Frequency division multiplexing multi-beam anti-interference method for shaped antenna

Abstract

The invention provides a frequency division multiplexing multi-beam anti-interference method for a shaped antenna, and the method comprises the steps: obtaining the signal data of each frequency band according to the multi-beam spatial domain data received by the shaped antenna and a filter of a frequency domain corresponding to the working frequency band of each beam; performing high-resolution spatial spectrum estimation on the signal data of each frequency band by adopting a MUSIC algorithm to obtain an incoming wave space direction of an interference signal; according to the incoming wave space direction of the interference signal, on the basis of a preset multi-beam zeroing anti-interference strategy, self-adaptive weights of all same-frequency beams are obtained in a spatial domain and frequency domain two-dimensional combined processing mode; the adaptive weights of the same-frequency beams are combined, a comprehensive directional diagram is obtained, and a multi-beam interference zeroing result is obtained, according to the method, the whole airspace is covered with the beams of different frequency points, and space-frequency combined suppression of airspace complex interference is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of signal processing, and particularly relates to a frequency division multiplexing multi-beam anti-jamming method for a shaped antenna. Background Art

[0002] A shaped antenna is an antenna with a special shape and structure, used for receiving and transmitting radio wave signals. Its beam shape can be designed according to specific application requirements to improve the efficiency of signal reception and transmission. In the field of communication, the simultaneous multi-beam technology applied to shaped antennas is an important research direction. In scenarios with high requirements for communication reliability and security, the multi-beam technology of shaped antennas can be used to form multiple beams covering different regions. Facing the increasingly complex electromagnetic environment of the contemporary era, it is required that communication satellites have a certain anti-jamming ability. By dynamically adjusting the response of the shaped antennas for reception and transmission, they can quickly respond to battlefield changes and improve the flexibility of command and dispatch. In the application of the simultaneous multi-beam anti-jamming technology for shaped antennas, under the background of a complex electromagnetic environment and interference, how to quickly and effectively select an anti-jamming strategy and take corresponding measures to form a null at the interference incident angle to adapt to the highly dynamic battlefield environment is the current research focus.

[0003] At present, the research on shaped antennas mainly focuses on single-beam traditional anti-jamming methods. There are mainly the least squares criterion, the linearly constrained minimum mean square error criterion, and the maximum signal-to-interference-plus-noise ratio criterion. Wu Meicheng proposed an improved adaptive multi-beam anti-jamming algorithm based on differential constraints based on the linearly constrained minimum variance criterion (Wu Meicheng, Zhao Yan, Wang Juan, et al. An Improved Adaptive Multi-Beam Anti-Jamming Algorithm Based on Differential Constraints [J]. Fire Control & Command Control, 2024, 49(4): 51-59, 70.). The literature "Multi-Beam Anti-Jamming Processing Method Combining Deep Coupling and Space-Time-Frequency" (Ren Qi, Wan Xiang, Zhao Lu, et al. Multi-Beam Anti-Jamming Processing Method Combining Deep Coupling and Space-Time-Frequency [J]. Electronic Design Engineering, 2023, 31(19): 6-9, 14.) proposed an adaptive space-time-frequency anti-jamming method based on satellite broadband signals, which can improve the anti-jamming performance of the array for each frequency point within the broadband. Zheng Yaping proposed a dynamic pointing multi-beam anti-jamming method based on the minimum variance distortionless response criterion. By obtaining the prior information of the satellite signal direction, using the K-means clustering algorithm, the clustering center of the signal direction is extracted, and this center is used as the constraint direction for beamforming to ensure that each satellite has a higher output SINR and improve the overall anti-jamming performance. (Zheng Yaping, Zhao Lulu, Gong Wenbin, et al. Multi-Beam Anti-Jamming Method for Satellite Navigation Signals Based on Dynamic Clustering [J]. Global Positioning System, 2021, 46(2): 32-36.). The paper "A Fixed Multi-Beam Anti-Jamming Method for Satellite Navigation Receivers" (Ma Yanxiu, Ma Zhongzhi, Li Xiaodong, et al. A Fixed Multi-Beam Anti-Jamming Method for Satellite Navigation Receivers [J]. Navigation Positioning and Timing, 2020, 7(1): 104-112.) proposed a fixed multi-beam forming anti-jamming method for satellite navigation antenna arrays aiming at the problems existing in engineering practice of adaptive beamforming, gave the engineering implementation architecture, analyzed the optimal beam space allocation scheme and satellite screening and positioning strategy, and verified the performance through simulation experiments.

[0004] Traditional multi-beam anti-jamming technologies have many limitations when facing interference. They can only handle relatively simple and single-type interference, and lack effective anti-jamming strategy selection in the face of different types of interference in a complex electromagnetic environment, and cannot play a good role. For example, for the multi-beam problem of shaped antennas, there is currently no research on eliminating complex interference in this context. Existing anti-jamming algorithms only target single-type interference. In the context of multi-beams, single interference belongs to different interference types for different beams, and using one method cannot effectively achieve multi-beam anti-jamming; for interference at different frequency points, existing methods need to divide it into multiple narrowband signals, which not only increases the computational complexity for simultaneous multi-beams, but also greatly reduces the anti-jamming effect for different frequency points. Summary of the Invention

[0005] To solve the above problems existing in the prior art, the present invention provides a frequency division multiplexing multi-beam anti-interference method for a shaped antenna, specifically including:

[0006] In a first aspect, the present invention provides a frequency division multiplexing multi-beam anti-interference method for a shaped antenna, including:

[0007] S1. Obtain signal data of each frequency band according to the multi-beam spatial domain data received by the shaped antenna and the filters in the frequency domain corresponding to the working frequency bands of each beam. The filter in the frequency domain corresponding to the working frequency band of the beam is a frequency domain filter with a bandwidth equal to the working frequency band of the beam;

[0008] S2. Use the MUSIC algorithm to perform high-resolution spatial spectrum estimation on the signal data of each frequency band to obtain the arrival space direction of the interference signal;

[0009] S3. According to the arrival space direction of the interference signal, obtain the adaptive weights of each co-frequency beam based on a preset multi-beam nulling anti-interference strategy;

[0010] S4. Perform a merging process on the adaptive weights of each co-frequency beam to obtain a combined radiation pattern and obtain the multi-beam interference nulling result.

