Mitigating Denial of Service Attacks in Fog-Based Wireless Sensor Networks Using Machine Learning Techniques
It addresses security issues in fog-based wireless sensor networks, which are critical for industrial applications, but the approach is incremental as it applies existing ML methods to a known problem.
This paper tackles the problem of detecting denial of service attacks in wireless sensor networks by proposing machine learning models, specifically decision trees and XGBoost, achieving a true positive rate of up to 99.01% and a false positive rate as low as 0.99% with XGBoost.
Wireless sensor networks are considered to be among the most significant and innovative technologies in the 21st century due to their wide range of industrial applications. Sensor nodes in these networks are susceptible to a variety of assaults due to their special qualities and method of deployment. In WSNs, denial of service attacks are common attacks in sensor networks. It is difficult to design a detection and prevention system that would effectively reduce the impact of these attacks on WSNs. In order to identify assaults on WSNs, this study suggests using two machine learning models: decision trees and XGBoost. The WSNs dataset was the subject of extensive tests to identify denial of service attacks. The experimental findings demonstrate that the XGBoost model, when applied to the entire dataset, has a higher true positive rate (98.3%) than the Decision tree approach (97.3%) and a lower false positive rate (1.7%) than the Decision tree technique (2.7%). Like this, with selected dataset assaults, the XGBoost approach has a higher true positive rate (99.01%) than the Decision tree technique (97.50%) and a lower false positive rate (0.99%) than the Decision tree technique (2.50%).