Time Synchronization in Network-Centric Sensor Networks

Time synchronization is a crucial component of infrastructure for wireless sensor networks (WSN). The design of new synchronization methods for WSNs needs to satisfy the unique requirements and constraints in terms of precision, lifetime, energy, and scope of the synchronization. In this paper we have designed and compared three techniques for achieving synchronization in network centric sensor networks, using sequential least squares, Kalman filter and fuzzy logic systems. These techniques achieve a highly accurate, long-term and adaptive synchronization by forming a time conversion scale between clocks of two nodes in the network. A synchronization protocol based on any of these techniques will be very energy-efficient, lightweight, multimodal, tunable, and even scalable