Adaptive interference suppression method and system for translation-invariant subspace matching decision

Abstract

The present invention belongs to the technical field of array signal processing, and discloses an adaptive interference suppression method and system of translation-invariant subspace matching decision. The target angle interval is roughly estimated by conventional beamforming method, and the current eigenvector is separated by comparing and analyzing the decision item constructed by the subspace of the sample covariance matrix and the orthogonal characteristics of the steering vector with the decision threshold. The property of the invariant subspace matching metric is used to analyze whether the current eigenvector belongs to the target separation, and finally the separated interference-noise subspace is reconstructed by orthogonal projection to perform spatial spectrum estimation, so as to suppress interference and noise outside the target angle interval. Compared with the existing subspace suppression method, the present invention improves the array output signal-to-interference ratio and reduces the sidelobe level. The actual sea trial data confirms the effectiveness of the present invention, which has higher engineering practical value and can be applied to strong interference suppression outside the target angle interval of underwater arrays and underwater weak target angle estimation.

Description

Technical Field

[0001] The invention belongs to the technical field of array signal processing, and in particular relates to an adaptive interference suppression method and system for translation-invariant subspace matching decision. Background Art

[0002] The underwater array is easily affected by the noise of the ship itself. There is strong interference from the ship in the end-fire direction of the array, so that the target signal received by the underwater array is covered by strong interference and noise.

[0003] In-depth research and analysis have been conducted on the problem of strong interference suppression, and many technical methods have been proposed. Spatial matrix filters are widely used. Matrix filters are used to extract the characteristic components of the signal of interest from the estimated signal plus interference subspace, and an efficient sub-region multi-source detection method is proposed. By using KL transformation to reduce the dimension of the matrix, a matrix filter with lower dimension is constructed. This type of method can suppress interference well, but it requires prior information of known or roughly estimated angles of the target and interference. Subspace-based methods are also developing. Brian F. Harrison proposed an interference suppression method based on characteristic component association. This method can maintain good interference suppression performance when the target signal angle is known, but it cannot handle strong interference in the array end-fire direction, and does not give a clear decision threshold. Reference 1 [Eigenanalysis-Based Adaptive Interference Suppression and Its Application in Acoustic Source Range Estimation [J]. IEEE Journal of Oceanic Engineering, 2015, 40 (4): 903-916.] proposed an adaptive interference suppression method based on characteristic analysis (EAAIS method). This method only needs to know the general angle of the target and can achieve good suppression of strong interference through adaptive decision conditions. However, the decision criterion of this method will leave more interference and noise components in the signal subspace in broadband processing, making its sidelobe level higher and the output signal-to-noise ratio lower, thereby affecting the resolution of weak target signals.

[0004] In order to better handle the large amount of interference and noise in the signal subspace in broadband processing, strong interference and noise at all angles are robustly suppressed to improve the array output signal-to-interference ratio and reduce the sidelobe level. In the actual underwater array, a more robust adaptive interference suppression method is needed to suppress strong interference and improve the detection capability of target signals.

[0005] Through the above analysis, the problems and defects of the prior art are as follows:

[0006] (1) Existing strong interference suppression methods based on spatial matrix filters require prior information of known or roughly estimated angles between the target and the interference.

[0007] (2) The existing subspace-based strong interference suppression method cannot handle the strong interference in the end-fire direction of the array, and does not give a clear decision threshold.

[0008] (3) The decision criteria of the existing adaptive interference suppression method based on feature analysis will leave more noise in the signal subspace in broadband processing, resulting in higher sidelobe levels and lower output signal-to-noise ratios. Summary of the invention

[0009] In order to overcome the problems existing in the related art, the disclosed embodiments of the present invention provide an adaptive interference suppression method and system for translation invariant subspace matching decision, and the technical solution is as follows:

[0010] The present invention is implemented in this way: the adaptive interference suppression method of translation invariant subspace matching decision comprises the following steps:

[0011] S1, establishing a data model for array reception according to the underwater array receiving target, noise and interference;

[0012] S2, using conventional beamforming to analyze the signals received by the sensor array, estimate the signal energy distribution at different angles, detect the target direction peak, determine the approximate angle range of the target signal, roughly estimate the target angle interval, and construct the decision term using the orthogonal characteristics of the eigenvector and the steering vector obtained by decomposing the sample covariance matrix; compare and analyze the decision term with the decision threshold to obtain the eigenvector corresponding to the target, interference or noise;

[0013] S3, judging and analyzing whether the current feature vector is the feature vector corresponding to the target according to the properties of the translation invariant subspace matching judgment, if it meets the judgment, the current feature vector is separated into the target subspace; if it does not meet the judgment, the current feature vector is separated into the interference-noise subspace;

[0014] S4, reconstruct the sample covariance matrix by orthogonally projecting the separated interference-noise subspace;

[0015] S5, using the reconstructed sample covariance matrix to calculate the power spectrum density of the signal, and normalize it, using the spatial spectrum to estimate the target information, and suppress the strong interference and noise components outside the target angle range.

[0016] In step S1, a data model of array reception is established according to underwater array reception targets, noise and interference, including: placing an underwater array consisting of multiple hydrophones evenly distributed at a certain interval in seawater at a certain depth to obtain array reception data.

