Formal verification of mobile agent based anomaly detection in Wireless Sensor Networks

Abstract

Mobile agent technology offers a number of advantages to resource constrained Wireless Sensor Networks (WSNs) by facilitating parallelism, code and data dissemination, localization, and distributed security services. As part of the second line of defense for WSNs, anomaly detection schemes are also benefiting from mobile agent technology. The formal verification can be used to validate the correctness of a system using formal specifications. This study employs Behavior Trees (BTs) for formal verification of mobile agent based anomaly detection system in WSNs. We employed BT based formal semantics to integrate the behavior projection of different components of anomaly detection system. The Symbolic Analysis Laboratory (SAL) tool is used to formally specify the behavior tree model. The analysis based on Linear Temporal Logic (LTL) theorems validates the completeness and consistency of the formal model of mobile agent based anomaly detection system.