Clock drift management using particle swarm optimization

Abstract

Time Synchronization is a common requirement for most network applications. It is particularly essential in Wireless sensor networks (WSN) to allow collective signal processing, proper correlation of diverse measurements taken from a set of distributed sensor elements and for an efficient sharing of the communication channel. The Flooding Time Synchronization Protocol (FTSP) was developed explicitly for time synchronization of wireless sensor networks. In this paper, FTSP is optimized for clock drift management using Particle Swarm Optimization (PSO). The paper estimates the clock offset, clock skew and generates linear line and optimizes the value of average time synchronization error using PSO. This paper presents implementation and experimental results that produce reduced average time synchronization error optimized using PSO compared to that of linear regression used in FTSP. © 2012 Springer-Verlag.