[0017] In step S2, using conventional beamforming to analyze the signal received by the sensor array includes: using conventional beamforming to obtain a rough estimate of the target angle, decomposing the sample covariance matrix features, and obtaining mutually orthogonal eigenvectors, then the decision item is:

[0018]

[0019] Where DF ii is the ratio of the maximum matching value of the lth eigenvector and the guidance vector in the target angle interval to the maximum matching value in all directions; θθ k is the target interval [θ S1 ,θ S2 ] is an angle value within θ S1 is the initial value in the angle interval, θ S2 is the final value in the angle interval, k is a value from 1 to P; P is the interval [θ S1 ,θ S2 ]The number of internal angles; a(θθ k ) is the steering vector corresponding to the rough estimated angle of the target; e(l) is the sample covariance matrix The lth eigencomponent of eigendecomposition; θ kk is the angle value in the range of 0°-180° in all directions; ii is an N×1 matrix; the superscript H is the conjugate transpose symbol.

[0020] Furthermore, the eigenvectors serve as an orthonormal basis for the steering vectors.

[0021] Furthermore, according to the ratio of the target angle interval to the maximum matching value of the entire angle interval, the decision threshold is set to 1, and the decision criteria are as follows: ii =1, it means that the current feature vector belongs to the target or noise, and the remaining feature vectors can form the interference-noise subspace; if DF ii <1, it means that the current eigenvector belongs to interference or noise, and the remaining eigenvectors are combined into an interference-noise subspace.

[0022] In step S3, judging and analyzing whether the current feature vector is a feature vector corresponding to the target according to the property of the translation invariant subspace matching judgment specifically includes:

[0023] By using the square of the F norm to analyze the metric distance of the two translation-invariant structures of each eigenvector, the state of the current eigenvector is analyzed according to the size of the metric distance, then:

[0024]

[0025]

[0026] In the formula, ||·|| F is the Froebenius norm, U is the eigenvector, U (1) is a characteristic subvector obtained by deleting the first row of elements, U (2) Another characteristic sub-vector obtained by deleting the last row of elements, is the characteristic subvector U (1) The projection matrix is ​​composed of is the characteristic subvector U (2) The projection matrix.

[0027] Furthermore, the ISMM metric distance is used to analyze whether the current feature vector belongs to the feature vector corresponding to the target. The processing method is as follows:

[0028] If the metric distance of the current eigenvector is less than the metric distance of the following eigenvector, the current eigenvector is considered to be the target component, and the current eigenvector is separated into the target subspace;

[0029] If the metric distance of the current eigenvector is greater than the metric distance of the subsequent eigenvector, the current eigenvector is considered to be an interference or noise component, and the current eigenvector is separated into the interference-noise subspace.

[0030] In step S4, reconstructing the sample covariance matrix includes:

[0031] The separated interference-noise subspace is reconstructed into the sample covariance matrix by orthogonal projection. I+N , the interference-noise subspace U I+N Perform orthogonal projection processing; wherein the orthogonal projection processing is realized by subtracting the projection of the interference-noise subspace from the unit matrix, then:

[0032]

[0033] Where P ⊥ is the orthogonal projection matrix, I v is the unit matrix, and all elements on the diagonal from the upper left corner to the lower right corner are 1; U I+N is the interference-noise subspace composed of the eigenvectors corresponding to the interference and noise, and the superscript H is the conjugate transpose symbol.

[0034] The sample covariance matrix is ​​reconstructed using the result of orthogonal projection processing of the interference-noise subspace, and the interference and noise components in the original sample covariance matrix are removed. Then:

[0035]

[0036] In the formula, is the reconstructed sample covariance matrix, is the covariance matrix of the original samples before reconstruction.

[0037] In step S5, the power spectrum density of the signal is calculated using the reconstructed sample covariance matrix, and normalized, and the target information is estimated using the spatial spectrum to suppress strong interference and noise components outside the target angle range. Specifically, the following steps are included:

[0038] By reconstructing the sample covariance matrix Substitute into the spatial power spectrum estimation, calculate the spectrum estimation value at each angle through angle scanning, find the peak in the angle scanning result, and estimate the energy distribution of the signal in different spatial directions; use the similarity of the weight vector formed at different angles and the reconstructed sample covariance matrix to distinguish and suppress the target signal and interference noise; when performing spatial spectrum estimation, process the weight vector of the target angle and the reconstructed sample covariance matrix to obtain a peak with a larger amplitude at the target angle; suppress the interference and noise components outside the target angle range by forming a null pit;

[0039]

[0040]

[0041] In the formula, P(θ) is the power spectrum value of each signal, ω(θ) is the weight vector, is the reconstructed sample covariance matrix, ω H (θ) is the conjugate transposed form of the weight vector ω(θ), θ is the angle, θ∈[0°, 180°], a(θ) is the steering vector matrix, and N is the number of array elements.

