Beamforming method, system and storage medium based on multi-static radar model

Abstract

The present application relates to radar communication and signal processing technical field, more particularly, to a kind of beam forming method based on multi-static radar model.For the beam forming optimization problem of multi-static radar radar model, the problem of large amount of calculation and high complexity is solved.The present application is based on multi-static radar model, constructs the signal-to-interference-plus-noise ratio problem of radar beam forming output, and gives the signal-to-noise ratio constraint and transmit power constraint of communication user beam forming output;Then using alternating optimization method, for non-convex quadratic inequality problem, non-convex problem is converted into convex problem by applying CCCP technology and Cauchy inequality, and is solved by CVX toolbox, and the optimal solution obtained is substituted back into the original optimization problem for alternating iteration until the convergence condition is reached.The purpose of widely applicable through optimizing the beam forming vector of transmitter and receiver to improve output signal-to-interference-plus-noise ratio is achieved.

Description

Technical Field

[0001] This invention relates to the field of radar communication and signal processing technology, and more specifically, to a beamforming method, system, and storage medium based on a multi-static radar model. Background Technology

[0002] Multistatic radar systems have attracted widespread attention due to their advantages in detection probability and tracking capabilities. Compared to monostatic radar, this system utilizes the diversity of target scattering to improve radar performance. However, traditional distributed multiple-input multiple-output (MIMO) radar, as a specific type of multistatic radar, sacrifices coherent processing gain on the transmitter side, which can lead to severe performance degradation in specific subsystems composed of individual transmitters and receivers. This phenomenon is obviously detrimental to subsequent detection and tracking processing. In recent years, receiver beamforming techniques have been proposed to improve the output signal-to-interference-plus-noise ratio (SNR) by optimizing the beamforming vectors of the transmitter and receiver.

[0003] Traditional beamforming techniques are mainly divided into fixed beamforming and adaptive beamforming. Fixed beamforming is an antenna array technique with fixed weights and phase settings. It points the primary receiving beam in a specific direction to enhance the speech signal from that direction and minimize interference signals from other directions. This fixed beam pointing may be determined based on a predefined sound source location or the desired signal direction. Therefore, fixed beamforming is only suitable for applications with fixed speech sources or requiring a specific transmission direction. Adaptive beamforming adjusts the weights and phase settings of the antenna array based on the real-time received speech signal and noise environment to provide better speech quality and microphone array performance. Its goal is to suppress interference noise to the greatest extent and enhance the desired speech signal. However, existing adaptive beamforming techniques simply determine adaptive weights through different criteria, and their effect on improving the output signal-to-interference-plus-noise ratio (SNR) is not significant.

[0004] Existing technology discloses a robust adaptive beamforming algorithm based on direct convex optimization modeling, belonging to the field of radar technology. This invention, based on the discovery that Robust-ADBF can reduce errors, discloses a robust adaptive beamforming algorithm based on direct convex optimization modeling. This algorithm further precisely controls the angle error range through signal amplitude response constraints and performs cross-iterative solving. This invention has a deeper notch in the interference direction, achieving an improvement of over 10dB compared to previous beamforming schemes; it also significantly improves the output SINR, resulting in superior system performance. However, in practical problems, non-convex problems are more common. This algorithm uses direct convex optimization modeling, which is relatively complex and difficult to apply directly in most cases, limiting its application scope. Summary of the Invention

[0005] The purpose of this invention is to disclose a beamforming method based on a multistatic radar model that can be widely applied to improve the output signal-to-interference-plus-noise ratio by optimizing the beamforming vectors of the transmitter and receiver.

[0006] To achieve the above objectives, the present invention provides a beamforming method based on a multi-static radar model, comprising the following steps:

[0007] S1: Construct a multistatic radar model that maximizes the SINR of radars while satisfying the minimum number of radars.

