Clustering network simulation: graph partitioning approach

Abstract

Scalability is one of the main challenges faced by every network simulator. Various techniques can be applied in order to improve scalability. The most common include increasing simulation speed or reducing the level of detail as a trade-off to network reality. Clustering the simulated network by applying graph partitioning algorithms is an alternative approach and also the main topic of this paper. The idea is to distribute the network simulation over the existing topology. We consider two graph partitioning algorithms available in literature. The algorithms were implemented and tested on randomly generated networks in order to compare their performances. Our motivation for research in this area is driven by the need for efficient techniques which could improve the simulation capabilities and increase the scalability of IMUNES. IMUNES is a network simulator which has been developed at the University of Zagreb, FER.