A meshless DOA estimation method based on antenna mutual coupling

By reconstructing the guide vector and constructing semi-positive planning problems using the properties of the mutual coupling matrix under the condition of antenna mutual coupling, the problems of large errors and grid mismatch in the condition of antenna mutual coupling are solved, and the conditions of antenna mutual coupling are realized Accurate DOA information acquisition and high-precision signal resolution.

CN117031390BActive Publication Date: 2024-11-12HARBIN ENG UNIV
View PDF 14 Cites 0 Cited by

Patent Information

Application Number
CN202311010755.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-08-11
Publication Date
2024-11-12
Estimated Expiration
2043-08-11

AI Technical Summary

Technical Problem

In the case of antenna mutual coupling, the traditional DOA estimation algorithm has large errors and grid mismatch problems, resulting in limited angle measurement accuracy. Especially the DOA estimation algorithm based on finite information rates has poor effect in this case.

Method used

A gridless DOA estimation method based on antenna mutual coupling is proposed. By using the properties of the mutual coupling matrix, a new equivalent guiding vector is obtained, a covariance matrix is ​​constructed and the properties of the Toplitz matrix are satisfied, and a semi-positive fixed planning problem is constructed to obtain a uniform signal. The sampled value is used to solve the feasible solution using a double iterative alternating minimization algorithm, and then calculate the signal incident angle.

Benefits of technology

Eliminate the influence of antenna mutual coupling effect on the received signal, realize accurate DOA information acquisition in the case of antenna mutual coupling, improve the calculation accuracy of multi-objective DOA and the resolution of adjacent target signals, and significantly improve the angle measurement performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117031390B_ABST
    Figure CN117031390B_ABST
Patent Text Reader

Abstract

A gridless DOA estimation method based on antenna mutual coupling, the present invention relates to a gridless DOA estimation method based on antenna mutual coupling. The purpose of the present invention is to solve the problem of large DOA estimation error of existing antenna arrays under the influence of mutual coupling effect. The process is: step one: based on the far-field incident narrowband signal in the space received by the array antenna, a new equivalent steering vector is obtained by using the properties of the mutual coupling matrix; step two: according to the new equivalent steering vector, the covariance matrix is ​​obtained, and it is proved that the covariance matrix satisfies the properties of the Toplitz matrix; step three: according to the covariance matrix, a semi-positive definite programming problem is constructed to obtain uniform sampling values ​​of the signal; step four: according to the uniform sampling values, a group of annihilation filter equations is constructed, and a feasible solution is obtained by using a double-iteration alternating minimization algorithm; step five: according to the feasible solution, the signal incident angle is calculated. The present invention belongs to the field of array direction finding.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a gridless DOA estimation method based on antenna mutual coupling, and belongs to the field of array direction finding. Background Art

[0002] In modern electronic warfare, electronic reconnaissance plays a vital role. In the radar passive direction finding system, the direction finding accuracy can be improved by increasing the number of reconnaissance antennas. However, increasing the number of antennas within a limited range will undoubtedly lead to denser antenna placement, which will make the antenna mutual coupling effect more obvious and seriously affect the normal operation of the antenna. How to estimate the direction of arrival (DOA) in the case of antenna mutual coupling has become a hot issue.

[0003] In addition, traditional DOA algorithms, such as l1-svd algorithm and MUSIC algorithm, all divide the grid in advance and select the optimal solution in the divided grid. Such algorithms have the problem of grid mismatch, which limits the angle measurement accuracy. The gridless DOA estimation algorithm can overcome this shortcoming and effectively improve the angle measurement accuracy. Among them, the DOA estimation algorithm based on finite rate of innovation (FRI) is widely used in conventional DOA estimation problems and polarization array estimation problems. However, the use of such algorithms is less effective in the case of antenna mutual coupling. Therefore, it is very important to propose a FRI-DOA estimation algorithm that can solve the mutual coupling. Summary of the invention

[0004] The purpose of the present invention is to solve the problem of large DOA estimation error of existing antenna arrays under the influence of mutual coupling effect, and propose a gridless DOA estimation method based on antenna mutual coupling.

