Modelling of network traffic usage using self-organizing maps techniques

Abstract

Currently, the Internet is one important part of the campus infrastructure that supports the teaching and learning activities. The important part of the internet facility is the provided bandwidth as Bandwidth management to teaching and learning is indispensable. In this study, an analysis and cluster of the university internet traffic is required as bandwidth management decision support. Therefore, Self-Organizing Maps (SOM) as a clustering algorithm bandwidth usage was implemented. The results showed that the SOM method can perform clustering. Furthermore, the clustering result could be a recommendation management bandwidth for network administrator.