[0011] In a second aspect, the present invention further provides a frequency division multiplexing multi-beam anti-interference device for a shaped antenna, including:

[0012] An acquisition module, configured to obtain signal data of each frequency band according to the multi-beam spatial domain data received by the shaped antenna and the filters in the frequency domain corresponding to the working frequency bands of each beam. The filter in the frequency domain corresponding to the working frequency band of the beam is a frequency domain filter with a bandwidth equal to the working frequency band of the beam;

[0013] A processing module, configured to use the MUSIC algorithm to perform high-resolution spatial spectrum estimation on the signal data of each frequency band to obtain the arrival space directions of the target signal and the interference signal; according to the arrival space direction of the interference signal, obtain the adaptive weights of each co-frequency beam based on a preset multi-beam nulling anti-interference strategy; perform a merging process on the adaptive weights of each co-frequency beam to obtain a combined radiation pattern and obtain the multi-beam interference nulling result.

[0014] In a third aspect, the present invention further provides an electronic device, including a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete mutual communication through the communication bus;

[0015] The memory is used to store a computer program;

[0016] The processor is configured to implement any method provided in the first aspect when executing the program stored in the memory.

[0017] Advantages of the present invention:

[0018] The frequency-division multiplexing multi-beam anti-jamming method for a shaped antenna provided by the present invention obtains signal data for each frequency band by using the multi-beam spatial domain data received by the shaped antenna and filters in the frequency domain corresponding to the working frequency bands of each beam; performs high-resolution spatial spectrum estimation on the signal data for each frequency band by using the MUSIC algorithm to obtain the incoming wave spatial directions of the interference signals; obtains the adaptive weights of each co-frequency beam according to the incoming wave spatial directions of the interference signals based on a preset multi-beam nulling anti-jamming strategy; performs a merging process on the adaptive weights of each co-frequency beam to obtain a combined radiation pattern, and obtains a multi-beam interference nulling result. This method realizes the multi-functional detection of multiple targets in the spatial domain by using beams with different frequency bandwidths to cover the entire spatial domain, realizes the effective detection and tracking of complex signals, and effectively suppresses complex interference.

[0019] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is a schematic flow chart of a frequency-division multiplexing multi-beam anti-jamming method for a shaped antenna provided by the present invention;

[0021] Figure 2 is a schematic structural diagram of a frequency-division multiplexing multi-beam anti-jamming device for a shaped antenna provided by the present invention;

[0022] Figure 3-25 is a series of schematic diagrams of simulation effects provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] The present invention will be further described in detail below with reference to specific embodiments, but the embodiments of the present invention are not limited thereto.

[0024] Since different interferences in a complex electromagnetic environment not only have different incoming wave directions but also different interference frequency bandwidths, in order to suppress the influence of interference on the system and solve the problems existing in the prior art, the present invention proposes a frequency-division multiplexing multi-beam anti-jamming method for a shaped antenna. The method framework is novel, which can improve the multi-beam anti-jamming ability of the shaped antenna against complex interferences with different frequency points and different incident angles, enhance the target signals in the spatial domain, and suppress the interference signals.

[0025] Figure 1 is a schematic flow chart of a frequency-division multiplexing multi-beam anti-jamming method for a shaped antenna provided by the present invention. As Figure 1 shown, the method includes:

[0026] S1. Obtain signal data for each frequency band by using the multi-beam spatial domain data received by the shaped antenna and filters in the frequency domain corresponding to the working frequency bands of each beam.

[0027] Among them, the filter in the frequency domain corresponding to the beam working frequency band is a frequency domain filter with a bandwidth equal to the beam working frequency band.

[0028] That is, based on a signal model with special requirements, the spatial domain data received by the shaped antenna is passed through filters of each beam working frequency band to obtain signal data of each frequency band.

[0029] In a possible implementation manner, step S1 includes: through fast Fourier transform, transforming the multi-beam spatial domain data received by the shaped antenna from the time domain to the frequency domain to obtain frequency domain spatial domain data; passing the frequency domain spatial domain data through filters in the frequency domain corresponding to each beam working frequency band to obtain multiple groups of narrowband frequency domain signals; through inverse Fourier transform, transforming the multiple groups of narrowband frequency domain signals from the frequency domain to the time domain to obtain signal data of each frequency band.

[0030] The signal data of each frequency band is expressed as:

[0031] X(f,t) = A(f,θ)S(f,t) + N(t),

[0032] where N(t) ∈ M×N represents zero-mean Gaussian white noise, N represents the number of time snapshots; S(f,t) represents the signal vector; A(f,θ) represents the matrix composed of the steering vectors of interference signals, f represents the center frequency of the signal, t represents time, and θ represents the angle in the spatial domain.

[0033] Specifically, in array signal processing, it is generally assumed that the signals received by the array antenna are narrowband signals and ideal far-field sources. This signal is a four-dimensional function of time, frequency, and space, and its complex form is:

[0034]

[0035] where k is the beam vector, t is the time variable, r is the position coordinate vector, A represents the signal amplitude, e represents a constant, f represents the signal frequency, T represents the transpose, r represents the position coordinate vector, k represents the beam vector.

[0036] Because the distance between the array elements and the signal source is very far, and the signal source is a narrowband signal, the signals arriving at each array element have approximately the same complex envelope, only with different phases.

[0037] For an equally spaced linear array, assuming there are M antenna array elements, and selecting the first array element (located at the origin of the rectangular coordinate system) as the reference point, the signal expressions arriving at each array element after spatial propagation are:

[0038]

[0039] Among them, α is the slow vector, w represents the angular frequency, and the phase difference between the array element and the reference array element is:

[0040]

[0041] Then the signal is expressed as:

[0042]

[0043] Therefore, the signals received by each array element are:

[0044]

[0045] The vector form of the array received signal is as follows:

[0046]

[0047] Among them, a(θ) is called the direction vector or steering vector, that is

[0048]

[0049] It can be seen from the formula that under the condition of narrowband signals, the steering vector is only related to the geometric position of the array elements and the propagation direction of the signal. Among them, the time delay of the m-th array element is: ω 0 τ m = 2π / λd(m - 1)sinθ.

[0050] The time delay τ m is related to the geometric position of the array element and the signal propagation direction. The time delay τ m when the signal arrives at each array element is:

[0051]

[0052] Among them, θ is the spatial azimuth angle, is the spatial elevation angle, (x m , y m , z m ) is the geometric position of the array element. For a linear array, it is only related to the spatial azimuth angle θ. For a planar array, it is related to both θ and are all related.

[0053] Suppose there are K signals incident on the antenna array, and the signal model is expressed as:

[0054] X(f, t) = A(f, θ)S(f, t) + N(t),

[0055] Among them, \(N(t)\in M\times N\) represents Gaussian white noise with zero mean, \(M\) represents the number of time array elements, \(N\) represents the number of time snapshots, \(S(f,t)\) represents the signal vector, \(A(f,\theta)\) represents the matrix composed of steering vectors, which is mainly affected by the array geometry, signal frequency, coupling degree between array elements (in this application, it is considered that the coupling effect is ignored under ideal conditions), etc. \(f\) represents the center frequency of the signal, \(t\) represents time, and \(\theta\) represents the angle in the spatial domain.

