NetBIOS DDoS Attacks Detection With Machine Learning Classification Algorithms

Abstract

Distributed Denial of Service (DDoS) is a cyberattack in which the attacker makes a system or network resources unavailable to the intended audience temporarily or permanently. NetBIOS DDoS is a reflection based DDoS attack makes the victim system unavailable to communication other NetBIOS hosts. Service unavailable makes huge impact in terms of financially and reputational. So DDoS attack detection at early stage is more important. This study proposed machine learning algorithms for NetBIOS DDoS attack detection. Experiments are performed on NetBIOS_DrDoS dataset, which is collected from CIC-DDoS2019 evaluation dataset. In order to reduce computational overheads features are selected by Correlation methods. This study uses Pearson, spearman and Kendall correlation methods to select uncorrelated features. This study evaluated Logistic regression, Decision tree, Random forest, Ada Boost, Gradient Boost, K-Nearest Neighbour, Naive-Bayes and Multilayer perceptron classification algorithms with Pearson, Spearman and Kendall uncorrelated feature subsets in order to classify attack and benign class labels. Multilayer perceptron with Pearson uncorrelated feature subset gives the best performance for NetBIOS DDoS attack detection.