Cyber–physical security in EV charging infrastructure: Components, vulnerabilities, and defense strategies

Abstract

The increasing number of electric vehicles (EVs) has led to a growing need to establish EV charging infrastructures (EVCIs) with fast charging capabilities to reduce congestion at EV charging stations (EVCS) and also provide alternative solutions for EV owners without residential charging facilities. This article presents a study of several components that model the EVCI as a cyber physical system (CPS). Various components of EVCI include energy storage systems, EV supply equipment, human–machine interfaces, communication systems, etc. These components are interconnected with real time processing units, communication networks, and sensors, making EVCI a CPS enabling a structured vulnerability and data flow analysis. It uniquely maps out the bi-directional interactions between electric vehicles, supply equipment, energy management systems, and communication/control networks. A detailed classification of EV charging technologies covering both onboard and offboard topologies and their communication dependencies revealing cybersecurity implications specific to each category. The study further examines a wide range of internationally adopted standards and protocols (e.g., ISO 15118, IEC 61851, SAE J2931, OCPP, IEEE 2030.5) and categorizes them by functionality, including charging interface, cybersecurity, and interoperability. It identifies critical vulnerabilities of EVCI that lead to various cyberattacks, such as spoofing, false data injection, man-in-the-middle, and denial-of-service attacks, by analyzing both cyber and physical layers. Real-world cyber incidents in the EVCI ecosystem are also reviewed to underscore the severity of these threats. For detection and mitigation, the paper introduces a comprehensive taxonomy of cyberattack defense mechanisms, extending across statistical analysis, machine learning (ML), signature-based, and hybrid intrusion detection systems (IDS). Novel approaches including HMM-based anomaly detection, event correlation frameworks, and optimization-driven defense strategies, are explored. The study highlights the socio-technical impacts of cyber incidents on power grid stability and public safety, making a strong case for proactive, intelligent cybersecurity solutions in modern EVCI systems.