Wideband adaptive beamforming method and system with pre-steering covariance matrix reconstruction

Abstract

The application belongs to the field of adaptive beam forming technology in array signal processing, and discloses a kind of robust broadband adaptive beam forming method and system based on pre-directing covariance matrix reconstruction, the method comprises: setting antenna array;Wideband signal frequency domain processing;Estimate pre-directing matrix using robust beam forming criterion;Desired signal power and noise power are obtained by STMV broadband spectrum estimation, and the desired signal part is removed in the newly estimated direct covariance matrix;Further optimized weight is obtained by solving the quadratic constraint quadratic programming problem;Calculate wideband adaptive beam forming weight vector.The application designs a kind of robust STMV algorithm for the direct covariance matrix (STCM) error and direct vector error in STMV algorithm, has the advantages, such as higher signal-to-interference-and-noise ratio of beam forming output under the condition of broadband signal with various errors, and desired signal is not distorted.

Description

Technical Field

[0001] This invention belongs to the field of adaptive beamforming technology in array signal processing, and particularly relates to a robust broadband adaptive beamforming method and system based on pre-guided covariance matrix reconstruction. Background Technology

[0002] Broadband adaptive beamforming is a signal processing technique used at the receiver to process signals received by multiple antennas to obtain the target signal and suppress interference signals. Based on adaptive filters and beamforming principles, this technique enhances the target signal and suppresses interference signals by weighting and phase-adjusting the signals received by each antenna. It is widely used in wireless communication, radar, and radio eavesdropping to improve signal quality and system performance.

[0003] When processing broadband signals, the signal waveform changes with frequency due to the large bandwidth, making traditional beamforming techniques unsuitable for direct application. Furthermore, extracting effective information from broadband signals requires high-resolution sampling and processing, increasing complexity and computation. One solution is subband beamforming, which decomposes the broadband signal into multiple narrowband signals and then performs beamforming on each. This method first divides the broadband signal into several overlapping narrowband signals, then performs beamforming on each. During this process, it's crucial to ensure overlap between adjacent narrowband signals to prevent signal loss. The beamforming results from each narrowband signal are then merged to obtain the broadband signal's beamforming result. Another solution is focused beamforming, which uses time-domain and frequency-domain signal processing to focus the beam onto the target, providing high resolution and low sidelobe levels while maintaining high computational efficiency. Coherent focusing methods, such as spatial resampling or coherent signal subspace methods, can be used to focus a narrowband correlation spectral matrix (CSDM) onto a single frequency over the bandwidth of a broadband signal before applying narrowband beamforming methods. Spatial resampling focuses sources from all angular sectors, but its application is limited to uniform linear arrays. The coherent signal subspace method (CSM) is applicable to arbitrary arrays, but typically requires an initial estimate of the DOA to estimate the focusing matrix. Subsequent Bayesian focusing methods, along with spatial interpolation methods, have addressed the impact of initial angle estimation errors to some extent. This approach is based on the observation that in a wavefield composed of plane waves, spatial and spectral parameters are correlated because plane waves depend only on frequency through a product. Therefore, the transformation of the array manifold between frequencies is equivalent to spatial interpolation of the array, where the new spatial coordinates are scaled according to the desired frequency ratio. Most focusing methods in the literature either belong to either directional methods requiring prior DOA knowledge or panoramic methods that do not. Several focusing methods attempt to compromise between directional and panoramic approaches.

[0004] Based on the above analysis, the problems and shortcomings of the existing technology are as follows:

[0005] (1) In the existing broadband adaptive beamforming of preguided structures, there are frequency domain guide vector errors and guide covariance matrix errors.

[0006] (2) In the existing broadband adaptive beamforming of the preguided structure, the influence caused by the presence of the desired signal in the guide covariance matrix is ​​not considered. Summary of the Invention

[0007] To address the problems existing in the prior art, this invention provides a robust broadband adaptive beamforming method and system based on pre-guided covariance matrix reconstruction.

[0008] This invention is implemented as follows: a robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction, wherein the robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction includes:

[0009] Step 1: Configure the antenna array;

[0010] Step 2: Broadband signal frequency domain processing;

[0011] Step 3: Pre-steering matrix estimation. The pre-steering matrix is ​​estimated using robust beamforming criteria.

[0012] Step 4: Steering Covariance Matrix (STCM) Reconstruction. The desired signal power and noise power are obtained by steering minimum variance (STMV) broadband spectrum estimation. The desired signal component is then removed from the newly estimated steering covariance matrix.

