A robust adaptive beamforming method and system based on sequential convex approximation

Through a robust adaptive beamforming method based on sequence convex approximation method, a signal-to-interference-noise ratio model is established and the error of the actual signal steering vector is introduced, which is converted into a second-order cone planning problem, and the pollution of the signal component of the traditional adaptive beamformer is solved. The problem of performance degradation in mismatch between the steering vector improves the signal-to-interference noise ratio and computing efficiency.

CN115642944BInactive Publication Date: 2024-11-22GUANGDONG UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202211282592.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-10-19
Publication Date
2024-11-22
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional adaptive beamformers have dramatically declined during signal component pollution and steering vector mismatch, and are unable to effectively suppress interference and noise, resulting in deterioration of array performance.

Method used

Using a robust adaptive beamforming method based on sequence convex approximation method, the optimization goal of maximizing the minimum signal-to-interference noise model in extreme cases is constructed by establishing a signal-to-interference noise ratio model and using the sampling matrix instead of the covariance matrix of interference plus noise. function, and introduce the error of the actual signal turning vector, transform it into a second-order cone planning problem, and solve it using the sequence convex approximation method to obtain the optimal weighted vector.

Benefits of technology

It effectively improves the output signal-to-interference-to-noise ratio capability of the antenna array, improves the computing efficiency, and overcomes the performance degradation caused by steering vector mismatch.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115642944B_ABST
    Figure CN115642944B_ABST
Patent Text Reader

Abstract

The present invention discloses a robust adaptive beamforming method and system based on a sequential convex approximation method, which relates to the technical field of array signal processing, including establishing a signal-to-interference-noise ratio model for robust adaptive beamforming; replacing the covariance matrix of interference plus noise in the signal-to-interference-noise ratio model with a sampling matrix, establishing an optimization objective function for maximizing the output of the minimum signal-to-interference-noise model in extreme cases; combining a number of small uncertainty sets to construct a large uncertainty set of actual signal steering vectors, substituting it into the optimization objective function, and establishing a new optimization objective function; converting the new optimization objective function into a SOCP problem with the same optimal solution; solving the SOCP problem using a sequential convex approximation method, obtaining the weighted vector of the optimal antenna array, and outputting the optimal robust adaptive beam. The present invention can avoid the problem of performance degradation of the beamformer due to significant mismatch of the steering vector, and improve the ability and computational efficiency of the antenna array beamformer to output the signal-to-interference-noise ratio.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of array signal processing, and more specifically, to a robust adaptive beamforming method and system based on a sequential convex approximation method. Background Art

[0002] Adaptive beamforming has a wide range of applications in radar, sonar, medical imaging, microphone array processing, wireless communications and other fields. Robust adaptive beamforming technology is a method that can significantly improve the performance of array output signal-to-noise ratio, main lobe width and sidelobe level, and robust adaptive beamforming technology is very effective for mismatch. Traditional adaptive beamformers assume that the desired signal components are free in the training data and the desired steering vector is precisely known. However, when the signal components contaminate the training samples and the steering vector mismatch occurs, the performance of the traditional beamformer will drop sharply. There are many methods to solve this problem, such as diagonal loading method, feature space-based method, and worst-case optimization method. However, when there is a significant mismatch between the actual steering vector and the assumed steering vector or other types of errors, a larger uncertainty set is required to model. In this case, the ability to suppress interference plus noise is reduced, which will lead to the deterioration of the performance of the beamformer array.

[0003] The prior art discloses a robust adaptive beamforming method for non-convex quadratic matrix inequality, including constructing a maximized SINR model; replacing the interference plus noise covariance matrix of the SINR model with a sampling covariance matrix, generating a new target optimization function and adding an error matrix to obtain a new target optimization problem; solving the target optimization problem to obtain an optimal solution or a suboptimal solution, and outputting a final beamforming vector; this method improves the output signal-to-noise ratio and overall performance of the array to a certain extent, but cannot overcome the problem that the significant mismatch of the steering vector causes performance degradation. Summary of the invention

[0004] In order to overcome the defect that the performance of the beamformer of the above-mentioned existing antenna array is degraded when the steering vector is significantly mismatched, the present invention provides a robust adaptive beamforming method and system based on sequential convex approximation, which avoids the problem of performance degradation of the beamformer due to significant mismatch of the steering vector and improves the ability and computational efficiency of the antenna array beamformer to output the signal-to-interference-noise ratio.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0006] The present invention provides a robust adaptive beamforming method based on a sequential convex approximation method, comprising:

