of the SAR signal, respectively, assuming i = 1, 2, ..., N. a and z = 1, 2, ..., N r , with S i (γ z Let ) represent the z-th sampling point of the i-th distance-direction pulse signal. Then, the Hankel matrix corresponding to this pulse can be expressed as:
[0099]
[0100] Where M = N r -N+1, from which the covariance matrix C corresponding to this pulse can be constructed. m =H m H m H N is the dimension of the covariance matrix, typically set to 16 to 128 in practice, depending on the situation. The constructed covariance matrix is then subjected to eigenvalue decomposition, and the largest eigenvalue is obtained. This process is repeated pulse-by-pulse to obtain a sequence of the largest eigenvalues of all pulses, represented as...
[0101]
[0102] In E max In the sequence, the eigenvalues corresponding to pulses without mutual interference are relatively small and fluctuate gently, while the maximum eigenvalues corresponding to pulses with mutual interference exhibit more anomalous behavior. To classify all range-direction pulses by presence or absence of mutual interference, a robust K-means clustering algorithm is used to perform binary segmentation on the sequence of maximum eigenvalues, yielding the final detection results:
[0103]
[0104] Where σ represents the threshold for adaptive clustering segmentation, a value of 1 indicates that the pulse contains mutual interference, and 0 indicates that it does not contain mutual interference.
[0105] Step S103: Based on the interference detection results in step S102, calculate the relative signal-to-interference ratio of the interference pulse;
[0106] Specifically, based on the correlation of SAR signals and combined with interference detection results, the average power of all pulses without mutual interference is calculated. This power can be used as a benchmark for further judging the signal-to-interference-plus-noise ratio (SNR) of pulses containing interference. Assume the number of pulses without mutual interference is N. m The number of pulses containing mutual interference can be expressed as N.a -N m The average power of signals without mutual interference is calculated as follows:
[0107]
[0108] Where ε represents the sequence number of the pulse containing mutual interference. P s This can be used as a basis for judging the signal-to-interference-plus-noise ratio (SIR), because when background noise is ignored, P s The power of the k-th pulse, which contains both mutual interference and the useful signal, can be expressed as:
[0109]
[0110] With P s Using the interference-free signal power as a reference, the relative signal-to-interference ratio of the k-th pulse can be calculated as follows:
[0111]
[0112] The relative signal-to-interference ratio (SINR) calculation was performed on each of the pulses containing mutual interference.
[0113] Step S104: Classify the interference pulses based on the relative signal-to-interference ratio (SIR) to obtain two parts of data: high SIR and low SIR.
[0114] Specifically, the calculated signal-to-interference ratio (SIR) is thresholded and segmented, as shown below.
[0115]
[0116] R(k) represents the final classification result, which categorizes pulses containing mutual interference into two classes, with values of 1 and 0 corresponding to high signal-to-interference-plus-noise ratio (SNR) data and low SNR data, respectively. This represents the classification threshold, which is an empirical value obtained from a large number of experiments.
[0117] like Figure 3 As shown, based on the bandwidth overlap of two SAR satellites in the same band, it can be seen that mutual interference signals are usually large-bandwidth signals, and the accuracy of previous methods such as frequency domain notch filtering and subband cancellation will decrease sharply. Combining the energy instability of interference, and making full use of the advantages of semi-parametric robust principal component analysis and non-parametric subspace projection methods, interference suppression can be achieved with high accuracy while effectively protecting the useful signal.
