Abstract: “In this paper, the design of a platform for the generation of NIDS evaluation datasets is proposed. The increasing sophistication of network attacks requires high volumes of high quality and realistic test data that can faithfully represent today's security threats to information systems. Our platform is based on the generation of aggregate traffic through the reproduction of real offline traffic and the injection of anomalous traffic. The use of free software tools in Linux environment is proposed. This solution provides an answer to the lack of datasets suitable for the study of Intrusion Detection Systems.”

Keywords: Datasets generation, NIDS, snort, test-bed