[0013] Step 5: Optimize the steering vector used for beamforming by solving a quadratic constraint quadratic programming problem to obtain further optimized weights;

[0014] Step 6: Calculate the broadband adaptive beamforming weight vector.

[0015] Further, in step one, a one-dimensional uniform linear array is set up with M array elements and an array spacing d = λ². Let M = 10 omnidirectional array elements form the uniform linear array. The interference is a linearly modulated frequency (LFM) signal, and the desired signal is generated by filtered Gaussian noise, where the noise power at each frequency may be different. The distance between two adjacent sensors is half the wavelength of the highest frequency. The signal bandwidth is assumed to be 30MHz, the signal carrier center frequency range is assumed to be 25GHz-27GHz, and the signal propagation speed is the speed of light in a vacuum. Regarding the incident angle, the desired signal originates from θ0 = 0°, a strong interference with INR = 30dB originates from θ1 = 40°, and a weak interference with INR = 10dB originates from θ2 = -40°. The desired signal is located in the angular sector Θ. s = [θ0-3°, θ0+3°], the interference signal is located at Θ i = [θ1-3°, θ1+3°]∪[θ2-3°, θ2+3°], the pure noise sector is Θ n .

[0016] Furthermore, step two includes:

[0017] (1) Obtain the array receiving time-domain signal X(n), and the time-domain output of the m-th array element is:

[0018] Xm (t)=S0(t-τ m (θ0))+N m (t)

[0019] Where S0(t) and τ m (θ0), N m (t) represents the desired signal waveform, the propagation delay of the m-th array element with respect to the corresponding direction θ0, and the noise and interference at the m-th sensor, respectively.

[0020] (2) Perform an L-point Fast Fourier Transform on the time-domain snapshot X(t) to generate a frequency-domain snapshot. For the l-th FFT unit, the m-th array element, and the n-th frequency-domain snapshot, we can write:

[0021]

[0022] x m,l (n), s 0,l (n), n m,l (n) represent the frequencies at f, respectively. l = (l-1) / LT s Fourier transform at T s For the sampling period, the array steering vector model of the center frequency signal can be written as:

[0023]

[0024] The nth frequency domain snapshot is written according to the signal structure as follows:

[0025]

[0026] (3) Calculation of cross-spectral density matrix (CSDM):

[0027]

[0028] (4) Calculation of the Steering Covariance Matrix (STCM):

[0029] For any angle of orientation, the covariance matrix is:

[0030]

[0031] The structure of STCM can be written as:

[0032]

[0033] in For the desired total signal power, Let be the power of the k-th interference signal at the l-th frequency. Let T be the total power of the Gaussian white noise, where T is the total power of the Gaussian white noise. l(θ) is the pre-steering matrix for the l-th frequency, in the form of:

[0034]

[0035] Furthermore, step three involves using the RCB method to correct the desired signal steering vector at each frequency point and correct the pre-steering matrix, including:

[0036] (1) Solve the problem of steering vector estimation of the desired signal at each frequency point:

[0037]

[0038] in For prior knowledge, the expected signal steering vector is the assumed signal; a l The desired signal steering vector to be estimated; the optimal solution to the optimization problem is in the form of:

[0039]

[0040] Where λ l To apply a diagonal loading factor λ obtained by the Lagrange multiplier method to the desired signal steering vector at the corresponding frequency, l To obtain it, follow these steps: First, Perform eigenvalue decomposition:

[0041]

[0042] Where U includes The eigenvectors are Γ, which is a diagonal matrix composed of the eigenvalues ​​of the corresponding eigenvectors, where the eigenvalues ​​are γ1≥γ2≥…≥γ. M ,make:

[0043]

[0044] Diagonal loading factor λ l It can be solved by the following equation:

[0045]

[0046] z m Let be the m-th element of vector z. The solution to the equation must satisfy the following inequality:

[0047]

[0048] Normalize the estimated steering vector:

[0049]

[0050] (2) Construct the pre-guided matrix:

[0051] The pre-steering matrix at the l-th frequency point is constructed as follows:

[0052]

[0053] Furthermore, step four includes:

[0054] (1) Use the pre-steering matrix obtained in step three to calculate the steering covariance matrix in the direction of the desired signal.

