Naïve bayes classifier to mitigate the DDoS attacks severity in Ad-Hoc networks

Abstract

Ad-Hoc networks are becoming more popular due to their unique characteristics. As there is no centralized control, these networks are more vulnerable to various attacks, out of which Distributed Denial of Service (DDoS) attacks consider as more severe attacks. DDoS attack detection and mitigation is still a challenging issue in Ad-Hoc Networks. The existing solutions find the fixed or dynamic threshold value to detect the DDoS attacks without any trained data. Very few existing solutions use machine learning algorithms to detect these attacks. However, existing solutions are inefficient to handle when DDoS attackers perform this attack through bursty traffic, packet size, and fake packets flooding. We have proposed DDoS attack severity mitigation solution. Out DDoS mitigation solution consists of a new network node authentication module and naïve Bayes classifier module to detect and isolate the DDoS attack traffic patterns. Our simulation results show that naïve Bayes DDoS attack traffic classification outperforms in the hostile environment and secure the legitimate traffic from DDoS attack. © 2020, Kohat University of Science and Technology.