Leveraging Deep Learning to Strengthen the Cyber-Resilience of Renewable Energy Supply Chains: A Survey

Abstract

Deep learning shows immense potential for strengthening the cyber-resilience of renewable energy supply chains. However, research gaps in comprehensive benchmarks, real-world model evaluations, and data generation tailored to the renewable domain persist. This study explores applying state-of-the-art deep learning techniques to secure renewable supply chains, drawing insights from over 300 publications. We aim to provide an updated, rigorous analysis of deep learning applications in this field to guide future research. We systematically review literature spanning 2020–2023, retrieving relevant articles from major databases. We examine deep learning’s role in intrusion/anomaly detection, supply chain cyberattack detection frameworks, security standards, historical attack analysis, data management strategies, model architectures, and supply chain cyber datasets. Our analysis demonstrates deep learning enables renewable supply chain anomaly detection by processing massively distributed data. We highlight crucial model design factors, including accuracy, adaptation capability, communication security, and resilience to adversarial threats. Comparing 18 major historical attacks informs risk analysis. We also showcase potential deep learning architectures, evaluating their relative strengths and limitations in security applications. Moreover, our review emphasizes best practices for renewable data curation, considering quality, labeling, access efficiency, and governance. Effective deep learning integration necessitates tailored benchmarks, model tuning guidance, and renewable energy data generation. Our multi-dimensional analysis motivates focused efforts on enhancing detection explanations, securing communications, continually retraining models, and establishing standardized assessment protocols. Overall, we provide a comprehensive roadmap to progress renewable supply chain cyber-resilience leveraging deep learning’s immense potential.