一种时变信道下基于粒子滤波的同频混合信号盲分离方法  

Blind Separation Algorithm of Co-Frequency Modulated Signals Based on Particle Filtering in Time-Varying Channel

作　　者：马欢[1] 江桦[1] 

机构地区：[1]信息工程大学,河南郑州450001

出　　处：《信息工程大学学报》2014年第2期181-186,共6页Journal of Information Engineering University

摘　　要：文章建立了时变信道下同频混合信号的基本模型,提出了利用粒子滤波的混合信号调制参数估计与盲分离算法。通过将混合信号的时变调制参数建模为一阶AR模型,重点推导了参数与符号序列的粒子更新公式,确定了粒子的权值更新方法,给出了相应的后验概率密度估计。实验仿真表明,文章算法能够较好地实现混合信号的调制参数估计和盲分离。A basic model is established for mixtures of two co-frequency modulated signals in time- varying channels, and an algorithm based on particle filtering is proposed to estimate the modulation parameters and transmitted symbols. By modeling the time-varying modulation parameters as AR model of the first order, formulae for particle update of parameters and symbol series are mainly de- rived, particle weights updating method is determined, and the estimation of a posterior distribution function is presented. Simulation results show that the proposed algorithm is feasible for the estima- tion and blind separation of the modulation parameters and transmitted symbols of the mixed signals.

关 键 词：同频混合信号 单通道盲分离 非线性状态估计 粒子滤波 

分 类 号：TN911.72[电子电信—通信与信息系统]