[0056] The spatial anti-jamming technology is to perform weighted summation processing on the outputs of all array elements, and at the same time reduce the gain of the array beam through the steering vector in the direction of the interference space to suppress the interference output as:

[0057] \(y(f,t)=w\) H \(X(f,t)=S(f,t)w\) H \(A(f,\theta)\),

[0058] Among them, \(P(f,\theta)=w\) H \(A(f,\theta)\) is called the radiation pattern, that is, the radiation pattern of the array. \(y(f,t)\) represents the output of the array, \(w\) represents the weighting of all array elements, and \(X(f,t)\) represents the data received by all array elements.

[0059] In the present invention, the multi-beam shaping antenna is a frequency-division multiplexing anti-jamming system, which can suppress narrowband interference and broadband interference in the spatial domain.

[0060] Narrowband interference is interference whose occupied frequency band is much smaller than the bandwidth of the spread-spectrum signal. It refers to interference whose power spectrum is concentrated in a very small frequency band near a certain center frequency, and this frequency band is much smaller than its center frequency. The frequency band is relatively narrow compared to the spread-spectrum bandwidth and only occupies a small part of the spread-spectrum transmission frequency band. Broadband interference refers to interference with a large ratio of bandwidth to the bandwidth of the transmitted signal or a large absolute bandwidth. It covers multiple frequency bands. Therefore, compared with narrowband interference, its impact on the signal processing system is more extensive and complex.

[0061] The signal obtained by the shaping antenna is transformed from the time domain to the frequency domain, and the frequency spectrum diagram of the signal is obtained through fast Fourier transform. To achieve the purpose of frequency-division multiplexing of the shaping antenna, by virtue of the spectral characteristics, the signal passes through frequency-domain filters with different bandwidths to segment the frequency spectrum. The narrowband frequency-domain signal remains a narrowband frequency-domain signal after passing through the filter, and the broadband frequency-domain signal becomes multiple groups of narrowband frequency-domain signals after passing through the filter. Finally, the signal is transformed from the frequency domain to the time domain through inverse Fourier transform to obtain the signal data of each frequency band, that is, the time-domain signal is obtained. Using the data of each frequency band, the multi-beam radiation pattern of the shaping antenna is nulled to achieve anti-jamming.

[0062] The frequency division multiplexing multi-beam anti-jamming algorithm for a shaped antenna provided by the present invention can enhance the system's ability to suppress complex interference from multiple dimensions through frequency domain filtering compared with traditional spatial domain anti-jamming methods, realizing the frequency division multiplexing multi-beam space-frequency joint anti-jamming of the shaped antenna.

[0063] The multi-beams of the shaped antenna cover the entire airspace with beams of different frequency bandwidths according to the working frequency bandwidth requirements, which can not only improve the system's multi-target detection and tracking ability in the airspace from the time perspective, but also effectively identify and eliminate the influence of interference with different bandwidths from the frequency domain perspective.

[0064] S2. Adopt the MUSIC algorithm to perform high-resolution spatial spectrum estimation on the signal data of each frequency band to obtain the incoming wave spatial directions of the target signal and the interference signal.

[0065] Among the signal incoming wave direction estimation algorithms (Direction Of Arrival, DOA) based on the eigenvalue decomposition of the antenna array covariance matrix, the MUSIC algorithm has general applicability. As long as the arrangement form of the antenna array is known, whether it is a linear array or a planar array, regardless of whether the array elements are equally spaced, high-resolution estimation results can be obtained. The array covariance matrix R x can obtain high-resolution estimation results. The array covariance matrix R x can be divided into two spaces, namely:

[0066]

[0067] Furthermore, it can be obtained that

[0068]

[0069] Among them, U S is a subspace composed of the larger (number of signal sources) eigenvectors among all the eigenvalues of R, called the signal subspace, and Σ S is a diagonal matrix composed of large eigenvalues; U N is a subspace composed of the smaller (number of array elements - number of signal sources) eigenvectors among all the eigenvalues of R, called the noise subspace, and Σ N is a diagonal matrix composed of small eigenvalues.

[0070] Exemplarily, if the number of signals is 1, for the M×M-dimensional matrix R x eigenvalue decomposition can obtain M eigenvalues and the corresponding eigenvectors. Σ S represents the diagonal matrix composed of large eigenvalues, and U S represents the eigenvectors corresponding to the large eigenvalues, representing the signal subspace; Σ N represents the diagonal matrix composed of small eigenvalues, and U NDenote the eigenvector corresponding to the smaller eigenvalue, which represents the noise subspace.

[0071] According to the above formula, we get

[0072] A(f,θ)RA H (f,θ)U N =0,

[0073] The matrix R is a full-rank matrix and non-singular, so its inverse exists.

[0074] Then the above formula can be transformed into A H (f,θ)U N =0, which indicates that each column vector in the matrix A(f,θ) is orthogonal to the noise subspace. Therefore, we have:

[0075]

[0076] From the orthogonal relationship between the noise eigenvector and the signal vector, the array spatial spectrum function is obtained

[0077]

[0078] Correspondingly, in a possible implementation manner, step S2 includes the following steps S21 - S24:

[0079] S21. According to the signal data X(f,t) of each frequency band received by the shaped antenna, estimate the covariance matrix, which is expressed as:

[0080]

[0081] where R represents the covariance matrix, N represents the number of time snapshots, n represents the time snapshot index, the superscript H represents the conjugate transpose of the matrix, and X(f,n) represents the n-point sampling value of X(f,t).

[0082] S22. Perform eigenvalue decomposition on the covariance matrix to obtain the eigenvalues, which are expressed as:

[0083] R = UΣU H ,

[0084] where U represents the matrix composed of eigenvectors after the eigenvalue decomposition of R, and Σ represents the diagonal matrix composed of eigenvalues after the eigenvalue decomposition of R.

[0085] S23. Arrange the eigenvectors corresponding to the largest eigenvalues with the same number of signals in descending order of eigenvalue magnitude, determine the eigenvectors corresponding to the largest eigenvalues as the signal subspace, and determine the eigenvectors corresponding to the remaining eigenvalues as the noise subspace, which is expressed as:

[0086]

[0087] where U SDenotes the signal subspace, i.e., the subspace composed of the eigenvectors corresponding to the larger eigenvalues among all the eigenvalues obtained by the eigenvalue decomposition of R, Σ S Denotes the diagonal matrix composed of large eigenvalues, U N Denotes the noise subspace, i.e., the subspace composed of the eigenvectors corresponding to the smaller (number of array elements - number of signal sources) eigenvalues among all the eigenvalues corresponding to R, Σ N Denotes the diagonal matrix composed of small eigenvalues.

