一种基于标准峭度的新型复数盲分离算法

在复值信号的盲分离算法中,经常采用信号的峭度最大化作为代价函数.以复数标准峭度代替复数峭度,将复数信号的标准峭度最大化作为新的代价函数,采用修正的复值拟牛顿迭代算法对代价函数进行优化,并运用该算法对混合QAM信号进行分离.仿真实验结果表明:改进后的算法具有很好的分离效果,相比于峭度最大化为代价函数的分离算法,收敛性能有明显提高.

A New Complex Blind Source Separation Algorithm Based on Standard Kurtosis

In the complex blind source separation algorithm， the complex signal kurtosis maximization is often used as the cost function. The complex standard kurtosis was used instead of complex kurtosis as the new cost function for optimimization， and a modified complex quasi-newton iterative algorithm was employed to optimize the cost function. The algorithm was applied to separate mixed QAM signal， and simulation results showed that the improved algorithm has a good separation effect. Compared with the algorithm of the kurtosis maximization as the cost function， the convergence performance was improved obviously.