重构分数低阶协方差的子空间拟合测向算法  

Direction finding of signal subspace fitting algorithm based on reconstructed fractional lower order covariance

作　　者：高洪元[1] 刁鸣[1] 

机构地区：[1]哈尔滨工程大学信息与通信工程学院,黑龙江哈尔滨150001

出　　处：《电波科学学报》2009年第4期729-734,共6页Chinese Journal of Radio Science

基　　金：黑龙江省科技攻关项目(GZ08A101)

摘　　要：基于重构的分数低阶矩阵,提出了重构分数低阶协方差的多重信号分类测向算法和信号子空间拟合测向算法。为了快速求解所提出的测向算法,设计了一种可进行多维搜索的自适应差分粒子群优化算法。利用粒子群算法和差分进化算法的优点,可以获得测向问题的全局最优解。Monte-Carlo仿真证明了所提测向算法可有效分辨相干源,并且其检测性能优于已有的一些经典算法。Based on reconstructed fractional lower order covariance（RFLOC）matrix, a RFLOC-multiple signal classification（MUSIC）algorithm and a RFLOC-signal subspace fitting （SSF）algorithm are proposed for direction finding. In order to quickly implement the proposed algorithms,an adaptive differential particle swarm optimization（ADPSO）algorithm is designed to search multi-dimensional optimal solution. The ADPSO is a global optimization algorithm for direction finding,which takes advantage of the merits of differential evolutionary algorithm and particle swarm algorithm. It proves that the proposed algorithms not only can find coherent sources, but also are superior to some previous classical algorithms in performance.

关 键 词：分数低阶协方差 信号子空间拟合 MUSIC算法 粒子群优化算法 差分进 化算法 

分 类 号：TN911.72[电子电信—通信与信息系统] TP301.6[电子电信—信息与通信工程]