[0088] S24. Determine the MUSIC spectral function based on the noise subspace, and determine the arrival space direction of the interference signal according to the peak value of the MUSIC spectral function, which is expressed as:

[0089]

[0090] P MCSIC (f,θ) denotes the MUSIC spectral function, U N Denotes the subspace composed of the eigenvectors corresponding to the smaller (number of array elements - number of signal sources) eigenvalues among all the eigenvalues of R, which is called the noise subspace, a(f,θ) denotes the steering vector, θ denotes the angle, and f denotes the center frequency of the signal.

[0091] Compared with the traditional spatial domain anti-jamming method, the frequency division multiplexing multi-beam anti-jamming algorithm for the shaped antenna provided by the present invention can effectively select a more appropriate anti-jamming strategy by determining the arrival direction of the interference signal through direction finding, so as to achieve the purpose of deeper nulls and lower side lobes, and realize the optimal selection of the spatial domain anti-jamming method for the shaped antenna.

[0092] S3. According to the arrival space direction of the interference signal, based on the preset multi-beam nulling anti-jamming strategy, obtain the adaptive weights of each co-frequency beam.

[0093] In a possible implementation manner, step S3 includes: adopting the OP adaptive anti-jamming algorithm based on the interference subspace to null the interference within the main lobe and obtain the adaptive beam weights for anti-main lobe interference;

[0094] Optionally, adopting the OP adaptive anti-jamming algorithm based on the interference subspace to null the interference within the main lobe and obtain the adaptive beam weights for anti-main lobe interference includes the following steps a1 - a5:

[0095] a1. Perform eigenvalue decomposition on the covariance matrix to obtain the target eigenvalues, which is expressed as:

[0096]

[0097] Among them, λ i Denotes the large eigenvalues after the eigenvalue decomposition of R, v i Denotes the eigenvectors after the eigenvalue decomposition of R, and P denotes the number of interferences, Denote the small eigenvalues after the eigen - decomposition of R.

[0098] a2. Arrange the target eigenvalues from large to small, and obtain the target eigen - vectors according to the arrangement result, denoted as:

[0099]

[0100] v 1 , v 2 ,..., v P , v P+1 ,..., v N ,

[0101] where v 1 , v 2 ,..., v P , v P+1 ,..., v N denote the target eigen - vectors.

[0102] a3. Determine the target interference subspace according to the eigenvalue magnitudes, denoted as:

[0103] U J = span{v 1 , v 2 ,…, v P},

[0104] where U J denotes the target interference subspace, span{} represents the vector space, and the target interference subspace is spanned by the vectors in the set {v 1 , v 2 ,…, v P}.

[0105] a4. Determine the orthogonal complement space of the interference according to the target interference subspace, denoted as:

[0106]

[0107] where, denotes the orthogonal complement space of the interference, and I represents the diagonal matrix with all elements being 1.

[0108] a5. Determine the adaptive beam weight for anti - main - lobe interference according to the orthogonal complement space of the interference, denoted as:

[0109]

[0110] where w opt denotes the adaptive beam weight (weight vector) for anti - main - lobe interference, a(f, θ 0 ) represents the steering vector of the beam main - lobe direction, and θ 0Indicates the direction of the main lobe of the beam.

[0111] In another possible implementation, step S3 includes: using the Least Mean Square (LMS) algorithm to nullify the interference within the sidelobe and obtain the adaptive beam weights for anti-sidelobe interference.

[0112] Optionally, using the Least Mean Square (LMS) algorithm to nullify the interference within the sidelobe and obtain the adaptive beam weights for anti-sidelobe interference, which is expressed as:

[0113] e(f,t) = y(f,t) - d(f,t),

[0114]

[0115] w(k + 1) = w(k) - 2μX(k)e * (k),

[0116] where y(f,t) represents the output after beamforming, d(f,t) represents the desired signal, δ represents the mean square error, E[·] represents taking the mathematical expectation, e(f,t) represents the error between the output y(f,t) after beamforming and the desired signal d(f,t), represents the gradient operator for w, represents the transient power error gradient, represents the partial derivative with respect to w, w represents the weighting of the array, w(k) represents the array weighting at the k-th iteration, μ represents the iteration step factor, R = E[X(f,t)X H (f,t)] represents the autocorrelation matrix of the input branches of each array element, r Xd = E[d(f,t)X H (f,t)] represents the cross-correlation matrix between X(f,t) and d(f,t), and the superscript * represents the conjugate, and e(k) represents the error between the array output y(f,t) and the desired signal d(f,t) after beamforming at the k-th iteration.

[0117] Specifically, the Least Mean Square (LMS) algorithm is widely used in the field of signal processing due to its small computational complexity and ease of implementation. In the field of array signal processing, the LMS algorithm can be used to implement adaptive beamforming. The LMS algorithm needs to generate a reference signal based on the prior knowledge of the useful signal, uses the Minimum Mean Square Error (MMSE) criterion as the optimization criterion, and adjusts the weights using the error between the reference signal and the output after beam weighting to achieve spatial filtering, strengthening the incoming wave in a specific direction or forming a notch. Minimum Mean Square Error (MMSE) criterion: In radar applications, the array covariance matrix usually contains the desired signal. Based on this situation, a criterion is proposed to minimize the mean square error between the array output and a certain desired response, without the need to know the direction of arrival of the desired signal.

[0118] Let the error between the desired signal d(t) and the output y(t) after beamforming be e(t), and the mean square error be δ, then:

[0119] e(f,t) = y(f,t) - d(f,t),

[0120] The mean square error δ is:

[0121]

[0122] where Re represents taking the real part, and R = E[X(f,t)X H (f,t)] is the autocorrelation matrix of the input branches of each array element, and r Xd = E[d(f,t)X H (f,t)] is the cross-correlation matrix between X(f,t) and d(f,t), and E[·] represents taking the mathematical expectation.

[0123] Using the steepest descent method, assume the initial value of w is w 0 . When adjusting the weight value along the direction of decreasing δ, the optimal weight vector w opt can be found. Because the negative gradient direction is the direction in which δ decreases fastest, take the gradient of δ with respect to w:

[0124]

[0125] Because the negative gradient direction is the direction in which δ decreases fastest, the recurrence formula for the optimal weight vector can be obtained:

[0126]

[0127] The μ in the formula is the iteration step factor, which affects the convergence rate and stability of the iteration process.

