An intrusion detection system in internet of things using grasshopper optimization algorithm and machine learning algorithms

Abstract

The Internet of Things (IoT) has emerged as a foundational paradigm supporting a range of applications, including healthcare, education, agriculture, smart homes, and, more recently, enterprise systems. However, significant advancements in IoT networks have been impeded by security vulnerabilities and threats that, if left unaddressed, could hinder the deployment and operation of IoT based systems. Detecting unwanted activities within the IoT is crucial, as it directly impacts confidentiality, integrity, and availability. Consequently, intrusion detection has become a fundamental research area and the focus of numerous studies. An intrusion detection system (IDS) is essential to the IoTs alarm mechanisms, enabling effective security management. This paper examines IoT security and introduces an intelligent two-layer intrusion detection system for IoT. Machine learning techniques power the system's intelligence, with a two layer structure enhancing intrusion detection. By selecting essential features, the system maintains detection accuracy while minimizing processing overhead. The proposed method for intrusion detection in IoT is implemented in two phases. In the first phase, the Grasshopper Optimization Algorithm (GOA) is applied for feature selection. In the second phase, the Support Vector Machine (SVM) algorithm is used to detect intrusions. The method was implemented in MATLAB, and the NSLKDD dataset was used for evaluation. Simulation results show that the proposed method improves accuracy compared to other approaches.