一种新型的基于ICA的盲源分离的多用户检测算法  

New Type of Multi-user Detection Algorithm Based on ICA Blind Source Seperation

作　　者：谢红 赵淼[1] 

机构地区：[1]哈尔滨工程大学,黑龙江哈尔滨150001

出　　处：《现代电子技术》2008年第4期124-125,共2页Modern Electronics Technique

摘　　要：提出一种基于独立的成份分析(ICA)的多用户检测方案,独立的成分分析只依靠时信号源独立性和非高斯性的假设,与传统独立成分分析方法相比更具灵活性和鲁棒性,实现了目标用户信号的提取。仿真实验表明,该算法的误码率性能在用户数量增大和远近效应严重的情况下都能较为理想地实现检测功能。This paper proposes a composition based on an Independent Component Analysis （ICA） of the multi-user detection program. Independent component analysis of the signal source only rely on the independence and the assumption of non-Gaussian,compared with the traditional independent component analysis ,it＇s more flexible and robust and can achieve the target user signal extraction. Simulation results show the algorithm performance and increase in the number of the users proximity effect more severe cases can achieve better detection function.

关 键 词：码分多址 独立分量分析 匹配滤波 盲多用户检测 

分 类 号：TN911[电子电信—通信与信息系统]