提出了一种非圆相干信号完全解相关算法。不同于传统阵列信号中每次快拍形成一组数据矢量,该算法每次快拍形成一组数据矩阵。通过形成的数据矩阵,在矩阵内部实现虚拟平滑以达到解相干目的,通过数学证明,有该数据矩阵形成的信号相关矩阵可以等效成入射端信号为独立源时形成的信号相关矩阵,这样使得相干源之间得到完全解相关。该方法无孔径损失,可检测N-1个相干源(N为阵元个数)。计算机仿真验证了该方法的有效性和优越性。
A new de correlation algorithm is presented to estimate the direction-of-arrival of non-circular coherent signals. Unlike the data vector in conventional array signal processing, a data matrix is formed for each snapshot. The correlation among inducing coherent signals is resolved by the virtual spatial smoothing in the data matrix which is formed for each snapshot. Analysis shows that the signal covariance matrix formed by data matrix is equivalent to that of independent inducing signals, thus the correlation among coherent signals can be completely de correlated. What is more, N-1 coherent signals can be detected without aperture loss (with N representing the number of sensor elements). Computer simulation results are provided to verify the validity and superiority.