Scholarlink

<p>The speedy expansion of the Internet of Things (IoT) has led to an increase in the number of interconnected devices, resulting in a heightened demand for effective and efficient intrusion detection systems (IDS). Traditional centralized IDS models may not be well-suited to address the unique challenges posed by IoT networks, particularly regarding device heterogeneity and data privacy. Applying Federated Learning (FL) directly to the devices may solve these problems, but at the cost of communication overhead and lower scalability. To overcome these issues, we propose a novel multiclass intrusion detection system using FL at the subnetwork level. We employ various machine learning models, including Artificial Neural Networks, Convolutional Neural Networks, and Long Short-Term Memory to build an effective IoT IDS. The results indicate that our approach (a) detects attacks with high classification performance, (b) has low communication overhead at the server; (c) is more scalable and, (d) achieves a better precision of 99.88%, showcasing its potential for real-world IoT environments.</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

Enhancing Scalability and Communication Efficiency in Sub-Network Based Federated Deep Learning for Multiclass Intrusion Detection in IoT Networks

  • 期刊论文

  • 发表日期：

    2026年7月16日

查看全文
  • Mamta Rawat,
  • Manan Suri,
  • Gaurav Singal

摘要

The speedy expansion of the Internet of Things (IoT) has led to an increase in the number of interconnected devices, resulting in a heightened demand for effective and efficient intrusion detection systems (IDS). Traditional centralized IDS models may not be well-suited to address the unique challenges posed by IoT networks, particularly regarding device heterogeneity and data privacy. Applying Federated Learning (FL) directly to the devices may solve these problems, but at the cost of communication overhead and lower scalability. To overcome these issues, we propose a novel multiclass intrusion detection system using FL at the subnetwork level. We employ various machine learning models, including Artificial Neural Networks, Convolutional Neural Networks, and Long Short-Term Memory to build an effective IoT IDS. The results indicate that our approach (a) detects attacks with high classification performance, (b) has low communication overhead at the server; (c) is more scalable and, (d) achieves a better precision of 99.88%, showcasing its potential for real-world IoT environments.