基于Bootstrap采样粒子滤波的WirelessHART时间同步算法

发布时间：2018-05-02 19:18

本文选题：无线HART传感网络 + 时间同步　；参考：《计算机应用研究》2017年03期

【摘要】：针对无线HART传感器网络时间同步精度较低、能耗过大等问题,提出了一种基于Bootstrap采样的粒子滤波时间同步算法。在未知网络延迟分布的情况下,为了减少观测次数,对发送端和接收端的时间戳观测值进行Bootstrap采样,采用混合粒子滤波算法,获得精确的时钟偏移,不仅降低了无线HART传感网络时间同步误差,而且使能耗减小。实验表明,对于无线HART网状分层网络,当观测量达到10以上时,粒子滤波算法获得的时间偏差的均方误差约是最大似然估计算法的50%,而基于Bootstrap采样的粒子滤波算法获得的时间偏差的均方误差约是最大似然估计算法的35%,验证了方法的可行性和有效性。
[Abstract]:Aiming at the problems of low precision and excessive energy consumption in wireless HART sensor networks, a particle filter time synchronization algorithm based on Bootstrap sampling is proposed. In the case of unknown network delay distribution, in order to reduce the number of observations, the time-stamp observations at the transmitter and receiver are sampled by Bootstrap, and a hybrid particle filter algorithm is used to obtain the accurate clock offset. Not only the time synchronization error of wireless HART sensor network is reduced, but also the energy consumption is reduced. The experimental results show that for wireless HART mesh layered networks, when the observed amount is more than 10, The mean square error of the time deviation obtained by the particle filter algorithm is about 50 of the maximum likelihood estimation algorithm, while the mean square error of the time deviation obtained by the particle filter algorithm based on Bootstrap sampling is about 35 percent of the maximum likelihood estimation algorithm. The feasibility and effectiveness of the method.
【作者单位】：重庆三峡学院计算机科学与工程学院;
【基金】：国家自然科学基金资助项目(61273219) 中国博士后科学基金资助项目(160560) 重庆市教委科学技术研究项目(KJ1710244,KJ1710257,KJ1601003,KJ1603303,KJ1401029,KJ1401002,KJ1401008) 重庆市科委基础与前沿项目(CSTC2014JCYJ1316) 重庆三峡学院科学研究项目(14QN30,15QN09)
【分类号】：TP212.9;TN929.5;TN713

【参考文献】