A Context-Aware on-board Intrusion Detection System

Abstract Modern vehicles are becoming increasingly attractive from the perspective of possible intruders. The main reasons are twofold: modern vehicles are now connected to the outside world via Wi-Fi, Bluetooth, and mobile connection, such as LTE and 5G, and the increasing complexity of the on-board software enlarges the attack surface. In this article, we introduce CAHOOTv2, a context-sensitive intrusion detection system (IDS) that uses the vehicle’s sensors to determine driver habits and gather information about the environment to detect intruders. We use hyperparameter tuning to increase detection accuracy. To demonstrate the validity of the algorithm, we collected driving data from both an Artificial Intelligence (AI) and 39 humans. We include the AI driver to demonstrate that CAHOOTv2 is able to detect intrusions when the driver is both a human or an AI. The dataset is obtained using a modified version of MetaDrive simulator where we consider also the presence of an intruder able to perform the following types of intrusions: denial of service, replay, spoofing, additive and selective attacks. We make several experiments showing the benefits of hyperparameters tuning. The results of CAHOOTv2 are promising on detection of intrusions.