Vehicular-Network-Intrusion Detection Based on a Mosaic-Coded Convolutional Neural Network

With the development of Internet Vehicles (IoV) technology, car is no longer a closed individual. It exchanges information with an external network, communicating through vehicle-mounted network (VMN), which, inevitably, gives rise to security problems. Attackers can intrude on VMN, using wireless or interface devices. To prevent such attacks, various intrusion-detection methods have been proposed, including convolutional neural (CNN) ones. However, existing CNN method was not able best use CNN’s capability, extracting two-dimensional graph-like data, and, at same time, reflect time connections among sequential data. Therefore, this paper proposed novel model, based Mosaic pattern coding, for anomaly detection. only make full ability extract grid data but also maintain relationship it. Simulations showed that could, effectively, distinguish attacks from normal vehicular improve reliability system’s discrimination, meet real-time requirement