Qualitative evaluation of denial of service datasets

• Carla O. Camargo Universidade Federal de Uberlândia
• Elaine R. Faria Universidade Federal de Uberlândia
• Bruno B. Zarpelão Universidade Estadual de Londrina
• Rodrigo S. Miani Universidade Federal de Uberlândia

Abstract

Denial of service attacks has shown to be one of the principal problems in the information security field on the Internet. Numerous works propose methodologies and tools to detect and mitigate DDoS attacks, and many of them use datasets to validate their proposal. The goal of this work is to identify undesirable, ideal and feasible characteristics of datasets used in DDoS researches. Moreover, an analysis of some datasets widely used in this research field in the light of the proposed characteristics will be made. The results have shown that datasets used broadly by researchers nowadays are outdated and unreproducible. On the other hand, there are more detailed and specific datasets on the attacks under discussion.

Keywords: Information Security, datasets, denial of service