Probabilistic time synchronization in wireless sensor networks

This paper presents a novel time synchronization method in real-time wireless sensor networks. Firstly, time synchronization algorithms in common distributed system are reviewed and essential requirements in the design of synchronization methods for wireless sensor network are proposed. Then, reference broadcast synchronization (RBS), a very efficient method, is introduced and the probabilistic technique is applied to RBS in order to improve the performance. After that, a new method, namely partial delay measurement time synchronization, is proposed in this paper to optimize RBS further. At last, the algorithm is implemented on MICA, a wireless sensor platform, and the results illustrate the feasibility and reliability of the new method. It is shown theoretically and experimentally that the probabilistic time synchronization introduced in this paper is able to acquire satisfactory accuracy with less computational costs than such algorithms developed in common distributed systems.