[0042] Another object of the present invention is to provide an adaptive interference suppression system for translation-invariant subspace matching decision, the system is used to control the adaptive interference suppression method for translation-invariant subspace matching decision, the system comprising:

[0043] A data acquisition module is used to establish a data model for array reception according to underwater array reception targets, noise and interference, and to obtain array reception data;

[0044] The feature vector acquisition module is used to roughly estimate the target angle interval using conventional beamforming, and construct the decision item using the orthogonal characteristics of the feature vector and the steering vector obtained by decomposing the sample covariance matrix; the decision item is compared and analyzed with the decision threshold to obtain the feature vector corresponding to the target, interference or noise;

[0045] The target separation judgment module is used to judge and analyze whether the current feature vector is the target corresponding feature vector according to the translation invariant subspace matching judgment property. If it meets the judgment, the current feature vector is separated into the target subspace; if it does not meet the judgment, the current feature vector is separated into the interference-noise subspace;

[0046] The spatial spectrum estimation module is used to calculate the normalized power spectrum using the reconstructed sample covariance matrix, estimate the target information using the spatial spectrum, and suppress strong interference and noise components outside the target angle range.

[0047] In combination with all the above technical solutions, the advantages and positive effects of the present invention are as follows: the present invention provides an adaptive interference suppression method for translation-invariant subspace matching judgment, and establishes a data model for array reception according to the underwater array receiving target, noise and interference; adopts conventional beamforming to roughly estimate the target angle interval, and constructs a decision item and a decision threshold for comparative analysis based on the orthogonal characteristics of the eigenvector obtained by sample covariance matrix decomposition and the steering vector, and separates the current eigenvector; judges and analyzes whether the current eigenvector is the eigenvector corresponding to the target according to the property of translation-invariant subspace matching judgment, and if it meets the judgment, the current eigenvector can be separated into the target subspace, and if it does not meet the judgment, the current eigenvector can be separated into the interference-noise subspace; reconstructs the sample covariance matrix by orthogonal projection of the separated interference-noise subspace; and calculates the normalized power spectrum with the reconstructed sample covariance matrix to suppress strong interference and noise outside the target angle interval, thereby improving the detection capability of the target, and can be used for underwater weak target angle estimation and strong interference suppression outside the target angle interval.

[0048] Different from conventional beamforming methods and existing subspace suppression methods (such as the EAAIS method in document 1), the present invention utilizes the orthogonal characteristics of the steering vector and the eigenvector to construct a decision item, and obtains the current eigenvector by comparing it with the decision threshold; according to the property of the translation-invariant subspace matching decision, it is determined and analyzed whether the current eigenvector is the eigenvector corresponding to the target, and the separated interference-noise subspace is reconstructed into a sample covariance matrix by orthogonal projection, and finally the spatial spectrum is estimated to suppress the interference and noise outside the target angle range, thereby improving the target detection capability and output signal-to-interference ratio.

[0049] In order to better improve the detection capability of target signals, the present invention adopts an adaptive interference suppression method of translation-invariant subspace matching judgment to suppress interference and noise outside the target angle interval, thereby improving the output signal-to-interference ratio and reducing the sidelobe level. Compared with the existing subspace suppression method (such as the EAAIS method), the adaptive interference suppression method of the present invention can suppress more noise and interference, improve the array output signal-to-interference ratio and reduce the sidelobe level; actual sea trial data confirms its effectiveness, has higher engineering practical value, and can be applied to strong interference suppression outside the underwater array target angle interval and underwater weak target angle estimation.

[0050] This technical solution will solve the current output signal-to-interference ratio problem in the field of underwater interference suppression, improve the resolution of weak target signals in a strong underwater interference suppression environment, and thus robustly perform passive detection of underwater arrays. This technical solution effectively overcomes the impact of interference signals on target signals in underwater environments and improves signal detection and transmission performance. This innovation fills the previous technical deficiencies in underwater interference suppression, provides new solutions for underwater applications, and also brings new development opportunities for scientific research and practical applications in underwater communications, marine exploration and other fields. This invention can effectively suppress interference and background noise, and provides new solutions and technical paths for underwater passive detection and other aspects. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings herein are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description, serve to explain the principles of the present disclosure;

[0052] Figure 1 It is a basic block diagram of the adaptive interference suppression method of translation invariant subspace matching decision provided by an embodiment of the present invention;

[0053] Figure 2 It is a specific implementation flow chart of the adaptive interference suppression method of translation invariant subspace matching decision provided by an embodiment of the present invention;

[0054] Figure 3 is a diagram of experimental data processing results of conventional beamforming provided by an embodiment of the present invention;

[0055] Figure 4 is a diagram of experimental data processing results of the EAAIS method provided by an embodiment of the present invention;

[0056] Figure 5 is a graph of experimental data processing results provided by an embodiment of the present invention;

[0057] Figure 6 It is a normalized power spectrum diagram of the conventional beamforming, EAAIS method and the method of the present invention provided by the embodiments of the present invention. DETAILED DESCRIPTION

[0058] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific implementation disclosed below.

