KS-entropy-based analysis of IP-network traffic in terms of time variation of dynamical properties

This paper presents an analysis of IP-network traffic in terms of the time variation of dynamical properties. To get a comprehensive view in analyzing IP-network traffic conditions, this paper used a self-organizing map, which is an effective tool to map high-dimensional data onto a low-dimensional domain. Furthermore, in checking time-varying properties of measured network traffic, this paper employed Kolmogorov–Shinai entropy (KS-entropy), which provides a quantitative measure of the degree of unpredictability of observed time-series signals. In applying this method to traffic analysis, this paper performed two kinds of traffic measurement: one based on IP-network traffic flowing into NTT Musashino R&D center (Tokyo, Japan) from the Internet and the other based on IP-network traffic flowing through at an interface point between an access provider (Tokyo, Japan) and the Internet. Based on sequential measurements of IP-network traffic at two locations, this paper derived the average throughput, the standard deviation of the average throughput, and the value of KS-entropy as input to self-organizing map. We visually confirmed that the traffic data could be projected onto the self-organizing map in accordance with the traffic properties, resulting in a combined depiction of multiple factors that describe dynamical properties of the target system.