Research and applications of blind signal separation based on kurtosis
Gexiang Zhang · Chinese Journal of Radio Science · 2008
Complex signal separation and detection is always an important and difficult issue.In this paper,an improved blind signal separation method was proposed.Based on independent component analysis(ICA),kurtosis's absolute value function is used as target function,and an improved Newton algorithm is used to look for the most solution for the target function.The effect of correlativity and change parameter of signal on separation result of signal is analyzed in detail.Theoretical analysis and experimental results show that the method is of good performance in the radar signal separation and detection.