Enhancing Ddos Attack Detection Using Machine Learning

Abstract Denial of Service (DoS) attacks aim to disrupt the availability of a machine or network resource, preventing legitimate users from accessing services. A more advanced variant, Distributed Denial of Service (DDoS), involves multiple compromised systems attacking a target, often overwhelming it with malicious traffic. DDoS attacks are executed for various purposes, including financial extortion, political motives, or simply for disruption. This paper presents an advanced DDoS detection system designed for Software-Defined Networking (SDN) environments, utilizing Deep Learning techniques. Specifically, our model leverages the Gated Recurrent Unit (GRU) algorithm to analyze network traffic patterns and effectively detect DDoS attacks. By utilizing the CICDDoS2019 dataset, our proposed approach demonstrates superior accuracy compared to traditional methods, reinforcing the security of SDN networks against evolving cyber threats.