[0128] The idea of applying the LMS algorithm in engineering practice is to use the transient value to replace the steady-state value, that is, use the gradient of the transient output power error to estimate the steady-state mean square error The transient power error gradient is:

[0129]

[0130] Use to replace the gradient Then sample the value to obtain the recurrence formula of the LMS algorithm:

[0131] w(k + 1) = w(k) - 2μX(k)e * (k),

[0132] The k in the formula is the number of iterations in the LMS algorithm.

[0133] It can be seen from the expression of the LMS algorithm that in the process of the transient weight vector approaching the optimal weight vector, a new weight vector is obtained by subtracting the product of the output error value and the input signal vector from the weight vector of the previous iteration. Using the instantaneous gradient value instead of the exact statistical gradient value is an unbiased estimate.

[0134] In another possible implementation, step S3 includes: adopting the linear constraint minimum variance (LCMV) criterion to nullify the interference in the main lobe and sidelobes to obtain the adaptive beam weights for anti-main and sidelobe combined interference.

[0135] Optionally, adopting the linear constraint minimum variance (LCMV) criterion to nullify the interference in the main lobe and sidelobes to obtain the adaptive beam weights for anti-main and sidelobe combined interference, which is expressed as:

[0136]

[0137] w opt =R -1 C(C H R -1 C) -1 F,

[0138] where C = [a(f,θ 0 ), af,(θ 1 ),…, a(f,θp - 1 )], a(f,θ 0 ) represents the steering vector of the main lobe direction, a(f,θ i ), i = 1,2,…, p - 1 represents the interference steering vector, R represents the covariance matrix, w opt represents the adaptive beam weights for anti-main and sidelobe combined interference.

[0139] The co-frequency beams of the shaped antenna, for the main lobe interference and sidelobe interference of different frequencies, pass through the filters of each beam working frequency band to eliminate the influence of interference from the frequency domain perspective; for the co-frequency interference, the influence of interference cannot be eliminated from the frequency domain, and different anti-interference strategies are adopted for different incident angles to eliminate the influence of interference from the spatial domain perspective, effectively combating the interference of different frequency bands and different incident angles in a complex environment.

[0140] S4, perform a merging process on the adaptive weights of each co-frequency beam to obtain the combined pattern and get the multi-beam interference nulling result.

[0141] Optionally, S4 is expressed as:

[0142] y(f,t) = w H X(f,t) = S(f,t)w H A(f,θ)

[0143] where P(f,θ) = W H A(f,θ) represents the combined radiation pattern, w represents the adaptive beam weight for anti-interference, A(f,θ) represents the array manifold vector of the signal, y(f,t) represents the signal after suppressing interference, X(f,t) represents the signal data of each frequency band, that is, the signal received by the array, and S(f,t) represents the signal vector.

[0144] Compared with the traditional spatial anti-interference method, the frequency-division multiplexing multi-beam anti-interference algorithm for shaped antennas provided by the present invention performs combined processing on the adaptive weights of co-frequency beams to obtain a conformal combined radiation pattern, forming effective nulls not only in the interference direction, suppressing the influence of different interferences on the system from both the spatial and frequency domains, while minimizing the main lobe offset of the nulling multi-beam radiation pattern to the greatest extent and ensuring that the radiation pattern is distortion-free to the greatest extent, and finally obtaining the frequency-division multiplexing multi-beam nulling anti-interference radiation pattern for shaped antennas.

[0145] The frequency-division multiplexing multi-beam anti-interference method for shaped antennas provided by the present invention obtains the signal data of each frequency band by using the multi-beam spatial domain data received by the shaped antenna and the filters in the frequency domain corresponding to the working frequency bands of each beam. The filters in the frequency domain corresponding to the working frequency bands of the beams are frequency domain filters with a bandwidth equal to the working frequency band of the beam; the MUSIC algorithm is used to perform high-resolution spatial spectrum estimation on the signal data of each frequency band to obtain the arrival space directions of the interference signals; according to the arrival space directions of the interference signals, based on a preset multi-beam nulling anti-interference strategy, the adaptive weights of each co-frequency beam are obtained; the adaptive weights of each co-frequency beam are combined and processed to obtain a combined radiation pattern, and the multi-beam interference nulling result is obtained. This method uses beams with different frequency bandwidths to cover the entire airspace. Compared with traditional single beams, multi-beams can make full use of time domain resources to achieve multi-functional detection of multiple targets in the airspace, while frequency-division multiplexing can make full use of frequency domain resources to achieve effective detection and tracking of complex signals in the frequency domain and effective suppression of complex interferences; the ability of the system to suppress complex interferences can be enhanced from multiple dimensions through frequency domain filtering, realizing the frequency-division multiplexing multi-beam space-frequency joint anti-interference of shaped antennas; by determining the arrival direction of the interference signal through direction finding, a more appropriate anti-interference strategy can be effectively selected, so as to achieve the purpose of deeper nulls and lower side lobes, and realize the optimal selection of the spatial anti-interference method of shaped antennas; the adaptive weights of co-frequency beams are combined and processed to obtain a conformal combined radiation pattern, forming effective nulls not only in the interference direction, but also minimizing the main lobe offset of the nulling multi-beam radiation pattern to the greatest extent.

[0146] Figure 2 is a schematic structural diagram of a frequency-division multiplexing multi-beam anti-interference device for a shaped antenna provided by the present invention, as Figure 2 shown, the device includes:

[0147] An acquisition module 21, configured to obtain signal data of each frequency band according to multi-beam spatial domain data received by a shaped antenna and filters in the frequency domain corresponding to the working frequency bands of each beam, where the filter in the frequency domain corresponding to the working frequency band of the beam is a frequency domain filter with a bandwidth equal to the working frequency band of the beam.

[0148] A processing module 22, configured to perform high-resolution spatial spectrum estimation on the signal data of each frequency band by using the MUSIC algorithm to obtain the incident spatial directions of target signals and interference signals; according to the incident spatial directions of the interference signals, based on a preset multi-beam nulling anti-interference strategy, obtain the adaptive weights of each co-frequency beam; perform a merging process on the adaptive weights of each co-frequency beam to obtain a combined pattern, and obtain a multi-beam interference nulling result.

