Enhancing DDoS Attack Detection deploying Machine Learning in Software Defined Networking Environment
Anshika Sharma, Himanshi Babbar · 2024
The Internet of Things, or IoT, has become essential to contemporary industrial operations due to the widespread use of devices that are connected to the Internet in industrial settings. However industrial networks are now more vulnerable to a range of cybersecurity risks, such as Distributed Denial of Service (DDoS) attacks, due to their growing connectedness. In this paper, a novel method that combines hybrid machine learning (ML) algorithms with Software Defined Networking (SDN) approaches has been provided to identify DDoS attacks in the IoT space. By utilising the Edge-IIoT dataset, which incorporates information from IoT devices and SDN infrastructure, one may improve the precision and resilience of DDoS detection by combining feature extraction techniques with hybrid ML models. The suggested technique effectively detects DDoS assaults with few false positives, as seen by the experimental findings. This improves the safety stance of industrial networks against the ever-evolving threat landscape. A comparison of the suggested hybrid ML model’s accuracy, precision, recall, and F1-score has been made with the different algorithms including KNN [1], RF [2] and LR [3]. The proposed accuracy rises to 99.10%, while the accuracy rates of [1], [2] and [3] are, respectively, 97.15%, 97.99% and 98.39%.