[0008] S2: Provide user SNR constraints and power constraints;

[0009] S3: Under the conditions of satisfying the user SNR and power constraints, establish the original optimization problem for the multistatic radar model constructed in step S1;

[0010] S4: The original optimization problem is optimized using an alternating optimization method to obtain the optimized problem;

[0011] S5: Use CCCP technology to convert non-convex problems in the optimized problem into convex problems;

[0012] S6: Solve the convex problem obtained in step S5 using the CVX toolbox to obtain the transmit beamforming vector or receive beamforming vector that meets the requirements. Determine whether the difference between the transmit beamforming vector or receive beamforming vector and the vector in the previous iteration is less than the preset parameter. If it is less, proceed to step S7; if it is greater, return to step S3.

[0013] S7: Outputs the transmit beamforming vector and the receive beamforming vector.

[0014] Furthermore, in step S1, for a multistatic radar system with transmitter Tx and Q receivers Rx and K communication users, a multistatic radar model that maximizes the radar SINR is established under the condition of satisfying the minimum number of radars.

[0015] The transmitter Tx has Mt components, and each receiver Rx has Mr components. The transmitter Tx transmits information to the communication user in a multicast manner. In addition, there are J interferences to the radar, a t (..) represents the radar transmit guidance vector, a r (.) represents the radar receiving guidance vector, and s(l) represents the radar waveform;

[0016] The signal transmitted by transmitter Tx to the target or jammer: Where t is the transmitted beamforming vector, and the received signal reflected by the transmitter Tx to the target and then to the q-th receiver Rx is:

[0017]

[0018] Where, α 0,q α j,q Let n be the amplitude, and nq be Gaussian white noise. θ 0,q θ represents the angle of incidence of the target to the receiver Rx. j,q θ represents the angle of incidence of the interference to the receiver Rx, and θ0 represents the angle of incidence of the transmitter Tx to the target. j Let E{|s(l)| represent the incident angle of the transmitter Tx to the interference. 2}=1

[0019] After the received signal from the q-th receiver Rx passes through the beamformer, we have:

[0020]

[0021] Among them, by get And based on their forms, the useful signal, interference signal, and noise are obtained respectively;

[0022] The multistatic radar model, obtained from the output of the adaptive beamformer, maximizes the SINR of the radar while satisfying the minimum number of radars.

[0023]

[0024] Furthermore, the received signal of the k-th user in step S2 is:

[0025] Where h k Let n be the channel between the radar transmitter and the k-th user. k It is noise, and The signal-to-noise ratio (SNR) for the k-th user is:

[0026] Furthermore, in step S3, under the conditions of satisfying the user SNR and power constraints, the original optimization problem is established for the multistatic radar model constructed in S1:

[0027]

[0028] stSNR k ≥E k k = 1, ..., K

[0029] ||t|| 2 ≤1

[0030] Where Ek is the threshold of the SNR of the communication user;

[0031] Furthermore, in step S4, the optimization problem is solved using an alternating optimization method;

[0032] First, fix the transmit beamforming vector and optimize the receive beamforming vector. Then the original problem becomes the following form:

[0033]

[0034] in

[0035]

[0036] Introducing U transforms the above problem into the following form:

[0037]

[0038]

[0039] The optimal solution to the above problem can be obtained. for The optimal value of the above problem is v*, which is the eigenvector corresponding to the largest eigenvalue.

[0040]

[0041] Pick Optimizing t, the original problem becomes the following form:

[0042]

[0043] stSNR k ≥E k k = 1, ..., K

[0044] ||t|| 2 ≤1

[0045] Introducing λ, such that SINR q ≥λ, that is: Where q = 1, ..., Q,

[0046]

[0047]

[0048] The original problem then becomes the following form:

[0049]

[0050] stSINR q ≥λ, q=1,...,Q

[0051] SNRk ≥E k k = 1, ..., K

[0052] ||t|| 2 ≤1.

[0053] Further, in step S5, for SINR q With the constraint ≥λ, we have: The constraint is transformed into a convex constraint using CCCP technology, that is:

[0054]

[0055] Among them, t (r) and λ (r) Let be the values ​​of t and λ in the r-th iteration.

