Network-based Intrusion Detection Datasets: A Survey

Abstract

The increasing reliance on computers and the internet has increased the need for protection against attacks and security threats. A network-based intrusion detection system (NIDS) is one of the promising research fields that could highly contribute to cyber-attacks detection and mitigation. Datasets are necessary for NIDS training and evaluation. This paper provides a comprehensive survey of publicly available NIDS datasets. The study covers the most used NIDS datasets, each dataset is briefly described along with its main properties and shortcomings. Moreover, the paper highlights the most important dataset characteristics and provides a clear datasets comparison accordingly. Finally, the paper discusses observations and provides some recommendations to help researchers identify future work.