[0118] Step S105: Reconstruct the interference signal under high signal-to-interference-plus-noise ratio conditions using the feature subspace projection method;
[0119] Specifically, when the R(k) value is marked as 1, that is, all pulse data corresponding to high signal-to-interference-plus-noise ratio, the mutual interference reconstruction is performed one by one using the subspace projection method. The subspace projection method performs eigenvalue decomposition on the covariance matrix constructed by the pulses, and then projects the interference and the target signal into different subspaces, which can be expressed as
[0120] C m = H m H m H = UΛU H
[0121] where Λ = diag(λ1, λ2,..., λ N ), λ corresponds to the eigenvalue sequence and is sorted from large to small. U = [μ1, μ2,..., μ N is the eigenvector corresponding to each eigenvalue. According to the maximum eigenvalue sequence in the interference detection step, first eliminate the eigenvalues containing interference pulses, and then fit the eigenvalues without interference to obtain the dynamic threshold. Mark the a (1 < a < N) eigenvalues higher than the threshold as the eigenvalues corresponding to interference, and the eigenvectors are the same by analogy. It is expressed as
[0122] Λ a = diag(λ1, λ2,..., λ a )
[0123] U a = [μ1, μ2,..., μ a
[0124] The interference subspace is formed by expanding the eigenvectors, which is expressed as span(U a ). Based on the Hankel matrix H m perform interference subspace projection:
[0125] S a = U a U a H H m
[0126] The interference subspace S a contains all the information of the mutual interference under the condition of high signal-to-interference-plus-noise ratio. Rearrange it to obtain a new vector, which is the interference reconstruction result under high signal-to-interference-plus-noise ratio.
[0127] Step S106: Use the robust principal component analysis method to extract the interference signal under the condition of low signal-to-interference-plus-noise ratio;
[0128] Specifically, based on the sparsity of the echo signals, the problem of mutual interference extraction can be transformed into an optimization problem using a semi-parametric method. Under low signal-to-interference-plus-noise ratio (SNR), the amplitudes of the SAR target signal and the mutual interference signal differ significantly. Therefore, the overall SAR signal can be modeled as a low-rank, smooth background, while the mutual interference, after range compression, can be modeled as sparse targets with energy anomalies. Thus, sparsity constraints are applied to the mutual interference, and a semi-parametric optimization problem is solved. For the observed echo signal matrix S, neglecting background noise, the above target optimization problem can be expressed as:
[0129]
[0130] Where λ represents the hyperparameter, and ||·||0 is the L0 norm. Because the target signal S r It possesses linear correlation and can be projected into a lower-dimensional subspace at low signal-to-interference-plus-noise ratios, representing it as a low-rank matrix. However, finding the rank of a matrix is non-convex, making it difficult to solve using correspondence optimization. This leads to its convex approximation optimization, i.e., the nuclear norm ‖·‖. * The above optimization problem can be updated to:
[0131]
[0132] Where ||·||1 is the L1 norm, and minimizing it is beneficial for sparse constraints on mutual interference signals. By balancing the low-rank and sparse terms with the weight parameter λ, when the optimization model meets the convergence condition or reaches the maximum number of iterations, strong-energy mutual interference can be effectively extracted even when the echo data has a low signal-to-interference-plus-noise ratio.
[0133] Step S107: Based on the interference signals extracted in steps S105 and S106, filter them out from the spaceborne synthetic aperture radar echo data, and finally perform inverse range compression processing on the data to restore the interference-free normal echo.
[0134] Specifically, by combining the reconstruction and extraction results of mutual interference under different signal-to-interference-to-noise ratio (SNR) data, the signals are arranged according to their original pulse order, with zeros filled in for pulses without mutual interference, to obtain the final extracted mutual interference echo. The extracted mutual interference is then transferred from the interference-compressed echo signal S... MF Elimination, i.e., suppression of mutual interference, is represented as...
[0135]
[0136] Where M E This represents the extracted mutual interference signal matrix. This is the final signal matrix after removing mutual interference. Additional processing is needed to recover the target signal affected by matched filtering, as range compression of mutual interference introduces unnecessary phase into the useful SAR signal. The matched filter design requires eliminating the corresponding phase to achieve this. express The distance spectrum is then represented as follows:
[0137]
[0138] Where M F * (f) represents the matched filter M F (f) complex conjugate, then with respect to S r (f) Performing an inverse Fourier transform will complete the data recovery and yield the final SAR signal after mutual interference suppression.
[0139] like Figure 4 The diagram shows a process flow diagram of the mutual interference method for spaceborne synthetic aperture radar (SAR) scattering waves based on matched filtering. This method starts from the original echo containing interference, and through interference range compression, interference detection, interference extraction and suppression, and finally echo data recovery, it can effectively suppress mutual interference between spaceborne SAR scattering waves.
