Securing IoT Networks: An Intrusion Detection Framework Using Convolutional Neural Networks
The proliferation of the Internet of Things (IoT) has significantly transformed human lives, embedding interconnected devices across various sectors such as industry, healthcare, agriculture, and transportation. However, this growth has also attracted cybercriminals, leading to a surge in cyber-attacks targeting IoT devices. Intrusion Detection Systems (IDS) have emerged as a critical defense mechanism against such attacks. In this paper, we propose a Convolutional Neural Network (CNN)-based framework for IoT intrusion detection. The framework employs CNNs for automatic feature selection from raw network traffic data, enhancing classification performance and reducing dimensionality. We utilize the IoTID20 benchmark dataset for training and testing the proposed framework. Experimental results demonstrate the effectiveness of our approach, achieving high accuracy of 99.93%, precision of 1.00%, recall of 99.92%, and F1-score of 99.98% for binary classification. For multi-class category classification, the framework achieves an accuracy of 99.51%, precision of 99.98%, recall of 99.87%, and F1-score of 99.93%. For multi-class sub-category classification, the accuracy is 92.46%, precision is 97.25%, recall is 90.54%, and F1-score is 92.83%. Our proposed framework outperforms existing methods, making it a promising solution for IoT network security
Download Paper