[0056] again Right now:

[0057] Furthermore, in step S5, the non-convex constraint of SINR for communication is transformed into a convex constraint using the properties of Cauchy's inequality, that is:

[0058]

[0059] but:

[0060] The original problem then becomes the following form:

[0061]

[0062]

[0063]

[0064] ||t|| 2 ≤1.

[0065] Further, in step S6, the solved transmit beamforming vector or receive beamforming vector is compared with the previous iteration. If the difference is less than the threshold, step S7 is executed; if it is greater than the threshold, the transmit beamforming vector or receive beamforming vector is alternately fixed, and the fixed transmit beamforming vector or receive beamforming vector is substituted into the original optimization model, and step S3 is executed to optimize the other vector.

[0066] Furthermore, this invention also provides a beamforming system based on a multi-static radar model, comprising:

[0067] Model building module: Constructs a multistatic radar model that maximizes the SINR of radars while satisfying the minimum number of radars.

[0068] Constraint module: Provides user SNR constraints and power constraints;

[0069] Primitive optimization problem establishment module: Under the conditions of satisfying user SNR and power constraints, the primitive optimization problem is established for the multistatic radar model constructed by the model building module;

[0070] Alternating Optimization Module: This module uses an alternating optimization method to optimize the original optimization problem.

[0071] Transformation module: Converts non-convex problems in the optimized problem into convex problems using CCCP technology;

[0072] Solving and Judgment Module: The convex problem obtained in the transformation module is solved using the CVX toolbox to obtain the transmit beamforming vector or receive beamforming vector that meets the requirements. It determines whether the difference between the transmit beamforming vector or receive beamforming vector and the vector in the previous iteration is less than the preset parameter. If it is less, the output module is executed; if it is greater, the original optimization problem establishment module is returned to be executed.

[0073] Output module: Outputs the transmit beamforming vector and the receive beamforming vector.

[0074] Furthermore, the present invention provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it performs the beamforming method based on a multi-static radar model. Compared with the prior art, the present invention has the following advantages:

[0075] This invention, based on a multi-static radar model, constructs a signal-to-interference-plus-noise ratio (SINR) problem for radar beamforming output, and provides SNR and transmit power constraints for communication user beamforming output. Then, an alternating optimization method is employed. For non-convex quadratic inequality problems, CCCP technology and Cauchy inequality are applied to transform the non-convex problem into a convex one. The CVX toolbox is used to solve the problem, and the optimal solution is substituted back into the original optimization problem for alternating iteration until convergence is achieved. This aims to maximize the initial objective of satisfying the minimum number of radar received signal-to-interference-plus-noise ratios.

[0076] Furthermore, compared to direct convex optimization modeling, this invention utilizes CCCP technology and Cauchy inequality to transform many real-world non-convex problems into easily solvable convex problems, and then uses the CVX toolbox to solve these convex problems. This avoids the situation where many real-world non-convex problems cannot be directly modeled as convex problems. It can be widely applied.

[0077] Attached Figures and Their Descriptions

[0078] Figure 1 This is a flowchart of the beamforming method based on a multi-static radar model according to the present invention.

[0079] Figure 2 This is a flowchart of the beamforming system based on a multi-static radar model according to the present invention.

[0080] Figure 3 This is a model diagram of the beamforming method based on a multi-static radar model according to the present invention. Detailed Implementation

[0081] Example 1

[0082] like Figure 1 As shown, a preferred embodiment of the present invention provides a beamforming method based on a multi-static radar model. The method includes the following steps:

[0083] S1: Construct a multistatic radar model that maximizes the SINR of radars while satisfying the minimum number of radars.

