Adaptive subspace detection method and system for extended target

Abstract

The invention belongs to the technical field of multi-channel signal detection, and particularly relates to a self-adaptive subspace detection method for an extended target, which comprises the following steps: S1, constructing a target signal subspace matrix and an interference signal subspace matrix which are respectively expressed by a matrix A and a matrix J, the dimensions of the matrixes are respectively N * p and N * q, (p + q) is less than or equal to N, and J is less than or equal to N; n represents the number of system channels; s2, using the training sample to construct a sampling covariance matrix S = YYH, Y representing the training sample, and a symbol (.) H representing conjugate transpose; according to the method, the detection statistic t is constructed, so that interference suppression, signal accumulation and extended target detection are integrally realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of multi-channel signal detection, and in particular relates to an adaptive subspace detection method and system for extended targets. Background Art

[0002] In the field of space-time adaptive detection (STAD), multi-channel signal detection utilizes the correlation characteristics of signals and noise in different channels (i.e., the spatial information of the target), and therefore has better detection performance than traditional single-channel signal detection. However, with the improvement of radar resolution, the method of treating the detected target as a single scattering point is obviously no longer applicable. In actual scenarios, many targets (such as large aircraft, ships, etc.) appear in the form of several isolated reflection points, that is, the target appears in a distributed form, which is called an extended target.

[0003] Most of the existing detectors for extended target detection in STAD assume that the target steering vector is completely known. However, in reality, it is usually difficult to obtain accurate target steering vector information due to array element errors, multipath effects, target maneuvers, and antenna shape distortion. The subspace model can effectively suppress the above-mentioned steering vector mismatch problem. In this model, the target steering vector is assumed to be located in a known subspace, but the corresponding coordinates are unknown. The emergence of the subspace model provides a new solution to the extended target detection problem.

[0004] At present, the adaptive subspace detection method for extended targets is mainly to design a detector based on the generalized likelihood ratio test (GLRT) criterion. This GLRT detector can obtain better detection performance in certain scenarios. However, since the noise covariance matrix, the subspace of the target signal and the subspace coordinates are unknown, the GLRT detector cannot obtain detection advantages in all parameter scenarios, and it is extremely necessary to develop new subspace detection. Based on this consideration, the present invention adopts the Durbin criterion to design a new adaptive subspace detection method and system for extended targets, providing a new way for extended target detection. Summary of the invention

[0005] The purpose of the present invention is to provide an adaptive subspace detection method for extended targets, which can realize interference suppression, signal accumulation and extended target detection in an integrated manner by constructing a detection statistic t.

[0006] The technical solution adopted by the present invention is as follows:

[0007] An adaptive subspace detection method for extended targets, the detection method comprising the following steps:

[0008] S1: Construct the target signal subspace matrix and the interference signal subspace matrix, represented by matrices A and J respectively, with the dimensions of the matrices being N×p and N×q respectively, where (p+q)≤N, and N represents the number of system channels;

[0009] In S1, the constructed target signal matrix A and interference signal matrix J are both column-full rank unitary matrices, and the unitary matrices A and J are linearly independent.

[0010] S2: Use training samples to construct the sampling covariance matrix S = YY H , where Y represents the training sample and the symbol (·) H represents conjugate transpose;

[0011] S3: Use the sampling covariance matrix S to construct a quasi-whitening matrix, and then use it to perform quasi-whitening processing on the target signal subspace matrix A, the interference signal subspace matrix J and the data matrix to be detected X, and obtain the whitening matrices and The dimension of the data matrix X to be detected is N×K, where K represents the extended dimension;

[0012] In S3, the quasi-whitening matrix constructed by the sampling covariance matrix S is used to perform quasi-whitening processing on the target signal subspace matrix A, the interference signal subspace matrix J and the to-be-detected data matrix X. The specific operation of the quasi-whitening processing is as follows: and

[0013] S4: Using the quasi-whitening matrix Construct the orthogonal projection matrix of the target signal And the corresponding complementary projection matrix At the same time, using the quasi-whitening matrix Constructing the orthogonal projection matrix of the interference signal And the corresponding complementary projection matrix

