Monitoring Routing Topology in Dynamic Wireless Sensor Network Systems
Rui Liu,Yao Liang,Xiaoyang Zhong,Yao LIang
DOI: https://doi.org/10.1109/icnp.2015.35
2015-11-01
Abstract:In large-scale multi-hop wireless sensor networks (WSNs) for data collection, the ability of monitoring per-packet routing paths at the sink is essential in better understanding network dynamics, and improving routing protocols, topology control, energy conservation, anomaly detection, and load balance in WSN deployments. In this study, we consider this important problem under tremendous WSN routing dynamics, which cannot be addressed by previous methods based on a routing tree model. We formulate the WSN topology inference as a novel optimization problem, and devise efficient decoding algorithms to effectively recover WSN routing topology at the sink in real-time using a small fixed-size path measurement attached to each packet. Rigorous complexity analysis of the devised algorithms is given. Performance evaluation is conducted via extensive simulations. The results reveal that our approach significantly outperforms other state-of-the-art methods including MNT, Pathfinder, and CSPR. Furthermore, we validate our approach intensively with a real-world outdoor WSN deployment running collection tree protocol for environmental data collection.