Abstract
Modern automobiles utilize numerous computerized systems to control and observe the state of the vehicle. These systems use in-vehicle networks to communicate with each other and make intelligent decisions. However, because of the increasing trend to offer connectivity between the in-vehicle network and external networks, automobiles are susceptible to several cyberattacks. This trend cannot continue unless the security of in-vehicle networks is enhanced to tackle all sorts of known and unknown attacks. This paper considers a multi-layered defense system for smart vehicles and models the in-vehicle security of the system. The model consists of a vehicle profile, communication profile, and considers several known defense mechanisms. The attacker profile is also captured based on its abilities and goals. We implement the proposed model in a probabilistic model checker and study the impact of a diverse attacker for a variety of defense profiles.