[0014] In S4, the quasi-whitening target signal matrix is ​​used The constructed orthogonal projection matrix and complementary projection matrix are expressed as and in(·) -1 represents the inverse of the matrix, I N is a unit matrix with dimension N×N; using the quasi-whitening interference signal matrix The constructed orthogonal projection matrix and complementary projection matrix are expressed as and

[0015]

[0016] S5: constructing a detection statistic using the quasi-whitened data matrix to be detected, the orthogonal projection matrix of the target signal and the orthogonal projection matrix of the interference signal;

[0017] In S5, the detection statistic is constructed by using the quasi-whitened data matrix to be detected, the orthogonal projection matrix of the target signal and the orthogonal projection matrix of the interference signal, which is expressed as:

[0018]

[0019] Here, tr(·) represents the trace of the matrix.

[0020] S6: Determine the detection threshold using the preset value of false alarm probability;

[0021] In S6, the detection threshold determined by the preset value of false alarm probability is:

[0022] η=t(n)

[0023] Where η represents the detection threshold, n = [Qμ], Q is the number of Monte Carlo simulations, μ represents the false alarm probability value set by the system, [·] represents the rounding operation, and t(n) is the sequence

[0024]

[0025] The nth largest value when arranged from largest to smallest, X(k) represents the k-th realization k=1, 2,…,Q of the data to be detected containing only interference and thermal noise components.

[0026] S7: Compare the detection statistic with the detection threshold, and determine whether the target exists.

[0027] In S7, the detection statistic is compared with the detection threshold to determine whether the target exists, and the determination is made in the following two cases:

[0028] If the detection statistic t is greater than or equal to the detection threshold η, the target is determined to exist;

[0029] If the detection statistic t is less than the detection threshold η, it is determined that the target does not exist.

[0030] An adaptive subspace detection system for extended targets includes the following modules:

[0031] Target and interference data matrix construction module: used to construct target signal subspace matrix and interference signal subspace matrix;

[0032] Sampling covariance matrix construction module: used to construct the sampling covariance matrix through training samples;

[0033] Data quasi-whitening module: used to perform quasi-whitening processing on the target signal subspace matrix, the interference signal subspace matrix and the data matrix to be detected, and obtain the quasi-whitened target signal subspace matrix, the quasi-whitened interference signal subspace matrix and the quasi-whitened data matrix to be detected respectively;

[0034] Detection statistic construction module: used to construct detection statistics through quasi-whitened target signal matrix, quasi-whitened interference signal orthogonal projection matrix and quasi-whitened to-be-detected data matrix;

[0035] Detection threshold calculation module: used to determine the detection threshold through detection statistics and false alarm probability preset value;

[0036] Target decision module: used to compare the detection statistic with the detection threshold and determine whether the target exists.

[0037] The technical effects achieved by the present invention are:

[0038] The adaptive subspace detection method and system for extended targets of the present invention realizes interference suppression, signal accumulation and extended target detection in an integrated manner by constructing a detection statistic t.

[0039] An adaptive subspace detection method and system for extended targets of the present invention uses a sampling covariance matrix S to perform quasi-whitening processing on a to-be-detected data matrix, a target signal and an interference signal matrix, so that the detector proposed by the present invention has a constant false alarm characteristic.

[0040] The adaptive subspace detection method and system for extended targets of the present invention can solve the problem of subspace adaptive detection of extended targets, and improve the radar target detection performance compared with the existing detection method based on the GLRT criterion. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a schematic diagram of the process of Embodiment 1 of the present invention;

[0042] Figure 2 is a structural framework diagram of a detection system according to Embodiment 2 of the present invention;

[0043] Figure 3 is a schematic diagram of the detection probability results of the extended target under different signal-to-noise ratios of the experimental example of the present invention;

[0044] Figure 4 It is a schematic diagram of the detection probability results of the experimental example of the present invention at different expansion degrees. DETAILED DESCRIPTION

[0045] In order to make the purpose and advantages of the present invention more clearly understood, the present invention is specifically described below in conjunction with embodiments. It should be understood that the following text is only used to describe one or several specific embodiments of the present invention, and does not strictly limit the scope of protection of the specific claims of the present invention.