[0059] Embodiment 1, as Figure 1 As shown, the adaptive interference suppression method for translation invariant subspace matching decision provided by the embodiment of the present invention includes the following steps:

[0060] S1, establishing a data model for array reception according to the underwater array receiving target, noise and interference;

[0061] S2, using conventional beamforming to analyze the signals received by the sensor array, estimate the signal energy distribution at different angles, detect the target direction peak, determine the approximate angle range of the target signal, roughly estimate the target angle interval, and construct the decision term using the orthogonal characteristics of the eigenvector and the steering vector obtained by decomposing the sample covariance matrix; compare and analyze the decision term with the decision threshold to obtain the eigenvector corresponding to the target, interference or noise;

[0062] S3, judging and analyzing whether the current feature vector is the feature vector corresponding to the target according to the properties of the translation invariant subspace matching judgment, if it meets the judgment, the current feature vector is separated into the target subspace; if it does not meet the judgment, the current feature vector is separated into the interference-noise subspace;

[0063] S4, reconstruct the sample covariance matrix by orthogonally projecting the separated interference-noise subspace;

[0064] S5, using the reconstructed sample covariance matrix to calculate the power spectrum density of the signal, and normalize it, using the spatial spectrum to estimate the target information, and suppress the strong interference and noise components outside the target angle range.

[0065] Preferably, the adaptive interference suppression method for translation invariant subspace matching decision provided in an embodiment of the present invention specifically comprises the following steps:

[0066] (1) Placing an underwater array consisting of multiple hydrophones evenly distributed at a certain interval in seawater at a certain depth to obtain array reception data;

[0067] (2) Use conventional beamforming to roughly estimate the target angle interval and construct the decision through the orthogonal characteristics of the eigenvector of the sample covariance matrix and the steering vector;

[0068] (3) Compare and analyze the decision item with the decision threshold to obtain the feature vector corresponding to the target, interference or noise;

[0069] (4) Using the translation-invariant subspace matching decision to measure the matching degree of the two translation-invariant structures of the current feature vector, analyzing whether the current feature vector belongs to the target feature vector, thereby constructing the target subspace and the interference-noise subspace;

[0070] (5) Reconstruct the sample covariance matrix by using orthogonal projection method on the separated interference-noise subspace;

[0071] (6) Spatial spectrum estimation is used to estimate target information and suppress interference and noise outside the target angle range.

[0072] Preferably, the embodiment of the present invention adopts conventional beamforming to roughly estimate the target angle interval, and constructs a decision through the orthogonal characteristics of the eigenvector of the sample covariance matrix and the steering vector, specifically including:

[0073] Since it is impossible to obtain the prior information of the target, conventional beamforming is used to obtain a rough estimate of the target angle. The sample covariance matrix is ​​eigendecomposed to obtain mutually orthogonal eigenvectors, and in an ideal case, the eigenvectors are used as the standard orthogonal basis of the steering vector. To this end, based on the property that the target eigenvector multiplied by the corresponding steering vector is the largest and multiplied by the remaining steering vectors within the target angle is 0, a decision term is constructed to facilitate the separation of the target eigenvector, namely:

[0074]

[0075] Where DF ii is the ratio of the maximum matching value of the lth eigenvector and the guidance vector in the target angle interval to the maximum matching value in all directions; θθ k is the target interval [θ S1 ,θ S2 ] is an angle value within θ S1 is the initial value in the angle interval, θ S2 is the final value in the angle interval, k is a value from 1 to P; P is the interval [θ S1 ,θ S2 ]The number of internal angles; a(θθ k ) is the steering vector corresponding to the rough estimated angle of the target; e(l) is the sample covariance matrix The lth eigencomponent of eigendecomposition; θ kk is the angle value in the range of 0°-180° in all directions; ii is an N×1 matrix; the superscript H is the conjugate transpose symbol.

[0076] According to the ratio of the target angle interval to the maximum matching value of the entire angle interval, the decision threshold can be set to 1, and the decision criteria are as follows: ii =1, it means that the current feature vector belongs to the target or noise, so the remaining feature vectors can form the interference-noise subspace; if DF ii <1, it means that the current eigenvector belongs to interference or noise, so the remaining eigenvectors are also combined into an interference-noise subspace.

[0077] Preferably, the embodiment of the present invention uses the translation invariant subspace matching judgment to measure the matching degree of the two translation invariant structures of the current feature vector, and analyzes whether the current feature vector belongs to the target feature vector, so as to form the target subspace and the interference-noise subspace, which specifically includes:

[0078] In broadband signal processing, the constructed decision item needs to judge the signal of each frequency point in the broadband, so that the separated target subspace contains not only the target component but also redundant noise components and interference components. To this end, the metric distance of the two translation-invariant structures of each eigenvector is analyzed by using the square of the F norm, and the state of the current eigenvector is analyzed according to the size of the metric distance, that is:

[0079]

[0080]

[0081] In the formula, ||·|| F is the Froebenius norm, U is the eigenvector, U (1) is a characteristic subvector obtained by deleting the first row of elements, U (2) Another characteristic sub-vector obtained by deleting the last row of elements, is the characteristic subvector U (1) The projection matrix is ​​composed of is the characteristic subvector U (2) The projection matrix.

[0082] The ISMM metric distance is used to analyze whether the current feature vector belongs to the feature vector corresponding to the target. The processing method is as follows:

[0083] If the metric distance of the current eigenvector is less than the metric distance of the following eigenvector, the current eigenvector is considered to be the target component and is separated into the target subspace;

[0084] If the metric distance of the current eigenvector is greater than the metric distance of the subsequent eigenvector, then the current eigenvector is considered to be an interference or noise component and is separated into the interference-noise subspace.

