Property Analysis of Heavy-tailed Traffic by Simulator

The scale-invariant burstiness or self-similarity has been found in real network. Relation between self-similarity and network and/or system parameter is mainly discussed in the context of end-to-end data transmission environment, and this self-similarity is mainly caused by the file size of Web servers or the duration of user sessions. On the other hand, the accesses to Web servers are not uniformly distributed, but are confined to specific Web servers such as search engine sites or portal sites, and other elements of network parameters are not clearly described in the context of self-similarity. In this paper, we have investigated the property of self-similar traffic varying the network environment using the network simulator. After analyzing the simulated results, following properties were extracted. Firstly, heavy-tailed distribution of file size induces self-similarity, especially in the case of small alpha. Secondly, distribution with power law of data rate emphasize self-similarity. Finally, greater error rate induces to activate the fluctuation of the throughput, and remains self-similar property