[0140] Comparison of effects before and after interference suppression Figure 5 As shown in the figure, the experimental data is real data from Sentinel-1, which is affected by mutual interference from scattered waves from Gaofen-3. By comparison, it can be seen that after processing with this invention, the SAR image, originally suppressed by cloud-like mutual interference artifacts, becomes clearly visible, the cloud-like effect of mutual interference almost disappears, and the quality of the SAR image is significantly improved. The mountains covered by white artifacts in the figure are clearly highlighted after interference suppression, significantly enhancing the application value of subsequent SAR images.
[0141] Exemplary System
[0142] The application embodiment also provides a spaceborne synthetic aperture radar scattering wave mutual interference suppression system based on matched filtering, including: an interference signal pulse compression unit configured to estimate the relevant parameters of the interference and perform range pulse compression on the interference signal in the echo data; an interference detection unit configured to use the maximum eigenvalue sequence detection method to detect whether all range pulses contain interference; a relative signal-to-interference ratio (SIR) calculation unit configured to calculate the relative SIR of all interference-containing pulses by power comparison based on the interference detection results; an interference SIR classification unit configured to classify the interference-containing pulses based on the obtained relative SIR to obtain two parts of data: high SIR and low SIR; an interference extraction unit configured to reconstruct and extract the interference signal under high SIR and low SIR conditions using the eigenspace projection method and robust principal component analysis method, respectively; and an interference suppression and data recovery unit configured to filter out the extracted interference signal from the echo data and then perform mutual interference inverse range compression processing on the data to restore the interference-free normal echo.
[0143] The matching filter-based spaceborne synthetic aperture radar scattering wave mutual interference suppression system provided in this application embodiment can realize any of the above-mentioned matching filter-based spaceborne synthetic aperture radar scattering wave mutual interference suppression steps and processes, and achieve the same technical effect, which will not be described in detail here.
[0144] Exemplary device
[0145] This application provides an electronic device, including a storage medium and a processor. The processor is suitable for executing various programs; the memory is used to store multiple programs. The device is characterized in that, when the memory executes the programs on the processor, it implements the aforementioned method for suppressing mutual interference of spaceborne synthetic aperture radar scattering waves based on matched filtering.
[0146] Since the steps of a method for suppressing mutual interference of spaceborne synthetic aperture radar scattering waves based on matched filtering have been described in detail in the specific implementation example, they will not be repeated here.
[0147] The processor includes a Central Processing Unit (CPU), a Network Processor (NP), etc., and can also be a digital signal processor, an application-specific integrated circuit, an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor.
[0148] The processor can be specifically configured as follows: First, based on prior orbit information and time-frequency relationship, the relevant parameters of mutual interference are estimated, and range pulse compression is performed on the interference signal in the echo data. Then, the maximum eigenvalue sequence detection method is used to detect whether there is interference in the range pulse. Based on the correlation of the echo signal and combined with the interference detection results, the relative signal-to-interference ratio (SNR) of the interference pulse is calculated. The interference pulse is classified by the relative SNR, resulting in two parts: high SNR and low SNR data. The eigenspace projection method and robust principal component analysis method are used to reconstruct and extract the interference signal under high SNR and low SNR conditions, respectively. Finally, the interference signal is filtered out from the echo data, and the data is then subjected to mutual interference inverse range compression to restore the interference-free normal echo. Through this invention, the compression gain of mutual interference can be effectively utilized, thereby more effectively extracting the mutual interference signal in the synthetic aperture radar echo, achieving high-precision interference suppression while better protecting the useful signal.
[0149] It should be noted that, depending on the implementation needs, the various components / steps described in the embodiments of this application can be broken down into more components / steps, or two or more components / steps or parts of the operation of components / steps can be combined into new components / steps to achieve the purpose of the embodiments of this application.