[0085] Preferably, the embodiment of the present invention reconstructs the sample covariance matrix of the separated interference-noise subspace by orthogonal projection, which specifically includes:

[0086] The separated interference-noise subspace is reconstructed into the sample covariance matrix by orthogonal projection. I+N , the interference-noise subspace U I+N Perform orthogonal projection processing; wherein the orthogonal projection processing is realized by subtracting the projection of the interference-noise subspace from the unit matrix, then:

[0087]

[0088] Where P ⊥ is the orthogonal projection matrix, I v is the unit matrix, and all elements on the diagonal from the upper left corner to the lower right corner are 1; U I+N is the interference-noise subspace composed of the eigenvectors corresponding to the interference and noise, and the superscript H is the conjugate transpose symbol;

[0089] The sample covariance matrix is ​​reconstructed using the result of orthogonal projection processing of the interference-noise subspace, and the interference and noise components in the original sample covariance matrix are removed. Then:

[0090]

[0091] In the formula, is the reconstructed sample covariance matrix, is the covariance matrix of the original samples before reconstruction.

[0092] By reconstructing the sample covariance matrix Substituting into the spatial power spectrum estimation, the weight vectors formed at different angles can be processed by the similarity with the reconstructed sample covariance matrix. This method is mainly used to distinguish and suppress the target signal and interference noise. When performing spatial spectrum estimation, the weight vector of the target angle and the reconstructed sample covariance matrix are processed to obtain a peak with a large amplitude at the target angle. This is because the target signal has obvious energy concentration in the reconstructed sample covariance matrix, so a peak with large power is formed at the target angle. When the weight vector of the interference and noise angles and the reconstructed sample covariance matrix are used for spatial spectrum estimation, some nulls will be formed at these angles. This is because the energy of interference and noise in the reconstructed sample covariance matrix is ​​small, resulting in lower spectrum estimation values ​​at these angles. By forming nulls, the interference and noise components outside the target angle range can be effectively suppressed, and the extraction accuracy and anti-interference ability of the target signal can be improved.

[0093]

[0094]

[0095] In the formula, P(θ) is the power spectrum value of each signal, ω(θ) is the weight vector, is the reconstructed sample covariance matrix, ω H (θ) is the conjugate transposed form of the weight vector ω(θ), θ is the angle, θ∈[0°, 180°], a(θ) is the steering vector matrix, and N is the number of array elements.

[0096] The adaptive interference suppression system for translation-invariant subspace matching decision provided by an embodiment of the present invention includes:

[0097] A data acquisition module is used to establish a data model for array reception according to underwater array reception targets, noise and interference, and to obtain array reception data;

[0098] The feature vector acquisition module is used to roughly estimate the target angle interval using conventional beamforming, and construct the decision item using the orthogonal characteristics of the feature vector and the steering vector obtained by decomposing the sample covariance matrix; the decision item is compared and analyzed with the decision threshold to obtain the feature vector corresponding to the target, interference or noise;

[0099] The target separation judgment module is used to judge and analyze whether the current feature vector is the target corresponding feature vector according to the translation invariant subspace matching judgment property. If it meets the judgment, the current feature vector is separated into the target subspace; if it does not meet the judgment, the current feature vector is separated into the interference-noise subspace;

[0100] The spatial spectrum estimation module is used to calculate the normalized power spectrum using the reconstructed sample covariance matrix, estimate the target information using the spatial spectrum estimation, and suppress the strong interference and noise components outside the target angle range.

[0101] Embodiment 2, the adaptive interference suppression method of translation invariant subspace matching decision provided by the embodiment of the present invention, uses the conventional beamforming method to roughly estimate the target angle interval, separates the current eigenvector by comparing and analyzing the decision item constructed by the subspace of the sample covariance matrix and the orthogonal characteristics of the steering vector with the decision threshold, uses the property of the invariant subspace matching metric to analyze whether the current eigenvector belongs to the target separation, and finally reconstructs the sample covariance matrix of the separated interference-noise subspace by orthogonal projection to perform spatial spectrum estimation, suppresses interference and noise outside the target angle interval, and improves the detection capability of the target. The specific algorithm flow chart is as follows Figure 2 shown.

[0102] Step 1: Place an underwater array consisting of N hydrophones evenly distributed at intervals of d in the seawater at a certain depth and monitor them from different directions. Receive Q signals, including 1 target and Q-1 interferences, then the data received by the array is modeled as:

[0103]

[0104]

[0105] A=[a(θ1)…a(θ q )…a(θ Q )]

[0106] S(t)=[s1(t)…s q (t)…s Q (t)] T

[0107] n(t)=[n(t1)…n(t n )] T

[0108] Where X(t) is the data received by the array, Q is the number of received signal sources, and a(θ q ) corresponds to θ q The array steering vector in the direction, θ q is an angle value within the angle range [0 to 180°], s q (t) is the qth incident signal source, n(t1) is the noise vector received by the array, N is the number of array elements, c is the speed of sound, f is the signal frequency, d is the array element spacing, A is the array steering vector matrix composed of all array steering vectors, S(t) is the incident signal component composed of all incident signal sources, n(t) is the noise matrix composed of all noise vectors, and the superscript T represents the matrix transpose symbol.

