基于多级维纳滤波的ESPRIT算法  

ESPRIT Algorithm Baesd on Multi-stage Winner Filtering

作　　者：张超[1] 黎仁刚[1] 顾军[1] 

机构地区：[1]中国船舶重工集团公司第七二三研究所,江苏扬州225001

出　　处：《舰船电子对抗》2017年第1期74-78,84,共6页Shipboard Electronic Countermeasure

摘　　要：因奇异值分解引入了庞大的计算量,子空间分解类的测向算法在工程中难以实现实时处理。多级维纳滤波(MSWF)算法,避免了奇异值分解运算,有效地减少了运算量。通过对多级维纳滤波算法进行改进,提出了一种信源数目的估计方法。与传统多级维纳滤波算法相比,该方法提高了测向精度。通过仿真证明了该算法的有效性。Because of the enormous calculating quantity introduced by singular value decomposition, the direction finding algorithm based on subspace decomposition is difficult to be realized in engi- neering in real time. The multi-stage Wiener filtering （MSWF） algorithm avoids the singular value decomposition calculation,reduces the calculating quantity effectively. This paper puts forward an estimation method of source number by improving MSWF algorithm. The algorithm improves the direction-finding accuracy compared with traditional MSWF algorithms. The validity of this algo- rithm is validated through simulation.

关 键 词：阵列信号处理 多级维纳滤波 信源数目估计 测向精度 

分 类 号：TN971.1[电子电信—信号与信息处理]