Application of Genetic Algorithm to Maximise Clean Energy Usage for Data Centres

Abstract: The ICT industry is quickly becoming a very significant consumer of global energy. This energy usage clearly comes at an environmental cost, and recent reports estimate that the industry now contributes 2% of the worlds CO2 emissions, or as much as the aviation industry. A significant part of this energy consumption can be attributed to data-centres, where huge numbers of energyintensive servers host a variety of internet services. Much of this energy consumption is driven by the popularity of the Internet, which continues to attract growing numbers of users who now rely on the Internet as part of their daily lives. A major factor behind this attraction is the multitude of services available on the Internet, ranging from web based services (e.g. facebook) to heavy power consuming services such as multimedia (e.g. youtube, IPTV). As a result of this increased energy usage, the research community has become focused on finding solutions to improve the energy efficiency and carbon footprint of data centres. In this paper we present our solution for delivering green data centres. We propose a Genetic Algorithm-based solution for determining the optimal placement of services in datacentre network, in order to maximize the overall renewable energy usage and minimize the cooling energy consumption. We then perform a series of experiments in order to evaluate our solution, incorporating varying service request profiles and actual weather and renewable energy production values.