Clock synchronization with partial timestamp information for wireless sensor networks

Abstract

Most existing clock synchronization algorithms in wireless sensor networks require full timestamp information to estimate clock skew and offset so as to maintain the same time scale between nodes. However, in many scenarios, with the purpose of decreasing energy consumption, reducing communication overhead, or improving security, timestamps are not exchanged in synchronization process, resulting in insufficient access to all time information by nodes to be synchronized. Hence, how to jointly estimate clock skew and offset with limited observation is a significant challenge. In order to solve this issue, this paper proposes a partial timestamp synchronization (PTS) scheme, which can implement complete synchronization function only according to the local timestamps and the timestamp relationships obtained by the basic synchronization protocol. Then, based on PTS, a joint Maximum Likelihood Estimator is developed to estimate clock skew and offset when stochastic delays follow Gaussian distribution. In addition, a Best Linear Unbiased Estimator for PTS with exponential random delays is presented, which can also estimate clock skew and offset simultaneously. Finally, simulations are conducted to illustrate the effectiveness of the proposed algorithms.