Applying machine learning algorithms to improve intrusion detection system in IoV

Since the last decade, Internet of vehicles (IoV) keeps attracting the interest of researchers and automakers with a growing variety of applications and services. The aim of these applications is to provide road safety and comfort for drivers and passengers. However, the quantities of data propagated in this network make its security a challenge. The big data generated, collected, and processed on IoV involve valuable and delicate information that malicious intruders might manipulate. Besides that, the specific characteristics of IoV make intrusion detection in this network complicated. The intrusion detection system (IDS) represents the most common approach for network protection. It monitors the traffic activity to identify signs of safety risks and generates alerts for any security anomalies detected. In order to ensure the accuracy of intrusion detection in IoV, the main purpose of this chapter is to apply machine learning (ML) algorithms for training the IDS with a sufficient dataset of security menaces. The IDS based on ML will be able to make crucial decisions in case of intrusion detection whereas continuing to learn about their highly dynamic environment. This chapter describes different security issues in IoV and presents diverse ML algorithms employed for constructing IDS toward protecting IoV from diverse cyber-attacks.