[0084] S2: Provide user SNR constraints and power constraints;

[0085] S3: Under the conditions of satisfying the user SNR and power constraints, establish the original optimization problem for the multistatic radar model constructed in step S1;

[0086] S4: The original optimization problem is optimized using an alternating optimization method to obtain the optimized problem;

[0087] S5: Use CCCP technology to convert non-convex problems in the optimized problem into convex problems;

[0088] S6: Solve the convex problem obtained in step S5 using the CVX toolbox to obtain the transmit beamforming vector or receive beamforming vector that meets the requirements. Determine whether the difference between the transmit beamforming vector or receive beamforming vector and the vector in the previous iteration is less than the preset parameter. If it is less, proceed to step S7; if it is greater, return to step S3.

[0089] S7: Outputs the transmit beamforming vector and the receive beamforming vector.

[0090] This embodiment constructs a multistatic radar model that maximizes the SINR of the radar while satisfying the minimum number of radars. It provides constraints on the signal-to-noise ratio (SNR) and transmit power of the user beamforming output. Then, an alternating optimization method is employed. For non-convex quadratic inequality problems, CCCP technology and Cauchy inequality are applied to transform the non-convex problem into a convex one. The CVX toolbox is used to solve the problem, and the optimal solution is substituted back into the original optimization problem for alternating iteration until convergence is achieved. This achieves the goal of improving the output SNR by optimizing the beamforming vectors of the transmitter and receiver, which can be widely applied.

[0091] Furthermore, compared to direct convex optimization modeling, this embodiment utilizes CCCP technology and Cauchy inequality to transform many real-world non-convex problems into easily solvable convex problems, and then uses the CVX toolbox to solve these convex problems. This avoids the situation where many real-world non-convex problems cannot be directly modeled as convex problems. It can be widely applied.

[0092] Example 2

[0093] like Figure 1 As shown, the preferred embodiment of the beamforming method based on a multi-static radar model of the present invention includes the following steps:

[0094] Consider a multistatic radar system with one transmitter Tx, Q receivers Rx, and K communication users. Each transmitter Tx has Mt elements, each receiver Rx has Mr elements, and the transmitter Tx transmits information to the communication users via multicast. Additionally, there are J interference signals affecting the radar. t (.) represents the radar transmit guidance vector, a r (.) represents the radar receiving guidance vector, and s(l) represents the radar waveform.

[0095] The signal transmitted by transmitter Tx to the target or jammer: Where t is the transmitted beamforming vector.

[0096] The signal received by the transmitter Tx after being transmitted to the target and reflected back to the q-th receiver Rx is:

[0097]

[0098] Where, α 0,q α j,q Let n be the amplitude, and nq be Gaussian white noise. θ 0,q θ represents the angle of incidence of the target to the receiver Rx. j,q θ represents the angle of incidence of the interference to the receiver Rx, and θ0 represents the angle of incidence of the transmitter Tx to the target. j Let E{|s(l)| represent the incident angle of the transmitter Tx to the interference. 2}=1

[0099] After the received signal from the q-th receiver Rx passes through the beamformer, we have:

[0100]

[0101] Where w q This is the receiving beamforming vector.

[0102] The signal-to-interference-plus-noise ratio (SINR) of the radar is:

[0103]

[0104] The received signal for the kth user is:

[0105] Where h k Let n be the channel between the radar transmitter and the k-th user. k It is noise, and The signal-to-noise ratio (SNR) of the k-th user is:

[0106] To maximize the SINR that satisfies the minimum number of radar receivers while satisfying the communication user SNR and power constraints, the primal optimization problem is as follows:

[0107]

[0108] stSNR k ≥E k k = 1, ..., K

[0109] ||t|| 2 ≤1

[0110] Where Ek is the threshold of the SNR of the communication user. We use the Alternative Optimization (AO) method to solve this optimization problem.