[0149] To further prove the beneficial effects of the present invention, the present invention also provides a series of simulation experiments, which are specifically as follows:

[0150] The airspace is divided into 52 regions according to different working frequency bands and covered by 52 beams. As Figure 3 shown, beams 1, 2, 3, and 4 are a group of co-frequency beams, beams 5, 17, 30, 34, 38, and 42 are a group of co-frequency beams, beams 6, 18, 29, 33, 37, and 41 are a group of co-frequency beams, beams 7, 19, 32, 36, 40, and 44 are a group of co-frequency beams, beams 8, 20, 31, 35, 39, and 43 are a group of co-frequency beams, beams 9, 13, 22, 26, 46, and 50 are a group of co-frequency beams, beams 9, 13, 22, 26, 46, and 50 are a group of co-frequency beams, beams 10, 14, 21, 25, 45, and 49 are a group of co-frequency beams, beams 11, 15, 24, 28, 48, and 52 are a group of co-frequency beams, and beams 12, 16, 23, 27, 47, and 51 are a group of co-frequency beams.

[0151] Experiment 1:

[0152] Considering the situation where the frequencies of different interferences in the airspace belong to different bandwidths, for interferences with the same bandwidth as the beam bandwidth, spatial nulling can be used to suppress the interference, while for interferences with different bandwidths from the beam bandwidth, frequency domain filtering is required. To verify that the signal in the algorithm is transformed into a frequency domain signal through discrete Fourier transform and the array anti-interference is still effective after frequency domain filtering, a simulation experiment is set up. The number of simulated interferences in Example 1 is 1, and the number of interferences set in this simulation is 2 for distinction.

[0153] Simulation experiment conditions: In this simulation experiment, the co-frequency beams are 11, 15, 24, 28, 48, and 52, and their operating frequency band is 1668 MHz - 1675 MHz. The signal parameters for the experiment are set as follows: The incident elevation angle of interference 1 is 30°, the azimuth angle is -55°, the incident angle in the u domain is 0.2877, the incident angle in the v domain is -0.4090, the center frequency point is 1668 MHz, the bandwidth of interference 1 is 7 MdB, and the signal-to-interference ratio is 30 dB.

[0154] Simulation experiment results: In the embodiment of the present invention, data is constructed based on the interference signal model, the incoming wave direction of the interference is detected, and then a suitable anti-interference strategy is selected for nulling, and finally the direction pattern of the co-frequency beam is synthesized.

[0155] For the spatial domain signal received by the shaped antenna, through the spectrogram of the fast Fourier transform, the frequency domain distribution of interference 1 in the spectrogram can be seen, as Figure 4 shown.

[0156] The spectrogram of the spatial domain signal received by the shaped antenna after being filtered by a frequency domain filter with a bandwidth of 7 MHz is as Figure 5 shown.

[0157] The MUSIC spectrogram obtained by direction finding using the MUSIC algorithm after frequency domain filtering is as Figure 6 shown.

[0158] It can be seen from Figure 6 that there is a spectral peak at the coordinates u = 0.2877 and v = -0.4090 in the MUSIC spectrogram, and the direction finding result is accurate.

[0159] The anti-main lobe interference direction pattern when the interference is incident in the beam with the spatial domain beam serial number of 11 is as Figure 7 shown.

[0160] The cross-sectional view of the anti-main lobe interference direction pattern in the u domain and v domain of the interference incident direction is as Figure 8 shown.

[0161] It can be seen from Figure 8 that the direction pattern of the beam with the serial number of 11 generates a null of -45.6158 dB in the incoming wave direction of interference 1.

[0162] The anti-side lobe interference direction pattern when the interference is incident in the beam with the spatial domain beam serial number of 15 is as Figure 9 shown.

[0163] The cross-sectional view of the anti-side lobe interference direction pattern in the u domain and v domain of the interference incident direction, as Figure 10 shown.

[0164] It can be seen from Figure 10It can be seen that a null of -46.0389 dB is generated in the direction of arrival of interference 1 in the synthesized radiation pattern.

[0165] The synthesized radiation patterns of co-frequency beams with spatial domain beam sequence numbers 11, 15, 24, 28, 28, and 32 are as Figure 11 shown.

[0166] Experiment 2:

[0167] Considering the situation where there is at least one interference in the spatial domain, for co-frequency beams, main lobe interference and sidelobe interference may occur simultaneously, and it is not applicable to the main lobe interference suppression algorithm or the sidelobe interference suppression algorithm. In order to verify that the algorithm is still effective when there are both main lobe interference and sidelobe interference, the number of interferences is increased for simulation experiments. The number of interferences in one simulation example is 1, which is different from the number of interferences set in this simulation, which is 2.

[0168] Simulation experiment conditions: In this simulation experiment, the co-frequency beams are 11, 15, 24, 28, 48, and 52, and their operating frequency band is 1668 MHz - 1675 MHz. The signal parameters set in the experiment are as follows: The incident elevation angle of interference 1 is 30°, the azimuth angle is -55°, the incident angle in the u domain is 0.2877, the incident angle in the v domain is -0.4090, the center frequency point is 1668 MHz, and the interference-to-noise ratio is 30 dB; the incident elevation angle of interference 2 is -30°, the azimuth angle is 40°, the incident angle in the u domain is -0.3816, the incident angle in the v domain is -0.3229, the center frequency point is 1670 MHz, and the interference-to-noise ratio is 30 dB.

[0169] Simulation experiment results: The inventive embodiment constructs data based on the interference signal model, detects the direction of arrival of the interference, then selects a suitable anti-interference strategy for nulling, and finally synthesizes the radiation pattern of the co-frequency beam.

[0170] For the spatial domain signal received by the shaped antenna, through the spectrogram of the fast Fourier transform, the frequency domain distribution of interference 1 in the spectrogram can be seen, as Figure 12 shown.

[0171] For the spatial domain signal received by the shaped antenna, the spectrogram after filtering by a frequency domain filter with a bandwidth of 7 MHz is as Figure 13 shown.

[0172] The MUSIC spectrogram obtained by direction finding using the MUSIC algorithm after frequency domain filtering is as Figure 14 shown.

[0173] From Figure 14 it can be seen that there is a spectral peak at the coordinates of u = 0.2877 and v = -0.4090 in the MUSIC spectrogram, and there is a spectral peak at the coordinates of u = -0.3816 and v = -0.3229, and the direction finding result is accurate.

[0174] The anti-main and sidelobe combined interference pattern that interferes with the beam with spatial domain beam sequence number 11 is as Figure 15 shown.

[0175] The cross-sectional views of the main lobe interference in the u-domain and v-domain of the interference incident direction are as Figure 16 shown.

[0176] From Figure 16 it can be seen that the pattern with beam sequence number 11 generates a null of -44.8872 dB in the incident direction of interference 1.