[0150] The methods described in the embodiments of this application can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium (such as CD ROM, RAM, floppy disk, hard disk, or magneto-optical disk), or as computer code downloaded over a network that is originally stored in a remote recording medium or a non-transitory machine storage medium and will be stored in a local recording medium. Thus, the methods described herein can be processed by software stored on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware (such as an ASIC or FPGA). It is understood that the computer, processor, microprocessor controller, or programmable hardware includes storage components (e.g., RAM, ROM, flash memory, etc.) capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, it implements the matched-filter-based spaceborne synthetic aperture radar scattering wave mutual interference suppression method described herein. Furthermore, when a general-purpose computer accesses the code used to implement the methods shown herein, the execution of the code transforms the general-purpose computer into a dedicated computer for executing the methods shown herein.
[0151] Those skilled in the art will recognize that the units and method steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application of the technical solution and the constraints involved. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of this application.
[0152] It should be noted that the various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for the device and system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments.
[0153] The device and system embodiments described above are merely illustrative. The units referred to as separate entities may or may not be physically separate. The entities mentioned as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.
[0154] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.
[0155] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification and claims of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.
[0156] Note that the above description is merely a preferred embodiment and application of the technical principles of the present invention. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the specific embodiments described herein, and may include many other effective embodiments without departing from the concept of the present invention. The scope of the present invention is determined by the scope of the appended claims.
Claims
1. A method for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering, characterized in that: Includes the following steps: Step S101: Estimate the relevant parameters of mutual interference based on the prior orbit information and time-frequency relationship, and perform range-direction pulse compression on the interference signal in the original echo data; Step S102: Use the maximum eigenvalue sequence detection method to detect whether there is interference in the range pulse; Step S103: Based on the interference detection results in step S102, calculate the relative signal-to-interference ratio of the interference pulse; Step S104: Classify the interference pulses by relative signal-to-interference ratio (SINR) to obtain two parts of data: high SINR and low SINR, namely, interference signals under high SINR conditions and interference signals under low SINR conditions. Step S105: Reconstruct the interference signal under high signal-to-interference-plus-noise ratio conditions using the feature subspace projection method; Step S106: Use robust principal component analysis to extract interference signals under low signal-to-interference-plus-noise ratio conditions; Step S107: Based on the interference signals obtained after reconstruction and extraction in steps S105 and S106, filter them out from the compressed synthetic aperture radar echo data. Finally, perform inverse range compression processing on the data after interference filtering to restore the interference-free normal echo.
2. The method for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering according to claim 1, characterized in that, Step S101 includes the following steps: Mutual interference signals are very similar to SAR signals, with D m B is the coefficient of the constant term. m The bandwidth represents mutual interference, and f represents the frequency. c and f m The carrier frequencies of the SAR satellite and the interference source satellite are respectively, K m For interference frequency modulation, t m To interfere with the time delay from transmission to reception; j is the imaginary unit. The subscripts are used to distinguish between the interference and target signals and have no actual meaning. The subscript c represents the target signal, and the subscript m represents the interference signal. The spectrum of the mutual interference signal can be represented as: The design of frequency-domain matched filters to address mutual interference can eliminate the quadratic phase in the model, expressed as: Carrier frequency f is estimated based on orbital data of the onboard SAR and the interfering satellite at the time of interference. m Relevant parameter B m and K m The estimation is then performed by combining the time-frequency relationship of strong interference in the echo; the carrier frequencies f of the two SAR satellites are... c and f m The magnitudes are similar or consistent; range compression of the interfering SAR signal is performed using the estimated parameters; ignoring the π / 4 phase term, the range compression of the mutual interference signal can be expressed as: The SAR echo time-domain model after frequency-domain matched filtering of the interference is as follows: Where D c t c K c and B c Let represent the amplitude, time delay, frequency modulation rate, and bandwidth of the normal SAR echo signal, respectively. The subscript 'c' represents the target signal, and the subscript 'm' represents the interference signal. Modeling after mutual interference range compression processing shows that the target SAR signal was not successfully compressed like the mutual interference signal due to mismatched filter parameters. The range compression effect of mutual interference is related to the carrier frequency offset of the two SAR satellites. The carrier frequency f that satisfies the mutual interference formation condition... c and f m If they are similar or identical, and if the two are the same or their minor differences are ignored, the mutual interference distance compression effect is the best.
