一种新型混合并行粒子滤波频率估计方法  

A Novel Parallel Particle Filter for Frequency Estimation

作　　者：王伟[1] 余玉揆 郝燕玲[1] 

机构地区：[1]哈尔滨工程大学自动化学院,哈尔滨150001

出　　处：《电子学报》2016年第3期740-746,共7页Acta Electronica Sinica

基　　金：国家自然科学基金(No.61571148);中国博士后科学基金(No.2014M550182);黑龙江省博士后特别资助(No.LBH-TZ0410);哈尔滨市科技创新人才资助课题(No.2013RFXXJ016)

摘　　要：针对高动态、低信噪比环境下的载波频率信号跟踪问题,提出一种新的混合并行粒子滤波算法(Multiple Extend Kalman Filter Independent Metropolis Hastings,M-E-IMH).该算法具有并行运算结构,实时性较基本粒子滤波有较大的提高.该算法直接利用同相支路(In-phase,I)和正交支路(Quadrature,Q)作为观测量,避免了传统方法中的鉴别器引入而引起的信噪比损耗.在高斯和非高斯环境下,与现有的载波跟踪方法如扩展卡尔曼滤波器(EKF),粒子滤波器(PF),卡尔曼滤波器(KF)等仿真对比表明,该方法在低信噪比下具有更高的跟踪精度.To improve the tracking accuracy of the carrier frequency in low signal-to-noise ratio（ SNR） and high dynamic environment,a new hybrid parallel particle filter algorithm,named multiple extend Kalman filter independent metropolis hastings（ M-E-IM H） is presented. The proposed algorithm has a parallel structure and is verified to be more efficient for the real time implementation compared w ith particle filter（ PF）. The method utilizes the output of the in-phase and quadrature（ IQ） branch as the observation directly to avoid the SNR loss caused by the discriminator. In both guass and non-guass environment,the simulations show that the proposed method has higher tracking accuracy at low SNR compared w ith the traditional methods,such as extended Kalman filter（ EKF）,particle filter（ PF） and Kalman filter（ KF） etc.

关 键 词：多普勒频率估计 并行粒子滤波 高动态 非高斯噪声 实时性 

分 类 号：TN966[电子电信—信号与信息处理]