Scholarlink

<p>In recent years, the Industrial Internet of Things (IIoT) has seen a significant surge in adoption, particularly within Industry 4.0 and 5.0. As IIoT is widely deployed in critical facilities and manufacturing plants, it has become an attractive target for cyber attackers seeking to exploit vulnerabilities, posing a major threat to industrial environments. Ensuring robust security mechanisms is therefore essential to protect these networks from malicious attacks. One prevalent threat is flooding attacks, often used in Distributed Denial of Service (DDoS) attacks to overwhelm communication and computational resources. This paper introduces a novel approach to detecting such attacks by combining fuzzy logic and machine learning (ML). The fuzzy logic component identifies suspicious behavior, while the centralized ML model refines detection by resolving uncertainties. Beyond detection, we propose a mitigation strategy to effectively block intruders. Simulations and experiments conducted on a public testbed validate the accuracy of our approach, with experimental results demonstrating approximately 90% accuracy in attack detection.</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

An ML/fuzzy based flooding attack mitigation mechanism for industrial internet of things

  • 期刊论文

  • 发表日期：

    2025年9月3日

查看全文
  • Reza Mohammadi,
  • Fabrice Theoleyre,
  • Mohammad Nassiri

摘要

In recent years, the Industrial Internet of Things (IIoT) has seen a significant surge in adoption, particularly within Industry 4.0 and 5.0. As IIoT is widely deployed in critical facilities and manufacturing plants, it has become an attractive target for cyber attackers seeking to exploit vulnerabilities, posing a major threat to industrial environments. Ensuring robust security mechanisms is therefore essential to protect these networks from malicious attacks. One prevalent threat is flooding attacks, often used in Distributed Denial of Service (DDoS) attacks to overwhelm communication and computational resources. This paper introduces a novel approach to detecting such attacks by combining fuzzy logic and machine learning (ML). The fuzzy logic component identifies suspicious behavior, while the centralized ML model refines detection by resolving uncertainties. Beyond detection, we propose a mitigation strategy to effectively block intruders. Simulations and experiments conducted on a public testbed validate the accuracy of our approach, with experimental results demonstrating approximately 90% accuracy in attack detection.