Résumé:
Intrusion Detection Systems (IDS) play a vital role in strengthening the security of digital environments, particularly in the Internet of Things (IoT), where connected devices are often vulnerable to increasingly sophisticated attacks. This work aims to design and evaluate an intelligent IDS capable of effectively identifying malicious activities within IoT networks using deep learning techniques. To achieve this, we developed a classification model based on a Convolutional Neural Network (CNN), trained on the Edge-IIoTset dataset—a recent and representative dataset of threats targeting industrial IoT systems. The proposed pipeline includes key steps such as data preprocessing and normalization, class re-balancing to address distribution imbalance, the design of an optimized CNN architecture, and performance analysis. Experimental results demonstrate that the proposed CNN model achieves high performance in terms of accuracy, recall, and F1-score, while maintaining a high detection rate and a low false alarm rate. This study highlights the effectiveness and robustness of CNN-based approaches for intrusion detection in IoT environments and suggests future improvements in terms of scalability and data privacy.
