极化域—空域联合的参数估计和干扰抑制方法研究

发布时间：2018-03-20 05:12

本文选题：极化域-空域　切入点：参数估计　出处：《电子科技大学》2016年硕士论文　论文类型：学位论文

【摘要】：相较于传统的标量阵列,极化阵列最大的优势是可以利用信号的极化信息,进一步提高参数估计精度和干扰抑制性能。但目前对极化阵列信号处理的方法大多是传统标量阵列信号处理方法的直接搬移,并没有太多针对极化阵列特殊结构的研究,且大部分的工作是进一步研究提高极化矢量阵列信号处理性能的方法。针对这一问题,本文将对极化域-空域联合的参数估计和波束形成方法进行研究,并基于四元数模型充分利用极化阵元分量间的正交结构,研究高效低复杂度的参数估计和干扰抑制新方法,并采用计算机仿真实验来加以验证。针对相干信号源的参数估计问题,基于四元数模型给出四元数空间平滑算法和四元数Toeplitz矩阵重构算法,在此基础上提出改进的四元数空间平滑算法和改进的四元数矢量重构算法,并给出两种算法实现解相干的推导过程。仿真实验结果表明,改进的解相干算法具有更好的参数估计性能和更好的空间分辨率,且矢量矩阵重构算法不以损失阵列有效孔径为代价,与空间平滑算法相比,可以实现更多相干信号的参数估计。针对非相干信号源的参数估计问题,由于极化信息的引入,阵列信号处理的维数从纯空域转变到极化域-空域的联合处理,导致算法的计算量明显增加。针对这一问题,研究降维MUSIC算法,在充分应用极化域信息和空域信息的前提下,将极化信息和空域信息剥离,分别采用MUSIC算法得到DOA估计和极化参数估计,在一定程度上降低算法的运算量。在此基础上,提出极化域-空域联合的模值约束降维Root-MUSIC算法,先采用Root-MUSIC算法求出波达方向角,再根据模值约束条件建立优化函数,通过闭合式解求得极化参数。将该算法应用于极化均匀线阵和L型阵列,通过对计算量的对比分析和仿真实验,验证该算法的优越性能。针对极化域-空域联合滤波,介绍联合滤波的原理和滤波性能的分析,给出最大输出信干噪比的表达式。基于特征子空间的思想,提出四元数投影波束形成算法,并给出幅相误差、耦合误差和位置误差的建模;在理想情况和存在误差的情况下,分析该算法的滤波性能。仿真实验对比四元数投影波束形成算法和四元数MVDR算法在理想情况和误差情况下的滤波性能,表明四元数投影波束形成算法的滤波性能优于四元数MVDR算法,能够在一定程度上减小误差带来的影响。
[Abstract]:Compared with the traditional scalar array, the biggest advantage of the polarization array is that it can make use of the polarization information of the signal. The precision of parameter estimation and the performance of interference suppression are further improved. However, at present, most of the signal processing methods of polarimetric array are the direct shift of the traditional scalar array signal processing methods, and there is not much research on the special structure of the polarized array. And most of the work is to further study the method to improve the signal processing performance of polarization vector array. In view of this problem, this paper will study the parameter estimation and beamforming method of polarization-spatial joint. Based on the quaternion model, a new method of parameter estimation and interference suppression with high efficiency and low complexity is studied by making full use of the orthogonal structure between the components of polarization array. To solve the parameter estimation problem of coherent signal source, a quaternion spatial smoothing algorithm and a quaternion Toeplitz matrix reconstruction algorithm are presented based on quaternion model. On this basis, an improved quaternion space smoothing algorithm and an improved quaternion vector reconstruction algorithm are proposed. The improved decoherence algorithm has better parameter estimation performance and better spatial resolution, and the vector matrix reconstruction algorithm does not take the loss of effective aperture of the array as the cost, compared with the spatial smoothing algorithm. For the parameter estimation of incoherent signal sources, the dimension of array signal processing changes from pure spatial domain to joint processing in polarization-spatial domain due to the introduction of polarization information. In view of this problem, the dimensionality reduction MUSIC algorithm is studied, and the polarimetric information and spatial information are separated under the premise of fully applying the polarization domain information and the spatial domain information. DOA estimation and polarization parameter estimation are obtained by using MUSIC algorithm respectively, which reduces the computational complexity of the algorithm to a certain extent. On this basis, a modular constraint reduced dimension Root-MUSIC algorithm based on polarization domain and spatial domain is proposed. First, the DOA of arrival is obtained by using Root-MUSIC algorithm. Then the optimization function is established according to the mode constraint condition, and the polarization parameters are obtained by the closed solution. The algorithm is applied to the polarization uniform linear array and the L-shaped array. Through the comparison and analysis of the calculation amount and the simulation experiment, the algorithm is applied to the polarization uniform linear array and the L-shaped array. The principle of joint filtering and the analysis of filtering performance are introduced, and the expression of maximum output signal-to-noise ratio is given, based on the idea of feature subspace. A quaternion projection beamforming algorithm is proposed, and the modeling of amplitude and phase error, coupling error and position error is given. The filtering performance of this algorithm is analyzed. The filtering performance of quaternion projection beamforming algorithm and quaternion MVDR algorithm is compared with that of quaternion MVDR algorithm under ideal and error conditions. It is shown that the filtering performance of the quaternion projection beamforming algorithm is better than that of the quaternion MVDR algorithm and can reduce the influence of the error to a certain extent.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN911.7

【参考文献】