[0111] First, fix t and optimize wq. The original problem then becomes the following form: in

[0112]

[0113] Introducing U transforms the above problem into the following form:

[0114]

[0115]

[0116] The optimal solution to the above problem can be obtained. for The eigenvector corresponding to the largest eigenvalue. Then the optimal value v* for the above problem is:

[0117]

[0118] Pick Optimizing t, the original problem becomes the following form:

[0119]

[0120] stSNR k ≥E k k = 1, ..., K

[0121] ||t|| 2 ≤1

[0122] Introducing λ, such that SINR q ≥λ, that is: Where q = 1, ..., Q,

[0123]

[0124] The original problem then becomes the following form:

[0125]

[0126] stSINR q ≥λ, q=1,…,Q

[0127] SNR k ≥E k k = 1, ..., K

[0128] ||t|| 2 ≤1

[0129] For SINR q With the constraint ≥λ, we have: Using the CCCP (Convex-Concave Procedure) technique, this constraint can be transformed into a convex constraint, that is:

[0130]

[0131] Among them, t (r) and λ (r) Let be the values ​​of t and λ in the r-th iteration.

[0132] again Right now:

[0133] And because:

[0134] but:

[0135] The original problem then becomes the following form:

[0136]

[0137]

[0138]

[0139] ||t|| 2 ≤1

[0140] This embodiment constructs a multistatic radar model that maximizes the SINR of the radar while satisfying the minimum number of radars. It provides constraints on the signal-to-noise ratio (SNR) and transmit power of the user beamforming output. Then, an alternating optimization method is employed. For non-convex quadratic inequality problems, CCCP technology and Cauchy inequality are applied to transform the non-convex problem into a convex one. The CVX toolbox is used to solve the problem, and the optimal solution is substituted back into the original optimization problem for alternating iteration until convergence is achieved. This achieves the goal of improving the output SNR by optimizing the beamforming vectors of the transmitter and receiver, which can be widely applied.

[0141] Non-convex problems are eventually transformed into convex problems, which can be solved directly using the CVX toolkit. Its advantage lies in the fact that by transforming it into a convex problem, the feasibility of the solution and the existence of a global optimum can be guaranteed. The CVX toolkit provides a concise and intuitive modeling language to simplify the problem description and solution process. Its flexibility and scalability make it suitable for various convex optimization problems, while efficient convex optimization algorithms and numerical optimization techniques ensure numerical stability and accuracy, thus making the modeling and solving of convex optimization problems more convenient, efficient and reliable.

[0142] Example 3:

[0143] like Figure 3 As shown, a preferred embodiment of the present invention provides a beamforming system based on a multi-static radar model, comprising:

[0144] Model building module: Constructs a multistatic radar model that maximizes the SINR of radars while satisfying the minimum number of radars.

[0145] Constraint module: Provides user SNR constraints and power constraints;

[0146] Primitive optimization problem establishment module: Under the conditions of satisfying user SNR and power constraints, the primitive optimization problem is established for the multistatic radar model constructed by the model building module;

[0147] Alternating Optimization Module: This module uses an alternating optimization method to optimize the original optimization problem.

[0148] Transformation module: Converts non-convex problems in the optimized problem into convex problems using CCCP technology;

[0149] Solving and Judgment Module: The convex problem obtained in the transformation module is solved using the CVX toolbox to obtain the transmit beamforming vector or receive beamforming vector that meets the requirements. It determines whether the difference between the transmit beamforming vector or receive beamforming vector and the vector in the previous iteration is less than the preset parameter. If it is less, the output module is executed; if it is greater, the original optimization problem establishment module is returned to be executed.

[0150] Output module: Outputs the transmit beamforming vector and the receive beamforming vector.

[0151] This embodiment constructs a multistatic radar model that maximizes the SINR of the radar while satisfying the minimum number of radars. It provides constraints on the signal-to-noise ratio (SNR) and transmit power of the user beamforming output. Then, an alternating optimization method is employed. For non-convex quadratic inequality problems, CCCP technology and Cauchy inequality are applied to transform the non-convex problem into a convex one. The CVX toolbox is used to solve the problem, and the optimal solution is substituted back into the original optimization problem for alternating iteration until convergence is achieved. This achieves the goal of improving the output SNR by optimizing the beamforming vectors of the transmitter and receiver, which can be widely applied.

