Supporting Energy-driven Adaptations in Distributed Environments
(INRIA Lille - Nord Europe, Project-team ADAM, University Lille 1 - LIFL CNRS UMR 8022, France)
The rise of the usage of digital devices and software services contribute to the increase of energy consumption of IT infrastructures. However, energy is still largely produced by limited resources. Therefore, optimizing and reducing its consumption is an economic and human necessity. Related works addressing energy optimization in computer science are widespread, but at the middleware layer, existing approaches are limited in their scope, adaptability and autonomous functioning. In this paper, we propose the foundations of a middleware architecture capable of handling various types of energy techniques and in different contexts. The distributed nature of our approach fits in a ubiquitous environment and covers the energy dimensions of both devices and software services. We also present the experimental results of a prototype implementing a subset of our proposed architecture. These results illustrate the potential of our energy-aware and autonomous approach.