IoT Intrusion Detection System Leveraging PSO And Sequential Forward Feature Selection

Abstract

The Internet of Things (IoT) plays an important role with wide application in various fields. However, the sustainability of IoT is limited by several challenges and security-related problems are the most dangerous. Therefore, in this work, the ability of two distinct feature selection methods including Particle Swarm Optimization (PSO)-based and sequential forward-based approaches are leveraged and compared to find the most appropriate feature extraction technique and the most valuable feature set of the IoT-23 dataset to construct an efficient Intrusion Detection System (IDS) validated by multiple Machine Learning (ML) algorithms. The prospect of solving problems in practical environments is demonstrated through the achieved results.