Abstract
The fuel cell is a promising power source for green data centers due to its high energy efficiency, low carbon emissions, and high reliability. However, because of the mechanical limitations related to fuel delivery, fuel cells are slow in adjusting power output when the energy demand quickly changes, which is called limited load following. Many recent work have studied to mitigate the limited load following by using energy storage to adjust energy supply, but achieves limited successes because of the constraint of energy storage size. In this paper, we address this challenge by changing both energy supply and demand, via joint workload scheduling and energy management. Specifically, we consider multiple geo-distributed data centers powered by both fuel cells and energy storage. An online algorithm has been proposed to minimize the gap between energy supply and demand by jointly managing the fuel cells output and migrating workloads among data centers. Simulations results based on real-world traces show that the proposed algorithms can achieve satisfactory performance.
| Original language | English |
|---|---|
| Article number | 8616794 |
| Pages (from-to) | 397-406 |
| Number of pages | 10 |
| Journal | IEEE Transactions on Green Communications and Networking |
| Volume | 3 |
| Issue number | 2 |
| DOIs | |
| Publication status | Published - 1 Jun 2019 |
Keywords
- cost minimization
- data center
- Fuel cell
- job scheduling
- Lyapunov optimization
ASJC Scopus subject areas
- Computer Networks and Communications
- Renewable Energy, Sustainability and the Environment
