On realistic network topologies for simulation

Abstract

Simulations are an important tool in network research. As the selected topology often influences the outcome of the simulation, realistic topologies are needed to produce realistic simulation results. We first discuss the different types of topologies and present our collection of real-world topologies that can be used for simulation. We then define several similarity metrics to compare artificially generated topologies with real world topologies. We use them to find out what the input parameter range of the topology generators of BRITE, TIERS and GTITM are to create realistic topologies. These parameters can act as a valuable starting point for researchers that have to generate artificial topologies.