A Systematic Literature Review on Machine Learning and Deep Learning Approaches for Detecting DDoS Attacks in...

Area of Science:

• Computer Science
• Network Security
• Artificial Intelligence

Background:

• Software-defined networking (SDN) offers flexibility but is vulnerable to distributed denial of service (DDoS) attacks.
• Existing DDoS detection methods in SDN face significant challenges, remaining an open research area.
• Machine learning (ML) and deep learning (DL) show promise for enhancing SDN security against sophisticated threats.

Purpose of the Study:

• To systematically review and analyze ML, DL, and hybrid approaches for DDoS attack detection in SDN.
• To identify trends, common methodologies, and prevalent datasets used in SDN DDoS detection research.
• To highlight existing challenges and open issues in the field.

Main Methods:

• A systematic literature review (SLR) protocol was followed, involving automatic and manual searches across eight databases.
• The review covered studies published between 2014 and 2022.
• Seventy primary studies were identified and analyzed.

Main Results:

• A significant increase in research on SDN DDoS detection using ML/DL approaches was observed in recent years.
• Ensemble, hybrid, and single ML-DL models are the predominant methods employed.
• Private synthetic and unrealistic datasets are commonly used for evaluating detection approaches.

Conclusions:

• The field of SDN DDoS detection using ML/DL is rapidly evolving.
• Current evaluation practices, particularly dataset usage, require improvement for more robust and realistic assessments.
• Further research is needed to address identified challenges and enhance the effectiveness of SDN security.