Radar clutter and interference simultaneous suppression method and system based on sparse recovery
Through sparse recovery technology and dimensionality reduction algorithm, the problem that airborne MIMO radar is difficult to suppress clutter and interference in an environment with insufficient interference samples is solved, effectively interference and clutter suppression is achieved, and system computing complexity and sample requirements are reduced.
Patent Information
- Application Number
- CN202411008856.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-12-06
AI Technical Summary
In complex electromagnetic environments, airborne MIMO radars are difficult to suppress clutter and interference simultaneously when there is insufficient interference samples, resulting in failure of adaptive beam formation.
Using a sparse recovery-based method, the OMP algorithm is used to estimate the interference subspace and interferection is suppressed through the projection matrix QJ, and the clutter suppression is achieved in combination with the dimensionality reduction algorithm to reduce the system's calculation complexity and sample requirements.
Effectively suppress interference and clutter in complex environments with insufficient interference samples, reducing the system's computing complexity and sample requirements, and is especially suitable for "pure" electromagnetic environments with insufficient interference samples.
Smart Images

Figure CN119087376A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of radar signal processing, and in particular to a method and system for simultaneously suppressing radar clutter and interference based on sparse recovery. Background Art
[0002] MIMO radar is applied to airborne platforms because of its more spatial degrees of freedom and higher resolution. Compared with airborne phased array radar, airborne MIMO radar can better suppress clutter and obtain better target detection performance improvement, but its demand for independent and identically distributed training samples and real-time computing capabilities are also significantly increased. The classic STAP method has the ability to suppress clutter and interference at the same time. It only needs to estimate the interference plus clutter covariance matrix to solve the space-time two-dimensional filter weights that suppress interference and clutter, but the classic STAP method requires the radar system to have enough degrees of freedom to cancel interference and clutter. Regarding how to achieve the problem of simultaneous suppression of clutter and interference under interference background, domestic and foreign scholars have conducted many useful explorations around this problem from different angles, but all of them revolve around how to quickly and effectively estimate the clutter subspace and reduce the demand for MIMO radar training samples, while assuming that the interference samples are sufficient, "pure" and easy to obtain. However, the key to interference suppression performance depends on whether enough interference samples can be obtained to estimate the spatial adaptive beam. In complex electromagnetic environments, "pure" interference samples are insufficient, which will lead to the failure of the adaptively formed spatial interference "zero-pitched" beam. How to achieve simultaneous suppression of airborne radar clutter interference in complex electromagnetic environments with insufficient interference samples deserves further study. Summary of the invention
[0003] The purpose of the present invention is to provide a method and system for simultaneously suppressing radar clutter and interference based on sparse recovery, so as to solve the problem raised in the background technology and realize simultaneous suppression of clutter and interference in a complex electromagnetic environment with insufficient interference samples.
[0004] To achieve the above object, the present invention provides the following solution: a method for simultaneously suppressing radar clutter and interference based on sparse recovery, the specific steps of which include the following:
[0005] S1. Acquire the transmission signal of the airborne MIMO radar;
[0006] S2. Based on the spatial sparse characteristics of interference, a sparse recovery OMP algorithm is used to estimate the interference subspace of the transmitted signal, and interference suppression is achieved using the MIMO STAP radar receiving degrees of freedom to obtain an echo signal after interference suppression;
[0007] S3. Perform matched filtering on the echo signal to obtain transmission degrees of freedom, and combine the transmission degrees of freedom with time domain degrees of freedom and a dimensionality reduction algorithm to perform clutter suppression.
[0008] Preferably, in step S2, the N echo signals received by the first pulse of the MIMO radar are projected using the projection matrix Q J Processing and adaptive suppression of interference.
[0009] Preferably, the projection matrix Q J The calculation formula is:
[0010]
[0011] Among them, U P ∈C N×P is the interference subspace matrix, which is composed of the left singular eigenvalue vectors corresponding to P large eigenvalues, I represents the unit matrix, Indicates U P The conjugate transposed matrix of .
[0012] Preferably, the upper limit of the rank of the MIMO radar clutter plus interference covariance matrix is the sum of the rank of the clutter covariance matrix and the rank of the interference covariance matrix minus the number of interferences.
