基于阵列结构的盲源分离算法

利用源信号的统计独立性提出了一种由混合信号和分离信号估计混合矩阵的算法,避免了一次矩阵求逆运算。同时,利用天线阵列的结构信息提出了一种直接重构混合矩阵的算法,该算法不需要估计源信号的载频和入射角。仿真结果表明,所提算法与基于来波方向估计的盲分离算法性能相当,在信号入射角间隔较大时,所提算法性能优于普通的盲分离算法,并且大大减少了运算量。

A Blind Source Separation Method Based on the Characteristic of Antenna Array

Based on the independency of source signals,an algorithm to estimate the mixing matrix with the mixed signals and the separated signals is proposed,in which one operation of inverse matrix is omitted.An algorithm to directly reconstruct the mixing matrix is also proposed according to the characteristic of the antenna array.The proposed algorithm does not need to estimate the carrier frequencies and the arrival directions of the source signals.Simulation results show that the proposed method can achieve mostly same performance as the blind source separation method based on the estimation of the arrival direction and perform better than the ordinary BSS methods when the intervals between the arrival directions of the source signals are comparatively large.Besides,the computational complexity is decreased greatly.