DDoS Detection in SDN using machine learning and feature selection methods
Year: 2025
Type: article
Abstract: There are comprehensive studies on DDoS attack detection and IDS systems, but the numbers are not as considerable in SDN(Software Defined Networks). This paper compared different feature selection approaches on SDN DDoS datasets using variousmachine learning algorithms. We first use famous machine learning algorithms, then we use feature selection techniques toreduce the training time, and after that, we compare the results with other related studies. Our results show that pairing theDecision Tree and Genetic Algorithm works excellently in both of our datasets. We obtained the accuracy of 0.999983 and 1using InSDN and DDoS SDN datasets, respectively. Also, the XG Boost algorithm has remarkable results, but the Decision Treeis the best when it comes to the training time, with an average training time of 0.33177 and 0.78960 using InSDN and DDoSSDN datasets, respectively.
Authors: Rahmani Zardak, Siroos, Rasekh, Amir Hossein
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7111014674
FWCI:
Cites:
Cited by:
Related to: