| dc.description.abstract | Time synchronization in wireless sensor networks (WSNs) is essential and significant for maintaining data consistency, coordination, and performing other fundamental operations, such as power management, security, and localization. Energy efficiency is the main concern in designing time synchronization protocols for WSNs
because of the limited and generally nonrechargeable power resources. In this dissertation, the problem of time synchronization is studied in three different aspects to achieve energy efficient time synchronization in WSNs.
First, a family of novel joint clock offset and skew estimators, based on the classical two-way message exchange model, is developed for time synchronization in WSNs. The proposed joint clock offset and skew correction mechanisms significantly increase the period of time synchronization, which is a critical factor in the over-all energy consumption required for global network synchronization. Moreover, the
Cramer-Rao bounds for the maximum likelihood estimators are derived under two different delay assumptions. These analytical metrics serve as good benchmarks for the experimental results thus far reported.
Second, this dissertation proposes a new time synchronization protocol, called the Pairwise Broadcast Synchronization (PBS), which aims at minimizing the number of message transmissions and implicitly the energy consumption necessary for global synchronization of WSNs. A novel approach for time synchronization is adopted in PBS, where a group of sensor nodes are synchronized by only overhearing the
timing messages of a pair of sensor nodes. PBS requires a far smaller number of timing messages than other well-known protocols and incurs no loss in synchronization accuracy. Moreover, for densely deployed WSNs, PBS presents significant energy saving.
Finally, this dissertation introduces a novel adaptive time synchronization protocol, named the Adaptive Multi-hop Timing Synchronization (AMTS). According to the current network status, AMTS optimizes crucial network parameters considering the energy efficiency of time synchronization. AMTS exhibits significant benefits
in terms of energy-efficiency, and can be applied to various types of sensor network applications having different requirements. | en |
