25-10-2012, 01:10 PM
Self-organizing Map-based Analysis of IP Packet Traffic in Terms of Time Variation of Self-Similarity
ABSTRACT
This paper describes an analysis of the IP-packet traffic in terms of time variation of self-similarity. To take compound
factors into account in analyzing IP-packet traffic, we used a self-organizing map, which provides a way to map
high-dimensional data onto a low-dimensional domain. Based on sequential measurements of IP-packet traffic flowing into
NTT Musashino R&D center from the Internet, we derived the corresponding values of the Hurst parameter, which can
indicate the degree of long-range dependence (LRD) or asymptotic similarity of network traffic. We visually confirmed that
the traffic data could be projected into the self-organizing map in accordance with the traffic properties, resulting in a combined
depiction of both effects of the degree of LRD and of the bandwidth utilization rates in a two-dimensional domain.