[0005] A meshless DOA estimation method based on antenna mutual coupling. The specific process is as follows:

[0006] Step 1: Based on the far-field incident narrowband signal in space received by the array antenna, a new equivalent steering vector is obtained by using the properties of the mutual coupling matrix;

[0007] Step 2: Obtain the covariance matrix based on the new equivalent steering vector, and prove that the covariance matrix satisfies the Toplitz matrix property;

[0008] Step 3: Construct a semi-positive definite programming problem based on the covariance matrix to obtain uniform sampling values ​​of the signal;

[0009] Step 4: Construct the zeroing filter equation group according to the uniform sampling value, and use the double iterative alternating minimization algorithm to find a feasible solution;

[0010] Step 5: Calculate the signal incident angle based on the feasible solution.

[0011] The beneficial effects of the present invention are:

[0012] The purpose of the present invention is to eliminate the influence of antenna mutual coupling effect on the antenna array receiving signal, so that the uniform array can obtain accurate DOA information under the condition of antenna mutual coupling.

[0013] The algorithm involved in the present invention utilizes the symmetric characteristics of the antenna mutual coupling matrix to compress several adjacent channels into one channel, thereby realizing the reconstruction of the steering vector.

[0014] The direction finding algorithm of the present invention is a FRI-DOA algorithm, which does not require grid division and has more accurate angle measurement performance.

[0015] The present invention realizes the decoupled FRI-DOA estimation algorithm based on channel compression. The channel compression equivalent model is as follows: Figure 1 The present invention can significantly improve the multi-target DOA calculation accuracy and the signal resolution capability of adjacent targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The equivalent change model diagram of the array element when the steering vector changes;

[0017] Figure 2 It is the DOA space spectrum of different algorithms;

[0018] Figure 3 This is the graph showing how the root mean square error of angle measurement changes with the signal-to-noise ratio under different algorithms;

[0019] Figure 4 This is a graph showing how the root mean square error of angle measurement changes with the number of snapshots under different algorithms;

[0020] Figure 5 This is a graph showing how the angular resolution probability changes with the angular interval under different algorithms;

[0021] Figure 6 This is a graph showing how the angular resolution probability changes with the mutual coupling factor under different algorithms;

[0022] Figure 7 This is a graph showing how the root mean square error of angle measurement changes with the mutual coupling factor under different algorithms. DETAILED DESCRIPTION

[0023] Specific implementation method 1: This implementation method is a gridless DOA estimation method based on antenna mutual coupling. The specific process is as follows:

[0024] Step 1: Steering vector conversion: Based on the far-field incident narrowband signal in space received by the array antenna, a new equivalent steering vector is obtained by using the properties of the mutual coupling matrix;

[0025] Step 2: Prove the properties of the covariance matrix. Obtain the covariance matrix based on the new equivalent steering vector and prove that the covariance matrix satisfies the properties of the Toplitz matrix.

[0026] Step 3: Restore the uniform sampling value, construct a semi-positive definite programming problem based on the covariance matrix, and obtain the uniform sampling value of the signal;

[0027] Step 4: Solve the signal parameters, construct the zeroing filter equations according to the uniform sampling values, and use the double iterative alternating minimization algorithm to find a feasible solution;

[0028] Step 5: Restore the angle and calculate the signal incident angle based on the feasible solution.