[0007] S1: Establishing the signal-to-interference-noise ratio model for robust adaptive beamforming;

[0008] S2: Replace the covariance matrix of interference plus noise in the signal-to-interference-noise ratio model with a sampling matrix, and establish an optimization objective function that maximizes the output of the minimum signal-to-interference-noise ratio model in extreme cases;

[0009] S3: Combine several small uncertainty sets to construct a large uncertainty set of the actual signal steering vector, substitute it into the optimization objective function, and establish a new optimization objective function;

[0010] S4: Convert the new optimization objective function into a second order cone programming (SOCP) problem with the same optimal solution;

[0011] S5: The SOCP problem is solved using the sequential convex approximation method to obtain the weight vector of the optimal antenna array and output the optimal robust adaptive beam.

[0012] Preferably, in step S1, the signal to interference noise ratio model of robust adaptive beamforming is specifically:

[0013]

[0014] Where SINR represents the signal-to-interference-to-noise ratio, σ s represents the power of the desired signal, a represents the actual steering vector of the received signal, and a H represents the transposed matrix of the actual steering vector of the received signal, w represents the weighted vector of the antenna array, w H Represents the transposed matrix of the weight vector of the antenna array, R i+n represents the covariance matrix of interference plus noise.

[0015] Preferably, in step S2, the sampling matrix is ​​specifically:

[0016]

[0017] In the formula, represents the sampling matrix, T represents the number of sampling times, y(t) represents the response of the antenna array to the received signal, y H (t) represents the conjugate transpose of the antenna array's response to the received signal, where:

[0018] y(t)=s(t)+i(t)+n(t)

[0019] Where s(t) represents the desired signal, i(t) represents the interference signal, and n(t) represents the noise component.

[0020] Preferably, in step S2, the optimization objective function for maximizing the minimum signal-to-interference-noise model output in an extreme case is specifically:

[0021] The optimization problem of maximizing the minimum signal-to-noise model output in extreme cases is:

[0022]

[0023] In the formula, Represents the uncertainty set of the actual steering vector of the received signal, specifically:

[0024]

[0025] In the formula, represents the estimated steering vector of the received signal, and N represents the number of antenna elements in the antenna array, ∥*∥ represents the Euclidean norm, and ∈ represents the vector threshold;

[0026] The above optimization problem is rewritten as the optimization objective function that maximizes the output of the minimum signal interference noise model under the following extreme conditions:

[0027]

[0028] In the formula, minimize means to obtain the minimization operation.

[0029] Preferably, in step S3, combining several small uncertainty sets to construct a large uncertainty set of the actual signal steering vector is specifically as follows:

[0030] Combine the M received signals to form a small uncertainty set of actual steering vectors, where:

[0031]

[0032] In the formula, represents the small uncertainty set of the actual steering vector of the mth received signal, represents the estimated steering vector of the mth received signal, ∈ m represents the mth vector threshold, and M represents the number of small uncertainty sets of the actual steering vectors of the received signal;

[0033] Then the large uncertainty set of the actual steering vector of the received signal is:

[0034]

[0035] In the formula, represents the large uncertain set of actual steering vectors of the received signal, and ∪ represents the union operation.

[0036] Preferably, in step S3, the newly established optimization objective function is specifically:

[0037] Using the large uncertainty set of the actual steering vector of the received signal Instead of receiving the signal, the actual steering vector uncertainty set Get the new optimization objective function:

[0038]

[0039] In the formula, Represents a small uncertainty set of the actual steering vector of the Mth received signal.

