PufCB-Auth: A lightweight continuous multi-factor authentication scheme integrated PUF with charging behavior features for EV charging.

Abstract

As electric vehicles (EVs) gain widespread adoption, interactions between EVs and charging infrastructure are increasing, driving the need for secure and efficient authentication methods. However, existing authentication approaches are inadequate to address the unique challenges of dynamic EV charging scenarios. Moreover, they often suffer from static credentials, high computational overhead, and limited adaptability to dynamic user behavior and environmental variability. To address these challenges, this paper proposed PufCB-Auth, a lightweight multi-factor authentication scheme that integrates hardware-level Physical Unclonable Functions (PUFs) with charging behavior features to generate a multi-modal digital fingerprint. To alleviate the negative effects rooting from EV user's charging behavior drift and PUF response fluctuations, the paper also proposed Enhanced PufCB-Auth by incorporating a fingerprint update mechanism. The proposed scheme achieves lightweight design, strong robustness, and continuous authentication capability, making it well-suited for dynamic and resource-constrained EV charging environments. Simulation results validate its effectiveness in improving authentication accuracy and robustness, with minimal system overhead, enabling practical deployment in real-world ChaoJi charging pile-EV interaction environments.