A Relative Risk Model for Anomaly Detection in Wireless Sensor Networks

Localization anomaly detection (LAD) for a wireless sensor network (WSN) plays a prominent role in defending against attacks on localization services. However, it is a difficult task to evaluate the effectiveness of the anomaly detection scheme according to complexity and uncertainty in various applications. This paper presents a relative risk model approach that was proposed for ecological risk assessment (ERA) to tackle this problem. The problem formulation and model-building process is described in detail. Our numerical simulation demonstrates that this relative risk model method is quite effective to the quantitative assessment of vulnerability analysis in wireless sensor networks.