[0109] In practice, since the true output covariance matrix R of the array cannot be obtained, the sample covariance matrix is ​​usually used By replacing the sample covariance Performing feature decomposition, we can obtain:

[0110]

[0111] In the formula, represents the original sample covariance matrix, X(t) represents the array received data, the superscript H represents the conjugate transpose symbol, Q represents the number of incident signals, N represents the number of array elements, L represents the number of snapshots, and λ i and e i The sample covariance matrix is The eigenvalues ​​and eigenvectors obtained by eigendecomposition are orthogonal to each other.

[0112] The eigenvalues ​​are distributed in descending order as follows:

[0113] λ1>…>λ Q >λ Q+1 >…>λ N (3)

[0114] The eigenvectors corresponding to the first Q large eigenvalues ​​form the main subspace U s , the eigenvectors corresponding to the remaining NQ small eigenvalues ​​constitute the noise subspace U N , which can be expressed as:

[0115] U s =[e1, e2…e Q ],U N =[e Q+1 ,e2…e N ] (4)

[0116] Step 2: Estimate the prior information of the target through the conventional beamforming method, roughly estimate the target angle θθ0, and preset the target angle range [θ S1 ,θ S2 ], satisfying 0°≤θ S1 <θθ0<θ S2 ≤180°, the grid size of the angle interval is 0.1°, a total of P = (θ S1 -θ S2 ) / 0.1+1 values. Construct P steering vectors, namely:

[0117]

[0118] θθ p ∈[θ S1 ,θ S2 ],p∈[1,P](5)

[0119] In the formula, a(θθ p ) is an N×1 complex vector, [θ S1 ,θ S2 ] The steering vector matrix of beam scanning can be formed within the interval:

[0120] AA=[a(θθ1),…,a(θθ p ),…,a(θθ P )] (6)

[0121] Where AA is an N×P complex matrix, a(θθ p ) is an N×1 complex vector, p∈[1…P].

[0122] A decision method based on the EAAIS method is adopted. By constructing a decision term in the estimated target angle interval, P steering vectors are traversed in two subspaces to find the target component in the angle interval. To this end, the decision term is set as follows:

[0123]

[0124] Where DF ii is the ratio of the maximum matching value of the lth eigenvector and the guidance vector in the target angle interval to the maximum matching value in all directions; θθ k is the target interval [θ S1 ,θ S2 ] is an angle value within θ S1 is the initial value in the angle interval, θ S2 is the final value in the angle interval, k is a value from 1 to P; P is the interval [θ S1 ,θ S2 ]The number of internal angles; a(θθ k ) is the steering vector corresponding to the rough estimated angle of the target; e(l) is the sample covariance matrix The lth eigencomponent of eigendecomposition; θ kk is the angle value in the range of 0°-180° in all directions; ii is an N×1 matrix; the superscript H is the conjugate transpose symbol.

[0125] According to the ratio of the target angle interval to the maximum matching value of the entire angle interval, the decision threshold can be set to 1, and the decision criteria are as follows: ii =1, it means that the current feature vector belongs to the target or noise, so the remaining feature vectors can form the interference-noise subspace; if DF ii <1, it means that the current eigenvector belongs to interference or noise, so the remaining eigenvectors are also combined into an interference-noise subspace.

[0126] Step 3: According to A = [a(θ1)…a(θ q )…a(θ Q )], the steering vector matrices of the two subarrays obey:

[0127] A (1) =A (2) D (8)

[0128]

[0129] Where D is a diagonal matrix.

[0130] In the absence of noise, U (i) The matrix composed of the first Q eigenvectors of is expressed as We can get:

[0131]

[0132] Combining equation (8) and equation (9), we get:

[0133]

[0134] In the formula, T=FDF -1 , which means and They form the same subspace, called the translation invariant subspace. This translation invariant subspace is characterized by the projection matrix on this subspace. express The projection matrix on :

[0135]

[0136] According to formula (10) and formula (11), we can get:

[0137]

[0138] When the data received by the array is noisy, the projection matrix and are no longer equal. To this end, the translation-invariant subspace matching decision method is introduced to measure the matching degree of the two projection matrices. The expression is as follows:

[0139]

[0140] By performing ISMM measurement on each feature vector, formula (13) can be written as formula (14):

[0141]

[0142] Where tr(·) represents the matrix trace operator, Represent the two translation-invariant structures of the z-th eigenvector, z∈[1,N], i∈1or2; for this, Then formula (14) becomes:

[0143]

[0144] The characteristic vector obtained by processing the decision term of formula (7) is analyzed by the properties of the invariant subspace theory to determine whether the characteristic vector obtained at each frequency point is the characteristic vector corresponding to the target. The processing method is as follows:

[0145] 1) Assume that the Jth eigenvector obtained by equation (7) is considered to be the eigenvector corresponding to the target. According to the principle of eigenvalue distribution from large to small, the JJ∈[J+1,N]th eigenvector is the eigenvector corresponding to the noise.

[0146] 2) Through the translation invariant signal subspace matching metric, we can obtain: Then the current Jth eigenvector is the eigenvector corresponding to the target; if Then the current Jth eigenvector is the eigenvector corresponding to the interference or noise.

[0147] Step 4: Use orthogonal projection to transform the obtained interference and noise subspace U I+N , from the sample covariance matrix Remove it and reconstruct the sample covariance matrix Right now:

[0148]

[0149] The sample covariance matrix will be reconstructed Substitute into the spatial power spectrum estimation, the spatial spectrum output result is:

[0150]

[0151] In the formula, ω(θ)=a(θ) / N, θ∈[0°, 180°].