[0040] Preferably, in step S4, the new optimization objective function is converted into a SOCP problem with the same optimal solution as follows:

[0041]

[0042] In the formula, represents the estimated steering vector of the mth received signal, ∈ m represents the mth vector threshold, and

[0043] Preferably, in step S5, the SOCP problem is solved by using a sequential convex approximation method to obtain a weighted vector of an optimal antenna array, and a specific method for outputting an optimal robust adaptive beam is:

[0044] S5.1: Rewrite the SOCP problem to obtain a new SOCP problem;

[0045] S5.2: Settings and auxiliary variable ξ;

[0046] S5.3: Define the iteration round l = 1, Randomly set w 0 , so that m=1,…,M; where w 0 represents the weight vector of the initial antenna array, w 0H The transposed matrix of the weight vector of the initial antenna array, v 0 represents the initial optimal value;

[0047] S5.4: w 0 Substitute into the new SOCP problem to solve and obtain the optimal solution w for the current round l , and calculate Among them, w l represents the weight vector of the optimal antenna array at the lth iteration, w lH represents the transposed matrix of the weight vector of the optimal antenna array at the lth iteration, v l represents the optimal value of the lth round;

[0048] S5.5: Let l = l + 1, and replace the optimal solution w of the previous round l Substitute into the new SOCP problem and solve it to obtain w l+1 and vl+1 ;

[0049] S5.6: Calculate the optimal value difference v ′ , v ′ =|v l-2 -v l-1 |; The optimal value difference v ′ Compared with the auxiliary variable ξ, if v ′ >ξ, then execute step S5.5; otherwise, execute step S5.7;

[0050] S5.7: v l-1 The optimal solution w for the corresponding round l-1 Output as a weighted vector of the optimal antenna array;

[0051] S5.8: Output an optimal robust adaptive beam according to the weight vector of the optimal antenna array.

[0052] Preferably, in step S5.1, the new SOCP problem is specifically:

[0053]

[0054] In the formula, represents the operation of obtaining the real part, express The transposed matrix, w l-1 represents the weight vector of the optimal antenna array at the l-1th iteration, and l represents the iteration round.

[0055] The present invention also provides a robust adaptive beamforming system based on sequential convex approximation method, which is used to implement the above-mentioned robust adaptive beamforming method based on sequential convex approximation method, comprising:

[0056] A model building module, used to build a signal-to-interference-noise ratio model for robust adaptive beamforming;

[0057] An optimization objective function establishment module is used to replace the covariance matrix of interference plus noise in the signal-to-interference-noise ratio model with a sampling matrix, and to establish an optimization objective function that maximizes the output of the minimum signal-to-interference-noise ratio model in extreme cases;

[0058] A new optimization objective function establishment module is used to construct a large uncertainty set of actual signal steering vectors, substitute it into the optimization objective function, and establish a new optimization objective function;

[0059] SOCP problem conversion module, used to convert the new optimization objective function into a SOCP problem with the same optimal solution;

[0060] The SOCP problem solving module uses the sequential convex approximation method to solve the SOCP problem, obtain the weight vector of the optimal antenna array, and output the optimal robust adaptive beam.

[0061] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0062] The present invention first establishes a signal-to-interference-noise ratio model for robust adaptive beamforming, and establishes an optimization objective function that maximizes the output of the minimum signal-to-interference-noise model under extreme conditions; considering the error of the actual signal steering vector, introducing a large uncertainty set of the actual signal steering vector, and reconstructing a new optimization objective function, the problem of performance degradation when the steering vector is significantly mismatched is solved; then it is converted into a SOCP problem with the same optimal solution, and solved using a sequential convex approximation method to obtain the weighted vector of the optimal antenna array, and output the optimal robust adaptive beam. The present invention effectively improves the ability of the antenna array beamformer to output the signal-to-interference-noise ratio, while improving the computational efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 This is a flowchart of a robust adaptive beamforming method based on sequential convex approximation as described in Example 1.

[0064] Figure 2 This is a schematic diagram of constructing a large uncertainty set of actual steering vectors of received signals as described in Example 2.

[0065] Figure 3 This is a schematic diagram of the structure of a robust adaptive beamforming system based on sequential convex approximation method described in Example 3. DETAILED DESCRIPTION

[0066] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;

[0067] In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;

[0068] It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0069] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0070] Example 1

[0071] This embodiment provides a robust adaptive beamforming method based on sequential convex approximation method, such as Figure 1 As shown, including:

[0072] S1: Establishing the signal-to-interference-noise ratio model for robust adaptive beamforming;

[0073] S2: Replace the covariance matrix of interference plus noise in the signal-to-interference-noise ratio model with a sampling matrix, and establish an optimization objective function that maximizes the output of the minimum signal-to-interference-noise ratio model in extreme cases;

[0074] S3: Combine several small uncertainty sets to construct a large uncertainty set of the actual signal steering vector, substitute it into the optimization objective function, and establish a new optimization objective function;

[0075] S4: Transform the new optimization objective function into a SOCP problem with the same optimal solution;

[0076] S5: The SOCP problem is solved using the sequential convex approximation method to obtain the weight vector of the optimal antenna array and output the optimal robust adaptive beam.

