Blind Separation Algorithm Based on Covariance Matrix
Xie Shengli · Jisuanji gongcheng · 2003
A novel linear mixture blind separation algorithm is proposed in this paper, which is based on the character that the covariance m atrix is diagonal matrix when signals are independent, and it is used as a separation contrast. The restriction of this algorithm is not too much and the computation is not large. Theory analysis and simulation results show that the separation effect is very good.