[0152] Figure 3 The figure shows the azimuth process of the conventional beamforming method for processing actual sea trial data. The target moves relatively from about 95° to 120°, and the signal-to-noise ratio is relatively small. Strong interference always exists in the end-fire direction of the array from 0° to 40° and from 160° to 180°, and the energy is relatively large. In addition to the end-fire direction of the array, there is a period of about 4 minutes of strong interference at about 130° between the 2nd minute and the 6th minute. It can be seen from the figure that the existence of strong interference affects the detection capability of the target to a certain extent.

[0153] Figure 4 The EAAIS method is used to process the actual sea trial data. Figure 3 According to the processing results, the EAAIS method improves the maximum output signal-to-interference ratio from -13dB to about 7dB in the array end-fire direction at the maximum intensity and the strong interference at other angles, which effectively suppresses the strong interference in the array end-fire direction.

[0154] Figure 5The figure shows the azimuth process diagram of the method of the present invention for processing actual sea trial data. The secondary decision method of the present invention blocks a part of the interference and noise to a large extent, ensures that there is only a small amount of interference and noise in the signal subspace, and suppresses most of the noise and interference. Compared with the EAAIS method, the maximum output signal-to-interference ratio is increased from about 8dB to about 16dB, and the sidelobe at the target position is reduced by more than 8dB at most.

[0155] Figure 6 The figure shows the normalized power spectra of the actual sea trial data processed by the method of the present invention, the EAAIS method and the conventional beamforming method. It can be seen from the normalized power spectra of the three methods at a single time point that compared with the EAAIS method, the sidelobe level of the SSJ method is reduced by more than 11dB at most, and the output signal-to-interference ratio in the end-fire direction is increased by about 7dB. The overall performance is better than the EAAIS method, and the strong interference at all angles is suppressed to a greater extent.

[0156] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0157] Since the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present invention, their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0158] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated in one processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment.

[0159] According to an embodiment of the present application, the present invention also provides a computer device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor implements the steps in any of the above-mentioned method embodiments when executing the computer program.

[0160] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0161] An embodiment of the present invention further provides an information data processing terminal, which, when executed on an electronic device, provides a user input interface to implement the steps in the above method embodiments. The information data processing terminal is not limited to mobile phones, computers, and switches.

[0162] An embodiment of the present invention further provides a server, which is used to provide a user input interface to implement the steps in the above method embodiments when executed on an electronic device.

[0163] An embodiment of the present invention further provides a computer program product. When the computer program product is run on an electronic device, the electronic device can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0164] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the camera / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0165] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0166] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered within the protection scope of the present invention.

Claims

1. An adaptive interference suppression method for translation-invariant subspace matching decision, characterized in that: The method comprises the following steps: S1, establishing a data model received by the array according to the underwater array receiving targets, noise and interference; S2, using conventional beamforming to analyze the signals received by the sensor array, estimate the signal energy distribution at different angles, detect the target direction peak, determine the approximate angle range of the target signal, roughly estimate the target angle interval, and construct the decision term using the orthogonal characteristics of the eigenvector and the steering vector obtained by decomposing the sample covariance matrix; compare and analyze the decision term with the decision threshold to obtain the eigenvector corresponding to the target, interference or noise; S3, judging and analyzing whether the current feature vector is the feature vector corresponding to the target according to the properties of the translation invariant subspace matching judgment, if it meets the judgment, the current feature vector is separated into the target subspace; if it does not meet the judgment, the current feature vector is separated into the interference-noise subspace; S4, reconstruct the sample covariance matrix by orthogonally projecting the separated interference-noise subspace; S5, using the reconstructed sample covariance matrix to calculate the power spectrum density of the signal, and normalize it, using the spatial spectrum to estimate the target information, and suppressing the strong interference and noise components outside the target angle range; In step S5, the power spectrum density of the signal is calculated using the reconstructed sample covariance matrix, and normalized, and the target information is estimated using the spatial spectrum to suppress strong interference and noise components outside the target angle range. Specifically, the following steps are performed: By reconstructing the sample covariance matrix Substitute into the spatial power spectrum estimation, calculate the spectrum estimation value at each angle through angle scanning, find the peak in the angle scanning result, and estimate the energy distribution of the signal in different spatial directions; use the similarity of the weight vector formed at different angles and the reconstructed sample covariance matrix to distinguish and suppress the target signal and interference noise; when performing spatial spectrum estimation, process the weight vector of the target angle and the reconstructed sample covariance matrix to obtain a peak with a larger amplitude at the target angle; suppress the interference and noise components outside the target angle range by forming a null; In the formula, P(θ) is the power spectrum value of each signal, ω(θ) is the weight vector, is the reconstructed sample covariance matrix, ω H (θ) is the conjugate transposed form of the weight vector ω(θ), θ is the angle, θ∈[0°,180°], a(θ) is the steering vector, and N is the number of array elements.

2. The adaptive interference suppression method of the shift-invariant subspace matching decision according to claim 1, characterized in that: In step S1, establishing an array reception data model according to underwater array reception targets, noise and interference includes: placing an underwater array consisting of a plurality of hydrophones evenly distributed at a certain interval in seawater at a certain depth to obtain array reception data.