[0013] Preferably, the process of step S3 is:
[0014] The K echo data in the coherent processing cycle pass through the projection matrix Q J After processing, vectorization is obtained:
[0015]
[0016] Where ρ represents the scattering coefficient, represents the time-oriented vector, represents the receive steering vector, represents the launch steering vector;
[0017] The clutter, interference suppression and noise are simultaneously passed through the projection matrix Q J The covariance matrix obtained by processing is:
[0018]
[0019] Then the covariance matrix used to estimate the STAP filter weights is:
[0020]
[0021] After being processed by the dimension reduction matrix T, the suboptimal weight is calculated as:
[0022]
[0023] in, T∈C MNK×D is the dimension reduction matrix, Represents the projection matrix Q J The processed clutter covariance matrix, Represents the projection matrix Q J The processed interference covariance matrix, Represents the projection matrix Q J The processed noise covariance matrix, σ ci represents the scattering coefficient of the i-th clutter block, σ p represents the amplitude of the pth interference, σ n represents the amplitude of Gaussian white noise, Represents the covariance matrix of airborne radar received data Covariance matrix after dimensionality reduction.
[0024] On the other hand, a radar clutter and interference simultaneous suppression system based on sparse recovery is provided, comprising a signal acquisition module, an interference suppression module, and a clutter suppression module; wherein,
[0025] The signal acquisition module is used to acquire the transmission signal of the airborne MIMO radar;
[0026] The interference suppression module is used to estimate the interference subspace based on the spatial sparse characteristics of the interference, adopt the sparse recovery OMP algorithm, and use the MIMO STAP radar receiving degree of freedom to achieve interference suppression, so as to obtain the echo signal after interference suppression;
[0027] The clutter suppression module is used to perform matched filtering on the echo signal to obtain the transmission degree of freedom, and to combine the transmission degree of freedom with the time domain degree of freedom and a dimensionality reduction algorithm to perform clutter suppression.
[0028] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: a two-stage airborne MIMO radar space-time adaptive processing method is proposed. Based on the spatial sparse characteristics of interference, the first stage adopts the OMP algorithm in the sparse recovery technology to achieve interference suppression with only one sample, and the second stage combines the residual degrees of freedom with the dimensionality reduction algorithm to suppress clutter. Simulation shows that this method can effectively suppress interference and clutter, while reducing the system calculation complexity and independent and identically distributed sample requirements, and is particularly suitable for complex electromagnetic environments where "pure" interference samples are insufficient. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0030] Figure 1is a flow chart of the method of the present invention;
[0031] Figure 2 The airborne MIMO STAP radar of the present invention has four characteristic spectra of suppressive noise interference;
[0032] Figure 3 The airborne MIMO STAP radar of the present invention has 6 characteristic spectra of suppressive noise interference;
[0033] Figure 4 There are 4 clutter power spectra under suppressive noise interference for the airborne MIMO STAP radar of the present invention;
[0034] Figure 5 There are 6 clutter power spectra under suppressive noise interference for the airborne MIMO STAP radar of the present invention;
[0035] Figure 6 This is a diagram showing the interference suppression performance of the present invention in four suppressive noise interference scenarios;
[0036] Figure 7 This is a diagram showing the interference suppression performance of the present invention in six suppressive noise interference scenarios;
[0037] Figure 8 This is a performance comparison diagram of the improvement factor of the classic dimensionality reduction algorithm of the present invention and the algorithm of the present invention under 4 suppressive noise interferences;
[0038] Fig. 9 This is a performance comparison chart of the improvement factors of the classic dimensionality reduction algorithm and the algorithm of the present invention under 6 suppressive noise interferences. DETAILED DESCRIPTION
[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0040] The purpose of the present invention is to provide a method for simultaneously suppressing radar clutter and interference based on sparse recovery, such as Figure 1 As shown, the STAP method includes two-stage suppression of clutter and interference. The specific steps are as follows:
[0041] S1. Acquire the transmission signal of the airborne MIMO radar;
[0042] S2. Based on the spatial sparse characteristics of interference, the sparse recovery OMP algorithm is used to estimate the interference subspace of the transmitted signal, and the MIMO STAP radar receiving degrees of freedom are used to achieve interference suppression, and the echo signal after interference suppression is obtained;
[0043] S3. Perform matched filtering on the echo signal to obtain the transmit degree of freedom, and combine the transmit degree of freedom and the time domain degree of freedom and combine with the dimensionality reduction algorithm to suppress clutter.
[0044] The time domain degree of freedom is obtained through delay, and one delay can obtain one time domain degree of freedom.
[0045] To address the problem of insufficient interference samples in complex electromagnetic environments, based on the spatial sparse characteristics of interference, a sparse recovery algorithm is first used to estimate the interference subspace with only one sample, and the MIMO STAP radar receiving degree of freedom is used to achieve interference suppression. Then, matched filtering is performed at the receiving end to obtain the transmitting degree of freedom. The transmitting degree of freedom and the time domain degree of freedom are combined with the dimensionality reduction algorithm to suppress clutter, thereby reducing the system computational complexity and sample requirements.