[0055]

[0056] (2) Calculate the STMV spectrum:

[0057]

[0058] Where 1 = [11...1] T The noise power and desired signal power are calculated using the estimated STMV spectrum. The integration in the algorithm is implemented using simulation, with a sampling point of 2500 for each angular sector. The Gaussian white noise power of each sector is then calculated. The estimate is:

[0059]

[0060] θ r Given the angle values ​​at the sampling points, the desired signal power is estimated. for:

[0061]

[0062] (3) After removing the desired signal component, the reconstructed STCM can be written as:

[0063]

[0064] Furthermore, in step five, an optimization problem is constructed to find the optimal steering vector solution. The post-guidance expected steering vector estimate for robust STMV is:

[0065]

[0066] Optimization problem:

[0067]

[0068] subject to 1 H e ⊥ =0

[0069]

[0070] This optimization problem is a convex problem, which can be solved using the convex optimization toolbox.

[0071] Furthermore, in step six, the broadband adaptive beamforming weight vector is calculated using the following formula:

[0072]

[0073] Another objective of this invention is to provide a robust broadband adaptive beamforming system based on pre-guided covariance matrix reconstruction, wherein the robust broadband adaptive beamforming system based on pre-guided covariance matrix reconstruction includes:

[0074] The pre-guide matrix estimation module is used to re-estimate the pre-guide matrix using robust beamforming criteria to obtain a more accurate guide covariance matrix.

[0075] The STCM reconstruction module is used to obtain the desired signal power and noise power through STMV broadband spectrum estimation, and remove the desired signal part from the newly estimated steering covariance matrix;

[0076] The steering vector optimization module is used to further optimize the steering vector used for beamforming.

[0077] The weight vector calculation module is used to calculate the broadband adaptive beamforming weight vector.

[0078] Another object of the present invention is to provide a computer device including a memory and a processor, the memory storing a computer program, which, when executed by the processor, causes the processor to perform the steps of the robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction.

[0079] Another object of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of the robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction.

[0080] Another objective of this invention is to provide an information data processing terminal for implementing the robust broadband adaptive beamforming system based on pre-guided covariance matrix reconstruction.

[0081] Based on the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solution to be protected by this invention are as follows:

[0082] First, this invention addresses the steering covariance matrix (STCM) error and steering vector error in the STMV algorithm by designing a robust STMV algorithm. This algorithm primarily estimates the more accurate steering vector of the desired signal at each frequency point using the robust beamforming (RCB) criterion, thereby constructing an accurate pre-steering matrix for each frequency point. This operation improves the accuracy of pre-steering and combats various types of steering vector errors. Then, using the STMV broadband spectrum estimation method, the desired signal power is estimated to reconstruct the interference plus noise steering covariance matrix. This effectively mitigates the self-zeroing effect caused by the desired signal in the STCM, resulting in better performance compared to other broadband STMV algorithms.

[0083] Second, considering the technical solution as a whole or from a product perspective, the technical effects and advantages of the technical solution to be protected by this invention are specifically described as follows:

[0084] By using the broadband STMV spectral integration method, the desired signal and noise are estimated and the desired signal component is subtracted from the steering covariance matrix. This prevents the broadband beamforming from the zero-point phenomenon, and even if the desired signal steering vector deviates, the output signal-to-interference-plus-noise ratio will not drop drastically.

[0085] Third, as supplementary evidence of the inventive step of the claims of this invention, it is also reflected in the following important aspects:

[0086] 1) Improve beamforming accuracy and precision: This method can effectively improve beamforming accuracy and precision by reconstructing the preguided covariance matrix and optimizing the guide vector weights, thereby improving the efficiency and precision of signal processing.

[0087] 2) Improve the robustness and reliability of the system: This method introduces a robust beamforming criterion, which can maintain good beamforming performance even in the presence of outliers or noise, thereby improving the robustness and reliability of the system.

[0088] 3) Adaptability to complex communication environments: The robustness and adaptability of this method are particularly suitable for complex communication environments, and it can still maintain good beamforming performance even when the signal is strongly interfered with and noiseed.

[0089] 4) Improve the real-time performance and operability of the system: This method has a fast computation speed and low computational complexity, which can realize real-time processing and fast response, thus improving the real-time performance and operability of the system. Attached Figure Description

[0090] Figure 1 This is a flowchart of the method provided in an embodiment of the present invention;

[0091] Figure 2This is a schematic diagram of the array setup provided in an embodiment of the present invention;

[0092] Figure 3 This is a schematic diagram showing the relationship between the signal-to-interference-plus-noise ratio (SNR) of the simulation array output under the line-of-sight error provided in the embodiment of the present invention.

