Maximise Clean Energy Usage for Data Centres

Pictured Left, L to R: Dr.Raymond Carroll, Dr. Sasitharan Balasubramaniam.

Researchers within the Telecommunications Software & Systems Group (TSSG) recently published a paper titled ‘Application of Genetic Algorithm to Maximise Clean Energy Usage for Data Centres’.  The paper was co-authored by researchers within the Bio-Inspired Networking Research Unit.

The focus of the paper is the investigation of the heavy energy usage within the ICT industry, with estimates that the industry now contributes 2% of the worlds CO2 emissions, which is equivalent to the worldwide aviation industry. The need for smart energy systems is driven by climate change and limited fossil fuel resources.

A significant part of this heavy ICT energy consumption can be attributed to data-centres, which house a large number of energy-intensive servers that host a variety of internet services. Much of this energy consumption is driven by the increasing popularity of the Internet, which continues to attract growing numbers of users annually. The internet has moved from being at the periphery to a central part of our daily lives. As a result of this increased energy usage, the research community has become focused on finding innovative solutions to improve the energy efficiency and sustainability of data centres.

Dr.Raymond Carroll, Dr. Sasitharan Balasubramaniam, Dr. William Donnelly and Dr. Dmitri Botvich respectively, presented their solution for delivering energy efficient green data centres. The proposed green data centre solution uses a Genetic Algorithm to place services in Data-Centres by targeting those in countries with the highest production of renewable energy and the best conditions for cooling. To validate this proposed solution some demonstrative simulations were carried out by gathering data regarding the renewable energy production and temperature profiles of each country, and implementing a genetic algorithm that aims to maximise the quantity of renewable energy consumed and minimise cooling energy expended. From these simulations they were able to demonstrate that by employing this technique, it is feasible to make significant improvements in the proportion of renewable energy utilised in data centre operation, thus reducing the quantity of fossil fuels burned and ultimately carbon emissions. It was also demonstrated that cooling energy could be reduced in circumstances where there is significant variance in the country’s temperature profiles. At the same time, the research carried out showed that this improved renewable energy utilisation did not come at an increased monetary cost for the operator.

Dr. Sasitharan Balasubramaniam, Research Unit manager for Bio-Inspired Networking was very optimistic about the work they had carried out “We hope that this research will create a new approach towards developing green ICT systems in the future, where systems can be aware of the source energy (renewable or fossil fuel based) and adapt its usage in order to minimize long term carbon emissions” he said.

The research carried out is sure to have an impact on the way we build and maintain data centres.

To view the full methodology and outcomes of the paper please click here

Additional Information

For more information on Bio-Inspired Networking Research Unit: click here

For more information about TSSG Active projects: click here