[0029] Specific implementation method 2: The difference between this implementation method and specific implementation method 1 is that the steering vector conversion in step 1 is based on the far-field incident narrowband signal in space received by the array antenna, and the properties of the mutual coupling matrix are used to obtain a new equivalent steering vector The specific process is:

[0030] The model of the far-field incident narrowband single snapshot signal in space received by a one-dimensional uniform array antenna is:

[0031] x(t)=as(t)+n(t)(1)

[0032] Where x(t) is the data vector received by the antenna in the single snapshot case, a is the steering vector of the antenna array in the single snapshot case, s(t) is the single snapshot spatial signal vector, and n(t) is the noise data vector of the array in the single snapshot case;

[0033] When the mutual coupling between antennas is taken into account, equation (1) should be rewritten as:

[0034] x(t)=Cas(t)+n(t) (2)

[0035] Where C is the array mutual coupling factor matrix, which can be expressed as:

[0036]

[0037] where c i is the mutual coupling factor, i = 0, 1…, P-1, P is the number of antennas affected unilaterally by the mutual coupling effect (the number of antennas affected to the left or right), c i It has the following properties:

[0038] c0=1>|c1|>|c2|>…>|c P-1 | (4)

[0039] |c l / c k |=k / l (5)

[0040]

[0041] in is the array antenna amplitude coupling coefficient; c l is the mutual coupling factor, c k is the mutual coupling factor, k=1…,P-1, l=1…,P-1; j is the imaginary unit, j 2 =-1; k0 is the phase mutual coupling coefficient of the array antenna;

[0042] Expanding the single snapshot case to the multi-snap case, the model of the far-field incident narrowband multi-snap signal in space received by the one-dimensional uniform array antenna considering the antenna mutual coupling can be expressed as:

[0043] X=CA(θ)S+N (7)

[0044] Where X is the data vector received by the antenna in the case of multiple snapshots, S is the spatial signal vector in the case of multiple snapshots, N is the noise data matrix of the array in the case of multiple snapshots, C is the array mutual coupling factor matrix, and A(θ) is the array receiving steering vector matrix in the case of multiple snapshots;

[0045] In the case of multiple snapshots, the array receiving steering vector matrix A(θ) is expanded as follows:

[0046]

[0047] Where d is the array element spacing, λ is the signal wavelength, and in a uniform array, d = λ / 2 is usually set, θ i′ is the signal incident angle, where i′=1,2,…n, n is the number of signals, and M is the number of array antennas;

[0048] Considering the influence of the array mutual coupling factor matrix C, an equivalent steering vector is generated. It is expressed as:

[0049]

[0050] Where P is the number of antennas affected unilaterally by the mutual coupling effect; is a variable;

[0051] According to the equivalent steering vector From the expansion of Therefore, the equivalent steering vectors from the P-1th to the MPth antenna elements are extracted. Introduce the array element selection matrix F;

[0052] Where F is as shown in formula (10):

[0053] F=[0 (M-2P)×(P-1)I M-2P 0 (M-2P)×(P-1) ] (10)

[0054] Among them, I M-2P is the M-2P dimensional identity matrix;

[0055] The far-field incident narrowband multi-quick-shot signal in the space received by the array antenna is multiplied by the array element selection matrix to obtain the receiving data of some array elements (caused by the array element selection matrix F), that is, the signal receiving form after the array element selection is as follows:

[0056]

[0057] in, is the antenna receiving data matrix after array element selection, A(θ) is the array receiving steering vector matrix in the case of multiple snapshots;

[0058] make Then we have:

[0059]

[0060] in, is an intermediate variable;

[0061] The signal receiving form after array element selection is changed to the following:

[0062]

[0063]

[0064]

[0065]

[0066] in, is the equivalent steering vector of the P-1th to MPth antenna array elements, that is, the new equivalent steering vector after array element selection, Γ(θ), is an intermediate variable, where Γ(θ) is a diagonal matrix with all elements except the diagonal ones being 0.

[0067] The other steps and parameters are the same as those in the first embodiment.

