Machine Learning for Advancing Electric Vehicles Security Leveraging CICEVSE2024

The emergence of Electric Vehicles (EVs) and related charging infrastructure marks a significant shift toward sustainable transportation solutions. However, the increasing deployment of Electric Vehicle Supply Equipment (EVSE) has introduced a new landscape for cybersecurity challenges. To address this critical issue, we utilize machine learning techniques for enhancing EVSE security systems leveraging CICEVSE2024, a new and comprehensive dataset tailored to advancing research on EVSE security. The dataset provides detailed records of EVSE activities, capturing both benign operations and attacks like Reconnaissance, Backdoor, Cryptojacking, Denial of Service (DoS), etc. in multiple states. Our research focuses on analyzing the performance of machine learning algorithms for securing EVSE systems. This study uses two techniques, namely Host Anomaly Detection Model (HADM) and Power Anomaly Detection Model (PADM), wherein machine learning algorithms are used to identify and classify attack scenarios with high performance. This research emphasizes the prospects of machine learning technologies in fortifying EVSE systems against cybersecurity threats, thereby contributing to the broader goal of ensuring the resilience of clean energy powered transportation solutions.