[0046] Take the airborne radar linear array as the research object, assuming that the number of array elements is N a , the number of pulses is N b , the corresponding system dimension N = N a ×N b The sample to be detected X occupies K echo units, and each echo unit is independent and identically distributed, so the dimension of X is N×K. In the H1 hypothesis, the sample to be detected X contains the target signal component T, the interference signal component A and the noise component N. The signal components T and A can be expressed as T=HΦ and A=JΨ respectively;

[0047] Among them, H and J are N×p-dimensional target signal subspace and N×q-dimensional interference signal subspace, respectively, both of which are column-full rank unitary matrices; Φ and Ψ are the corresponding unknown subspace coordinates, and the corresponding dimensions are p×1 and q×1, respectively. i ,i=1,2,…,K satisfies the mean value of 0 N×1 , the covariance is R t The complex Gaussian distribution, that is, n i ~CN N (0 N×1 ,R t ).

[0048] In the H0 hypothesis, the sample to be tested X only contains the interference signal component A and the noise component N. In summary, the binary hypothesis detection problem for extended target detection can be expressed as:

[0049]

[0050] Among them, X L is a training sample set with dimension N×L, where L is equal to the total number of training samples; N L Then it is the training sample X L The noise component in each column n e,i , i=1,2,…,L all satisfy the mean value of 0 N×1 , a complex Gaussian distribution with covariance R, that is, n i ~CN N (0 N×1 ,R).

[0051] Embodiment 1:

[0052] like Figure 1As shown, an adaptive subspace detection method for extended targets, the detection method comprises the following steps:

[0053] S1: Construct the target signal subspace matrix and the interference signal subspace matrix, represented by matrices A and J respectively, with the dimensions of the matrices being N×p and N×q respectively, where (p+q)≤N, and N represents the number of system channels;

[0054] In S1, the constructed target signal matrix A and interference signal matrix J are both column-full rank unitary matrices, and the unitary matrices A and J are linearly independent.

[0055] S2: Use training samples to construct the sampling covariance matrix S = YY H , where Y represents the training sample and the symbol (·) H represents conjugate transpose;

[0056] S3: Use the sampling covariance matrix S to construct a quasi-whitening matrix, and then use it to perform quasi-whitening processing on the target signal subspace matrix A, the interference signal subspace matrix J and the data matrix to be detected X, and obtain the whitening matrices and The dimension of the data matrix X to be detected is N×K, where K represents the extended dimension;

[0057] In S3, the quasi-whitening matrix constructed by the sampling covariance matrix S is used to perform quasi-whitening processing on the target signal subspace matrix A, the interference signal subspace matrix J and the to-be-detected data matrix X. The specific operation of the quasi-whitening processing is as follows: and

[0058] S4: Using the quasi-whitening matrix Construct the orthogonal projection matrix of the target signal And the corresponding complementary projection matrix At the same time, using the quasi-whitening matrix Constructing the orthogonal projection matrix of the interference signal And the corresponding complementary projection matrix

[0059] In S4, the quasi-whitening target signal matrix is ​​used The constructed orthogonal projection matrix and complementary projection matrix are expressed as and in(·) -1 represents the inverse of the matrix, I N is a unit matrix with dimension N×N; using the quasi-whitening interference signal matrix The constructed orthogonal projection matrix and complementary projection matrix are expressed as and

[0060]

[0061] S5: constructing a detection statistic using the quasi-whitened data matrix to be detected, the orthogonal projection matrix of the target signal and the orthogonal projection matrix of the interference signal;

[0062] In S5, the detection statistic is constructed by using the quasi-whitened data matrix to be detected, the orthogonal projection matrix of the target signal and the orthogonal projection matrix of the interference signal, which is expressed as:

[0063]

[0064] Here, tr(·) represents the trace of the matrix.

