Generating statistical insights into network behavior using SKETURE

Article type: Research Article

Authors: Al-Haj Baddar, Sherenaz^{a; *} | Merlo, Alessio^b | Migliardi, Mauro^c

Affiliations: [a] KASIT – The University of Jordan, 11942, Amman, Jordan. E-mail: [email protected] | [b] DIBRIS – University of Genoa, 16145, Genoa, Italy. E-mail: [email protected] | [c] DEI – University of Padua, Padua, 35131, Italy. E-mail: [email protected]

Correspondence: [*] Corresponding author. E-mail: [email protected].

Abstract: Tools that generate informative and efficient statistical summaries of nodes’ activities in a given network have become crucial for robust behavioral anomaly detection. Yet, addressing network abnormalities and threats should not be done at the expense of users’ privacy. In this study we illustrate the use of SKETURE, a packet analysis tool leveraging a sketch-based architecture, in summarizing the behavior of nodes in a real campus network for a whole month, without breaching users’ privacy. Moreover, we share some insights into this network which were compiled using SKETURE.

Keywords: Traffic analysis tools, statistical modeling, streaming algorithms, sketch-based analysis

DOI: 10.3233/JHS-160539

Journal: Journal of High Speed Networks, vol. 22, no. 1, pp. 65-76, 2016