摘要
针对通信信号的特点,着重研究了适合于通信信号盲源分离的联合对角化方法.首先研究了两种自适应盲源分离算法——自然梯度算法和FastICA算法,分析了算法在处理通信信号时存在分离性能低和收敛速度慢等问题的原因;然后用非正交联合对角化方法进行改进,给出了一种基于非正交联合对角化(FAJD)的盲源分离方法.仿真结果表明改进的算法在低信噪比条件下的分离性能有明显的提高.
Based on the characteristic of communication signals,the paper emphatically investigated the joint diagonalization method,which was suitable for the sepatation of communication signals.Two adaptive blind source separation algorithms were studied firstly and the problem when aplied to communication signals separation was analyzed.Then,we realized the improved algorithm based on non-orthogonal joint diagonalization of second order statistics,which imporved the performance especially in noisy environment.Simulation results shown effectiveness of the improved algorithm.
出处
《陕西科技大学学报（自然科学版）》
2011年第4期77-81,共5页
Journal of Shaanxi University of Science & Technology
关键词
盲源分离
二阶统计量
联合对角化
blind source separation（BSS）
second order statistics
joint diagonalization
