Mitigating DDoS Attacks in SDN Using Machine Learning and Deep Learning: Challenges and Opportunities

EasyChair Preprint 15323

Abstract

Distributed Denial of Service (DDoS) attacks represent a significant and evolving threat within the realm of cybersecurity. In Software-Defined Networking (SDN), leveraging Machine Learning (ML) and Deep Learning (DL) techniques has proven to be a promising strategy for detecting and mitigating these attacks. This systematic literature review (SLR) provides a comprehensive analysis of current research in this field. The findings illustrate the versatility of ML and DL models in adapting to various attack vectors, their capacity for real-time decision-making, and their resilience against adversarial threats. However, challenges remain, including optimizing performance, ensuring scalability, enhancing resource efficiency, improving model interpretability, and addressing ethical considerations. The SLR highlights the critical importance of having labeled datasets, fostering ethical and legal awareness, and preparing network administrators for collaborative engagement with ML and DL-based DDoS mitigation systems. As the cybersecurity landscape continues to evolve, this review underscores the ongoing effort required to fully exploit the potential of ML and DL in protecting SDN networks against DDoS threats

Keyphrases: DDoS attacks, Deep Learning Algorithms, Machine Learning Algorithms, SDN

@booklet{EasyChair:15323,
  author    = {G.B Veeresh and Vanita Jaitly and V Lokeswara Reddy},
  title     = {Mitigating DDoS Attacks in SDN Using Machine Learning and Deep Learning: Challenges and Opportunities},
  howpublished = {EasyChair Preprint 15323},
  year      = {EasyChair, 2024}}