[0065] S6: Determine the detection threshold using the preset value of false alarm probability;

[0066] In S6, the detection threshold determined by the preset value of false alarm probability is:

[0067] η=t(n)

[0068] Where η represents the detection threshold, n = [Qμ], Q is the number of Monte Carlo simulations, μ represents the false alarm probability value set by the system, [·] represents the rounding operation, and t(n) is the sequence

[0069]

[0070] The nth largest value when arranged from largest to smallest, X(k) represents the k-th realization k=1, 2,…,Q of the data to be detected containing only interference and thermal noise components.

[0071] S7: Compare the detection statistic with the detection threshold, and determine whether the target exists.

[0072] In S7, the detection statistic is compared with the detection threshold to determine whether the target exists, and the determination is made in the following two cases:

[0073] If the detection statistic t is greater than or equal to the detection threshold η, the target is determined to exist;

[0074] If the detection statistic t is less than the detection threshold η, it is determined that the target does not exist.

[0075] Embodiment 2:

[0076] like Figure 2 As shown, an adaptive subspace detection system for extended targets includes the following modules:

[0077] Target and interference data matrix construction module: used to construct target signal subspace matrix and interference signal subspace matrix;

[0078] Sampling covariance matrix construction module: used to construct the sampling covariance matrix through training samples;

[0079] Data quasi-whitening module: used to perform quasi-whitening processing on the target signal subspace matrix, the interference signal subspace matrix and the data matrix to be detected, and obtain the quasi-whitened target signal subspace matrix, the quasi-whitened interference signal subspace matrix and the quasi-whitened data matrix to be detected respectively;

[0080] Detection statistic construction module: used to construct detection statistics through quasi-whitened target signal matrix, quasi-whitened interference signal orthogonal projection matrix and quasi-whitened to-be-detected data matrix;

[0081] Detection threshold calculation module: used to determine the detection threshold through detection statistics and false alarm probability preset value;

[0082] Target decision module: used to compare the detection statistic with the detection threshold and determine whether the target exists.

[0083] Experimental example:

[0084] Assume that the number of radar system channels is N = 10 and the false alarm probability is 1 × 10 -4 , the element at the (i, j)th position of the covariance matrix R of the clutter and noise is set to R(i, j) = 0.95 |i-j| , i, j = 1, 2,…, N, where the symbol |·| indicates taking the absolute value. Figure 3 The detection probability of the target under different signal-to-noise ratios (SNRs) is given, where the subspace dimensions of the target signal and the interference signal are p=3 and q=3, K=4, the number of training samples is L=2N, the interference-to-noise ratio (INR) is 5dB, and the INR and SNR are

[0085] INR=ψ H J H R -1 Jψ

[0086] and

[0087] SNR=φ H A H R -1 Aφ

[0088] from Figure 3 It can be seen that when the signal-to-noise ratio is higher than 17dB, the detection probability of the proposed method for extended targets is higher than 80%. Compared with the existing GLRT criterion detector, the proposed detection method has better detection performance. In addition, Figure 4The detection probability of the proposed method for targets with different extension degrees is given, where p = 5, q = 4, signal-to-noise ratio SNR = 15dB, and interference-to-noise ratio INR = 5dB. From the results, it can be seen that when the extension degree is low, the detection performance of the proposed Durbin criterion detector for extended targets is significantly better than that of the existing GLRT detector.

[0089] The above is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be considered as the protection scope of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention shall be implemented according to the conventional means in the art unless otherwise specified and limited.

