Heuristics for dynamic topologies to reduce power consumption of networks

Energy consumption of communication networks is an important contributor to the ICT sector's greenhouse gas emission footprint. This research project focuses on power consumption reduction of communication networks by dynamically adapting network configuration to traffic demands. This is promising as networks are often under utilised over long periods. In the context of this work, dynamic topologies refers to a method of changing network links and nodes according to traffic loads. In this paper preliminary results are introduced and two simple heuristics are investigated: the Lightest Node First and the Least Loaded Node algorithms. Both generate reduced topologies for given traffic loads with smaller energy footprints than unmodified networks. Initial numerical results are presented for a small sample network of eight nodes with a large set of traffic demands. Depending on overall network utilisation, the algorithms are able to reduce the average network power consumption by up to 50% for this sample network.