Covert Attacks Through Adversarial Learning: Study of Lane Keeping Attacks on the Safety of Autonomous Vehicles

Road management systems are to improve in terms of integrity, mobility, sustainability, and safety by the adoption of artificial intelligence and Internet of Things services. This article introduces the concept of covert attacks on autonomous vehicles which can jeopardize the safety of passengers. Covert attacks are designed to manipulate outputs of cyber physical systems through network channels in a way that while the changes are not easily perceptible by human beings, systems are negatively affected in the long run. We argue that future smart vehicles are vulnerable to viruses which can use adversarial learning methods to adapt themselves to hosts and remain stealth for a long period. As a case study, we design a covert attack on the lane keeping system of autonomous vehicles. In the scenario, an intelligent adversary manipulates sensor readings (lane position, curvature, etc.) in order to deceive the controller to drive the vehicle closer to the boundaries. The virus/attacker interactively learns the host vehicle behaviors in terms of lateral deviation and maneuverability and tries to increase the errors to the extent that remains unnoticeable to the driver. This process is carried out through actor-critic learning based on the Newton--Raphson method. We additionally design an intrusion detection system for such covert attacks. We use the GPS data and offline maps to reconstruct road curves and match them against readings. A simulation testbed is developed based on the map of Nurburgring-Grand Prix track to evaluate our models. Results confirm the validity and effectiveness of the proposed models.