Scholarlink

<p>The increasing complexity and interconnectivity of smart grid (SG) systems have exposed them to a wide array of cybersecurity threats. This review paper critically surveys recent advancements in federated learning (FL) as a privacy-preserving machine learning technique for addressing these challenges. The objective of this review is to analyze how FL can support secure, decentralized anomaly detection and mitigate attacks such as False Data Injection (FDI) and Distributed Denial of Service (DDoS) in smart grid infrastructures. We explore major cyber threats targeting smart grid architectures and evaluate FL-based and non-FL-based solutions in terms of performance metrics such as accuracy, recall, and <i>F</i>1-score. Practical considerations for FL deployment, including device heterogeneity, communication constraints, and adversarial machine learning risks, are also discussed. The paper highlights critical gaps and outlines future research directions for improving smart grid resilience using federated intelligence.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Federated intelligence for smart grids: a comprehensive review of security and privacy strategies

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月16日

查看全文
  • Raseel Z. Alshamasi,
  • Dina M. Ibrahim

摘要

The increasing complexity and interconnectivity of smart grid (SG) systems have exposed them to a wide array of cybersecurity threats. This review paper critically surveys recent advancements in federated learning (FL) as a privacy-preserving machine learning technique for addressing these challenges. The objective of this review is to analyze how FL can support secure, decentralized anomaly detection and mitigate attacks such as False Data Injection (FDI) and Distributed Denial of Service (DDoS) in smart grid infrastructures. We explore major cyber threats targeting smart grid architectures and evaluate FL-based and non-FL-based solutions in terms of performance metrics such as accuracy, recall, and F1-score. Practical considerations for FL deployment, including device heterogeneity, communication constraints, and adversarial machine learning risks, are also discussed. The paper highlights critical gaps and outlines future research directions for improving smart grid resilience using federated intelligence.