Secure Cloud Assisted Fog Computing Framework for IoT Applications Using Ensemble Learning

Secure Cloud Assisted Fog Computing Framework for IoT Applications Using Ensemble Learning

연구 분야: Cryptography

학회: SN Computer Science

The Internet of Things (IoT) paradigm has revolutionized various domains by interconnecting physical devices to the internet, enabling seamless communication and data exchange. The pervasive deployment of IoT devices introduced significant security level challenges, making them susceptible to various cyber-attacks in the IoT networks. Traditional centralized security mechanisms often failed to provide adequate protection due to dynamic, unpredictable, and heterogeneous environment of IoT networks. Hence, decentralized and fog-based attack detection system, has come out as promising solutions to enhance IoT security. This paper proposed the framework with aims to detect and mitigate various types of attacks in real-time ensuring integrity and trustworthiness of IoT systems. The proposed work operates by collecting and analysing data from IoT devices at the fog nodes, where local decision-making and anomaly detection take place. The secure framework can identify deviations from normal behaviour patterns, indicative of potential attacks or anomalies using Machine Learning (ML) Model. The NSL-KDD dataset is utilized to train and assess the performance of our model with respect to influential parameters such as accuracy, F-score, precision, and recall. In addition, our proposed model provides a more accurate classification of Denial of Service (DoS), Root to Local (R2L) and user to root(U2R), and Probe attacks. The proposed model achieved an accuracy rate of 99.80%, recall rate of 99.70%, precision rate of 99.85%, F1 score of 99.77%, and the lowest False Alarm rate of 0.7396%. Consequently, the integrated Ensemble model generated the fewest false alarms attack on the extensive data, surpassing the baseline methods in terms of both recall and precision.