Systematic Literature Review on Distributed Denial of Service Attack

Abstract

In recent years, advanced threats against the internet and systems, such as Distributed Denial of Service (DDoS) attacks, are increasing with traditional machine learning-based intrusion detection systems (IDSs) often fail to efficiently detect such attacks when unbalanced datasets are used for IDS training. This work provides a system literature review (SLR) on Distributed Denial of Service (DDoS) attack detection, by presenting a detailed assessment of the approaches and methodologies taken throughout the nine years, emphasizing machine learning and deep learning techniques. The review examines various approaches, including token embedding for feature extraction, transformer-based models, and hybrid detection techniques. Despite improvements, the study highlights ongoing challenges, such as computational complexity and the need for enhanced solutions like blockchain-based detection systems. Open research gaps and future directions, including the refinement of detection algorithms for evolving DDoS tactics, are also discussed, offering a comprehensive resource for researchers aiming to improve DDoS mitigation.