SentinelIoT: An AI-Powered Intrusion Detection System for IoT Networks Using Random Forest Classification
Year: 2026
Type: preprint
Abstract: The rapid growth of Internet of Things (IoT) devices has introduced significant cybersecurity challenges due to their limited computational resources, heterogeneous communication protocols, and exposure to network-based attacks. This paper presents SentinelIoT, an AI-powered intrusion detection system designed to identify suspicious network activity using machine learning techniques. The proposed system employs a Random Forest classifier to distinguish between normal and malicious network traffic. Its architecture incorporates dataset-based traffic analysis, data preprocessing, feature extraction, machine learning classification, alert generation, and a web-based security dashboard. The system is designed using Python, Scikit-learn, FastAPI, SQLite, and modern web technologies. The research discusses publicly available IoT intrusion detection datasets, including CICIoT2023 and IoT-23, and presents a methodology for evaluating classification performance using accuracy, precision, recall, F1-score, and confusion matrices. It also examines the technical challenges of ensuring compatibility between offline dataset features and features extracted from live network traffic. SentinelIoT aims to provide a modular and extensible framework for IoT network security monitoring, with potential future integration into Raspberry Pi, ESP32, and other IoT environments. The paper presents the proposed architecture, implementation methodology, evaluation framework, and limitations without claiming unverified experimental performance. Keywords: Internet of Things, IoT Security, Intrusion Detection System, Machine Learning, Random Forest, Network Security, Cybersecurity, Traffic Classification, SentinelIoT.
Authors: Sompura, Avsar, Gopalani, Divij
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7236840689