[0046] 1st level interference suppression:
[0047] First, the N echo signals received by the MIMO radar for the first pulse are projected using the projection matrix Q J Processing, as the first level of interference suppression, The specific construction method is as follows:
[0048] When the radar is in a silent period (receiving signals only but not transmitting signals, without clutter and target signals), the radar receiving signal contains only interference signals and noise signals, x = x J +x N Assume that there are P incoherent suppression interferences, and the spatial frequency of the pth interference is The interference signal can be expressed as:
[0049]
[0050] Among them, ρ p represents the amplitude of the pth interference signal, represents the receive steering vector of the pth interference signal.
[0051]
[0052] The received spatial frequency is uniformly quantized to N S =η S N resolution units, η S Representing the resolution scale, the echo signal can be rewritten as:
[0053] x=Ψ s Υ s +xN ;
[0054] in, represents the supercomplete basis of spatial steering vectors, Υ s =[γ1,γ2,...,γ Ns ] T ∈C Ns×1 is the echo signal x in the basis Ψ s The distribution on the spatial spectrum of the suppressive noise interference is sparse in the entire spatial domain. s The significant value corresponds to the spatial frequency of the blocking interference, and the number of significant values is the number of interferences P. s It is obtained by solving the following minimum l0 norm optimization problem:
[0055] min||γ||0,st||x-Ψγ||2≤ε;
[0056] The optimization problem is an NP-hard problem. It has been proved that it can be converted into a norm to solve it. There are many sparse recovery algorithms to solve this kind of problem. Considering the amount of calculation, the OMP algorithm is selected in this embodiment. Based on the OMP algorithm, the interference support set of the echo signal can be accurately recovered with only one sample. The interference support set matrix can be expressed as:
[0057] A=Ψ s (:,supp(x));
[0058] Where, supp(x)={i|s.tΥ s (i)≠0,i∈[1,…,P]} is Υ s The set of non-zero elements of the vector. Performing SVD decomposition on the interference support set matrix A, we can get:
[0059]
[0060] Among them, U∈C N×N represents a unitary matrix consisting of N-order left singular vectors, V∈C P×P represents a unitary matrix consisting of right singular vectors of order P, Λ∈C N×P is an N×P matrix. P ∈C N×P is the interference subspace matrix, which is composed of the left singular eigenvalue vectors corresponding to P large eigenvalues. P The column vector of the interference space domain guide vector forms the same subspace, so the projection matrix Q is used J Processing the received data is equivalent to adaptively suppressing interference:
[0061]
[0062] Assuming that the suppression noise interference is stable within a coherent processing cycle, the projection matrix Q is used for each pulse echo data within the cycle. J to process, and then the echo data after interference suppression is processed at the second level.
[0063] Level 2 clutter suppression:
[0064] The K echo data in the coherent processing cycle pass through the projection matrix Q J After processing, vectorization is obtained:
[0065]
[0066] The covariance matrix obtained by performing similar processing on clutter, suppressed interference and noise is:
[0067]
[0068] Therefore, the covariance matrix used to estimate the STAP filter weights is:
[0069]
[0070] After being processed by the dimension reduction matrix T, the suboptimal weight is calculated as:
[0071]
[0072] in, T∈C MNK×D is the dimension reduction matrix, Represents the projection matrix Q J The processed clutter covariance matrix, Represents the projection matrix Q J The processed interference covariance matrix, Represents the projection matrix Q J The processed noise covariance matrix, σ ci represents the scattering coefficient of the i-th clutter block, σ p represents the amplitude of the pth interference, σ n represents the amplitude of Gaussian white noise, Represents the covariance matrix of airborne radar received data Covariance matrix after dimensionality reduction.
[0073] Considering P suppressive noise interference scenarios, the computational complexity and training sample requirements are shown in Table 1:
[0074] algorithm Computational complexity Sample requirements Full-dimensional STAP <![CDATA[O((MNK) 3 )]]> 2MNK Method of the present invention <![CDATA[O(PNN s )+O(P 3 )+O(D 3 )]]> 2D
[0075] Table 1
[0076] Among them, O(PNNs ) is the computational complexity of finding the support set of the suppressive noise interference, O(P 3 ) is the computational complexity of SVD decomposition, O(D 3 ) is the computational complexity of the weights. According to the RMB criterion, in order to make the clutter suppression performance loss caused by the clutter covariance matrix estimation error less than 3dB, the number of independent and identically distributed samples required is at least 2 times the system degrees of freedom. D is the system degrees of freedom after dimensionality reduction. As can be seen from Table 1, the method proposed in the present invention has a significant performance improvement in terms of computational complexity and sample requirements compared with the full-dimensional STAP method.
