Title: Data-Driven Misbehavior Generation and Detection Methodologies for Intelligent Transportation Systems
Abstract:
Intelligent Transportation Systems (ITSs) is a major application field for Cyber-Physical Systems. Future public transportation system will feature autonomous vehicles where different data sources (e.g., traffic counts, in-vehicle driving data, and so on) will be utilized to increase roadway safety and traffic efficiency. It is possible to introduce different misbehavior or false information attack generation and detection methods in different ITS scenarios by exploiting traffic data and maneuver-wise Controller Area Network (CAN)-bus data. In this research, we introduce a data-driven location privacy attack in auto-insurance programs and a misbehavior detection system and an adversarial attack for the driving maneuver classifier in Connected Autonomous Vehicles (CAV) systems. We conclude that such misbehavior generation and detection methods in ITS scenarios will aid in developing secure and trustworthy applications toward increasing roadway safety and traffic efficiency while maintaining user privacy. This proposal provides an overview of the scope of misbehavior generation and detection research, some of the key challenges in building CAV systems, hypothesized contributions, potential solutions, an evaluation plan, and some preliminary results.
Committee Members:
- Andrew Grimshaw (Chair)
- Haiying Shen (Advisor)
- Lu Feng
- Yuan Tian
- Laura Barnes