摘要
独立分量分析方法无需完整的系统先验知识,可用于多天线移动通信系统多用户盲接收机。为了改善收敛速度,采用修正的具有三阶收敛速度的牛顿迭代法的近似公式,提出了一种改进的复FastICA算法。并将该算法应用于多输入多输出系统中进行仿真,结果表明新算法在与传统的复FastICA算法分离效果相当的情况下,迭代次数减少了16.5%,加快了收敛速度。
Independent component analysis can be used to be a blind receiver for multiple-input multiple-output (MIMO) mobile communication systems. ICA-based multi-user detectors do not require complete knowledge of the system like other traditional detectors. An improved CFICA algorithm was proposed. A new Newton’s iteration method was used which had three order of convergence to speedup the convergence speed. The simulation results of MIMO system which use the new algorithm show that the iterations of improved algorithm reduce 16.5% with the correspondent separation performance compared with the original algorithm.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2008年第12期3333-3335,共3页
Journal of System Simulation
关键词
独立分量分析
复FastICA
牛顿迭代法
多输入多输出(MIMO)
independent component analysis
FastICA of Complex Valued Signals
Newton’s iteration method
MIMO (Multiput-Input Multiple-Output)
作者简介
李文元(1964-),男,陕西岐山人,副教授,研究方向为通信信号处理,移动通信抗干扰等;
金凤(1979-),女,陕西西安人,研究方向为军事移动通信等。