摘要
简要回顾了基于粒子滤波的同频混合信号盲分离算法。着重研究了混合信号的定时问题,提出了基于Gardner算法的定时方法并从理论上分析了其有效性。在此基础上提出了一种Gardner定时+粒子滤波+Viterbi序列检测的同频混合信号实时盲解调方案,其中Viterbi算法可以根据粒子滤波估计出的信号参数,对信号符号作进一步估计以达到更好的解调效果。仿真结果验证了该方案的可行性。
The particle filtering based blind separation algorithm of co-frequency signals was briefly reviewed at first. Timing of co-frequency signals was investigated, and then a solution was proposed based on well-known Gardner algorithm, the validity of this method was also analyzed from a theoretical aspect. Based on the study above, a real-time blind demodulation scheme was presented which comprises Gardner timing, particle filtering and Viterbi sequence detection. By Viterbi algorithm, symbol detection can be refined utilizing the signal parameters estimated by the particle filter. Simulation results verify the feasibility of this scheme.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2009年第19期6203-6206,6210,共5页
Journal of System Simulation