[0077] In the specific implementation process, this embodiment first establishes a signal-to-interference-noise ratio model for robust adaptive beamforming, and establishes an optimization objective function that maximizes the output of the minimum signal-to-interference-noise model in extreme cases; considers the error of the actual signal steering vector, introduces a large uncertainty set of the actual signal steering vector, and reconstructs a new optimization objective function; then converts it into a SOCP problem with the same optimal solution, and uses the sequential convex approximation method to solve it, obtains the weighted vector of the optimal antenna array, and outputs the optimal robust adaptive beam. This embodiment effectively improves the ability of the antenna array beamformer to output the signal-to-interference-noise ratio, while improving the computational efficiency.

[0078] Example 2

[0079] This embodiment provides a robust adaptive beamforming method based on a sequential convex approximation method, including:

[0080] S1: Establish a signal-to-interference-noise ratio model for robust adaptive beamforming; specifically:

[0081]

[0082] Where SINR represents the signal-to-interference-to-noise ratio, σ s represents the power of the desired signal, a represents the actual steering vector of the received signal, and a H represents the transposed matrix of the actual steering vector of the received signal, w represents the weighted vector of the antenna array, w H Represents the transposed matrix of the weight vector of the antenna array, R i+n represents the covariance matrix of interference plus noise.

[0083] S2: Replace the covariance matrix of interference plus noise in the signal-to-interference-noise ratio model with a sampling matrix, and establish an optimization objective function that maximizes the output of the minimum signal-to-interference-noise ratio model in extreme cases;

[0084] The sampling matrix is ​​specifically:

[0085]

[0086] In the formula, represents the sampling matrix, T represents the number of sampling times, y(t) represents the response of the antenna array to the received signal, y H (t) represents the conjugate transpose of the antenna array's response to the received signal, where:

[0087] y(t)=s(t)+i(t)+n(t)

[0088] Where s(t) represents the desired signal, i(t) represents the interference signal, and n(t) represents the noise component.

[0089] The optimization problem of maximizing the minimum signal-to-noise model output in extreme cases is:

[0090]

[0091] In the formula, Represents the uncertainty set of the actual steering vector of the received signal, specifically:

[0092]

[0093] In the formula, represents the estimated steering vector of the received signal, and N represents the number of antenna elements in the antenna array, ∥*∥ represents the Euclidean norm, and ∈ represents the vector threshold;

[0094] The above optimization problem is rewritten as the optimization objective function that maximizes the output of the minimum signal interference noise model under the following extreme conditions:

[0095]

[0096] In the formula, minimize means to obtain the minimization operation.

[0097] S3: Combine several small uncertainty sets to construct a large uncertainty set of the actual signal steering vector, substitute it into the optimization objective function, and establish a new optimization objective function;

[0098] like Figure 2 As shown in Figure 1, when a large uncertainty set is modeled in a real scenario, a large mismatch in the direction of arrival and other types of errors occur, the robust adaptive beamforming problem becomes more conservative. In this case, instead, it is assumed that there are multiple possible hypothesized steering vectors, and multiple small uncertainty sets are combined to form a dozen uncertainty sets to solve the problem of significant mismatch in the steering vector, that is:

[0099] Combine the M received signals to form a small uncertainty set of actual steering vectors, where:

[0100]

[0101] In the formula, represents the small uncertainty set of the actual steering vector of the mth received signal, represents the estimated steering vector of the mth received signal, ∈ m represents the mth vector threshold, and M represents the number of small uncertainty sets of the actual steering vectors of the received signal;

[0102] Then the large uncertainty set of the actual steering vector of the received signal is:

[0103]

[0104] In the formula, represents the large uncertain set of actual steering vectors of the received signal, and ∪ represents the union operation.

[0105] The newly established optimization objective function is specifically:

[0106] Using the large uncertainty set of the actual steering vector of the received signal Instead of receiving the signal, the actual steering vector uncertainty set Get the new optimization objective function:

[0107]

[0108] In the formula, Represents a small uncertainty set of the actual steering vector of the Mth received signal.

