稳健的自适应波束形成算法研究
发布时间：2019-03-08 09:10
【摘要】：自适应波束形成(Beamforming)是阵列信号处理领域中的一个重要研究方向,其利用空间多传感器阵列所构成的处理系统对空间信号进行发射或接收,在雷达、声纳、通信系统、智能家电以及智能会议系统中有着广泛的应用。但是,自适应波束形成算法的性能依赖于对入射信号、阵列以及环境的假设,对假设的准确性非常敏感。在实际的工程应用中,误差总是存在的,因此研究对误差稳健的波束形成算法很有必要。 目前比较流行的稳健算法主要分为三类：特征子空间类、对角加载类以及凸优化类。第一类算法在低信噪比条件下性能损失严重,并且需要已知信号源的数目；第二类算法的缺点主要在于加载因子和实际误差的上下限没有可靠的确定方法；最后一类算法是近几年研究得最多的,这类算法的提出使波束形成算法的性能得到了很大的提升,但仍然存在一些缺陷,比如对阵元位置以及多径等误差比较敏感。本文的研究主要集中在特征子空间类和凸优化类算法上,针对实际应用中存在的误差提出了三种稳健的波束形成算法,主要内容和创新点如下： 1.针对多约束线性约束最小方差算法输出信干噪比损失的问题,提出了基于复数约束的线性约束最小方差稳健波束形成算法。算法采用复数约束,并且约束是一个变量,最优的约束值可以在最大信噪比准则下求解得到,文中对算法进行了详细的推导。通过计算机仿真实验验证了该方法的有效性,算法有良好的输出性能,并且计算复杂度和常规的稳健波束形成算法相当。 2.协方差矩阵估计是波束形成算法需要解决的关键问题之一,本文针对低快拍条件下,协方差矩阵估计误差较大的问题,提出了一种基于先验知识的稳健自适应波束形成算法,这种算法先利用矩阵重构的方法构造出不包含期望信号的协方差矩阵,再利用采样数据协方差矩阵来联合估计理论最优的协方差矩阵。对比实验表明该算法在低快拍条件下具有良好的输出性能。 3.目前有学者提出了一种迭代的稳健波束形成算法,但当干扰的干噪比比期望信号的输入信噪比大时,算法可能收敛到干扰的方向。针对这一问题,本文提出了一种新算法,算法采用一种新的信号子空间估计方法,无需已知信号源的数目,并且在低信噪比条件仍然有效。用此方法可以得到干扰加噪声投影矩阵,然后利用期望信号与干扰加噪声空间的正交性对期望信号导向矢量进行估计,避免算法收敛到干扰方向。在迭代过程中,采用放宽的约束,从而保证期望信号在约束空间中,避免算法收敛不到最优解。最后通过仿真实验,验证了所提算法的有效性,并对算法进行了对比分析和总结。
[Abstract]:Adaptive beamforming (Beamforming) is an important research direction in the field of array signal processing. It uses a processing system composed of spatial multi-sensor arrays to transmit or receive space signals in radar, sonar and communication systems. Intelligent home appliances and intelligent conference systems have a wide range of applications. However, the performance of adaptive beamforming algorithm depends on the assumption of incident signal, array and environment, and is very sensitive to the accuracy of hypothesis. In practical engineering applications, the error always exists, so it is necessary to study the error robust beamforming algorithm. At present, the popular robust algorithms are divided into three categories: feature subspace class, diagonal loading class and convex optimization class. Under the condition of low signal-to-noise ratio (SNR), the first kind of algorithm has serious performance loss and needs the number of known signal sources, and the disadvantage of the second type algorithm is that there is no reliable method to determine the upper and lower bound of loading factor and actual error. The last kind of algorithm has been studied most in recent years. The performance of beamforming algorithm has been greatly improved by this kind of algorithm, but there are still some shortcomings, such as the error of element position and multipath is sensitive. In this paper, we mainly focus on the feature subspace class and convex optimization algorithm, and propose three robust beamforming algorithms for the errors existing in practical applications. The main contents and innovations are as follows: 1. A linear constrained minimum variance robust beamforming algorithm based on complex constraints is proposed to solve the problem of output signal-to-noise ratio loss of multi-constraint linear constrained minimum variance algorithm. The complex constraint is used in the algorithm, and the constraint is a variable. The optimal constraint value can be obtained under the maximum signal-to-noise ratio criterion. In this paper, the algorithm is deduced in detail. The effectiveness of the proposed method is verified by computer simulation. The algorithm has good output performance and the computational complexity is similar to that of the conventional robust beamforming algorithm. 2. Covariance matrix estimation is one of the key problems to be solved in beamforming algorithm. In this paper, a robust adaptive beamforming algorithm based on prior knowledge is proposed to solve the problem of large error of covariance matrix estimation under the condition of low fast beat. In this algorithm, the covariance matrix without the desired signal is constructed by the method of matrix reconstruction, and then the covariance matrix of sampled data is used to jointly estimate the theoretical optimal covariance matrix. The experimental results show that the proposed algorithm has good output performance under the condition of low speed beat. 3. At present, an iterative robust beamforming algorithm is proposed. However, when the interference noise ratio is larger than the expected signal-to-noise ratio, the algorithm may converge to the direction of interference. In order to solve this problem, a new algorithm is proposed in this paper. The algorithm adopts a new method of signal subspace estimation without the number of known signal sources, and is still valid under low SNR conditions. The projection matrix of interference plus noise can be obtained by this method, and then the guidance vector of desired signal can be estimated by using the orthogonality of expected signal and interference plus noise space to avoid convergence to the direction of interference. In the iterative process, a relaxed constraint is used to ensure that the desired signal is in the constraint space, so that the algorithm does not converge to the optimal solution. Finally, the validity of the proposed algorithm is verified by simulation experiments, and the algorithm is compared and summarized.
