A Lightweight Machine Learning–Based Intrusion Detection System for Resource-Constrained IoT Networks

The rapid expansion of Internet of Things (IoT) devices in smart environments has significantly increased the vulnerability of network infrastructures to cyberattacks. Due to limited computational power, memory, and energy constraints, traditional security mechanisms are unsuitable for IoT ecosystems. Intrusion Detection Systems (IDS) based on conventional techniques often incur high computational overhead and fail to provide real-time protection. This paper proposes a lightweight machine learning–based intrusion detection framework specifically designed for resource-constrained IoT networks. The proposed approach integrates efficient feature selection techniques with lightweight machine learning classifiers to achieve high detection accuracy while minimizing computational complexity. Experiments conducted using the UNSW-NB15 dataset demonstrate that the proposed framework outperforms traditional IDS approaches in terms of accuracy, detection time, and false positive rate, making it suitable for real-time IoT security applications.