Clusters-based Low-overhead Time Synchronization Technology In Wireless Sensor Network

With the innovation of sensor technology, embedded system technology and wireless communication technology, wireless sensor network research has entered a boom phase. However, the monitoring environment of wireless sensor network is becoming increasingly complex, and monitoring tasks of wireless sensor network are becoming more arduous. There are many difficulties need to be overcome, and the wireless sensor network time synchronization is one of the major problems.Although the traditional time synchronization protocols provide high-precision time service, the wireless sensor network nodes are obviously unable to use those high-overhead, low-adaptability time service to achieve synchronization. In fact, wireless sensor network urgently needs a new time synchronization protocol which is designed according to the characteristics of its own.In this paper, we focus on the existing weaknesses of wireless sensor network time synchronization protocols, especially in low communication overhead, low power consumption wireless sensor network. We propose an innovative cluster-head broadcasting synchronization model, and design a low-overhead time synchronization protocol that based on clusters. The main topics of this paper are as follows.(1) Propose an innovative cluster-head broadcasting synchronization model. In this paper, we use cluster division technology to improve the performance of one-way communication time synchronization model. Through introducing beacon nodes which send beacon information, we successfully remove the offset between the cluster-head nodes and the cluster-member nodes. We also use the broadcasting synchronization to replace traditional point-to-point synchronization, which obviously reduces whole network time synchronization communication overhead and significantly improve synchronization efficiency.(2) Design a clusters-based low-overhead time synchronization protocol. This protocol takes advantages of the cluster structure and the cluster-head broadcasting synchronization model. In multi-hop network, this synchronization protocol is improved by optimizing the logic tree and cluster partitioning algorithm HeeD. At the same time, this time synchronization protocol designs a series of important mechanisms to ensure the reliability and stability of the protocol. Above designs make performance of this time synchronization protocol better than the traditional time synchronization protocols on synchronization accuracy, communication overhead, energy consumption, convergence and robustness.(3) Carry out experiments in a real wireless sensor network platform. Through building a wireless sensor network experimental platform and running clusters-based low-overhead time synchronization protocol, we assess the performance of this protocol on synchronization accuracy, communication overhead, energy consumption, convergence and robustness. Meanwhile, in the same experimental platform, we also run the recognized classic time synchronization protocol (FTSP) as a control protocol. The experiments have objectively demonstrated the advantages of the cluster-based low-overhead time synchronization protocol.