Intrusion Detection In-Vehicle Networks using Bio-Inspired Approaches

Vehicular Ad Hoc Networks (VANETs) are significant for providing services, applications, and communication to vehicles, connecting them to the outside world and supporting Intelligent Transportation Systems (ITS). Inside the car, the invehicle network (IVN), also known as CAN (Controller Area Network), handles communication between Electronic Control Units (ECUs) and sensors, ensuring the vehicle’s functionality and safety. However, external connections make this internal network vulnerable to unauthorized and malicious access. This work presents a comparative study of two bio-inspired metaheuristics (Bat and Ant Colony Optimization) for use in Intrusion Detection System (IDS) models. It focuses on machine learning-based classifiers to detect and classify anomalous and malicious traffic on the CAN bus.