基于非正交联合对角化的盲波束形成

Blind Beamforming Based on Non-orthogonal Joint Diagonalization

该文提出了一种基于非正交联合对角化的盲波束形成算法。该算法不需要预白化,直接对阵列接收信号的四阶累积量矩阵进行联合对角化,得到非正交的分离矩阵,避免了由白化引入的误差。仿真实验表明,新算法在低信噪比下相比于JADE算法有更好的分离效果,而且该算法还可以通过增加采样数据点长度来提高分离性能。

A blind beamforming algorithm based on non - orthogonal joint diagonalization is proposed, which jointly diagonalization the fourth - order eumulant matrix of the received signals directly to obtain non - orthogonal separation matrix without pre - whitening, thus exeluding errors induced by whitening. Simulations show that the proposed algorithm outperforms JADE algorithm under low signal to noise ratios. Increasing the number of data can improve the performance of the algorithm further.