基于信息融合的分布式时间同步算法

A distributed time synchronization algorithm based on information fusion

时间同步算法作为无线传感器网络的支撑技术,成为近些年研究的一个热点。其中集中式时间同步算法一直以来被广泛研究,它依靠参考节点通过广播、泛洪或者多跳的方式达到全网时间同步,取得了良好的同步精度,但同时鲁棒性和可扩展性差。提出了一种分布式时间同步算法,通过融合全网节点的时间信息来达到时间同步,这种算法消除了网络拓扑变化和节点密度变化对集中式算法的影响。实验结果表明,针对典型的网络拓扑,分布式时间同步算法实现了全网时间同步,并且对于网络变化和节点数目变化具有鲁棒性。

As a support technology of wireless sensor networks, time synchronization algorithms become a research hotspot in recent years. Centralized time synchronization algorithms have been widely studied, which use the reference node via broadcast, flooding or multi hop to obtain time synchronization of the whole network. They achieve good accuracy, but robustness and scalability are poor. We propose a distributed time synchronization algorithm, which realizes time synchronization by fusing the time information of the whole network nodes. The algorithm can eliminate the influence of the transformation of network topology and the change of nodes＂ density on centralized algorithms. Experimental results show that the distributed time synchronization algorithm achieves the whole network time synchronization on several typical network topologies and is robust to the change of network topology and the number of nodes.