瞬时混合盲信号分离问题的自适应算法比较

Comparison Among the Adaptive Algorithms of Instantaneous Mixed Blind Signal Separation

盲信号分离始于瞬时混合盲信号分离,本文通过论述盲信号分离的典型自适应算法——EAS(ILMS)算法和RLS算法,利用这些算法分别对合成信号进行实验,并对仿真结果进行分析比较,得到一些结论:EAS(ILMS)算法与RLS算法相比,其收敛性能更差,但稳态性能更好,实际应用时需选择合适的方法和参数。

Blind signal separation begins with the instantaneous mixed blind signal separation. In this paper,by introducing the typical adaptive algorithms of blind signal separation -- EASI （LMS） algorithm and RLS algorithm and by doing experiments on synthetic data, some conclusions is gotten by analysis and comparison of the simulation results. The result is that EASI（LMS） algorithm has worse convergence and better stability than RLS algorithm. Appropriate methods and parameters should be chosen in practical application.