An ensemble-based IDS for edge computing network
Amit Kumar · International Journal of Internet of Things and Cyber-Assurance · 2023
Advanced and sophisticated cyber-attacks are rapidly evolving in the context of edge computing and IoT networks. The distributed nature of these networks, the large number of connected devices, and the potential heterogeneity of the devices and protocols involved create an expanded attack surface for cybercriminals. Hence, this research aims to design and develop efficient and robust ensemble-based intrusion detection system (IDS) for detecting attacks in edge computing networks. The proposed model employs a voting classifier in four different classification models: decision tree, random forest, extra tree and K-nearest neighbours. This paper uses the recursive feature elimination technique to select the relevant and practical features. The novelty of the paper is the detection of new types of intrusion or attacks in network traffic. Therefore, the recently released dataset CIC-IDS-2017 was used to evaluate the proposed IDS model. Analytical results showed that the proposed model has achieved high accuracy of 99.92%, recall of 99.93%, the precision of 99.97%, f1-score of 99.95%, and a false-positive rate of 0.001%, as compared to the previous studies.