Integration of Deep Learning, Machine Learning and Blockchain for VANETs

Abstract

This paper deals with security of Vehicular Ad Hoc Networks (VANETs) based on novel technologies such as artificial intelligence and Blockchain technology. VANET enable communication between vehicles or with infrastructure located at the roadsides to improve road safety or provide Internet access for passengers. To set up such a network, certain electronic equipment must be installed in vehicles (computers, network cards, sensors, a GPS location system, and of course, a processing platform). Our goal is to improve road safety of VANETs. However, these networks face significant security challenges. They are particularly vulnerable to various attacks that can disrupt communications between vehicles, affect real-time decision-making, and harm overall network security. We generate our own dataset, and then we try to integrate deep learning (DL) and machine learning (ML) to perform an intrusion detection system (IDS). Moreover, IDS will be combined with blockchain technology, ML/DL approaches can detect and structure attacks using powerful algorithms adapted to the different types of data collected in VANETs. Logistic regression, for example, has shown notable effectiveness in cases where a clear binary classification can be applied. Although limited by its linear nature, it remains a reference model for identifying suspicious behavior quickly and interpretably