基于矩阵填充的互质阵列欠定DOA估计方法

发布时间：2018-06-03 04:51

本文选题：阵列信号处理 + 欠定DOA估计　；参考：《工程科学与技术》2017年05期

【摘要】：为了解决现有基于互质阵列的DOA估计方法舍弃差联合阵列中非均匀虚拟阵元而导致最大可估计信号数损失的问题,提出了一种基于矩阵填充的DOA估计方法。首先,根据差联合阵列与波程差一一对应的特性,构造一个部分元素缺失的Toeplitz化的阵列协方差矩阵,建立了基于矩阵填充的DOA估计模型,并验证了该模型满足零空间性质;然后,根据低秩矩阵填充理论,将DOA估计问题转化为矩阵核范数最小化问题进行求解,通过不定点延续算法将该协方差矩阵中的零元素进行填充恢复为完整协方差矩阵;最后,对协方差矩阵进行奇异值分解,转化为多项式求根,得到DOA的估计。仿真实验结果验证了本文方法的有效性和优越性。实验结果表明,本文方法能够对差联合阵列中的空洞部分进行有效填充,增加了可利用的阵列自由度,提高了可估计信号数,同时能够有效避免传统稀疏重构算法中由于角度域离散化导致的基不匹配问题对估计性能的影响,提高了估计精度和分辨力。
[Abstract]:In order to solve the problem that the existing DOA estimation methods based on the mutual-mass array abandon the non-uniform virtual array elements of the differential joint array and result in the loss of the maximum estimable signal number, a DOA estimation method based on matrix filling is proposed. Firstly, according to the one-to-one correspondence between the difference joint array and the wave path difference, a partial element missing Toeplitz array covariance matrix is constructed, and the DOA estimation model based on matrix filling is established, and the zero space property of the model is verified. Then, according to the filling theory of low rank matrix, the DOA estimation problem is transformed into a matrix kernel norm minimization problem to be solved, and the zero element in the covariance matrix is filled into the complete covariance matrix by the unfixed point continuation algorithm. The singular value decomposition of covariance matrix is carried out, which is transformed into polynomial to find the root, and the estimate of DOA is obtained. Simulation results show the effectiveness and superiority of the proposed method. The experimental results show that the proposed method can effectively fill the voids in the differential joint array, increase the available degree of freedom of the array, and increase the number of estimable signals. At the same time, it can effectively avoid the influence of the base mismatch caused by the angle domain discretization in the traditional sparse reconstruction algorithm, and improve the estimation accuracy and resolution.
【作者单位】：解放军电子工程学院网络系;
【基金】：国家自然科学基金资助项目(61171170) 安徽省自然科学基金资助项目(1408085QF115)
【分类号】：TN911.7

【相似文献】