AI-based IoT Intrusion Detection Using Machine Learning

Abstract

The development of smart home technology offers an unprecedented level of convenience, fundamentally transforming traditional home environments, but it also introduces the system to various cybersecurity risks such as unauthorized access and data leakage. With the expansion of the application scope of the Internet of Things network and the increase in the number of device deployments, traditional static defense methods have proven inadequate in the timely identification of complex intrusion behaviors and protecting network security. In response to this situation, this study proposes an intrusion detection system based on artificial intelligence, which can autonomously identify abnormal network traffic and isolate infected devices. The study used the CIC-IoT2023 and IoT-23 datasets to train and evaluate three machine learning models: logistic regression, support vector machine (RBF kernel), and random forest. The research results indicate that all three models exhibit high performance in terms of accuracy, robustness, and response speed, verifying the feasibility of applying artificial intelligence and machine learning to the security monitoring system of smart home systems. This study proposes a novel approach for constructing intelligent Internet of Things systems with self-defense capabilities.