Performance of conventional beamforming algorithms used to separate sources decreases when the injecting direction of interesting signal and interference is close. In this paper a high angular resolution sources separation algorithm is proposed to solve this problem. An array extrapolation way based on auto-regressive model which fits for all arrays is derived first, and then a robust separation algorithm of independent component analysis is presented. Combined with the strong points of array extrapolation and blind signal separation, a high angular resolution blind separation algorithm is brought forward and this algorithm is able to get higher resolution for the array without adding actual antennas. The real data testing reveals that the proposed algorithm is able to separate the signals as close as 0.5 degree on 9 elements uniform circular array, which has high angular resolution.%传统的波束形成算法在期望信号和干扰信号来向相近的情况下性能下降,不能很好地完成信源分离,针对此问题提出了一种高角度分辨率的信源分离算法.首先选取一种适合任意阵型的AR模型预测方法进行阵列扩展,然后利用基于独立分量分析的鲁棒性分离算法完成信号分离,通过结合阵列扩展和盲源分离的优点,本文提出的分离算法在不增加实际阵元数目的前提下具备比原阵列更高的角度分辨率.实际数据测试表明提出的算法能够在9元均匀圆阵上分离方位角来向间隔为0.5度的信号.
展开▼
