Implementation of Machine Learning Algorithms to Detect DoS Attack in WSN

This paper presents a comprehensive work on the detection of Denial of Service (DoS) attacks in Wireless Sensor Networks (WSNs) using machine learning algorithms. It provides an overview of WSNs, the nature of DoS attacks, and the specific challenges of DoS attacks in WSN environments. Utilizing the WSN-DS dataset, the implementation includes data preprocessing techniques such as Principal Component Analysis (PCA) and Min-Max Scaling. The approach uses Convolutional Neural Networks (CNN) and XGBoost algorithms forming four combination algorithms: CNN coupled with PCA, CNN, XGBoost with PCA, and XGBoost to classify and detect DoS attacks. Among the models tested, XGBoost with Min-Max Scaling demonstrated superior accuracy of 99.7% compared to the other models.