Network Utilization: The Flow View
Venue
IEEE INFOCOM 2013, IEEE, Turin, Italy
Publication Year
2013
Authors
Avinatan Hassidim, Danny Raz, Michal Segalov, Ariel Shaqed (Scolnicov)
BibTeX
Abstract
Building and operating a large backbone network can take months or even years, and
it requires a substantial investment. Therefore, there is an economical drive to
increase the utilization of network resources (links, switches, etc.) in order to
improve the cost efficiency of the network. At the same time, the utilization of
network components has a direct impact on the performance of the network and its
resilience to failure, and thus operational considerations are a critical aspect of
the decision regarding the desired network load and utilization. However, the
actual utilization of the network resources is not easy to predict or control. It
depends on many parameters like the traffic demand and the routing scheme (or Traffic
Engineering if deployed), and it varies over time and space. As a result it is very
difficult to actually define real network utilization and to understand the reasons
for this utilization. In this paper we introduce a novel way to look at the network
utilization. Unlike traditional approaches that consider the average link
utilization, we take the flow perspective and consider the network utilization in
terms of the growth potential of the flows in the network. After defining this new
Flow Utilization, and discussing how it differs from common definitions of network
utilization, we study ways to efficiently compute it over large networks. We then
show, using real backbone data, that Flow Utilization is very useful in identifying
network state and evaluating performance of TE algorithms.
