Scholarlink

Supervisory Control and Data Acquisition (SCADA) systems are critical control systems that ensure the smooth and efficient operation of processes within the energy industry. With the emergence of the Smart Grid, which integrates highly advanced information technology in power grids, SCADA systems are more vulnerable to cyber-attacks. Stealthy False Data Injection (FDI) attacks remain particularly challenging to address as they can manipulate grid data without immediate detection, potentially leading to power outages. It is crucial to implement efficient mitigation techniques that reduce the risk of successful FDI attacks while maintaining the continuity of real-time grid operations. Moving Target Defense (MTD) is an emerging strategy that enhances security by continuously changing the attack surface, making it more difficult for attackers to exploit vulnerabilities. Software Networking (SDN) is useful for deploying the MTD strategy centrally and rapidly. This paper proposes an innovative SDN-based Moving Target Defense (MTD) approach to dynamically reconfiguring the network of a Smart Grid SCADA system to counter stealthy FDI attacks. We also determine the optimal network reconfiguration strategy and frequency based on real-time network state using the Proximal Policy Optimization (PPO) Deep Reinforcement Learning algorithm. Our results demonstrate that this method can prevent up to 92% of stealthy FDI attacks while maintaining SCADA system performance.

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

Leveraging Network Reconfiguration to Mitigate Stealthy FDI Attacks in Smart Grid SCADA Systems by Exploiting Attacker Uncertainty

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Aurélie Kpoze,
  • Jules Degila,
  • Arnaud Ahouandjinou

摘要

Supervisory Control and Data Acquisition (SCADA) systems are critical control systems that ensure the smooth and efficient operation of processes within the energy industry. With the emergence of the Smart Grid, which integrates highly advanced information technology in power grids, SCADA systems are more vulnerable to cyber-attacks. Stealthy False Data Injection (FDI) attacks remain particularly challenging to address as they can manipulate grid data without immediate detection, potentially leading to power outages. It is crucial to implement efficient mitigation techniques that reduce the risk of successful FDI attacks while maintaining the continuity of real-time grid operations. Moving Target Defense (MTD) is an emerging strategy that enhances security by continuously changing the attack surface, making it more difficult for attackers to exploit vulnerabilities. Software Networking (SDN) is useful for deploying the MTD strategy centrally and rapidly. This paper proposes an innovative SDN-based Moving Target Defense (MTD) approach to dynamically reconfiguring the network of a Smart Grid SCADA system to counter stealthy FDI attacks. We also determine the optimal network reconfiguration strategy and frequency based on real-time network state using the Proximal Policy Optimization (PPO) Deep Reinforcement Learning algorithm. Our results demonstrate that this method can prevent up to 92% of stealthy FDI attacks while maintaining SCADA system performance.