Green communications through network redesign

Abstract

We have extended our redesign network methodology with the capability to explore the redesign space for an enterprise network (EN) to meet the green communication constraints. These constraints attempt to save energy/power those results in efficient usage of existing network resources. The power consumption in EN is gaining importance due to its contribution towards making EN environmental friendly. In this paper, we have viewed EN as a flexible infrastructure, where each node in EN can be moved to join a cluster with high association with its nodes rather than remaining in its original cluster. We have proposed a number of redesign operations at a node-level to change the existing traffic flow by reducing the traffic at the backbone. We have modeled the enterprise network redesign problem as an optimization problem and utilized Simulated Annealing to search for enhanced association between nodes and clusters. We have estimated the power saving through the power spectral density (PSD) of the utilized polar non-return to zero (NRZ) encoding signal. The simulation results demonstrate an overall drop in the backbone traffic, which estimates by 10% power saving. © 2011 Springer-Verlag.