Development of Intrusion detection system for VANET using Machine Learning

Abstract

With the proliferation of vehicular technology, modern vehicles are fortified with ever more electronic maneuvers. This advancement ratifies the evolution of Intelligent Transportation Systems (ITS)to provide services like shared travel, smart driving, on-the-go Internet, etc. As a traditional application of ITS, a Vehicular Ad hoc Network (VANET) enables smart communication between vehicle nodes and network infrastructures to provide various expedient services such as road safety, data sharing, traffic management, parking assistance, entertainment, route recommendation, mobile payment, and even cloud applications. The high-speed mobile nodes (vehicles) in VANET perform very differently from other wireless communication networks and have a set of distinct features such as frequent link disconnection, highly dynamic topology, limited coverage area, and heterogeneous system architecture that may affect the performance and service quality of the VANET significantly. With this motivation, this work attempts to develop distributed cooperative cluster-based IDS to identify potential cyberattacks in VANET effectively. Firstly, this work develops a stable and reliable clustering method named Fuzzy Logic-based Clustering (FLC) to create a collaborative and reliable communication environment. A new fuzzy logic-based CH node selection algorithm is also developed based on node degree, average velocity difference, and relative velocity of the vehicle to create a more robust clustering structure with minimum cost in VANET