Scholarlink

The fast growth of IoT networks present harsh cyberspace issues, particularly in limited resources and evolving circumstances. To achieve this, a new context-aware adaptive cybersecurity framework is offered, which combines reinforcement learning and meta-learning to dynamically tune the process of detecting threats to devices, location, and time. The framework is tested on CICIDS2017 and IoT-23 datasets, with a detection accuracy of 95.9, a low latency of 5.46 ms, and high precision in detecting attacks like Mirai and Sybil than traditional, respectively. The solution offers a scalable, efficient, and real-time middleware to enhance the security of nonhomogeneous IoT environments, enabling more efficient defense of energy-conscious and informatics systems and infrastructure.

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

Enhancing Dynamic Threat Detection in IoT Networks Through Context-Aware Adaptive Cybersecurity Algorithms for Energy and Informatics Optimization

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Amer Ibrahim,
  • Lateef Abd Zaid Qudr,
  • Abdul Samad Bin Shibghatullah,
  • Safwan Nadweh,
  • Ali Imad Naji,
  • Ahmed Dheyaa Radhi,
  • Zahraa A. Jaaz,
  • Reyad Omran Essa,
  • Dima Haider Rasheed,
  • Salwan S. Hatif

摘要

The fast growth of IoT networks present harsh cyberspace issues, particularly in limited resources and evolving circumstances. To achieve this, a new context-aware adaptive cybersecurity framework is offered, which combines reinforcement learning and meta-learning to dynamically tune the process of detecting threats to devices, location, and time. The framework is tested on CICIDS2017 and IoT-23 datasets, with a detection accuracy of 95.9, a low latency of 5.46 ms, and high precision in detecting attacks like Mirai and Sybil than traditional, respectively. The solution offers a scalable, efficient, and real-time middleware to enhance the security of nonhomogeneous IoT environments, enabling more efficient defense of energy-conscious and informatics systems and infrastructure.