摘要
提出一种信号子空间拟合的新方法。推导出信号子空间拟合一个新的基本公式,进而建立信号子空间拟合一个新的准则函数。分析表明,信号子空间可以由多级维纳滤波器的若干个匹配滤波器有效地构成,而且匹配滤波器只需要多级维纳滤波器的若干步前向递推,所以该方法具有小运算量和低复杂度的特点。最后,计算机仿真验证了该方法的有效性。
A new method for signal subspace fitting is presented. Firstly, a novel criterion of signal subspace fitting is derived. It is shown that the correct knowledge of the number of signals is unnecessary since the columns of the signal subspace can be much lower than the number of signals. The proposed method does not resort to the estimate of the covariance matrix or its eigendecomposition, thereby implying much lower computational complexity. Finally, the effectiveness of the proposed approach is verified by numerical results.
出处
《电子与信息学报》
EI
CSCD
北大核心
2005年第8期1197-1200,共4页
Journal of Electronics & Information Technology
基金
国家自然科学基金(60102005)资助课题
关键词
波达方向
信号子空间拟合
降维
多级维纳滤波器
Direction Of Arrival （DOA）, Signal subspace fitting, Rank reduction, Multi-Stage Wiener Filter （MSWF）
