This paper aims to investigate how wireless sensor networks (WSNs) play a pivotal role in modern communication systems but remain highly vulnerable to denial-of-service (DoS) attacks due to their limited computational and energy resources.
This paper presents a lightweight, scalable framework that integrates fuzzy C-means clustering with a Variable Selection Ensemble Machine Learning Algorithm to detect and mitigate various types of DoS attacks, including black hole, grey hole, flooding and scheduling attacks.
The proposed framework uses effective feature selection and soft clustering to enhance detection accuracy while minimising resource overhead. Using the WSN-DS data set, the model demonstrates improved performance in terms of accuracy, packet delivery ratio, control overhead and network delay, outperforming conventional detection techniques.
The experimental evaluation confirms that the proposed method is both efficient and robust, making it suitable for deployment in resource-constrained WSN environments.
