A Comprehensive Survey on Low-rate DDoS Attacks Detection Based on Deep Learning
Zengguang Liu, Xiaochun Yin, Deyong Liu · 2024
Low-rate Distributed Denial-of-Service attacks, abbreviated as LDDoS, are experiencing an explosive and continuous growth in recent years. Meanwhile, people worked hard for making great contributions to prevent LDDoS attacks by deep learning methods. However, the existing surveys don't give detailed summaries on all aspects of neural network for LDDoS attacks. This paper investigates the trend and the challenges of LDDoS attacks. And then, it classifies the neural network methods into DL-based cluster, RL-based cluster, FL-based cluster and GNN-based cluster. And it presents the comprehensive survey along with state-of-the-art detection solutions for each of them. At last, the future research directions about developing an LDDoS detection system with broad-spectrum capabilities, cost-effective mechanisms and robustness against adversarial attacks are given.