An Optimized Ensemble IDS for DDoS Detection in IoT Environment

Abstract

Due to resource scarcity, efficient and accurate Intrusion Detection Systems (IDS) are crucial for safeguarding Internet of Things (IoT) settings from Distributed Denial of Service (DDoS) attacks. This study introduces an optimised ensemble (OE-IDS) framework that is both efficient and lightweight. It employs Recursive Feature Elimination (RFE) for feature selection and Principal Component Analysis (PCA) for feature extraction. The ensemble-based eXtreme Gradient Boosting (XGB) classifier outperforms conventional ensemble algorithms in detection rates when supplied with the appropriate information. The proposed model has been evaluated on many DDoS attack detection datasets, and experimental findings indicate that it outperforms contemporary strategies in recall, precision, F1-score, and false positive rate.The model has an efficacy of over 99% accuracy and a 99% true positive rate, making it a suitable option for DDoS attack detection.