Threat Detection Using MLP for IoT Network

Threat Detection Using MLP for IoT Network

연구 분야: Safety

학회: World Congress in Computer Science, Computer Engineering & Applied Computing

While the popularity of IoT networks has grown significantly, they remain highly vulnerable to various cyber-attacks. These attacks can disrupt services, compromise sensitive data, and damage the integrity of IoT ecosystems. Machine learning (ML) techniques, particularly deep learning (DL) models, have been employed to effectively detect and mitigate such threats by identifying abnormal patterns in network traffic. In this paper, we propose utilizing a multilayer perceptron (MLP)-based machine learning model to detect cyber-attacks by using the NF-ToN-IoT dataset, which contains a diverse set of cyber-attacks. The MLP model has been trained and optimized to distinguish between normal and malicious activity. Our results demonstrate the effectiveness of this approach, with the MLP achieving a training accuracy of approximately 97% and a test accuracy of around 95–97%. This high accuracy indicates the model’s capability to generalize well across different attacks while creating a robust solution for real-time threat detection and mitigation in IoT networks.