[0177] The cross-sectional views of the sidelobe interference in the u-domain and v-domain of the interference incident direction are as Figure 17 shown.

[0178] From Figure 17 it can be seen that the pattern with beam sequence number 11 generates a null of -70.3313 dB in the incident direction of interference 2.

[0179] The combined pattern of co-frequency beams with spatial domain beam sequence numbers 11, 15, 24, 28, 28, and 32 is as Figure 18 shown.

[0180] Experiment 3:

[0181] Considering the interference of different frequencies in the spatial domain, for co-frequency beams, there may be interference frequencies outside the working frequency band. To verify that the algorithm can handle interference outside the working frequency band, interference outside the working frequency band is now added for simulation experiments. The frequencies simulated in Example 1 are 1668 MHz and 1670 MHz, and the interference frequencies set in this simulation are 1068 MHz and 1668 MHz for distinction.

[0182] Simulation experiment conditions: In this simulation experiment, the co-frequency beams are 11, 15, 24, 28, 48, and 52, and their working frequency band is 1668 MHz - 1675 MHz. The signal parameters set in the experiment are as follows: The incident elevation angle of interference 1 is 30°, the azimuth angle is -55°, the incident angle in the u-domain is 0.2877, the incident angle in the v-domain is -0.4090, the center frequency point is 1068 MHz, and the interference-to-noise ratio is 30 dB; The incident elevation angle of interference 2 is -30°, the azimuth angle is 40°, the incident angle in the u-domain is -0.3816, the incident angle in the v-domain is -0.3229, the center frequency point is 1668 MHz, and the interference-to-noise ratio is 30 dB.

[0183] Simulation experiment results: The invention example constructs data based on the interference signal model, detects the incident direction of the interference, then selects a suitable anti-interference strategy to null it, and finally synthesizes the pattern of the co-frequency beams.

[0184] The spatial domain signal received by the shaped antenna, through the spectrogram of the fast Fourier transform, can show the frequency domain distribution of interference 1 in the spectrogram, as Figure 19 shown.

[0185] The spectrogram of the spatial domain signal received by the shaped antenna after being filtered by a frequency domain filter with a bandwidth of 7 MHz is as Figure 20 shown.

[0186] From Figure 20 it can be seen that in the spectrum after the green bottle, there is only a signal within the working frequency band, and the signals outside the working frequency band have been filtered out.

[0187] The MUSIC spectrogram obtained by direction finding through the MUSIC algorithm after frequency domain filtering is as Figure 21 shown.

[0188] From Figure 21 it can be seen that there is a spectral peak at the coordinates of u = -0.3816 and v = -0.3229 in the MUSIC spectrogram, and the direction finding result is accurate.

[0189] The anti-main and side lobe combined interference direction pattern of the interference incident on the beam with the spatial domain beam serial number of 11 is as Figure 22 shown.

[0190] The cross-sectional views of the main lobe interference in the u-domain and v-domain of the interference incident direction are as Figure 23 shown.

[0191] From Figure 23 it can be seen that the gain of the direction pattern with the beam serial number of 11 in the direction of interference 1 is 15.54 dB, and no null is generated.

[0192] The cross-sectional views of the side lobe interference in the u-domain and v-domain of the interference incident direction are as Figure 24 shown.

[0193] From Figure 24 it can be seen that the direction pattern with the beam serial number of 11 generates a null of -75.3746 dB in the direction of arrival of interference 2.

[0194] The combined direction pattern of the co-frequency beams with the spatial domain beam serial numbers of 11, 15, 24, 28, 28, and 32 is as Figure 25 shown.

[0195] The present invention also provides a structure of an electronic device, including a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete mutual communication through the communication bus,

[0196] The memory is used to store computer programs;

[0197] A processor, when executing a program stored in a memory, implements the steps provided in the above method embodiments.

[0198] A communication interface is used for communication between the above electronic device and other devices.

[0199] The method provided in the embodiments of the present invention can be applied to an electronic device. Specifically, the electronic device can be: a desktop computer, a portable computer, a smart mobile terminal, a server, etc. There is no limitation here, and any electronic device that can implement the present invention belongs to the protection scope of the present invention.

[0200] The present invention also provides a computer-readable storage medium, in which a computer program is stored, and when the computer program is executed by a processor, the steps provided in the above method embodiments are implemented.

[0201] For the device / electronic device / storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple. For the specific content, beneficial effects, and other related aspects, please refer to the partial description of the method embodiments.

[0202] The terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, "a plurality" means two or more, unless otherwise specifically defined.

[0203] The above content is a further detailed description of the present invention in combination with specific preferred embodiments. It cannot be determined that the specific implementation of the present invention is only limited to these descriptions. For those of ordinary skill in the technical field to which the present invention belongs, without departing from the concept of the present invention, several simple deductions or substitutions can still be made, and all should be regarded as belonging to the protection scope of the present invention.

Claims

1. A frequency division multiplexing multi-beam anti-interference method for a shaped antenna, characterized in that: include: S1, obtaining signal data of each frequency band according to the multi-beam spatial domain data received by the shaped antenna and the frequency domain filter corresponding to the working frequency band of each beam, wherein the frequency domain filter corresponding to the working frequency band of the beam is a frequency domain filter having a bandwidth of the working frequency band of the beam; S2, using the MUSIC algorithm to perform high-resolution spatial spectrum estimation on the signal data of each frequency band to obtain the spatial direction of the interference signal; S3, according to the spatial direction of the interference signal, based on a preset multi-beam nulling anti-interference strategy, obtaining adaptive weights of each co-frequency beam; S4, combining the adaptive weights of the same-frequency beams to obtain a comprehensive directional diagram and a multi-beam interference nulling result.

2. The method according to claim 1, characterized in that: Said S1 comprises: By fast Fourier transform, the multi-beam spatial domain data received by the shaped antenna is transformed from the time domain to the frequency domain to obtain frequency domain spatial domain data; The frequency domain and spatial domain data are passed through filters in the frequency domain corresponding to the working frequency bands of each beam to obtain multiple groups of narrowband frequency domain signals; The plurality of groups of narrowband frequency domain signals are transformed from the frequency domain to the time domain by inverse Fourier transform to obtain the signal data of each frequency band; The signal data of each frequency band is expressed as: X(f,t)=A(f,θ)S(f,t)+N(t), Where N(t)∈M×N represents zero-mean Gaussian white noise, M represents the number of time array elements, N represents the number of time snapshots, S(f,t) represents the signal vector, A(f,θ) represents the matrix composed of steering vectors, f represents the central frequency of the signal, t represents time, and θ represents the angle in the spatial domain.

