Application of Probabilistic Data Structures for Detecting Distributed Denial of Service (DDoS) attacks

Abstract

This paper investigates the utilization of probabilistic data structures as a novel approach for enhancing the detection and mitigation of DDoS attacks. Traditional methods often struggle to keep pace with the evolving nature of DDoS attacks, leading to high false positive rates and scalability challenges. In contrast, probabilistic data structures offer efficient, scalable, and memory efficient solutions for analyzing large volumes of network traffic and identifying DDoS related patterns and anomalies. Key probabilistic data structures include Bloom filters, Count Min Sketches, and HyperLogLog, each providing unique capabilities for detecting DDoS attacks based on set membership, frequency estimation, and cardinality approximation, respectively. This paper examines the strengths, limitations, and practical considerations of leveraging probabilistic data structures for DDoS detection through a comprehensive analysis of methodology, experimental results, case studies, challenges, and future directions. By exploring the application of probabilistic data structures, this research aims to provide valuable insights and recommendations for cybersecurity practitioners, researchers, and stakeholders involved in combating DDoS attacks and safeguarding critical digital assets.