【学位授予单位】：中国科学技术大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.7
[Abstract]:Adaptive beamforming (Beamforming) is an important research direction in the field of array signal processing. It uses a processing system composed of spatial multi-sensor arrays to transmit or receive space signals in radar, sonar and communication systems. Intelligent home appliances and intelligent conference systems have a wide range of applications. However, the performance of adaptive beamforming algorithm depends on the assumption of incident signal, array and environment, and is very sensitive to the accuracy of hypothesis. In practical engineering applications, the error always exists, so it is necessary to study the error robust beamforming algorithm. At present, the popular robust algorithms are divided into three categories: feature subspace class, diagonal loading class and convex optimization class. Under the condition of low signal-to-noise ratio (SNR), the first kind of algorithm has serious performance loss and needs the number of known signal sources, and the disadvantage of the second type algorithm is that there is no reliable method to determine the upper and lower bound of loading factor and actual error. The last kind of algorithm has been studied most in recent years. The performance of beamforming algorithm has been greatly improved by this kind of algorithm, but there are still some shortcomings, such as the error of element position and multipath is sensitive. In this paper, we mainly focus on the feature subspace class and convex optimization algorithm, and propose three robust beamforming algorithms for the errors existing in practical applications. The main contents and innovations are as follows: 1. A linear constrained minimum variance robust beamforming algorithm based on complex constraints is proposed to solve the problem of output signal-to-noise ratio loss of multi-constraint linear constrained minimum variance algorithm. The complex constraint is used in the algorithm, and the constraint is a variable. The optimal constraint value can be obtained under the maximum signal-to-noise ratio criterion. In this paper, the algorithm is deduced in detail. The effectiveness of the proposed method is verified by computer simulation. The algorithm has good output performance and the computational complexity is similar to that of the conventional robust beamforming algorithm. 2. Covariance matrix estimation is one of the key problems to be solved in beamforming algorithm. In this paper, a robust adaptive beamforming algorithm based on prior knowledge is proposed to solve the problem of large error of covariance matrix estimation under the condition of low fast beat. In this algorithm, the covariance matrix without the desired signal is constructed by the method of matrix reconstruction, and then the covariance matrix of sampled data is used to jointly estimate the theoretical optimal covariance matrix. The experimental results show that the proposed algorithm has good output performance under the condition of low speed beat. 3. At present, an iterative robust beamforming algorithm is proposed. However, when the interference noise ratio is larger than the expected signal-to-noise ratio, the algorithm may converge to the direction of interference. In order to solve this problem, a new algorithm is proposed in this paper. The algorithm adopts a new method of signal subspace estimation without the number of known signal sources, and is still valid under low SNR conditions. The projection matrix of interference plus noise can be obtained by this method, and then the guidance vector of desired signal can be estimated by using the orthogonality of expected signal and interference plus noise space to avoid convergence to the direction of interference. In the iterative process, a relaxed constraint is used to ensure that the desired signal is in the constraint space, so that the algorithm does not converge to the optimal solution. Finally, the validity of the proposed algorithm is verified by simulation experiments, and the algorithm is compared and summarized.
【学位授予单位】：中国科学技术大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.7
【共引文献】
相关期刊论文 前10条
1 王登伟;吕英华;张博;刘兵;;基于光纤无线电的优化自适应波束形成算法[J];半导体光电;2007年05期
2 赵红;刘桥;;一种低旁瓣的子阵级数字波束形成方法[J];兵工自动化;2009年11期
3 杨小明;陶然;;基于分数阶Fourier变换和ESPRIT算法的LFM信号2D波达方向估计[J];兵工学报;2007年12期
4 梁国龙;张锴;付进;张瑶;李利;;单矢量水听器的高分辨方位估计应用研究[J];兵工学报;2011年08期
5 卢海杰;章新华;熊鑫;;流形分离在非均匀圆阵上的应用[J];兵工学报;2011年09期
6 谭渊;杨勇;袁乃昌;;超分辨谱估计法在测速雷达中的应用[J];兵工学报;2011年11期
7 卞红雨;王s，
本文编号：2436656
本文链接：https://www.wllwen.com/kejilunwen/wltx/2436656.html
