Energy-aware network resource allocation in a cloud environment

Abstract

In the world of computing, one major statement has always been true: the request for computing power is always greater than the available resources. For this reason, scientists and researchers started considering the concept of sharing resources. In the last decade, the world of computing has seen a big leap forward with the emergence of cloud computing. In this new paradigm, computing resources of data centers spread all over the world can be shared to provide increased power and availability to users. However, with this increase in computing power came along an exponential increase in energy consumption, thus leading to an age where the Information and Communication Technology (ICT) ranks as one of the greatest contributors to Global Warming. In parallel to making computers powerful, research topics today concentrate on energy efficient computing. The most discussed topic is the energy efficiency of data center, where huge amounts of energy are wasted and dissipated in the form of heat and cooling overheads. Few articles discuss the energy problem at the level of the cloud network; even though, it is equally power demanding. The huge number of messages, going back and forth through the cloud, needs a huge infrastructure that uses vast amounts of energy. These messages, whether data, requests, or response of services are ever increasing in number and the anticipated increase in the industry’s carbon dioxide footprint is tremendous. The energy efficiency of the ICT plays an important role in sustaining the advancement of Global Warming. The steps towards green computing started with simple guidelines for hardware manufacturers and evolved into complex energy aware management of networks and data centers. Traditional routing algorithms have mainly focused on maximizing network resources utilization. This approach is not appropriate for a cloud system, due to the huge amount of traffic at hand.