R.A. Dunaytsev, A.V. Krendzel, Y.A. Koucheryavy, and J.J. Harju (Finland)
Web traffic estimation, user behavior.
Providing an access to the Internet and support of high speed data transfer nowadays are the key issues for the most telecommunication operators. One of the perspective trends today is providing an Internet access for home users over home (campus) networks. At the stage of planning and preparation of system project for deployment of a network it is necessary to estimate traffic volume which will be generated by Web users in this network. Unfortunately a large diversity of Web content and Web users' behavior make the estimation problem quite complex. In this paper we present estimation method of Web traffic volume based on a probabilistic model of events initiated by Web users in a home network. It is proposed to make both a decomposition of Web content into some types and a distribution of Web users into some groups in accordance with a demand for different Web content. This approach allows segregating the common Web traffic into several segments and calculating the data traffic generated by different types of Web users in a home network while requesting different types of Web content.
Important Links:
Go Back
