Improving data delivery performance in wireless sensor networks

Yunhao Liu,Jiliang Wang
DOI: https://doi.org/10.14711/thesis-b1155191
2011-01-01
Abstract:A wireless sensor network (WSN) typically consists of a large number of resource constrained sensor motes spanning in a large field for data collection. Data delivery is usually achieved with multi-hop transmission along a sequence of nodes. Thus multi-hop data delivery is a fundamental issue in WSNs. Based on a real world environment monitoring sensor network project GreenOrbs, this thesis addresses four key aspects for improving data delivery performance from different layers in wireless sensor networks. At network layer, this thesis presents a comprehensive path quality estimation metric and introduces an optimal packet scheduling algorithm to balance workloads among sensor motes for a low-duty-cycled network in which motes periodically wake up to save energy. At MAC layer, this thesis introduces an approach to improve channel efficiency by combining multiple packets and determining an appropriate sending time and presents a method to alleviate packet losses by considering receiver-side collisions. Through intensive simulations and real world implementations, I evaluate the performance of the proposed methods in a real system and verify the applicability. The results show that the proposed methods are effective and efficient.
What problem does this paper attempt to address?