基于延时相关处理的ESPRIT算法

摘要

为了提高旋转不变子空间(estimation of signal parameters via rotational invariance techniques, ESPRIT)算法的分辨力和测角精度,充分利用非零延迟相关函数中信号入射角度的信息,提出了基于延时相关处理的ESPRIT算法.根据所有阵列间延时相关信息,构造新的阵列输出矩阵,并且得到新的协方差矩阵.对新的协方差矩阵进行特征值分解得到特征向量,通过将特征向量划分得到含有入射角度信息的子阵,最终求得信源的入射角度.仿真结果表明,该算法的分辨力和测角精度均优于原ESPRIT算法,并且在小角度间距情况下也有较好的分辨性能.%In order to improve the resolution performance and angular precision of the estimation of signal parameters via rotational invariance techniques (ESPRIT) algorithm and make full use of the information of the directions of arrival (DOA) contained in the delay correlation functions, an ESPRIT algorithm based on delay correlation is proposed. According to the delay correlation information among the arrays, a new array output matrix is presented, and a new covariance matrix is obtained. The submatrices that contain the information of the DOA are acquired by dividing the eigenvectors, which could be determined by the eigenvalue decomposition of the new covariance matrix. Finally, the DOA could be obtained. Simulation results show that the resolution performance and angular precision of the proposed algorithm are better than the ESPRIT algorithm, and has better resolution performance at the condition of small angle spacing.