Date of this Version
Troy Green, “A STACKING-BASED MISBEHAVIOR DETECTION SYSTEM IN VEHICULAR COMMUNICATION NETWORKS ”. M.S. thesis, University of Nebraska-Lincoln (2022)
Over the past few decades communication systems for vehicles have continued to advance. Communications between these vehicles can be classified into safety related and non safety related messages. An example of a safety related message would be one vehicle warning others of an icy road it encountered, where a non safety related communication would be a passenger streaming a movie. In either case it's important to secure the communications so that the system continues to behave as expected. In this thesis we propose a Misbehavior Detection System (MDS), which is a system that monitors messages sent between vehicles, and detects misbehaviors for possible attacks. In our approach we use a stacking based machine learning algorithm to determine if vehicles are misbehaving. We then compare this approach to other MDS to determine if our approach makes a measurable difference. In the analysis and comparison section of this thesis, we evaluate the simulation and performance data, showing that our protocol has an accuracy of 91.8%.
Advisor: Yi Qian