Intrusion Detection System using Modified Arithmetic Optimization Algorithm for Large-scale IoT

This paper introduces an Intrusion Detection System (IDS) designed to enhance security in expansive IoT networks. The proposed IDS leverages one of thee most popular Machine Learning method with a modified Arithmetic Optimization Algorithm (AOA) to identify the optimal feature set and deal with the binary nature of this optimization problem. This set of features is then employed utilizing the K-Nearest Neighbors Algorithm. Our study utilized a dataset, called BotIoT2018. Notably from the results, the proposed IDS using the modified AOA demonstrated a remarkably precise and resilient detection model tailored for IoT networks.