一种非理想阵列的快速宽带信号方位估计算法

针对宽带信号方向估计算法需要进行初次估计和进行多次奇异值分解的问题,提出了在传感器阵列存在误差的情况下,基于Prony-Lanczos方法的快速宽带信号方位估计算法。方法将接收信号转换到频率域内,分离阵列流形中信号方向和阵列参数;其次,在考虑阵列互耦的情况下,采用快速奇异值分解的方法求取奇异向量,并利用奇异向量构造聚焦矩阵;最后,利用聚焦求和窄带方位估计对宽带信号进行方法估计。仿真试验表明,存在阵列误差的情况下,本算法具有较好的分辨率和较低的计算复杂度和鲁棒性。

Fast Algorithm for Direction of Arrival Estimation of Wideband Signal Based on Nonideal Array

Due to the need of preliminary angle estimates and large numbers of singular value decompositions(SVD)in wideband direction of arrival(DOA)estimation,a new fast wideband signal DOA estimates algorithm based on Prony-Lanczos method was proposed in the presence of the nonideal arbitrary sensor array and the mutual coupling between array elements.In the algorithm,firstly,the received wideband signal was transformed into frequency domain,and the DOA and the others array parameters were separated in the array matrix manifold.Secondly,the singular vectors of the array manifold were derived by fast SVD based on Prony-Lanczos method in the condition of errors of the actual positions,gains and phases.Moreover,the singular vectors constructed the focusing matrices.Lastly,the DOAs of the wideband signals were estimated by sum of focusing matrices and narrow estimation method.The computer simulation results illustrate that the new algorithm has low computational complexity and the system stability.