I. Introduction
In The recent years, the increasing number and length of the traffic congestion caused by limited highway capacity has become a main concern. With the development of wireless communication and sensing technologies, CAVs are expected to enhance the safety, efficiency and throughput of our intelligent transportation systems via sharing its local sensor measurements with the other surrounding vehicles through vehicular networks [1]. For instance, by exchanging intervehicle data via wireless communications, Cooperative Adaptive Cruise Control (CACC) allows groups of vehicles to drive on the highway in a closely-coupled platoon form so that traffic throughput and safety are increased and fuel consumption and CO2 emissions are reduced. However, the sensors of modern vehicles are usually designed without enough security concerns and hence remain vulnerable to cyberattacks; moreover, because of a lack of centralized administration, vehicular networks are also providing chances for attackers to disrupt the normal operation of the transportation systems, see, e.g., [2]--[15], and references therein. In [6], researchers showed that they were able to stop the engine of the vehicle remotely while it was driving down a busy highway. The vulnerabilities of millimeter-wave radars, ultrasonic sensors and forward-looking cameras to attacks is investigated in [16]. The authors of [17] show that various types of attacks can be performed using commonly affordable commodity hardware. A simpler attacks on vehicular networks might be a malicious vehicle transmitting fraudulent data about road conditions or vehicle positions for its own profits, which might cause traffic jams or catastrophic consequences. Cyberattacks on a CAV are severely threatening not only the normal operation of our transportation systems, but also the lives of the drivers, the passengers and the pedestrians. This has shown that great efforts needs to be made to provide strategic mechanism for identifying and mitigating cyberattacks on CAVs.