Intrusion Detection Systems: Survey, Taxonomy and Challenges

Network threats and hazards have been rapidly developing in recent years. Intrusion Detection System (IDS) is an effective and powerful network security system used to detect unauthorized and abnormal network traffic. Due to the current research mainly based on public datasets, however, public datasets have many limitations. The purpose of this article is to discuss various machine learning methods and common datasets used for intrusion detection, and to provide some suggestions for future IDS research. For this purpose, this article discusses the concept and taxonomy of intrusion detection systems, commonly used intrusion detection systems, citation-based analysis of benchmark datasets, machine learning techniques for intrusion detection, and generic method of researches for intrusion detection systems. Finally, the challenges faced by IDS and future research directions were proposed.