Security Analysis Of Modern Automotive Based On Controller Area Network

Controller Area Network(CAN),the de facto standard in-vehicle network protocol,prompts modern automobile an integrated system that achieves real-time interactions with roads,vehicles and people.Yet such connectivity makes it feasible to illegally access,or even attack the CAN,causing not only privacy disclosure,property damage,but also life threat.In order to identify the most critical threat and thus determine the appropriate defense strategy,we take a first step toward modeling security issues in modern automotive based on CAN from both technical and economic perspectives.From a technical perspective,we first analyze intrinsic weakness in CAN protocol that is mostly exploited by attackers and comprehensively survey the existing attacks based on CAN interfaces.Then we use Markov chain to propose a dynamic attackdefense tree based model,which is temporal-aware,to characterize an attack effort made by an attacker and a corresponding countermeasure responded by defender.Finally,we simulate steady state when altering the difficulty of attack as well as defense.And results demonstrate that enhancing the defense of a difficult attack would to some extent lower the security level of the entire system.From an economic perspective,we use single-stage static game to reflect how attacker and defender simultaneously choose the strategies,where both rational participants tend to select Nash Equilibrium in mixed strategy game guaranteeing the utility balance of the other.For better defense,we analyze the dominant strategy of the defender under different scenarios.In addition,we utilize reinforcement learning to model how defender learns to maximize his average payoff in a repeated game.