Detecting malicious behavior in a vehicular DTN for public transportation

In the past we presented a system that uses public buses and trams in the city of Braunschweig to monitor air pollution in a distributed way. This system has been built on the principles of delay tolerant networking. Since reliability and resilience is an important factor in such systems, in this work we propose a Misbehavior Detection System (MDS) to defend the Braunschweig network against the interference of faulty and malicious nodes. By collecting and securely exchanging data of previous encounters, a node can use cluster analysis to assess the trustworthiness of other nodes in order to detect malicious behavior. By evaluating our method through different routing protocols in Delay- and Disruption-Tolerant Networks (DTNs), we will show that our MDS is able to efficiently protect the system with low overhead.