Scholarlink

With the rapid development of IoT technology, network environments have become increasingly complex, posing significant challenges to the accuracy and real-time performance of traditional intrusion detection systems. To address the complex graph-structured features of device communications, this paper proposes a hybrid intrusion detection model combining GCN and BiLSTM. The model constructs a time- and protocol-based interaction graph, leveraging GCN to extract high-order structural features and BiLSTM to capture the temporal dynamics of communication behavior, enabling more effective detection of IoT attacks. Experimental results show that the proposed method outperforms traditional models in accuracy and other aspects, providing theoretical and technical support for IoT security.

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

Intrusion Detection Method of Internet of Things Dynamic Graph Structure Based on GCN-BiLSTM Hybrid Network

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Changqin Xu,
  • Guangfu Wu,
  • Zhihui Cai

摘要

With the rapid development of IoT technology, network environments have become increasingly complex, posing significant challenges to the accuracy and real-time performance of traditional intrusion detection systems. To address the complex graph-structured features of device communications, this paper proposes a hybrid intrusion detection model combining GCN and BiLSTM. The model constructs a time- and protocol-based interaction graph, leveraging GCN to extract high-order structural features and BiLSTM to capture the temporal dynamics of communication behavior, enabling more effective detection of IoT attacks. Experimental results show that the proposed method outperforms traditional models in accuracy and other aspects, providing theoretical and technical support for IoT security.