3. The method for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering according to claim 1, characterized in that, Step S102 includes the following steps: The maximum eigenvalue sequence is obtained through matrix eigenvalue decomposition, and then a threshold is set to obtain the final detection result of whether there is mutual interference; based on the signal after mutual interference distance compression, N a and N r Let i represent the number of sampling points in the azimuth and range directions of the SAR signal, respectively, assuming i = 1, 2, ..., N. a and z = 1, 2, ..., N r , with S i (γ z Let ) represent the z-th sampling point of the i-th distance-direction pulse signal. Then, the Hankel matrix corresponding to this pulse can be represented as: Where M = N r -N+1, from which the covariance matrix C corresponding to this pulse can be constructed. m =H m H m H N is the dimension of the covariance matrix; the covariance matrix is decomposed into eigenvalues, and the largest eigenvalue is obtained; this process is repeated pulse by pulse to obtain the sequence of the largest eigenvalues of all pulses, denoted as: In E max In the sequence, the eigenvalues corresponding to pulses without mutual interference are relatively small and fluctuate gently, while the maximum eigenvalues corresponding to pulses with mutual interference exhibit more anomalous behavior. For all range-direction pulses, we classify them by presence or absence of mutual interference, and use a robust K-means clustering algorithm to perform binary segmentation on the sequence of maximum eigenvalues to obtain the final detection results. Where σ represents the threshold for adaptive clustering segmentation, a value of 1 indicates that the pulse contains mutual interference, and 0 indicates that it does not contain mutual interference.
4. The method for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering according to claim 1, characterized in that, Step S103 specifically includes the following steps: Based on the correlation of SAR signals and combined with interference detection results, the average power of all pulses without mutual interference is calculated. This can be used as a benchmark for further judging the signal-to-interference-plus-noise ratio (SNR) of pulses containing interference. Assume the number of pulses without mutual interference is N. m The number of pulses containing mutual interference can be expressed as N. a -N m The average power of signals without mutual interference is calculated as follows: Where ε represents the sequence number of the pulse containing mutual interference; P s This can be used as a basis for judging the signal-to-interference-plus-noise ratio (SIR), because when background noise is ignored, P s The power of the k-th pulse, which contains both mutual interference and the useful signal, can be expressed as: With P s Using the interference-free signal power as a reference, the relative signal-to-interference ratio of the k-th pulse can be calculated as follows: The relative signal-to-interference ratio (SINR) calculation was performed on each of the pulses containing mutual interference.
5. The method for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering according to claim 1, characterized in that, Step S105 specifically includes the following steps: When R(k) is marked as 1, i.e., all pulse data corresponding to high signal-to-interference-plus-noise ratio (SINR) are reconstructed one by one using the subspace projection method. The subspace projection method performs eigenvalue decomposition on the covariance matrix constructed from the pulses, thereby projecting the interference and target signals into different subspaces, represented as... C m =H m H m H =UΛU H where Λ = diag(λ1, λ2, …, λ N ), λ corresponds to the eigenvalue sequence and is sorted from large to small; U = [μ1, μ2, …, μ N is the eigenvector corresponding to each eigenvalue; according to the maximum eigenvalue sequence in the interference detection step, first eliminate the eigenvalues containing interference pulses, then fit the eigenvalues without interference to obtain the dynamic threshold, and mark the a eigenvalues higher than the threshold as the eigenvalues corresponding to interference, 1 < a < N, and the eigenvectors are类推; denoted as Λ a = diag(λ1, λ2, …, λ a ) U a =[μ1,μ2,…,μ a ] The interference subspace is formed by expanding the eigenvectors, denoted as span(U a Based on the Hankel matrix H m Perform interference subspace projection: WITH a =U a IN a H H m Interference subspace S a It contains all the information about mutual interference under high signal-to-interference-plus-noise ratio conditions; Rearrange them to obtain a new vector, which is the interference reconstruction result under high signal-to-interference-plus-noise ratio.
