USB-IDS-1: a Public Multilayer Dataset of Labeled Network Flows for IDS Evaluation

Abstract

The scarceness of publicly available data from real-life operational networks is a long-standing problem for the security research community. Many public intrusion detection datasets have spread in the literature; however, they may lack relevant details concerning the victim servers and applications. This paper describes USB-IDS-1, a novel public intrusion detection dataset developed at the University of Sannio at Benevento, Italy. The dataset considers both network traffic and application-level facets, such as performance measurements, configuration and defense modules of the victim server under attack. The paper describes the collection environment, provides key insights into traffic data and demonstrates the impact of the attacks adopted against the server in hand.