Socially-responsible load scheduling algorithms for sustainable data centers over smart grid

Abstract

Power consumption of data centers accounts for a significant portion of operational cost for service providers. However, it is not sufficient to minimize electricity cost only. Instead, service providers should also consider the sustainability of data centers. In this paper, we investigate the problem of socially-responsible load scheduling for sustainable data centers in a smart grid environment. We formulate the joint optimization of electricity cost, performance cost and social cost into a min-cost network flow model, which takes the diversity of regional electricity prices, user request latency, usage of renewable energy into account. We further propose both offline and online load scheduling algorithms to achieve the optimality of the objective cost function. Finally, a series of simulation experiments are conducted to evaluate the effectiveness of our proposed algorithms.