[0077] In this embodiment, experimental simulation is carried out to verify the beneficial effects of the method of the present invention. Consider an airborne MIMO radar configuration with co-location of transmission and reception, and the simulation parameters are shown in Table 2. Consider two scenarios with 4 suppressive noise interferences and 6 suppressive noise interferences. In the first scenario, the spatial frequencies of the interferences are set to -0.2, -0.1, 0.1 and 0.2 respectively, and in the second scenario, the spatial frequencies of the interferences are set to -0.3, -0.2, -0.1, 0.1, 0.2 and 0.3 respectively. The spatial and temporal discrete grids are divided into 50, N S =N d =50.
[0078]
[0079] Table 2
[0080] from Figure 2 , Figure 3 It can be seen that the ranks of the clutter plus interference covariance matrix of the airborne MIMO STAP radar in the three cases of no interference, 4 interferences and 6 interferences are 46, 1066 and 1576 respectively. Each increase of one interference will bring an increase of 255 large eigenvalues. It can be seen that when the airborne MIMO STAP radar faces multiple interferences, the system needs a large degree of freedom for interference cancellation.
[0081] Clutter power spectrum distribution of airborne MIMO STAP radar under suppressive noise interference. Figure 4 and Figure 5 The power spectrum distribution in two scenarios is shown below: Figure 4It can be seen that the suppressive noise interference is continuous in the time domain and fills the instantaneous bandwidth of the radar system, but it is distributed in several different directions in the spatial domain, showing a point-like distribution, which meets the sparsity requirements of the sparse recovery algorithm. As analyzed above, for every additional suppressive noise interference, the number of large eigenvalues of the radar echo covariance matrix increases by KM-1, and the classic STAP algorithm needs to increase KM-1 system degrees of freedom to cancel the interference. In practice, for MIMO STAP radars that use dimensionality reduction processing, these degrees of freedom are difficult to provide, so it is particularly necessary to consider "hierarchical" suppression and suppress the interference in the spatial domain first.
[0082] This embodiment provides the suppression performance of suppressive noise interference based on the OMP algorithm in two scenarios. Figure 6 , Figure 7 It can be seen that in the two scenarios of 4 interferences and 6 interferences, the proposed method achieves adaptive nulling at the spatial frequency where the interference is located, and the main lobe and side lobe performance are good. Compared with the traditional adaptive beamforming method, the proposed method uses the OMP algorithm to accurately recover the interference subspace and suppress interference with only one sample, which is especially suitable for complex electromagnetic environments where "pure" interference samples are scarce.
[0083] Signal-to-Clutter-to-Noise Ratio (SCINR) is often used to verify the target detection performance of airborne radar systems. The improvement factor in the presence of clutter and interference is defined as the output signal-to-clutter-to-noise ratio (SCINR out ) than the input signal-to-interference-to-noise ratio (SCINR in )for:
[0084]
[0085] This embodiment takes the three classic dimensionality reduction algorithms, ACR algorithm, JDL algorithm and GMB algorithm, as examples, and further gives the performance comparison of the improvement factor between the classic dimensionality reduction algorithm and the proposed algorithm under suppressive noise interference. In the simulation, the GMB algorithm and the JDL method, the transmitting channel, receiving channel and Doppler channel are uniformly set to. Figure 8 , Fig. 9 It can be seen that the performance of the proposed "hierarchical" suppression method is significantly better than the STAP method for simultaneous suppression of interference and clutter. The "hierarchical" JDL algorithm is close to the optimal processing performance, and the GMB algorithm has poor interference suppression performance due to the relatively low system freedom. Since the proposed method uses the MIMO radar receiving freedom to first suppress interference in the airspace, it reserves sufficient freedom for subsequent clutter suppression, so it can obtain near-optimal clutter suppression performance and signal improvement capabilities.
[0086] In view of the problems that insufficient interference samples cannot effectively form a spatial "null trap" in an active interference environment, and when clutter and interference are suppressed simultaneously, the system consumes a lot of degrees of freedom and has a large computational complexity, the present invention proposes a two-stage clutter interference simultaneous suppression method based on sparse recovery, which uses a sparse recovery OMP algorithm to suppress interference with only one sample, and then combines a dimensionality reduction algorithm to further reduce the sample requirements and computational complexity of airborne MIMO radar. Simulation experiments show that compared with the dimensionality reduction space-time adaptive processing method that uses clutter and interference simultaneous suppression, the proposed method has fewer sample requirements and lower computational complexity, and is particularly suitable for complex battlefield environments where interference samples are difficult to obtain.
