A Systematic Review of Machine Learning-Based Models for IoT Network Security in Intrusion Detection Systems

Abstract

The rapid growth of Internet of Things (IoT) devices has increased exposure to cyber threats such as DDoS, botnets, flooding, and brute-force attacks, making intrusion detection systems (IDS) essential for network security reinforcement. This systematic review analyzes recent studies (2022–2024) focusing on machine learning, ensemble, and hybrid deep learning IDS models. The review highlights widely adopted techniques including Random Forest, XGBoost, Extremely Randomized Trees, deep neural networks, autoencoders, and CNN–LSTM architectures. Results show that ensemble methods often achieve the highest accuracy (up to 99.7%), while hybrid deep learning improves spatio-temporal traffic analysis. The study also identifies key gaps such as outdated datasets, limited real-world deployment, computational overhead, and lack of explainability, providing future research directions for scalable and efficient IoT security.