AI-Powered intrusion detection system for IoT networks using hybrid learning models

With the explosive growth in Internet of Things (IoT) networks, it has become increasingly difficult to ensure network security because connected devices are heterogeneous and distributed. Conventional security systems frequently lack effectiveness in detecting advanced and inuring threats. This paper introduces an IoT network-specific intrusion detection system (IDS) based on artificial intelligence (AI), employing a hybrid learning model that integrates the benefits of ensemble methods and deep learning models. It adopts a stacked model consisting of XGBoost and long short-term memory (LSTM) networks for effective real-time detection of anomalies in network traffic. Training and performance analysis are conducted based on publicly sourced CICIDS2017 and TON\_IoT datasets. Experimental findings show that the baseline classifiers are significantly surpassed by the proposed model in terms of detection precision, recall, and F1-score. It has potential application in real-world IoT settings as a scalable and adaptability-driven proactive threat detection solution.