Self-Adaptive IDS in VANETs: A Game Theory and Deep Q-Learning Network Based Generic Scheme.

Because of the nature of high mobility and dynamic network topology, Intrusion Detection Systems (IDSs) in Vehicular Ad-hoc Networks (VANETs) face the challenge in balancing the accuracy and efficiency of detection. Two crucial problems are remained unsolved in existing studies: 1) how to perceive the environmental change in the perspective of an IDS? 2) how to make the IDS adaptive in different scenarios? In this paper, a self-adaptive scheme is proposed for IDSs in VANETs based on Bayesian game theory and Deep Q-learning Network (DQN). In the scheme, the interactions between an IDS and attackers are formulated as a dynamic intrusion detection game, in which the proposed scheme decides either to just adjust or to completely retrain the IDS. The Nash Equilibria (NE) of the game is derived to reveal how the optimal decision of the IDS depends on the detection performance and road conditions. Moreover, a DQNAdjustment is proposed to realize the self-adaptation of the IDS in the dynamic game. Simulation results show that the IDS with the proposed self-adaptive scheme has better performance than other existing IDSs with higher detection rate as well as lower detection time and overhead.