[0093] Figure 4 This is a schematic diagram showing the relationship between the signal-to-interference-plus-noise ratio (SIR) of the array output and the number of snapshots under the line-of-sight error provided in the simulation experiment of the embodiment of the present invention.

[0094] Figure 5 This is a schematic diagram showing the relationship between the signal-to-interference-plus-noise ratio (SNR) of the simulation array output under array geometric error conditions, provided by an embodiment of the present invention.

[0095] Figure 6 This is a schematic diagram showing the relationship between the signal-to-interference-plus-noise ratio (SNR) of the simulation array output and the number of snapshots under the condition of array geometric error, provided by the embodiment of the present invention. Detailed Implementation

[0096] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0097] like Figure 1 As shown, the robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction provided in this embodiment of the invention includes the following steps:

[0098] S101, Configure the antenna array;

[0099] S102, wideband signal frequency domain processing;

[0100] S103, Pre-guided matrix estimation, using robust beamforming criteria to estimate the pre-guided matrix;

[0101] S104, STCM reconstruction, obtains the desired signal power and noise power through STMV broadband spectrum estimation, and removes the desired signal part from the newly estimated steering covariance matrix;

[0102] S105, optimize the steering vector used for beamforming, and obtain further optimized weights by solving a quadratic constraint quadratic programming problem;

[0103] S106, Calculate the broadband adaptive beamforming weight vector.

[0104] The following are specific embodiments of the robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction provided by the present invention:

[0105] Step 1: Configure the antenna array

[0106] Before implementing this method, an antenna array needs to be set up at the receiving end to receive broadband signals.

[0107] Step 2: Frequency Domain Processing of Broadband Signals

[0108] The received broadband signal is processed in the frequency domain to obtain the frequency response matrix.

[0109] Step 3: Pre-guided matrix estimation

[0110] The pre-guidance matrix is ​​estimated using robust beamforming criteria. In this embodiment, this step can be implemented using the SOCP (SecondOrder Cone Programming) algorithm.

[0111] Step 4: STCM Refactoring

[0112] The desired signal power and noise power are obtained through STMV broadband spectral estimation, and the desired signal component is removed from the newly estimated steering covariance matrix. In this embodiment, this step can be implemented using the Capon algorithm or the MLE (Maximum Likelihood Estimation) algorithm.

[0113] Step 5: Optimize the steering vector weights

[0114] The weights are further optimized by solving a quadratic constrained quadratic programming problem. In this embodiment, the LMS (Least Mean Squares) algorithm or the RLS (Recursive Least Squares) algorithm can be used to implement this step.

[0115] Step 6: Calculate the broadband adaptive beamforming weight vector

[0116] Based on the optimized steering vector weights, calculate the broadband adaptive beamforming weight vector.

[0117] By following the steps described above, a robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction can be achieved. In practical applications, parameters can be adjusted and optimized according to the specific communication environment and signal characteristics to achieve the best beamforming effect.

[0118] Step one provided in this embodiment of the invention involves setting up a... Figure 2The diagram shows a one-dimensional uniform linear array with M elements and an array spacing of d = λ². Assume M = 10 omnidirectional elements forming a uniform linear array. The interference is a linearly modulated frequency (LFM) signal, and the desired signal is generated by filtered Gaussian noise, where the noise power at each frequency may differ. The distance between two adjacent sensors is half the wavelength of the highest frequency. The signal bandwidth is assumed to be 30MHz, the signal carrier center frequency range is assumed to be 25GHz-27GHz, and the signal propagation speed is the speed of light in a vacuum. Regarding the incident angle, the desired signal originates from θ0 = 0°, a strong interference with INR = 30dB originates from θ1 = 40°, and a weak interference with INR = 10dB originates from θ2 = -40°. The desired signal is located in the angular sector Θ. s = [θ0-3°, θ0+3°], the interference signal is located at Θ i = [θ1-3°, θ1+3°]∪[θ2-3°, θ2+3°], the pure noise sector is Θ n .

[0119] Step two provided in this embodiment of the invention includes:

[0120] (1) Obtain the array receiving time-domain signal X(n), and the time-domain output of the m-th array element is:

[0121] X m (t)=S0(t-τ m (θ0))+N m (t)

[0122] Where S0(t) and τ m (θ0), N m (t) represents the desired signal waveform, the propagation delay of the m-th array element with respect to the corresponding direction θ0, and the noise and interference at the m-th sensor, respectively.

