2007 International Symposium on Applications and the Internet (SAINT'07)
On the Effect of Scale-Free Structure of Network Topology on End-to-End Performance
Hiroshima, Japan
January 15-January 19
ISBN: 0-7695-2756-6
In recent years, it has been reported that several existing networks including the Internet have scale-free structure. In this paper, through a simple numerical analysis, we investigate effect of the scale-free structure of communication networks on their end-to-end performance. As network topologies, a random network and a scale-free network with the equal number of nodes and the equal number of links are used. We compare end-to-end performance of flows (i.e., throughput) in both random and scale-free networks. Consequently, we show that when the average degree of a network is small (i.e., whah the number of links is small), a scale-free network shows better end-to-end performance. On the contrary, when the average degree of a network is large (i.e., when the number of links is large), we show that a random network shows better end-to-end performance.
Index Terms:
Random Network, Scale-free Network, BA (Barabasi-Albert) Model, End-to-End Performance
Citation:
Hiroyuki Ohsaki, Koutaro Yagi, Makoto Imase, "On the Effect of Scale-Free Structure of Network Topology on End-to-End Performance," saint, pp.12, 2007 International Symposium on Applications and the Internet (SAINT'07), 2007