Distributed Fault-Tolerant Detection in Wireless Sensor Networks

Event detection, boundary detection, and faulty sensor detection are important research topics in wireless sensor networks. Finite power supplies and limited communication bandwidth within wireless sensor networks mean applications must use these resources efficiently. With that in mind, this book presents the concept of using minimal neighbor sensors in the detection process. This concept can be used when deploy sensors strategically in building sensor networks. Reliability of small sensors is questionable, especially when the sensors are deployed in harsh environments. This book will explore the ways to improve detection accuracy by incorporating sensor reliability explicitly in the fault-tolerant detection process. An adaptive detection system is also presented for event boundary detention. Finally, a hypothesis testing method is used for faulty sensor detection.

Dr. Xuanwen Luo is a senior engineer in control system and software in Sandvik Mining USA. He was an electrical engineer in Beijing Capital Engineering and Research Inc. He also worked as a research engineer for GM and as a senior project engineer for Ford. His research interests include sensor networks, machine learning and distributed detection.