A Privacy-Preserving Cybersecurity Framework for AI-Driven Green Mobility Ecosystems

Abstract

The integration of Artificial Intelligence (AI), Internet of Things (IoT), and Vehicle-to-Everything (V2X) technologies in green mobility systems introduces new cybersecurity and privacy challenges. This paper proposes a lightweight cybersecurity framework that integrates compact convolutional neural networks (CNNs) for real-time anomaly detection at the edge, federated learning for decentralized model training, and blockchain-based decentralized identity management with zero-knowledge proofs. These mechanisms collectively ensure sub-100 ms threat detection latency, reduced communication overhead, and GDPR-compliant privacy preservation. Simulation results demonstrate a 60% reduction in latency, 45% lower communication costs, 30% energy savings at edge nodes, and a detection accuracy of 93.4% compared to traditional cloud-centric models.