[0087] On the other hand, a radar clutter and interference simultaneous suppression system based on sparse recovery is provided, comprising a signal acquisition module, an interference suppression module, and a clutter suppression module; wherein,
[0088] A signal acquisition module, used to acquire the transmission signal of the airborne MIMO radar;
[0089] The interference suppression module is used to estimate the interference subspace based on the spatial sparse characteristics of interference, use the sparse recovery OMP algorithm, and use the MIMO STAP radar receiving degrees of freedom to achieve interference suppression and obtain the echo signal after interference suppression;
[0090] The clutter suppression module is used to perform matched filtering on the echo signal to obtain the transmission degree of freedom, and to suppress clutter by combining the transmission degree of freedom and the time domain degree of freedom and combining the dimensionality reduction algorithm.
[0091] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of the present invention. At the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.
Claims
1. A method for simultaneous suppression of radar clutter and interference based on sparse recovery, It is characterized in that The specific steps include the following: S1. Acquire the transmission signal of the airborne MIMO radar; S2. Based on the spatial sparse characteristics of interference, a sparse recovery OMP algorithm is used to estimate the interference subspace of the transmitted signal, and interference suppression is achieved using the MIMO STAP radar receiving degrees of freedom to obtain an echo signal after interference suppression; S3. Perform matched filtering on the echo signal to obtain transmission degrees of freedom, and combine the transmission degrees of freedom with time domain degrees of freedom and a dimensionality reduction algorithm to perform clutter suppression.
2. The method for simultaneously suppressing radar clutter and interference based on sparse recovery according to claim 1, It is characterized in that In step S2, the N echo signals received by the first pulse of the MIMO radar are projected using the projection matrix Q J Processing and adaptive suppression of interference.
3. The method for simultaneously suppressing radar clutter and interference based on sparse recovery according to claim 2, It is characterized in that The projection matrix Q J The calculation formula is: Among them, U P ∈C N×P is the interference subspace matrix, which is composed of the left singular eigenvalue vectors corresponding to P large eigenvalues, I represents the unit matrix, Indicates U P The conjugate transposed matrix of .
4. The method for simultaneously suppressing radar clutter and interference based on sparse recovery according to claim 1, It is characterized in that The upper limit of the rank of the clutter plus interference covariance matrix of the MIMO radar is the sum of the rank of the clutter covariance matrix and the rank of the interference covariance matrix minus the number of interferences.
5. The method for simultaneously suppressing radar clutter and interference based on sparse recovery according to claim 2, It is characterized in that The process of step S3 is: The K echo data in the coherent processing cycle pass through the projection matrix Q J After processing, vectorization is obtained: Where ρ represents the scattering coefficient, represents the time-oriented vector, represents the receive steering vector, represents the launch steering vector; To suppress clutter, interference and noise, the projection matrix Q J The covariance matrix obtained by processing is: Then the covariance matrix used to estimate the STAP filter weights is: After being processed by the dimension reduction matrix T, the suboptimal weight is calculated as: in, T∈C MNK×D is the dimension reduction matrix, Represents the projection matrix Q J The processed clutter covariance matrix, Represents the projection matrix Q J The processed interference covariance matrix, Represents the projection matrix Q J The processed noise covariance matrix, σ ci represents the scattering coefficient of the i-th clutter block, σ p represents the amplitude of the pth interference, σ n represents the amplitude of Gaussian white noise, It represents the covariance matrix of the airborne radar received data after dimensionality reduction.
6. A radar clutter and interference simultaneous suppression system based on sparse recovery, It is characterized in that It includes a signal acquisition module, an interference suppression module, and a clutter suppression module; among which, The signal acquisition module is used to acquire the transmission signal of the airborne MIMO radar; The interference suppression module is used to estimate the interference subspace based on the spatial sparse characteristics of the interference, adopt the sparse recovery OMP algorithm, and use the MIMO STAP radar receiving degree of freedom to achieve interference suppression, so as to obtain the echo signal after interference suppression; The clutter suppression module is used to perform matched filtering on the echo signal to obtain the transmission degree of freedom, and to combine the transmission degree of freedom with the time domain degree of freedom and a dimensionality reduction algorithm to perform clutter suppression.