[0068] Specific implementation method three: This implementation method is different from specific implementation method one or two in that: in step 2, the covariance matrix property is proved, the covariance matrix is ​​obtained according to the new equivalent steering vector, and it is proved that the covariance matrix satisfies the Toplitz matrix property; the specific process is:

[0069] According to the new equivalent guidance vector Get the covariance matrix; expressed as:

[0070]

[0071] in, is the signal noise power, I M-2P is the M-2P dimensional identity matrix, R ss () is the signal autocorrelation matrix, H is the conjugate transpose;

[0072] The elements in the αth row and βth column of the covariance matrix can be expressed as:

[0073]

[0074]

[0075]

[0076] where χ α,β is the array characteristic function, H(θ i′ ) is the equivalent array amplitude, * is the conjugate operator symbol, is the signal-to-noise power;

[0077] By observing equation (18), without considering the noise, the element at position (α, β) and the element at position (β, α) on the covariance matrix are conjugate to each other; it can be seen that under the ideal noise-free, uniform linear array conditions, the covariance matrix is It is a conjugate symmetric Toeplitz matrix, and the entire matrix can be determined by a row or a column of elements.

[0078] The other steps and parameters are the same as those in the first or second embodiment.

[0079] Specific implementation method 4: This implementation method is different from one of the specific implementation methods 1 to 3 in that: in step 3, the uniform sampling value is restored, and a semi-positive definite programming problem is constructed according to the covariance matrix to obtain the uniform sampling value of the signal; the specific process is:

[0080] Using the covariance matrix fitting criterion formula 17, the covariance matrix that satisfies the Toeplitz matrix form is restored

[0081] According to the Hermitian property of the covariance matrix, the objective function of the covariance matrix fitting criterion can be expressed as:

[0082]

[0083]

[0084] Among them, ||·|| Frepresents the F norm of the matrix, tr(·) represents the trace of the matrix, Toep(b) represents the Toeplitz matrix consisting of the b vector as the first row, represents the covariance matrix, represents the covariance matrix that satisfies the Toeplitz matrix form, ≥ means that the diagonal elements on the matrix are greater than or equal to;

[0085] The problem of solving the minimum value can be regarded as a semi-positive programming problem, as follows:

[0086]

[0087]

[0088] in, represents the transformation between optimization problems, represents a vector set, M represents the number of array antennas;

[0089] Introduce the matrix Z so that Then the final equivalent of formula (22) is:

[0090]

[0091] Among them, Z represents the introduced matrix, represents a set of matrices, represents the optimized and restored uniform sampling value with noise, Indicated by The vector is the Toeplitz matrix formed by the first row;

[0092] Use the CVX toolbox to solve the optimization problem formula (23) and obtain the optimization result of the optimization variable b is the uniformly sampled value of the complex exponential signal.

[0093] The other steps and parameters are the same as those in Specific Embodiments 1 to 3.

[0094] Specific implementation method 5: This implementation method is different from any one of the specific implementation methods 1 to 4 in that: in step 4, the signal parameters are solved, a zeroing filter equation group is constructed according to the uniform sampling value, and a double-iteration alternating minimization algorithm is used to find a feasible solution; the specific process is:

[0095] According to the FRI signal reconstruction theory, the uniform sampling values ​​are used to construct the annihilation filter equations to restore the signal parameters. The covariance matrix satisfying the Toeplitz form obtained in step 3 is equivalent to the uniformly sampled data.

[0096] According to the uniform sampling value Constructing the annihilation filter equations

[0097] The two constraints in equation (24) are solved by using a double-iteration alternating minimization algorithm. and ) is a feasible solution, and the feasible solution u is obtained;

[0098]

[0099] in is a vector set, is a vector set, ε is the error tolerance;

[0100] The intermediate variable R(u) is:

[0101]

[0102] where u i″ represents the coefficient of the zeroing filter, i″=0,1,2,…n, n is the number of signals;

[0103] u0 to u in the intermediate variable R(u) n That is the feasible solution u, u=(u0 u1 … u n ).

[0104] The other steps and parameters are the same as those in Specific Implementation 1 to 4-1.