[0123] (2) Perform an L-point Fast Fourier Transform (FFT) on the time-domain snapshot X(t) to generate a frequency-domain snapshot. For the l-th FFT unit, the m-th array element, and the n-th frequency-domain snapshot, we can write:

[0124]

[0125] x m,l (n), s 0,l (n), n m,l (n) represent the frequencies fl = (l-1) / LT, respectively. s Fourier transform at T s For the sampling period, the array steering vector model of the center frequency signal can be written as:

[0126]

[0127] The nth frequency domain snapshot is written according to the signal structure as follows:

[0128]

[0129] (3) Calculation of cross-spectral density matrix (CSDM):

[0130]

[0131] (4) Calculation of the Steering Covariance Matrix (STCM):

[0132] For any angle of orientation, the covariance matrix is:

[0133]

[0134] The structure of STCM can be written as:

[0135]

[0136] in For the desired total signal power, Let be the power of the k-th interference signal at the l-th frequency. This represents the total power of the Gaussian white noise.

[0137] Where T l (θ) is the pre-steering matrix for the l-th frequency, in the form of:

[0138]

[0139] In step three of this embodiment of the invention, due to errors, the pre-steering matrix cannot perfectly rotate the desired steering vector at each frequency point to an all-1 vector. Here, the RCB method is used to correct the desired signal steering vector at each frequency point, correcting the pre-steering matrix, including:

[0140] (1) Solve the problem of steering vector estimation of the desired signal at each frequency point:

[0141]

[0142] in This represents prior knowledge, and is the hypothetical expected signal steering vector. l Let be the steering vector of the true desired signal to be estimated. The optimal solution to the optimization problem is in the form of:

[0143]

[0144] Where λ l To apply a diagonal loading factor λ obtained by the Lagrange multiplier method to the desired signal steering vector at the corresponding frequency, l To obtain it, follow these steps: First, Perform eigenvalue decomposition:

[0145]

[0146] Where U includes The eigenvectors are Γ, which is a diagonal matrix composed of the eigenvalues ​​of the corresponding eigenvectors, where the eigenvalues ​​are γ1≥γ2≥…≥γ. M .make:

[0147]

[0148] Diagonal loading factor λ l It can be solved by the following equation:

[0149]

[0150] z m Let be the m-th element of vector z. The solution to the equation must satisfy the following inequality:

[0151]

[0152] Since the magnitude of the corrected steering vector changes, affecting the construction of the pre-steering matrix, the estimated steering vector is normalized:

[0153]

[0154] (2) Construct the pre-guided matrix:

[0155] The pre-steering matrix at the l-th frequency point is constructed as follows:

[0156]

[0157] Step four provided in this embodiment of the invention includes:

[0158] (1) Use the pre-steering matrix obtained in step three to calculate the steering covariance matrix in the direction of the desired signal.

[0159]

[0160] (2) Calculate the STMV spectrum:

[0161]

[0162] Where 1 = [11...1] T The noise power and desired signal power are calculated using the estimated STMV spectrum. The integration in the algorithm is implemented using simulation. The sampling point is set to 2500 for each angular sector. The Gaussian white noise power for each sector is calculated. The estimate is:

[0163]

[0164] θ r The angle value is the sampling point. Desired signal power estimation. for:

[0165]

[0166] (3) After removing the desired signal component, the reconstructed STCM can be written as:

[0167]

[0168] Step five of this embodiment of the invention involves constructing an optimization problem to find the optimal steering vector solution. The post-guidance expected steering vector estimate for robust STMV is as follows:

[0169]

[0170] Optimization problem:

[0171]

[0172] subject to 1 H e ⊥ =0

[0173]

[0174] This optimization problem is a convex problem, which can be solved using the convex optimization toolbox.

[0175] Step six of this embodiment of the invention involves calculating the broadband adaptive beamforming weight vector, using the following formula:

[0176]

[0177] The robust broadband adaptive beamforming system based on pre-guided covariance matrix reconstruction provided in this invention includes:

[0178] The pre-guide matrix estimation module is used to re-estimate the pre-guide matrix using robust beamforming criteria to obtain a more accurate guide covariance matrix.

