Self-organizing time synchronization inwireless sensor network

Time synchronization is fundamentally necessary to many applications and protocols in wireless sensor network, such as data aggregation, localization and so on. In this paper, inspired by the synchronicity flashes of fireflies, a Self-organizing Time Synchronization (STS) method was proposed based on the original firefly algorithm. Theoretical analyses of its convergence in all-to-all and line topology are given and a distributed sequence number mechanism is used to prevent the emergence of clusters in multi-hop network in the presence of loop. Synchronization is also achieved from synchronicity by synchronizing the number of periods. STS is completely self-organized and has good scalability. It doesn't need unique ID for each node, neither network-wide structures nor special reference node in the network. Furthermore, its performance doesn't depend on any optimization of parameters or fire functions to work in a specific network, so it can be easily deployed under any environment. Both simulation and experiment on real sensor Micaz mote are implemented and the result shows that STS can synchronize the network quickly with good scalability, robustness and suitability.