[0109] S4: Convert the new optimization objective function into a SOCP problem with the same optimal solution; specifically:

[0110]

[0111] In the formula, represents the estimated steering vector of the mth received signal, ∈ m represents the mth vector threshold, and

[0112] S5: Solve the SOCP problem using the sequential convex approximation method, obtain the weight vector of the optimal antenna array, and output the optimal robust adaptive beam; the specific method is:

[0113] S5.1: Rewrite the SOCP problem to obtain a new SOCP problem;

[0114] because Satisfy the following formula:

[0115]

[0116] Then the new SOCP problem is:

[0117]

[0118] In the formula, represents the operation of obtaining the real part, express The transposed matrix, w l-1 represents the weight vector of the optimal antenna array at the l-1th iteration, and l represents the iteration round.

[0119] S5.2: Settings and auxiliary variable ξ;

[0120] S5.3: Define the iteration round l = 1, Randomly set w 0 , so that m=1,…,M; where w 0 represents the weight vector of the initial antenna array, w 0H The transposed matrix of the weight vector of the initial antenna array, v 0 represents the initial optimal value;

[0121] S5.4: w 0 Substitute into the new SOCP problem to solve and obtain the optimal solution w for the current round l , and calculate Among them, w l represents the weight vector of the optimal antenna array at the lth iteration, w lH represents the transposed matrix of the weight vector of the optimal antenna array at the lth iteration, v l represents the optimal value of the lth round;

[0122] S5.5: Let l = l + 1, and replace the optimal solution w of the previous round l Substitute into the new SOCP problem and solve it to obtain w l+1 and v l+1 ;

[0123] S5.6: Calculate the optimal value difference v ′ , v ′ =|v l-2 -v l-1 |; The optimal value difference v ′ Compared with the auxiliary variable ξ, if v ′ >ξ, then execute step S5.5; otherwise, execute step S5.7;

[0124] S5.7: v l-1 The optimal solution w for the corresponding round l-1 Output as a weighted vector of the optimal antenna array;

[0125] S5.8: Output an optimal robust adaptive beam according to the weight vector of the optimal antenna array.

[0126] Example 3

[0127] This embodiment provides a robust adaptive beamforming system based on sequential convex approximation method, such as Figure 3 As shown, including:

[0128] A model building module, used to build a signal-to-interference-noise ratio model for robust adaptive beamforming;

[0129] An optimization objective function establishment module is used to replace the covariance matrix of interference plus noise in the signal-to-interference-noise ratio model with a sampling matrix, and to establish an optimization objective function that maximizes the output of the minimum signal-to-interference-noise ratio model in extreme cases;

[0130] A new optimization objective function establishment module is used to construct a large uncertainty set of actual signal steering vectors, substitute it into the optimization objective function, and establish a new optimization objective function;

[0131] SOCP problem conversion module, used to convert the new optimization objective function into a SOCP problem with the same optimal solution;

[0132] The SOCP problem solving module uses the sequential convex approximation method to solve the SOCP problem, obtain the weight vector of the optimal antenna array, and output the optimal robust adaptive beam.

[0133] The same or similar reference numerals correspond to the same or similar components;

[0134] The terms used in the drawings to describe positional relationships are only used for illustrative purposes and should not be construed as limiting this patent;

