Scholarlink

<p>The Internet of Medical Things (IoMT) enhances healthcare through real-time patient monitoring but is vulnerable to flooding attacks like SYN and UDP floods, which disrupt critical communications and threaten patient safety. This paper proposes an AI-driven mechanism to detect and mitigate such attacks in Software-Defined Networking (SDN)-based 5G IoMT networks. Machine learning models, trained on a comprehensive IoMT traffic dataset, identify malicious patterns in real-time. Upon detection, an Intrusion Detection System (IDS) alerts the SDN controller, which enforces mitigation policies at edge switches. Implemented in a realistic NS-3 and Mininet simulation environment with a Ryu SDN controller, our approach outperforms traditional methods by achieving high detection accuracy, reduced attack impact, and efficient resource utilization. Unlike prior studies relying on offline datasets, our simulation-based evaluation captures real-world network dynamics, ensuring practical applicability. This work demonstrates a scalable, efficient solution for securing IoMT ecosystems, with potential for broader attack mitigation in future healthcare networks.</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

AI-driven intrusion detection and mitigation of flooding attacks in SDN-based 5G IoMT networks

  • 期刊论文

  • 发表日期：

    2025年10月7日

查看全文
  • Reza Mohammadi

摘要

The Internet of Medical Things (IoMT) enhances healthcare through real-time patient monitoring but is vulnerable to flooding attacks like SYN and UDP floods, which disrupt critical communications and threaten patient safety. This paper proposes an AI-driven mechanism to detect and mitigate such attacks in Software-Defined Networking (SDN)-based 5G IoMT networks. Machine learning models, trained on a comprehensive IoMT traffic dataset, identify malicious patterns in real-time. Upon detection, an Intrusion Detection System (IDS) alerts the SDN controller, which enforces mitigation policies at edge switches. Implemented in a realistic NS-3 and Mininet simulation environment with a Ryu SDN controller, our approach outperforms traditional methods by achieving high detection accuracy, reduced attack impact, and efficient resource utilization. Unlike prior studies relying on offline datasets, our simulation-based evaluation captures real-world network dynamics, ensuring practical applicability. This work demonstrates a scalable, efficient solution for securing IoMT ecosystems, with potential for broader attack mitigation in future healthcare networks.