摘要
提出了一种基于主分量分析(PCA)的CSM类宽带DOA估计自聚焦算法,利用子空间投影变换将信号分离后应用PCA算法快速估计信号DOA,通过不断更新聚焦方向实现自聚焦。与已有算法相比,该算法不受DOA初始值的影响,有更好的聚焦精度。聚焦矩阵更新过程中无需再做奇异值分解,用PCA迭代算法替代特征分解过程,计算量小。仿真实验结果表明,该算法以较小的计算代价实现了较好的估计精度。
A CSM based auto-focusing algorithm for wideband DOA estimation using principal component analysis （PCA） is proposed. Signals are separated by means of subspace projection and iterative algorithm based on PCA is used to estimate its DOA. While compared with the previous works, it is more robust to initial DOA estimation errors. Since no more SVD is needed during iteration and PCA is used instead of eigendecomposition, the computational burden is reduced. The experimental results show that proposed method has better performance and less computation.
出处
《计算机工程与设计》
CSCD
北大核心
2008年第22期5751-5754,共4页
Computer Engineering and Design
关键词
波达方向估计
宽带信号
主分量分析
相关子空间法
自聚焦
DOA estimation
wideband signal
principal component analysis
coherent signal-subspace method
auto-focusing
