The rapid growth of the Internet of Things (IoT) has introduced considerable security challenges, as the large number of interconnected devices presents multiple vulnerabilities that can be exploited by malicious actors. Network Intrusion Detection Systems (NIDS) are crucial in detecting and mitigating these threats, but current detection systems face difficulties in achieving a balance between maintaining high detection accuracy and minimizing false-positive rates. Many traditional methods for feature selection in NIDS are inefficient due to selecting too many irrelevant or redundant features, leading to poor performance. This paper addresses these challenges by proposing a hybrid feature selection method that combines the Arithmetic Optimization Algorithm (AOA) and the Equilibrium Optimizer (EO), termed AOAE, specifically designed for IoT environments. The main objective is to enhance the detection capabilities of NIDS while reducing the number of features used, thus improving system efficiency. The proposed method is evaluated on multiple benchmark datasets, including NSL-KDD, UNSW-NB15, Bot-IoT, and CICIDS2017, among others. The results indicate that AOAE substantially enhances detection accuracy, lowers false alarm rates, and optimizes feature selection by identifying a minimal yet effective subset of features. The proposed solution outperforms several state-of-the-art methods in terms of accuracy, precision, and computational efficiency, making it a promising and practical approach for improving IoT security systems, especially under simulated real-world conditions.
Makhadmeh et al. (Sun,) studied this question.