[0179] The STCM reconstruction module is used to obtain the desired signal power and noise power through STMV broadband spectrum estimation, and remove the desired signal part from the newly estimated steering covariance matrix;

[0180] The steering vector optimization module is used to further optimize the steering vector used for beamforming.

[0181] The weight vector calculation module is used to calculate the broadband adaptive beamforming weight vector.

[0182] The robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction provided in the application embodiment of the present invention is applied to a computer device. The computer device includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor performs the steps of the robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction.

[0183] The robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction provided in the application embodiment of the present invention is applied to an information data processing terminal, which is used to implement the robust broadband adaptive beamforming system based on pre-guided covariance matrix reconstruction.

[0184] The embodiments of the present invention have achieved some positive results during the research and development or use process, and have indeed great advantages compared with the prior art. The following content describes them in conjunction with the data, charts and other information of the experimental process.

[0185] To verify the performance of this invention, two sets of simulation experiments were designed under different error conditions. The Monte Carlo simulations were performed 500 times each. The interference-to-noise ratio (INR) was set to 30 dB. The algorithms compared to this were RSS unitary focus, AFT-REB, WBRCB, and ISM, respectively. When calculating the pre-steering matrix for each frequency point, ε was set to 0.3.

[0186] UnitaryFocus algorithm: MADoron,AJWeiss.On focusing matrices for wide-band array processing[J].IEEE Transactions on Signal Processing,1992,40(6):1295-1302

[0187] AFT-REB algorithm: Chen, P., Wang, W., & Gao, J. (2021, June). Focusing-BasedWideband Adaptive Beamforming Using Covariance Matrix Reconstruction. InICASSP 2021-2021IEEE International Conference on Acoustics, Speech and SignalProcessing (ICASSP) (pp.4510-4514). IEEE.

[0188] WBRCB algorithm: Somasundaram, SD (2012). Wideband robust capon beamforming for passive sonar. IEEE Journal of Oceanic Engineering, 38(2), 308-322.

[0189] ISM algorithm: BD Carlson, "Covariance matrix estimation errors and diagonal loading in adaptive arrays," IEEE Transactions on Aerospace and Electronicsystems, vol.24, no.4, pp.397–401, Apr.1988

[0190] The first set of simulation experiments was conducted under the condition of random incoming line-of-sight error. The line-of-sight error, with the expected signal and interference signal having an average DOA error distribution in the range of [-4°, 4°], was used to verify the SNR of different inputs. In the experiments verifying different snapshot numbers, the SNR was set to 30dB, and the changes in the array output SINR were observed. The simulation results are as follows: Figure 3 and Figure 4 As shown.

[0191] The second set of simulation experiments was conducted under the condition of array geometric errors. Errors in the antenna array geometry are typically modeled as sensor position errors, which are uniformly plotted from the wavelength-measured interval [-0.025, 0.025].

[0192] In the experiment verifying different input SNRs, the number of snapshots was set to 500, and the SNR simulation range was -30dB to 40dB. The changes in the array output SINR were observed. In the experiment verifying different number of snapshots, the SNR was set to 30dB, and the changes in the array output SINR were observed. The simulation results are as follows: Figure 5 and Figure 6 As shown.

[0193] It should be noted that embodiments of the present invention can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of the above-described hardware circuitry and software, such as firmware.

[0194] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction, characterized in that, The robustness of beamforming is improved by estimating the pre-guide matrix using a robust beamforming criterion; the desired signal power and noise power are obtained through STMV broadband spectrum estimation, and the desired signal component is removed from the newly estimated guide covariance matrix to further optimize the pre-guide matrix; the weights are further optimized by solving a quadratic constrained quadratic programming problem, which improves the performance of beamforming; and the broadband adaptive beamforming weight vector is calculated to achieve effective processing of broadband signals and improve the overall performance of beamforming. include: Step 1: Configure the antenna array; Step 2: Broadband signal frequency domain processing; Step 3: Pre-steering matrix estimation. The pre-steering matrix is ​​estimated using robust beamforming criteria. Step 4, STCM reconstruction: The desired signal power and noise power are obtained through STMV broadband spectrum estimation, and the desired signal part is removed from the newly estimated steering covariance matrix. Step 5: Optimize the steering vector used for beamforming by solving a quadratic constraint quadratic programming problem to obtain further optimized weights; Step 6: Calculate the broadband adaptive beamforming weight vector; In step one, a one-dimensional uniform linear array is set up, comprising: Each array element, array spacing ,set up A uniform linear array composed of omnidirectional elements; the interference is a linearly modulated frequency (LFM) signal; and the desired signal is generated by filtered Gaussian noise, where the noise power at each frequency may differ. The distance between two adjacent sensors is half the wavelength of the highest frequency. The signal bandwidth is 30MHz by default, the signal carrier center frequency range is 25GHz - 27GHz by default, the signal propagation speed is the speed of light in a vacuum, and the desired signal, with respect to the angle of incidence, originates from... A strong interference with an INR of 30 dB comes from A weak interference with INR=10 dB comes from The desired signal is located in the corner sector. The interference signal is located at Pure noise sector is .

