Time Synchronization In Wireless Sensor Network Applications

In most sensor network applications, events are time stamped with node's local time. However, energy is highly constrained resource in sensor networks. The purpose of this paper is to present a time-synchronization algorithm for sensor networks that aims at reducing the computation and communication energy expended by the algorithm. We use MAC-layer time stamping and estimate the clock drift rate and the offset in order to obtain high precision performance. Our algorithm works in two steps. In the first step, a spanning tree is built in the sensor network. In the second step, all nodes in the network synchronize their clocks to their parent nodes. We analyze and implement our time synchronization algorithm on Berkeley MicaZ platform and show that it can synchronize a pair of neighboring motes to an average accuracy of around one microsecond with communication complexity of O(log n).