3. The method according to claim 2, characterized in that The S2 comprises: According to the signal data X(f, t) of each frequency band received by the shaped antenna, the covariance matrix estimated is expressed as: Where R represents the covariance matrix, the superscript H represents the conjugate transpose of the matrix, and X(f,n) represents the n-point sampling value of X(f,t); The covariance matrix is ​​subjected to eigenvalue decomposition to obtain eigenvalues, which are expressed as: R=UΣU H , Among them, U represents the matrix composed of eigenvectors after eigendecomposition of R, and Σ represents the diagonal matrix composed of eigenvalues ​​after eigendecomposition of R; According to the order of the eigenvalues, the eigenvector corresponding to the largest eigenvalue with equal number of signals is determined as the signal subspace, and the eigenvectors corresponding to the remaining eigenvalues ​​are determined as the noise subspace, which is expressed as: Among them, U S represents the signal subspace, Σ S represents the diagonal matrix composed of large eigenvalues, U N represents the noise subspace, Σ N represents a diagonal matrix composed of small eigenvalues; According to the orthogonal relationship between the noise subspace and the signal subspace, the MUSIC spectrum function is determined, and the spatial direction of the interference signal is determined according to the peak value of the MUSIC spectrum function, which is expressed as: P MCSIC (f,θ) represents the MUSIC spectrum function, θ represents the angle, f represents the center frequency of the signal, and a(f,θ) represents the steering vector.

4. The method according to claim 3, characterized in that The S3 includes: Adopting the OP adaptive anti-interference algorithm based on interference subspace, the interference in the main lobe is zeroed to obtain the adaptive beam weights to resist the main lobe interference; and / or, The minimum mean square error (LMS) algorithm is used to zero the interference in the side lobe and obtain the adaptive beam weights that resist side lobe interference. and / or, The linear constrained minimum variance (LCMV) criterion is used to zero the interference in the main lobe and side lobes, and obtain the adaptive beam weights that can resist the joint interference of the main lobe and side lobes.

5. The method according to claim 4, characterized in that The OP adaptive anti-interference algorithm based on the interference subspace is used to zero the interference in the main lobe to obtain the adaptive beam weight for resisting the main lobe interference, including: The covariance matrix is ​​subjected to eigenvalue decomposition to obtain the target eigenvalue, which is expressed as: Among them, λ i represents the large eigenvalue of R after eigendecomposition, v i represents the eigenvector of R after eigendecomposition, P represents the number of interferences, Represents the small eigenvalue of R after eigendecomposition; The target eigenvalues ​​are arranged from large to small, and the target eigenvector is obtained according to the arrangement result, which is expressed as: v1,v2,...,v P ,v P+1 ,...,v N , Among them, v1,v2,...,v P ,v P+1 ,...,v N represents the target feature vector; According to the size of the eigenvalue, the target interference subspace is determined, which is expressed as: U J =span{v1,v2,…,v P }, Among them, U J represents the target interference subspace, span{} represents the vector space, which consists of the set {v1,v2,…,v P The vectors in} form the target interference subspace; According to the target interference subspace, the orthogonal complementary space of the interference is determined, which is expressed as: in, represents the orthogonal complement space of interference, and I represents a diagonal matrix of all 1s; According to the orthogonal complementary space of the interference, the adaptive beam weight for resisting main lobe interference is determined, which is expressed as: Among them, w opt represents the adaptive beam weight for resisting main lobe interference, a(f,θ0) represents the steering vector pointing to the main lobe of the beam, and θ0 represents the pointing direction of the main lobe of the beam.

6. The method according to claim 4, characterized in that The minimum mean square error LMS algorithm is used to zero the interference in the side lobe to obtain the adaptive beam weight for resisting side lobe interference, which is expressed as: e(f,t)=y(f,t)-d(f,t), w(k+1)=w(k)-2μX(k)e * (k), Where y(f, t) represents the output after beamforming, d(f, t) represents the expected signal, δ represents the mean square error, E[·] represents the mathematical expectation, and e(f, t) represents the error between the output y(f, t) after beamforming and the expected signal d(f, t). represents the gradient operator for w, represents the transient power error gradient, represents the partial derivative of w, w represents the weight of the array, w(k) represents the array weight of the kth iteration, μ represents the iteration step factor, R = E[X(f, t)X H (f, t)] represents the autocorrelation matrix of each input element branch, r Xd =E[d(f,t)X H (f, t)] represents the cross-correlation matrix between X(f, t) and d(f, t), and the superscript * represents conjugation, e * (k) represents the error between the array output y(f, t) after the kth iterative beamforming and the desired signal d(f, t).

7. The method according to claim 4, characterized in that The linear constrained minimum variance LCMV criterion is used to zero the interference in the main lobe and the side lobe to obtain the adaptive beam weights for resisting the joint interference of the main lobe and the side lobe, which is expressed as: w opt =R -1 C(C H R -1 C) -1 F, Among them, C=[a(f,θ0),a(f,θ1),…,a(f,θ p-1 )], a(f,θ0) represents the steering vector pointing to the main lobe, a(f,θ i ), i = 1, 2, ..., p-1 represents the interference steering vector, w opt Represents the adaptive beam weight for resisting main-lobe and side-lobe combined interference.

8. The method according to claim 7, characterized in that The S4 is expressed as: y(t)=w H X(f,t)=S(f,t)w H A(f,θ), Where P(θ) = w H A(f,θ) represents the comprehensive directional pattern, w represents the adaptive beam weight for anti-interference, A(f,θ) represents the array manifold vector of the signal, y(f,t) represents the signal after interference suppression, X(f,t) represents the signal data of each frequency band, and S(f,t) represents the signal vector.

9. A frequency division multiplexing multi-beam anti-interference device for a shaped antenna, characterized in that: include: An acquisition module is used to obtain signal data of each frequency band according to the multi-beam spatial domain data received by the shaped antenna and the frequency domain filter corresponding to the working frequency band of each beam, wherein the frequency domain filter corresponding to the working frequency band of the beam is a frequency domain filter with a bandwidth of the working frequency band of the beam; The processing module is used to use the MUSIC algorithm to perform high-resolution spatial spectrum estimation on the signal data of each frequency band to obtain the spatial directions of the incoming waves of the target signal and the interference signal; according to the spatial directions of the incoming waves of the interference signal, based on the preset multi-beam nulling anti-interference strategy, the adaptive weights of each co-frequency beam are obtained; the adaptive weights of each co-frequency beam are combined to obtain a comprehensive directional diagram, and the multi-beam interference nulling result is obtained.

10. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, for implementing any of the methods described in claims 1-8 when executing a program stored in a memory.

Images