Model-based clock synchronization protocol for wireless sensor networks

In a wireless sensor network, nodes communicate time-stamped packets in order to synchronize their clocks, i.e., estimate each other's time display. We introduce and analyze a parametrized stochastic model for clocks and use it to calculate, at any given time, the Maximum-Likelihood (ML) estimate of the relative skew/offset between two communicating nodes. For network synchronization, event-based Kalman-Bucy filtering gives rise to a centralized scheme, and we propose an efficient distributed suboptimal algorithm. We study the performance both analytically and experimentally and provide provable guarantees. We summarize our findings into defining a new distributed model-based clock synchronization protocol (MBCSP), and present a comparative simulation study of its accuracy versus prior art to showcase improvements.