摘要
对于自适应盲分离算法,选择步长参数以达到理想的分离性能是必要的。结合基于估计函数的可变步长自适应优化方法,提出了基于负熵的自适应算法,有效提高了算法的收敛速度,降低了算法性能对步长的依赖性。仿真实验证明了此改进算法具有较好的分离效果。
Careful selection of step size parameters is often necessary to obtain good performance from gradient-based adaptive algorithms for blind source separation.Combining an adaptive variable step-size based on the estimated function,a new adaptive algorithm based on negative entropy is given.Computer simulation results show that this improved adaptive algorithm has better separation efficiency and fast convergence.
出处
《电子科技》
2009年第12期83-87,共5页
Electronic Science and Technology
关键词
独立分量分析
盲源分离
可变步长
independent component analysis
blind signal processing
variable step-size
