Communication models with distributed transmission rates and buffer sizes

Abstract

The paper is concerned with the interplay between network structure and traffic dynamics in a communications network, from the viewpoint of end-to-end performance of packet transfer. We use a model of network generation that allows the transition from random to scale-free networks. Specifically, we are able to consider three different topologycal types of networks: (a) random; (b) scale-free with γ=3; (c) scale free with γ=2. We also use an LRD traffic generator in order to reproduce the fractal behavior that is observed in real world data communication. The issue is addressed of how the traffic behavior on the network is influenced by the variable factors of the transmission rates and queue length restrictions at the network vertices. We show that these factors can induce drastic changes in the throughput and delivery time of network performance and are able to counter-balance some undesirable effects due to the topology.