一种实域空间平滑解相干算法

A Real Domain Spatial Smoothing Decorrelation Algorithm

特征子空间类算法在处理相干信号时,需要预先进行空间平滑来解相干,这一过程通常是在复数域进行的,而且一次平滑只能处理一个相干信号。论文提出实域空间平滑算法从窄带信号的正交接收模型出发,运用三角和差化积公式对I/Q通道输出单独进行简单加、减运算,得到了实值阵列流形,在此基础上采用前向空间平滑进行解相干,一次平滑能够处理两个相干信号,且算法完全在实域进行。以MUSIC为例进行的仿真试验表明:算法能够取得与FBSS算法相似的性能,且运算量较之FBSS算法大大降低。

Eigen-subspace algorithms have to take spatial smoothing to do decorrelation when deal with coherent signals,which always been taken in the complex domain,and one decorrelation operation only can deal with one coherent signal.Based on orthogonal receiver model,this paper took addition and subtraction operations on output of channel I and Q to get real array manifold through triangle formulas,then forward space smoothing was used for decorrelation,and one decorrelation operation could deal with two coherent signals,which took place in real domain.Simulations for this method by MUSIC indicated this method had similar performance of FBSS algorithm,and computational complexity of the method was reduced greatly.