WSEAS Transactions on Information Science and Applications
Print ISSN: 1790-0832, E-ISSN: 2224-3402
Volume 22, 2025
IoT Network Security based on Intrusion Detection System using Stacked Ensemble
Authors: , , ,
Abstract: The rapid evolution of IoT networks has led to an increasing number of devices connecting to the internet, exposing them to various cyber threats. Detecting intrusions in IoT environments is essential but challenging. Network Intrusion Detection Systems are vital in analyzing network traffic to differentiate normal and malicious activities without compromising security. However, the abundance of benign traffic complicates accurate detection. To overcome this challenge, we propose an Ensemble-based Network Intrusion Detection Systems framework, where five Machine Learning classifiers are combined through a Stacking approach and with nature-inspired feature selection techniques to enhance the detection effectiveness. The performance of the proposed model was compared to four base models - Random Forest, Extra Trees, AdaBoost, and Gradient Boosting - in terms of several metrics. The experimental results on the CICIoT2023 dataset reveal that the proposed stacking model consistently outperforms the base classifiers across all evaluation metrics.
Search Articles
Keywords: Stacked Ensemble, Intrusion Detection, Internet of Things, Cybersecurity, Machine Learning, Deep Learning
Pages: 466-473
DOI: 10.37394/23209.2025.22.38