Efficient DDoS Attack Detection through Lightweight Deep Learning Model in Cloud Computing Environment
Brij Bhooshan Gupta, Akshat Gaurav, Varsha Arya, Kwok Tai Chui · 2024
In the era of cybersecurity, Distributed Denial of Service (DDoS) attacks continue to pose a substantial threat to network availability and reliability. To counter these attacks, sophisticated detection mechanisms are imperative. This paper presents an innovative approach to DDoS attack detection using a lightweight deep learning model. Leveraging the KDDCup dataset, a well-established benchmark in intrusion detection research, we have achieved remarkable results with a training accuracy of 99.61% and a testing accuracy of 99.56%. Our proposed model emphasizes efficiency without compromising accuracy, addressing the resource constraints often associated with real-time attack detection. By employing a carefully designed architecture, we reduce computational complexity while maintaining a high level of discriminatory power between normal and attack traffic. This lightweight deep learning model not only demonstrates superior performance but also showcases its potential for rapid deployment in scenarios demanding swift and reliable attack detection.