[0135] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A robust adaptive beamforming method based on sequential convex approximation, characterized in that: include: S1: Establish a signal-to-interference-noise ratio model for robust adaptive beamforming, specifically: Where SINR represents the signal-to-interference-to-noise ratio, σ s represents the power of the desired signal, a represents the actual steering vector of the received signal, and a H represents the transposed matrix of the actual steering vector of the received signal, w represents the weighted vector of the antenna array, w H Represents the transposed matrix of the weight vector of the antenna array, R i+n represents the covariance matrix of interference plus noise; S2: Replace the covariance matrix of interference plus noise in the signal-to-interference-noise ratio model with a sampling matrix, and establish an optimization objective function that maximizes the output of the minimum signal-to-interference-noise ratio model in extreme cases; The sampling matrix is ​​specifically: In the formula, represents the sampling matrix, T represents the number of sampling times, y(t) represents the response of the antenna array to the received signal, y H (t) represents the conjugate transpose of the antenna array's response to the received signal, where: y(t)=s(t)+i(t)+n(t) In the formula, s(t) represents the desired signal, i(t) represents the interference signal, and n(t) represents the noise component; The optimization objective function for maximizing the minimum signal interference noise model output in the extreme case is specifically: The optimization problem of maximizing the minimum signal-to-noise model output in extreme cases is: In the formula, Represents the uncertainty set of the actual steering vector of the received signal, specifically: In the formula, represents the estimated steering vector of the received signal, and N represents the number of antenna elements in the antenna array, ∥*∥ represents the Euclidean norm, and ∈ represents the vector threshold; The above optimization problem is rewritten as the optimization objective function that maximizes the output of the minimum signal interference noise model under the following extreme conditions: In the formula, minimize means to obtain the minimization operation; S3: Combine several small uncertainty sets to construct a large uncertainty set of the actual signal steering vector, substitute it into the optimization objective function, and establish a new optimization objective function; The method of combining several small uncertainty sets to construct a large uncertainty set of actual signal steering vectors is specifically as follows: Combine the M received signals to form a small uncertainty set of actual steering vectors, where: In the formula, represents the small uncertainty set of the actual steering vector of the mth received signal, represents the estimated steering vector of the mth received signal, ∈ m represents the mth vector threshold, and M represents the number of small uncertainty sets of the actual steering vectors of the received signal; Then the large uncertainty set of the actual steering vector of the received signal is: In the formula, represents the large uncertain set of the actual steering vector of the received signal, ∪ represents the union operation; The newly established optimization objective function is specifically: Using the large uncertainty set of the actual steering vector of the received signal Instead of receiving the signal, the actual steering vector uncertainty set Get the new optimization objective function: In the formula, represents the small uncertainty set of the actual steering vector of the Mth received signal; S4: Transform the new optimization objective function into a second-order cone programming problem with the same optimal solution, specifically: In the formula, represents the estimated steering vector of the mth received signal, ∈ m represents the mth vector threshold, and S5: Use the sequential convex approximation method to solve the second-order cone programming problem, obtain the weight vector of the optimal antenna array, and output the optimal robust adaptive beam, including: S5.1: Rewrite the second-order cone programming problem to obtain a new second-order cone programming problem: In the formula, represents the operation of obtaining the real part, express The transposed matrix, w l-1 represents the weight vector of the optimal antenna array at the l-1th iteration, and l represents the iteration round; S5.2: Settings and auxiliary variable ξ; S5.3: Define the iteration round l = 1, Randomly set w 0 , so that m=1,…,M; where w 0 represents the weight vector of the initial antenna array, w 0H Represents the transposed matrix of the weight vector of the initial antenna array, v0 represents the initial optimal value; S5.4: w 0 Substitute it into the new second-order cone programming problem to solve and obtain the optimal solution w for the current round l , and calculate Among them, w l represents the weight vector of the optimal antenna array at the lth iteration, w lH represents the transposed matrix of the weight vector of the optimal antenna array at the lth iteration, v l represents the optimal value of the lth round; S5.5: Let l = l + 1, and replace the optimal solution w of the previous round l Substitute it into the new second-order cone programming problem and solve it to obtain w l+1 and v l+1 ; S5.6: Calculate the optimal value difference v′, v′=|v l-2 -v l-1 |; Compare the optimal value difference v′ with the auxiliary variable ξ, if v′>ξ, execute step S5.5; otherwise, execute step S5.7; S5.7: v l-1 The optimal solution w for the corresponding round l-1 Output as a weighted vector of the optimal antenna array; S5.8: Output an optimal robust adaptive beam according to the weight vector of the optimal antenna array.

