TOWARDS SAFE AND SECURE URBAN TRANSPORTATION INTRUSION DETECTION SYSTEM FOR CONNECTED VEHICLES IN SMART CITIES

Abstract

The rapid development of Smart Cities relies heavily on Connected Vehicles (CVs) and Vehicle-to-Everything (V2X) communication to achieve efficient and safe urban transportation. However, this extensive connectivity dramatically expands the cyberattack surface, exposing both the in-vehicle network (IVN), particularly the CAN bus, and the external communication infrastructure to severe security threats. Cyberattacks on CVs can lead to vehicular malfunction, data theft, and catastrophic physical harm, fundamentally undermining the safety and public trust required for smart city adoption. This paper addresses these challenges by proposing a novel, multi-layered Intrusion Detection System (IDS) specifically tailored for the dynamic and resource-constrained environment of connected urban transport. Our system leverages Machine Learning (ML) and real-time traffic analysis to effectively monitor both internal CAN bus activity for localized attacks (e.g., DoS, spoofing) and V2X data flows for external threats (e.g., man-in-the-middle). The proposed IDS architecture aims for high detection accuracy and low latency, demonstrating superior performance in identifying zero-day and sophisticated intrusion patterns compared to existing solutions. The ultimate goal is to establish a robust cybersecurity framework that ensures the safety and security of connected vehicles, paving the way for trustworthy and resilient smart urban mobility.