[0105] Specific implementation method 6: This implementation method is different from one of the specific implementation methods 1 to 5 in that: in step 5, the angle is restored and the signal incident angle is calculated according to the feasible solution; the specific process is:

[0106] By constructing a polynomial and solving it, the DOA information of the incident signal can be restored. The calculation formula is:

[0107]

[0108] Among them, roots(·) represents the root function, arg(·) represents the angle calculation based on complex numbers, is the signal incident angle, Represents the unknown number to be solved.

[0109] The other steps and parameters are the same as those in Specific Implementation Methods 1 to 5-1.

[0110] The following examples are used to verify the beneficial effects of the present invention:

[0111] Embodiment 1:

[0112] Figure 1 The equivalent array element diagram representing the change of the steering vector, the mutual coupling model can be equivalent to a new array element obtained by weighted accumulation of p-1 array elements on the left and right with a single array element as the center, and the weighted coefficient is the mutual coupling coefficient.

[0113] Figure 2 The figure shows the DOA spatial spectra of different algorithms under signal mutual coupling, where the straight lines represent the true angles. It can be observed that the traditional MUSIC algorithm is completely ineffective under this method, with only one spectrum peak being correct; the traditional FRI algorithm has an obvious deviation from the standard intersection angle; the l1-svd algorithm suitable for decoupling has many pseudo peaks; and the proposed FRI algorithm has accurate angle estimation, which proves the feasibility of the algorithm.

[0114] Figure 3 The angle measurement accuracy curves of different algorithms as the signal-to-noise ratio changes. The number of snapshots is 100, the mutual coupling factor is 0.6, and 500 Monte Carlo experiments are performed. It can be seen that the original FRI algorithm has poor performance. The FRI algorithm proposed in this patent has the best performance and has higher angle measurement accuracy.

[0115] Figure 4 The angle measurement accuracy curves of different algorithms as snapshots change. The signal-to-noise ratio is 10dB, the mutual coupling factor is 0.6, and 500 Monte Carlo experiments are performed. It can be seen that the original FRI algorithm has poor performance. The FRI algorithm proposed in this patent has the best performance and has higher angle measurement accuracy.

[0116] Figure 5 The curves of success probability of angle measurement of different algorithms with the change of angle interval. The signal-to-noise ratio is 10dB, the number of snapshots is 100, the mutual coupling factor is 0.6, and 500 Monte Carlo experiments are performed. It can be seen that the performance of the original FRI algorithm is poor, and the FRI algorithm proposed in this patent has the best performance and a higher success probability of angle resolution.

[0117] Figure 6 The curves of the success probability of angle measurement of different algorithms with the change of mutual coupling factor are shown. The signal-to-noise ratio is 10dB, the number of snapshots is 100, the mutual coupling factor is 0.6, the angle interval is 2°, and 500 Monte Carlo experiments are performed. It can be seen that the performance of the original FRI algorithm is poor, and the FRI algorithm proposed in this patent has the best performance, and there is still a high probability of successful angle resolution even when the angle interval is small.

[0118] Figure 7 The angle measurement error curves of different algorithms as the mutual coupling factor changes. The signal-to-noise ratio is 10dB, the number of snapshots is 100, the mutual coupling factor is 0.6, and 500 Monte Carlo experiments are performed. It can be seen that the performance of the original FRI algorithm is poor, and the FRI algorithm proposed in this patent has the best performance and is less affected by the mutual coupling factor.

[0119] In summary, the present invention proposes a decoupling FRI-DOA algorithm based on channel compression for one-dimensional uniform arrays. This algorithm is based on gridless DOA estimation and combines the mutually coupled antenna array. First, the original mutually coupled steering vector matrix is ​​transformed, and the equivalent steering vector is generated by utilizing the particularity of the mutually coupled matrix. Based on the equivalent steering vector, a new covariance matrix is ​​constructed, and the equivalent uniformly sampled signal is restored through a convex optimization algorithm. Finally, a group of annihilation filter equations is constructed, and the final angle is obtained using a double iteration algorithm. The unoptimized FRI-DOA algorithm and other traditional decoupling algorithms are compared, and the advantages of the algorithm of the present invention in estimation accuracy are verified through simulation experiments.

