Securing the road ahead: Machine learning-driven DDoS attack detection in VANET cloud environments

Abstract

Vehicular ad-hoc network (VANET) technology has gained prominence, especially in the context of the emerging field of VANET Cloud as an integral part of connected and autonomous vehicles. The automotive industry’s move towards automation and the integration of vehicles into the digital ecosystem has revolutionized wireless network communications. Nevertheless, security remains a paramount concern in these advanced technological landscapes. Safeguarding system integrity and data privacy is of utmost importance before the widespread adoption of VANET Cloud solutions. This study addresses the critical challenge of security within the context of VANET Cloud. Specifically, the focus is on anticipating and mitigating Distributed Denial of Service (DDoS) attacks, which can potentially disrupt the functioning of connected vehicles and associated cloud-based services. To tackle this issue, an innovative architectural framework is proposed to capture and analyze network flows within the VANET Cloud environment. Additionally, it leverages machine learning techniques for classification and predictive analytics with an accuracy of 99.59%. The architecture presented in this research offers the potential to significantly enhance security measures in VANET Cloud deployments. Its adaptability ensures practical applicability to real-world systems, enabling timely responses to security threats and breaches.

Keywords

• DDoS
• VANET
• Cloud
• NS2
• Machine learning
• Fuzzification