Dynamic Adaptive Security Framework for Software-Defined Autonomous Vehicles in Edge Computing Environments

Abstract

Security and accountability must be prioritized in the ever-increasing deployment of autonomous vehicles (AVs). This work aims to suggest an all-encompassing strategy for addressing these challenges. Our primary contribution will be to enhance cybersecurity by implementing a solid mechanism for identifying authorized users within AVs in response to the increasing importance of speaker identification. An ensemble-based strategy that makes use of speaker verification techniques is offered as a means of preventing the threat of voice spoofing. This approach ensures that user commands are genuine. Furthermore, in the event of incidents involving autonomous vehicles, there is a requirement for precise accountability and the distribution of liability liabilities. We propose a novel application of blockchain technology to address this issue. This application makes it possible to create an event recording system that guarantees transparent and tamper-proof records. EC-CAVs, connected autonomous vehicles powered by edge computing, have the potential to meet the growing demand for intelligent transportation on a worldwide scale. Any number of advantages, including proactive vehicle monitoring, traffic management, and services that offer completely autonomous driving, could result from incorporating such advances into transportation ecosystems. When deploying EC-CAVs, there must be strict adherence to safety and security protocols to avoid any possibility of vehicle immobilization, road accidents, data leaks, or other problems. A standard attack taxonomy for the EC-CAVs ecosystem will be derived from this study's current and future cyber threats investigation. This article presents the EC-CAV reference architecture. The following part provides an in-depth analysis of security measures that can ensure passengers' safe and secure transportation from one place to another by utilizing an ecosystem of EC-CAVs.