A Deep Learning-Based Model for IoT Traffic Intrusion Detection
Jingwei Fan, Xiaochong Tian, Shicheng Yan, Shanshan Dong, Xiaobo Huang, Zhibo Yang · 2024
With the rapid proliferation of Industrial Internet of Things (IIoT) devices, the increasingly diverse and complex network traffic presents significant challenges for intrusion detection. To address this issue, this paper proposes a deep learning model that integrates Temporal Convolutional Networks (TCN), Bidirectional Long Short-Term Memory (BiLSTM), and an attention mechanism. TCN excels at capturing long-term dependencies in data, BiLSTM enhances the model's understanding of sequential information, and the attention mechanism focuses on extracting and weighting key features, significantly improving detection accuracy. We conducted multiclass classification experiments on the Edge-IIoTset dataset, evaluating the model's performance using accuracy, precision, recall, and F1-score metrics. The experimental results demonstrate that the proposed model exhibits outstanding performance in IIoT traffic intrusion detection tasks.