[0152] Transforming non-convex problems into easily solvable convex problems offers several advantages. First, convex optimization problems guarantee a global optimum, while non-convex problems may have multiple local optima. By transforming them into convex problems, we can ensure finding a global optimum or a near-optimal solution, avoiding the trap of getting stuck in local optima. Second, convex optimization problems possess favorable mathematical properties, such as the existence of first and second derivatives of convex functions and the properties of convex sets. These properties make the solution of convex optimization problems more stable and reliable, and allow for the application of a series of efficient convex optimization algorithms. Furthermore, the modeling and solving of convex optimization problems are relatively simplified, with concise mathematical expressions that are easy to describe and understand. Finally, convex optimization problems exhibit high algorithmic feasibility and scalability, with various efficient solution methods available, and the efficiency can be further improved through parallelization and distributed computing techniques.

[0153] In summary, this invention provides a scheduling optimization method and system based on peak information age under dynamic channel conditions. This embodiment first constructs a multi-static radar model that maximizes the SINR of the radars while satisfying the minimum number of radars, and provides constraints on the signal-to-noise ratio (SNR) and transmit power of the user beamforming output. Then, an alternating optimization method is used. For non-convex quadratic inequality problems, CCCP technology and Cauchy inequality are applied to transform the non-convex problem into a convex problem, which is then solved using the CVX toolbox. The optimal solution is substituted back into the original optimization problem for alternating iteration until convergence is achieved. This achieves the goal of improving the output SNR by optimizing the beamforming vectors of transmitters and receivers, which can be widely applied.

[0154] Furthermore, compared to direct convex optimization modeling, this embodiment utilizes CCCP technology and Cauchy inequality to transform many real-world non-convex problems into easily solvable convex problems, and then uses the CVX toolbox to solve these convex problems. This avoids the situation where many real-world non-convex problems cannot be directly modeled as convex problems. It can be widely applied.

[0155] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and substitutions can be made without departing from the technical principles of the present invention, and these improvements and substitutions should also be considered within the scope of protection of the present invention.

Claims

1. A beamforming method based on a multi-static radar model, characterized in that, Includes the following steps: S1: Construct a multistatic radar model that maximizes the SINR of radars while satisfying the minimum number of radars; S2: Provide user SNR constraints and power constraints; S3: Under the conditions of satisfying the user SNR and power constraints, establish the original optimization problem for the multistatic radar model constructed in step S1; S4: The original optimization problem is optimized using an alternating optimization method to obtain the optimized problem; S5: Convert non-convex problems in the optimized problem into convex problems using CCCP technology; S6: Solve the convex problem obtained in step S5 using the CVX toolbox to obtain the transmit beamforming vector or receive beamforming vector that meets the requirements. Determine whether the difference between the transmit beamforming vector or receive beamforming vector and the vector in the previous iteration is less than the preset parameter. If it is less, proceed to step S7; if it is greater, return to step S3. S7: Output transmit beamforming vector and receive beamforming vector; In step S4, the optimization problem is solved using an alternating optimization method; First, fix the transmit beamforming vector and optimize the receive beamforming vector. Then the original problem becomes the following form: ， in Introducing U transforms the above problem into the following form: The optimal solution to the above problem can be obtained. , for The optimal value of the above problem is v*, which is the eigenvector corresponding to the largest eigenvalue. Pick If we optimize t, the original problem becomes the following form: Introduction , making ,Right now: Where q=1, Q, , The original problem then becomes the following form: ； In step S5, for The constraints are: The constraint is transformed into a convex constraint using CCCP technology, that is: in, again ,Right now: ; In step S5, the non-convex constraint of SINR for communication is transformed into a convex constraint using the property of Cauchy's inequality, that is: but: , The original problem then becomes the following form: 。 2. The beamforming method based on a multi-static radar model according to claim 1, characterized in that, In step S1, for a multistatic radar system with transmitter Tx, Q receivers Rx, and K communication users, a multistatic radar model is constructed that maximizes the radar SINR while satisfying the minimum number of radars. The transmitter Tx has Mt components, and each receiver Rx has Mr components. The transmitter Tx transmits information to the communication user in a multicast manner. In addition, there are J interferences affecting the radar. Indicates the radar transmission guidance vector. Indicates the radar receiving guidance vector. Represents radar waveform; The signal transmitted by transmitter Tx to the target or jammer: Where t is the transmitted beamforming vector, and the received signal reflected by the transmitter Tx to the target and then to the q-th receiver Rx is: in, , For amplitude, It is Gaussian white noise, and , , , This represents the angle of incidence of the target onto the receiver Rx. This represents the angle of incidence of the interference on the receiver Rx. Indicates the angle of incidence of the transmitter Tx target. This represents the angle of incidence of interference at transmitter Tx. After the received signal from the q-th receiver Rx passes through the beamformer, we have: Among them, by get And based on their forms, the useful signal, interference signal and noise are obtained respectively; The multistatic radar model, obtained from the output of the adaptive beamformer, maximizes the SINR of the radar while satisfying the minimum number of radars. 。 3. The beamforming method based on a multi-static radar model according to claim 1, characterized in that, The received signal of the k-th user in step S2 is: in This is the channel between the radar transmitter and the k-th user. It is noise, and Then the signal-to-noise ratio (SNR) of the k-th user is: .