3. The adaptive interference suppression method of shift-invariant subspace matching decision according to claim 1, characterized in that: In step S2, using conventional beamforming to analyze the signal received by the sensor array includes: using conventional beamforming to obtain a rough estimate of the target angle, decomposing the sample covariance matrix features, and obtaining mutually orthogonal eigenvectors, then the decision item is: Where DF ii is the ratio of the maximum matching value of the lth eigenvector and the guidance vector in the target angle interval to the maximum matching value in all directions; k is the target interval [θ S1 ,θ S2 ] is an angle value within θ S1 is the initial value in the angle interval, θ S2 is the final value in the angle interval, k is a value from 1 to P; P is the interval [θ S1 ,θ S2 ]The number of internal angles; a(θθ k ) is the steering vector corresponding to the rough estimated angle of the target; e(l) is the sample covariance matrix The lth eigencomponent of eigendecomposition; θ kk is the angle value in the range of 0°-180° in all directions, kk is the number of angles in the direction range of 0°-180°; ii is an N×1 matrix; the superscript H is the conjugate transpose symbol.

4. The adaptive interference suppression method of the shift-invariant subspace matching decision according to claim 3 is characterized in that: The eigenvectors serve as an orthonormal basis for the steering vectors.

5. The adaptive interference suppression method of shift-invariant subspace matching decision according to claim 3, characterized in that: According to the ratio of the target angle interval to the maximum matching value of the entire angle interval, the decision threshold is set to 1, and the decision criteria are as follows: ii =1, it means that the current eigenvector belongs to the target, and the remaining eigenvectors can form the interference-noise subspace; if DF ii <1, it means that the current eigenvector belongs to interference or noise, and the remaining eigenvectors are combined into an interference-noise subspace.

6. The adaptive interference suppression method of shift-invariant subspace matching decision according to claim 1, characterized in that: In step S3, judging and analyzing whether the current feature vector is a feature vector corresponding to the target according to the property of the translation invariant subspace matching judgment specifically includes: By using the square of the F norm to analyze the metric distance of the two translation-invariant structures of each eigenvector, the state of the current eigenvector is analyzed according to the size of the metric distance, then: In the formula, ||·|| F is the Froebenius norm, U is the eigenvector, U (1) is a characteristic subvector obtained by deleting the first row of elements, U (2) Another characteristic sub-vector obtained by deleting the last row of elements, is the characteristic subvector U (1) The projection matrix is ​​composed of is the characteristic subvector U (2) The projection matrix.

7. The adaptive interference suppression method of shift-invariant subspace matching decision according to claim 6, characterized in that: The processing method of analyzing whether the current feature vector belongs to the feature vector corresponding to the target by measuring the distance through ISMM is as follows: If the metric distance of the current eigenvector is less than the metric distance of the following eigenvector, the current eigenvector is considered to be the target component, and the current eigenvector is separated into the target subspace; If the metric distance of the current eigenvector is greater than the metric distance of the subsequent eigenvector, the current eigenvector is considered to be an interference or noise component, and the current eigenvector is separated into the interference-noise subspace.

8. The adaptive interference suppression method of shift-invariant subspace matching decision according to claim 1, characterized in that: In step S4, reconstructing the sample covariance matrix includes: The separated interference-noise subspace is reconstructed into the sample covariance matrix by orthogonal projection. I+N , the interference-noise subspace U I+N Perform orthogonal projection processing; wherein the orthogonal projection processing is realized by subtracting the projection of the interference-noise subspace from the unit matrix, then: Where P ⊥ is the orthogonal projection matrix, I v is the unit matrix, and all elements on the diagonal from the upper left corner to the lower right corner are 1; U I+N is the interference-noise subspace composed of the eigenvectors corresponding to the interference and noise, and the superscript H is the conjugate transpose symbol; The sample covariance matrix is ​​reconstructed using the result of orthogonal projection processing of the interference-noise subspace, and the interference and noise components in the original sample covariance matrix are removed. Then: In the formula, is the reconstructed sample covariance matrix, is the covariance matrix of the original samples before reconstruction.

9. An adaptive interference suppression system with translation-invariant subspace matching decision, characterized in that: The system is used to control the adaptive interference suppression method of the translation invariant subspace matching decision according to any one of claims 1 to 8, and the system comprises: A data acquisition module is used to establish a data model for array reception according to underwater array reception targets, noise and interference, and to obtain array reception data; The feature vector acquisition module is used to roughly estimate the target angle interval using conventional beamforming, and construct the decision item using the orthogonal characteristics of the feature vector and the steering vector obtained by decomposing the sample covariance matrix; compare and analyze the decision item with the decision threshold to obtain the feature vector corresponding to the target, interference or noise; The target separation judgment module is used to judge and analyze whether the current feature vector is the target corresponding feature vector according to the translation invariant subspace matching judgment property. If it meets the judgment, the current feature vector is separated into the target subspace; if it does not meet the judgment, the current feature vector is separated into the interference-noise subspace; The spatial spectrum estimation module is used to calculate the normalized power spectrum using the reconstructed sample covariance matrix, estimate the target information using the spatial spectrum, and suppress strong interference and noise components outside the target angle range.

Images