2. The robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction according to claim 1, characterized in that, Step two includes: (1) Obtain the array receive time domain signal , No. The time-domain output of each array element is: in , , These represent the desired signal waveforms, respectively. Each array element corresponds to a direction The delay in propagation, and the first Noise and interference at each sensor; (2) Take a snapshot of the time domain conduct Point Fast Fourier Transform, generating frequency domain snapshot, for the ... The FFT cell, the first Each array element and the first A frequency domain snapshot can be written as follows: , , They represent the frequencies respectively. Fourier transform at the point, For the sampling period, the array steering vector model of the center frequency signal can be written as: No. Each frequency domain snapshot is written according to the signal structure as follows: (3) Calculation of cross-spectral density matrix (CSDM): (4) Calculation of the Steering Covariance Matrix (STCM): For any angle of orientation, the covariance matrix is: The structure of STCM can be written as: in For the desired total signal power, For the first The interference signal at the first Power at each frequency The total power of Gaussian white noise is given by , where For the first The pre-steering matrix for each frequency is in the form of: 。 3. The robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction according to claim 1, characterized in that, Step three involves using the RCB method to correct the desired signal steering vector at each frequency point and to correct the pre-steering matrix, including: (1) Solve the problem of estimating the steering vector of the desired signal at each frequency point: in It represents prior knowledge and is the expected signal steering vector of the hypothesis. The desired signal steering vector to be estimated; the optimal solution to the optimization problem is in the form of: in The diagonal loading factor obtained by the Lagrange multiplier method is used to guide the desired signal at the corresponding frequency. To obtain it, follow these steps: First, Perform eigenvalue decomposition: in Include eigenvectors, Let be a diagonal matrix consisting of the eigenvalues ​​of the corresponding eigenvectors, where the eigenvalues ​​are... ,make: Diagonal loading factor It can be solved by the following equation: For vectors The For a given set of elements, the solution to the equation must satisfy the following inequalities: Normalize the estimated steering vector: (2) Construct the pre-guidance matrix: In the The pre-guidance matrix for each frequency point is constructed as follows: 。 4. The robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction according to claim 1, characterized in that, Step four includes: (1) Use the pre-steering matrix obtained in step three to calculate the steering covariance matrix in the direction of the desired signal. : (2) Calculate the STMV spectrum: in The noise power and desired signal power are calculated using the estimated STMV spectrum. The integration in the algorithm is implemented using simulation, with a sampling point of 2500 for each angular sector. The Gaussian white noise power of each sector is then calculated. The estimate is: Given the angle values ​​at the sampling points, the desired signal power is estimated. for: (3) After removing the desired signal component, the reconstructed STCM can be written as: 。 5. The robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction according to claim 1, characterized in that, Step five involves constructing an optimization problem to find the optimal steering vector solution. The post-steering expected steering vector estimate for robust STMV is as follows: Optimization problem: This optimization problem is a convex problem, which can be solved using the convex optimization toolbox.

6. The robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction according to claim 1, characterized in that, Step six involves calculating the broadband adaptive beamforming weight vector, using the following formula: 。 7. A system for implementing the robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction as described in any one of claims 1-6, characterized in that, The system of the robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction includes: The pre-guide matrix estimation module is used to re-estimate the pre-guide matrix using robust beamforming criteria to obtain a more accurate guide covariance matrix. The STCM reconstruction module is used to obtain the desired signal power and noise power through STMV broadband spectrum estimation, and remove the desired signal part from the newly estimated steering covariance matrix; The steering vector optimization module is used to further optimize the steering vector used for beamforming. The weight vector calculation module is used to calculate the broadband adaptive beamforming weight vector.

8. A computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the robust broadband adaptive beamforming method based on pre-guided covariance matrix reconstruction as described in any one of claims 1-6.

Images