Metrics for Intrusion Detection in Wireless Sensor Networks

Wireless sensor networks are commonly deployed for collecting information about environments by using sensors. The nodes are restricted in their resources and dynamically maintain the network structure by themselves. Therefore, they are vulnerable to many intrusions or network attacks. In this book a novel approach for intrusion detection in wireless sensor networks is developed. Common performance and network metrics are considered to be indicators for a local intrusion detection on wireless sensor nodes. The mechanism does not need additional data collection, because most of the local metrics are already collected on each node. For the analysis the metrics are measured in different scenarios and evaluated in a real wireless sensor network deployment at the Technical University in Darmstadt. The considered and implemented network attacks for the intrusion detection are jamming and blackhole. The evaluation shows that several metrics are suitable for the detection. The results can be used for lightweight implementations of intrusion detection systems in wireless sensor networks.

Daniel Thies, Master of Science: Born 1984 in Germany. Study of Computer Science at the Technical University in Darmstadt (Major Subject: IT Security, Minor Subject: Psychology). Scholarship of "Stiftung Industrieforschung". Nominated for CAST e.V. Sponsorship Award IT Security 2013. He currently works as an IT security consultant.