Hybrid Ensemble Deep Learning Framework for Efficient DDoS Attack Detection in Software-Defined Networks

Abstract

In the ever-evolving digital era, robust network security and efficient intrusion detection mechanisms are paramount. With the widespread integration of technology into everyday activities and the increasing sophistication of cyber threats, innovative strategies are essential to protect network infrastructures. Distributed Denial of Service (DDoS) attacks, in particular, have become increasingly prevalent and sophisticated, posing a significant risk to network frameworks. Existing intrusion detection systems (IDSs) struggle to keep pace with these evolving threats, highlighting the need for innovative solutions to maintain network security. This paper proposes a hybrid deep learning (DL) framework that integrates multiple algorithms, including Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), Hierarchical Temporal Memory (HTM), and Convolutional Neural Networks (CNN), into an ensemble architecture to significantly enhance the accuracy and efficiency of DDoS attack detection. This novel ensemble strategy has the potential to strengthen security in Software-Defined Networks (SDNs) and combat DDoS attacks effectively. The framework's efficacy was validated using a widely recognized benchmark dataset, CIC-DDoS2019. Our experimental results demonstrated that the proposed algorithm achieved a maximum accuracy of 99.17%, outperforming other methods.