Traffic-based network clustering

Network clustering is traditionally approached just relying on the topology of the network, and neglecting the information on the traffic intensity between the nodes. In this paper we propose traffic-aware clustering, whereby networks are clustered on the basis of their traffic matrices. We redefine two clustering metrics for the context of traffic matrices, and perform an exploratory analysis by comparing four well known algorithms against two real-world datasets, each made of 1000 traffic matrices, respectively from Abilene and Géant networks. The Spectral Filtering algorithm appears as the best performer. However, in the Géant network dataset the two metrics provide different rankings for the algorithms under examination, and Newman's algorithm can perform marginally better under one of the two metrics.