4. The beamforming method based on a multi-base radar model according to claim 1, characterized in that, In step S3, under the conditions of satisfying the user SNR and power constraints, the primal optimization problem is established for the multistatic radar model constructed in S1: in This is the threshold for the SNR of communication users.

5. The beamforming method based on a multi-static radar model according to claim 1, characterized in that, In step S6, the solved transmit beamforming vector or receive beamforming vector is compared with the previous iteration. If the difference is less than the threshold, step S7 is executed; if it is greater than the threshold, the transmit beamforming vector or receive beamforming vector is alternately fixed, and the fixed transmit beamforming vector or receive beamforming vector is substituted into the original optimization model. Then, step S3 is executed to optimize the other vector.

6. A beamforming system based on a multi-static radar model, characterized in that, include: Model building module: Constructs a multistatic radar model that maximizes the SINR of radars while satisfying the minimum number of radars. Constraint module: Provides user SNR constraints and power constraints; Primitive optimization problem establishment module: Establishes the primal optimization problem for the multistatic radar model constructed by the model building module, while satisfying user SNR and power constraints; Alternating Optimization Module: This module uses an alternating optimization method to optimize the original problem, resulting in an optimized problem. Transformation module: Converts non-convex problems in the optimized problem into convex problems using CCCP technology; Solving and Judgment Module: The convex problem obtained in the transformation module is solved using the CVX toolbox to obtain the transmit beamforming vector or receive beamforming vector that meets the requirements. It determines whether the difference between the transmit beamforming vector or receive beamforming vector and the vector in the previous iteration is less than the preset parameter. If it is less, the output module is executed; if it is greater, the original optimization problem establishment module is returned to be executed. Output module: Outputs the transmit beamforming vector and the receive beamforming vector; In the alternating optimization module, the optimization problem is solved using the alternating optimization method; First, fix the transmit beamforming vector and optimize the receive beamforming vector. Then the original problem becomes the following form: ， in Introducing U transforms the above problem into the following form: The optimal solution to the above problem can be obtained. , for The optimal value of the above problem is v*, which is the eigenvector corresponding to the largest eigenvalue. Pick If we optimize t, the original problem becomes the following form: Introduction , making ,Right now: Where q=1, Q, , The original problem then becomes the following form: ； In the conversion module, for The constraints are: The constraint is transformed into a convex constraint using CCCP technology, that is: in, again ,Right now: ; In the conversion module, the non-convex SINR constraint for communication is transformed into a convex constraint using the properties of Cauchy's inequality, i.e.: but: , The original problem then becomes the following form: 。 7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the beamforming method based on a multi-static radar model as described in any one of claims 1 to 5.

Images