Security games for vehicular networks

Vehicular ad-hoc networks (VANETs) enabling vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications is an emerging research field aiming to improve transportation security, reliability, and management. To better understand networking and security aspects of VANETs, we have been investigating network connectivity issues and mappings of car networks to the underlying road topology. Using this mapping and various metrics, we locate hot-spots in vehicular networks to determine the most vulnerable points for jamming. We also use these to optimize the placement of roadside units. To this end, we first develop quantitative measures for assessment of the importance of road segments in the context of security and traffic flow. We then use game theoretic analysis to investigate the effects of possible malicious users on the system for increased reliability and better management of resources. For example, we study the optimal deployment of traffic control and security infrastructure both in the static (roadside units) and dynamic case (law enforcement units). We use realistic simulation data, obtained from traffic scientists, as input to our models as well as to evaluate the effectiveness of countermeasures.