[0120] The present invention may also have many other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art may make various corresponding changes and modifications based on the present invention, but these corresponding changes and modifications should all fall within the scope of protection of the claims attached to the present invention.

Claims

1. A gridless DOA estimation method based on antenna mutual coupling, Features: The specific process of the method is: Step 1: Based on the far-field incident narrowband signal in space received by the array antenna, a new equivalent steering vector is obtained by using the properties of the mutual coupling matrix; Step 2: Obtain the covariance matrix based on the new equivalent steering vector, and prove that the covariance matrix satisfies the Toplitz matrix property; Step 3: Construct a semi-positive definite programming problem based on the covariance matrix to obtain uniform sampling values ​​of the signal; Step 4: Construct the zeroing filter equation group according to the uniform sampling value, and use the double iterative alternating minimization algorithm to find a feasible solution; Step 5: Calculate the signal incident angle based on the feasible solution; In the step 1, based on the far-field incident narrowband signal in space received by the array antenna, a new equivalent steering vector is obtained by using the properties of the mutual coupling matrix; the specific process is: The model of the far-field incident narrowband single snapshot signal in space received by a one-dimensional uniform array antenna is: x(t)=as(t)+n(t) (1) Where x(t) is the data vector received by the antenna in the single snapshot case, a is the steering vector of the antenna array in the single snapshot case, s(t) is the single snapshot spatial signal vector, and n(t) is the noise data vector of the array in the single snapshot case; When the mutual coupling between antennas is taken into account, equation (1) should be rewritten as: x(t)=Cas(t)+n(t) (2) Where C is the array mutual coupling factor matrix, expressed as: where c i is the mutual coupling factor, i=0,1…,P-1, P is the number of antennas affected by the mutual coupling effect, c i It has the following properties: c 0 =1>|c 1 |>|c 2 |>…>|c P-1 | (4) |c l / c k |=k / l (5) in is the array antenna amplitude coupling coefficient; c l is the mutual coupling factor, c k is the mutual coupling factor, k=1…,P-1, l=1…,P-1; j is the imaginary unit, j 2 =-1; k 0 is the phase mutual coupling coefficient of the array antenna; Expanding the single snapshot case to the multi-snap case, the model of the far-field incident narrowband multi-snap signal in space received by the one-dimensional uniform array antenna considering the antenna mutual coupling is expressed as: X=CA(θ)S+N (7) Where X is the data vector received by the antenna in the case of multiple snapshots, S is the spatial signal vector in the case of multiple snapshots, N is the noise data matrix of the array in the case of multiple snapshots, C is the array mutual coupling factor matrix, and A(θ) is the array receiving steering vector matrix in the case of multiple snapshots; In the case of multiple snapshots, the array receiving steering vector matrix A(θ) is expanded as follows: Where d is the array element spacing, λ is the signal wavelength, and θ i′ is the signal incident angle, where i′=1,2,…n, n is the number of signals, and M is the number of array antennas; Considering the influence of the array mutual coupling factor matrix C, an equivalent steering vector is generated. It is expressed as: Where P is the number of antennas affected unilaterally by the mutual coupling effect; is a variable; According to the equivalent steering vector From the expansion of Therefore, the equivalent steering vectors from the P-1th to the MPth antenna elements are extracted. Introduce the array element selection matrix F; Where F is as shown in formula (10): F=[0 (M-2P)×(P-1) I M-2P 0 (M-2P)×(P-1) ] (10) Among them, I M-2P is the M-2P dimensional identity matrix; The far-field incident narrowband multi-quick-shot signal in the space received by the array antenna is multiplied by the array element selection matrix to obtain the receiving data of some array elements, that is, the signal receiving form after the array element selection is as follows: in, is the antenna receiving data matrix after array element selection, A(θ) is the array receiving steering vector matrix in the case of multiple snapshots; make Then we have: in, is an intermediate variable; The signal receiving