Enhancing Cybersecurity in smart grid: a review of machine learning approaches

Originally designed for isolated operation using proprietary and insecure protocols, SCADA systems are increasingly integrated with IoT technologies to enhance smart grid (SG) efficiency and flexibility. This integration significantly expands the attack surface, creating new cybersecurity vulnerabilities. The objective of this study is to address these challenges by analyzing real-world SG attack scenarios and evaluating the effectiveness of existing machine learning (ML)-based threat detection methods. A comprehensive review of ML techniques, including their applications and limitations, was conducted, with a focus on issues such as data sensitivity and vulnerability to adversarial attacks. Key findings indicate that while ML-based methods demonstrate significant potential for detecting cyber threats, they face challenges in scalability, accuracy, and robustness. To mitigate these issues, the study provides actionable recommendations, including the integration of blockchain for data integrity, advanced anomaly detection techniques, and optimization strategies for resource-constrained environments. These findings contribute to future research directions aimed at strengthening SG cybersecurity in the IoT era, ultimately supporting the protection of critical infrastructure.