基于粒子滤波的跳频信号频率跟踪方法  

Frequency-Hopping Signal Tracking Based on Particle Filtering

作　　者：程曙晖 王斌[1] 张俊凯[2] 

机构地区：[1]信息工程大学,河南郑州450001 [2]西安通信学院,陕西西安710106

出　　处：《信息工程大学学报》2014年第1期34-40,共7页Journal of Information Engineering University

基　　金：国家自然科学基金资助项目(61201381)

摘　　要：针对跳频信号的参数估计问题,提出一种基于粒子滤波的跳频信号频率实时跟踪方法。首先建立以跳频瞬时频率为系统状态的状态空间模型,然后通过基本粒子滤波算法序贯重要性重采样(sequential importance resampling,SIR)实现了对跳频信号频率的后验概率密度估计,进而得到频率的实时估计;为进一步提高粒子滤波的跟踪性能,提出一种基于ESPRIT辅助的粒子滤波(auxiliary esprit particle filtering,AESPRIT-PF)算法。仿真实验分析了在不同信噪比和不同粒子数目下算法的跟踪性能,结果表明该算法具有稳健的跳频频率实时跟踪能力,是一种有效的跳频信号参数估计方法。A method of tracking frequency-hopping signals based on particle filtering is proposed to estimate the parameters of frequency-hopping signals. First, a state-space model with instantaneous frequency system state is structured, which leads to implementing the posterior probability density estimating of the frequency, then the real-time frequency tracking is realized by using the basic particle filter algorithm, which is called sequential importance resampling （SIR）. To improve the tracking ability, an auxiliary particle filter approach based on ESPRIT （AESPRIT-PF） is devel- oped. The tracking ability of the algorithm under different SNR and particle number is analyzed by simulation. The results show that the algorithm could robustly track the hop frequency within real time, and is effective for estimating the parameters of frequency-hopping signals.

关 键 词：跳频信号 粒子滤波 序贯重要性重采样 辅助粒子滤波 

分 类 号：TN911.7[电子电信—通信与信息系统]