Claims

1. An adaptive subspace detection method for extended targets, characterized by: The detection method comprises the following steps: S1: Construct the target signal subspace matrix and the interference signal subspace matrix, represented by matrices A and J respectively, with the dimensions of the matrices being N×p and N×q respectively, where (p+q)≤N, and N represents the number of system channels; S2: Use training samples to construct the sampling covariance matrix S = YY H , where Y represents the training sample and the symbol (·) H represents conjugate transpose; S3: Use the sampling covariance matrix S to construct a quasi-whitening matrix, and then use it to perform quasi-whitening processing on the target signal subspace matrix A, the interference signal subspace matrix J and the data matrix to be detected X, and obtain the whitening matrices and The dimension of the data matrix X to be detected is N×K, where K represents the extended dimension; S4: Using the quasi-whitening matrix Construct the orthogonal projection matrix of the target signal And the corresponding complementary projection matrix At the same time, using the quasi-whitening matrix Constructing the orthogonal projection matrix of the interference signal And the corresponding complementary projection matrix S5: constructing a detection statistic using the quasi-whitened data matrix to be detected, the orthogonal projection matrix of the target signal and the orthogonal projection matrix of the interference signal; S6: Determine the detection threshold using the preset value of false alarm probability; S7: Compare the detection statistic with the detection threshold, and determine whether the target exists.

2. The adaptive subspace detection method for extended targets according to claim 1, characterized in that: In S1, the constructed target signal matrix A and interference signal matrix J are both column-full rank unitary matrices, and the unitary matrices A and J are linearly independent.

3. The adaptive subspace detection method for extended targets according to claim 1, characterized in that: In S3, the quasi-whitening matrix constructed by the sampling covariance matrix S is used to perform quasi-whitening processing on the target signal subspace matrix A, the interference signal subspace matrix J and the to-be-detected data matrix X.

4. The adaptive subspace detection method for extended targets according to claim 3, characterized in that: The specific operation of the quasi-whitening process is as follows: and 5. The adaptive subspace detection method for extended targets according to claim 1, characterized in that: In S4, the quasi-whitening target signal matrix is ​​used The constructed orthogonal projection matrix and complementary projection matrix are expressed as and in(·) -1 represents the inverse of the matrix, I N is the identity matrix of dimension N×N; Using the quasi-whitening interference signal matrix The constructed orthogonal projection matrix and complementary projection matrix are expressed as and 6. The adaptive subspace detection method for extended targets according to claim 1, characterized in that: In S5, the detection statistic is constructed by using the quasi-whitened data matrix to be detected, the orthogonal projection matrix of the target signal and the orthogonal projection matrix of the interference signal, which is expressed as: Here, tr(·) represents the trace of the matrix.

7. The adaptive subspace detection method for extended targets according to claim 1, characterized in that: In S6, the detection threshold determined by the preset value of false alarm probability is: η=t(n) Where η represents the detection threshold, n = [Qμ], Q is the number of Monte Carlo simulations, μ represents the false alarm probability value set by the system, [·] represents the rounding operation, and t(n) is the sequence The nth largest value when arranged from largest to smallest, X(k) represents the k-th realization k=1, 2,…,Q of the data to be detected containing only interference and thermal noise components.

8. The adaptive subspace detection method for extended targets according to claim 1, characterized in that: In S7, the detection statistic is compared with the detection threshold to determine whether the target exists, and the determination is made in the following two cases: If the detection statistic t is greater than or equal to the detection threshold η, the target is determined to exist; If the detection statistic t is less than the detection threshold η, it is determined that the target does not exist.

9. An adaptive subspace detection system for extended targets, characterized in that: Includes the following modules: Target and interference data matrix construction module: used to construct target signal subspace matrix and interference signal subspace matrix; Sampling covariance matrix construction module: used to construct the sampling covariance matrix through training samples; Data quasi-whitening module: used to perform quasi-whitening processing on the target signal subspace matrix, the interference signal subspace matrix and the data matrix to be detected, and obtain the quasi-whitened target signal subspace matrix, the quasi-whitened interference signal subspace matrix and the quasi-whitened data matrix to be detected respectively; Detection statistic construction module: used to construct detection statistics through quasi-whitened target signal matrix, quasi-whitened interference signal orthogonal projection matrix and quasi-whitened to-be-detected data matrix; Detection threshold calculation module: used to determine the detection threshold through detection statistics and false alarm probability preset value; Target decision module: used to compare the detection statistic with the detection threshold and determine whether the target exists.

Images