form after array element selection is changed to the following: in, is the equivalent steering vector of the P-1th to MPth antenna array elements, that is, the new equivalent steering vector after array element selection, Γ(θ), is the intermediate variable, where Γ(θ) is a diagonal matrix, except for the elements on the diagonal, all other positions are 0; In the step 2, the covariance matrix is ​​obtained according to the new equivalent steering vector, and it is proved that the covariance matrix satisfies the Toplitz matrix property; the specific process is: The covariance matrix is ​​obtained according to the new equivalent steering vector; it is expressed as: in, is the signal noise power, I M-2P is the M-2P dimensional identity matrix, R ss () is the signal autocorrelation matrix, H is the conjugate transpose; The elements in the αth row and βth column of the covariance matrix can be expressed as: where χ α,β is the array characteristic function, H(θ i′ ) is the equivalent array amplitude, * is the conjugate operator symbol, is the signal-to-noise power; By observing equation (18), without considering the noise, the element at position (α, β) and the element at position (β, α) on the covariance matrix are conjugate to each other; it can be seen that under the ideal noise-free, uniform linear array conditions, the covariance matrix is is a conjugate symmetric Toeplitz matrix; In step 3, a semi-positive definite programming problem is constructed according to the covariance matrix to obtain a uniform sampling value of the signal; the specific process is: According to the Hermitian property of the covariance matrix, the objective function of the covariance matrix fitting criterion is expressed as: Among them, ||·|| F represents the F norm of the matrix, tr(·) represents the trace of the matrix, Toep(b) represents the Toeplitz matrix consisting of the b vector as the first row, represents the covariance matrix, represents the covariance matrix that satisfies the Toeplitz matrix form, Indicates that the diagonal elements of the matrix are greater than or equal to; The problem of solving the minimum value can be regarded as a semi-positive programming problem, as follows: in, represents the transformation between optimization problems, represents a vector set, M represents the number of array antennas; Introduce the matrix Z so that Then the final equivalent of formula (22) is: Among them, Z represents the introduced matrix, represents a set of matrices, represents the optimized and restored uniform sampling value with noise, Indicated by The vector is the Toeplitz matrix formed by the first row; Use the CVX toolbox to solve the optimization problem formula (23) and obtain the optimization result of the optimization variable b is the uniformly sampled value of the complex exponential signal; In the step 4, a zeroing filter equation group is constructed according to the uniform sampling value, and a feasible solution is obtained by using a double-iteration alternating minimization algorithm; the specific process is: According to the uniform sampling value Constructing the annihilation filter equations The dual-iteration alternating minimization algorithm is used to solve the feasible solutions of the two constraints in equation (24) and obtain the feasible solution u; in is a vector set, is a vector set, ε is the error tolerance; The intermediate variable R(u) is: where u i″ represents the coefficient of the zeroing filter, i″=0,1,2,…n, n is the number of signals; u in the intermediate variable R(u) 0 to u n That is the feasible solution u, u=(u 0 u 1 … u n ); In step 5, the signal incident angle is calculated according to the feasible solution; the specific process is: By constructing a polynomial and solving it, the DOA information of the incident signal can be restored. The calculation formula is: Among them, roots(·) represents the root function, arg(·) represents the angle calculation based on complex numbers, is the signal incident angle, Represents the unknown number to be solved.

Citation Information

Patent Citations

  • Wave arrival direction estimation method under array error

    CN104965188A

  • Single-base MIMO radar angle estimation method based on covariance vector sparse representation under cross coupling condition

    CN105068041A

  • Sub-sampling antenna elements

    CN107787534A

  • Frequency diverse array based mutual coupling array DOA estimation method

    CN110109051A

  • DOA estimation method for moving target echoes under multiple external radiation sources

    CN110967665A