2. A robust adaptive beamforming system based on sequential convex approximation, used to implement the robust adaptive beamforming method based on sequential convex approximation according to claim 1, characterized in that: include: The model building module is used to build a signal-to-interference-noise ratio model for robust adaptive beamforming, specifically: Where SINR represents the signal-to-interference-to-noise ratio, σ s represents the power of the desired signal, a represents the actual steering vector of the received signal, and a H represents the transposed matrix of the actual steering vector of the received signal, w represents the weighted vector of the antenna array, w H Represents the transposed matrix of the weight vector of the antenna array, R i+n represents the covariance matrix of interference plus noise; An optimization objective function establishment module is used to replace the covariance matrix of interference plus noise in the signal-to-interference-noise ratio model with a sampling matrix, and to establish an optimization objective function that maximizes the output of the minimum signal-to-interference-noise ratio model in extreme cases; The sampling matrix is ​​specifically: In the formula, represents the sampling matrix, T represents the number of sampling times, y(t) represents the response of the antenna array to the received signal, y H (t) represents the conjugate transpose of the antenna array's response to the received signal, where: y(t)=s(t)+i(t)+n(t) In the formula, s(t) represents the desired signal, i(t) represents the interference signal, and n(t) represents the noise component; The optimization objective function for maximizing the minimum signal interference noise model output in the extreme case is specifically: The optimization problem of maximizing the minimum signal-to-noise model output in extreme cases is: In the formula, Represents the uncertainty set of the actual steering vector of the received signal, specifically: In the formula, represents the estimated steering vector of the received signal, and N represents the number of antenna elements in the antenna array, ∥*∥ represents the Euclidean norm, and ∈ represents the vector threshold; The above optimization problem is rewritten as the optimization objective function that maximizes the output of the minimum signal interference noise model under the following extreme conditions: In the formula, minimize means to obtain the minimization operation; A new optimization objective function establishment module is used to combine several small uncertainty sets to construct a large uncertainty set of the actual signal steering vector, substitute it into the optimization objective function, and establish a new optimization objective function; The method of combining several small uncertainty sets to construct a large uncertainty set of actual signal steering vectors is specifically as follows: Combine the M received signals to form a small uncertainty set of actual steering vectors, where: In the formula, represents the small uncertainty set of the actual steering vector of the mth received signal, represents the estimated steering vector of the mth received signal, ∈ m represents the mth vector threshold, and M represents the number of small uncertainty sets of the actual steering vectors of the received signal; Then the large uncertainty set of the actual steering vector of the received signal is: In the formula, represents the large uncertain set of the actual steering vector of the received signal, ∪ represents the union operation; The newly established optimization objective function is specifically: Using the large uncertainty set of the actual steering vector of the received signal Instead of receiving the signal, the actual steering vector uncertainty set Get the new optimization objective function: In the formula, represents the small uncertainty set of the actual steering vector of the Mth received signal; The second-order cone programming problem conversion module is used to convert the new optimization objective function into a second-order cone programming problem with the same optimal solution, specifically: In the formula, represents the estimated steering vector of the mth received signal, ∈ m represents the mth vector threshold, and The second-order cone programming problem solving module uses the sequential convex approximation method to solve the second-order cone programming problem, obtain the weight vector of the optimal antenna array, and output the optimal robust adaptive beam, including: Rewrite the second-order cone programming problem to obtain a new second-order cone programming problem: In the formula, represents the operation of obtaining the real part, express The transposed matrix, w l-1 represents the weight vector of the optimal antenna array at the l-1th iteration, and l represents the iteration round; set up and auxiliary variable ξ; Define the iteration round l = 1, Randomly set w 0 , so that Among them, w 0 represents the weight vector of the initial antenna array, w 0H Represents the transposed matrix of the weight vector of the initial antenna array, v0 represents the initial optimal value; w 0 Substitute it into the new second-order cone programming problem to solve and obtain the optimal solution w for the current round l , and calculate Among them, w l represents the weight vector of the optimal antenna array at the lth iteration, w lH represents the transposed matrix of the weight vector of the optimal antenna array at the lth iteration, v l represents the optimal value of the lth round; Let l = l + 1, and replace the optimal solution w from the previous round l Substitute it into the new second-order cone programming problem and solve it to obtain w l+1 and v l+1 ; Calculate the optimal value difference v′, v′=|v l-2 -v l-1 |; Compare the optimal value difference v′ with the auxiliary variable ξ. If v′>ξ, set l=l+1 and replace the optimal solution w l Substitute it into the new second-order cone programming problem and solve it to obtain w l+1 and v l+1 ; Otherwise, execute the following steps; v l-1 The optimal solution w for the corresponding round l-1 Output as a weighted vector of the optimal antenna array; The optimal robust adaptive beam is output according to the weight vector of the optimal antenna array.