6. The method for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering according to claim 1, characterized in that, Step S106 specifically includes the following steps: Based on the sparsity of echo signals, the problem of mutual interference extraction can be transformed into an optimization problem of semi-parametric method: Under low signal-to-interference-plus-noise ratio, the amplitudes of SAR target signals and mutual interference signals differ greatly. Therefore, the overall SAR signal can be modeled as a low-rank smooth background, while the mutual interference after range compression can be modeled as a sparse target with energy anomalies. Thus, the mutual interference is subject to sparsity constraints, and a semi-parametric optimization problem is solved. For the observed echo signal matrix S, neglecting background noise, the above optimization problem can be expressed as: Where λ represents the hyperparameter, and ||·||0 is the L0 norm; because the target signal S r It possesses linear correlation and can be projected into a lower-dimensional subspace at low signal-to-interference-plus-noise ratios, represented as a low-rank matrix. However, finding the rank of the matrix is non-convex, making it difficult to solve using correspondence optimization. This leads to its convex approximation optimization, i.e., the nuclear norm ‖·‖. * The above optimization issues have been updated to: Where ||·||1 is the L1 norm, and minimizing it is beneficial for sparse constraints on mutual interference signals; By balancing the low-rank and sparse terms with the weighting parameter λ, when the optimization model meets the convergence condition or reaches the maximum number of iterations, it is possible to effectively extract strong-energy mutual interference in the case of low signal-to-interference-plus-noise ratio of echo data.
7. The method for suppressing mutual interference of scattered waves from spaceborne synthetic aperture radar based on matched filtering according to claim 6, characterized in that, Step S107 specifically includes the following steps: By combining the reconstruction and extraction results of mutual interference under different signal-to-interference-to-noise ratio (SNR) data, the signals are arranged according to their original pulse order, with zeros filled in for pulses without mutual interference, to obtain the final extracted mutual interference echo. The extracted mutual interference is then extracted from the interference-compressed echo signal S. MF Elimination, i.e., suppression of mutual interference, is represented as... Where M E This represents the extracted mutual interference signal matrix. To obtain the final signal matrix after removing mutual interference, additional processing is needed to recover the target signal affected by matched filtering. This is because range compression of mutual interference introduces unnecessary phase into the useful SAR signal. The matched filter design requires eliminating this phase. express The distance spectrum is then represented as follows: Where M F * (f) represents the matched filter M F (f) complex conjugate, then with respect to S r (f) Performing an inverse Fourier transform will complete the data recovery and yield the final SAR signal after mutual interference suppression.
8. A spaceborne synthetic aperture radar scattering wave mutual interference suppression system based on matched filtering, characterized in that, include: An interference signal pulse compression unit is configured to estimate the relevant parameters of the interference and perform range-directed pulse compression on the interference signal in the echo data. The interference detection unit is configured to use the maximum eigenvalue sequence detection method to detect whether there is interference in all range pulses; The relative signal-to-interference ratio (SINR) calculation unit is configured to calculate the relative SINR of all interfering pulses based on power comparison of the detection results based on interference. The interference signal-to-interference-plus-noise ratio (SINR) classification unit is configured to classify interference pulses based on the obtained relative SINR, and obtain two parts of data: high SINR and low SINR. The interference extraction unit is configured to reconstruct and extract interference signals under high signal-to-interference-plus-noise ratio (SNR) and low SNR conditions using the feature subspace projection method and robust principal component analysis method, respectively. The interference suppression and data recovery unit is configured to filter out the extracted interference signals from the echo data and then perform mutual interference inverse range compression processing on the data to restore the normal echo without interference.
9. A storage medium storing a plurality of programs, characterized in that, The program application is loaded and executed by a processor to implement the mutual interference suppression method for spaceborne synthetic aperture radar scattering waves based on matched filtering as described in any one of claims 1-7.
10. An electronic device, characterized in that, Includes storage media and processor; the processor is used to execute various programs. The memory is used to store multiple programs; its characteristic is that when the memory executes the program on the processor, it implements the mutual interference suppression method of spaceborne synthetic aperture radar scattering waves based on matched filtering as described in any one of claims 1-7.