Abstract | ||
---|---|---|
This paper investigates the energy aware long term optimization problem of the data center network which is equipped with energy storage. The network optimizes its operation on workload distribution and energy supply to reduce the operational cost. Because of the uncertainty of workload and electricity prices, on which we do not assume any distribution, a stochastic network optimization problem is formulated. By introducing virtual queues for the batteries and constructing virtual workload queues, the problem fits into the framework of Lyapunov optimization, which leads to a dynamic and distributed algorithm. The proposed algorithm guarantees the worst case delay of the workload is no larger than a constant. Extensive simulation results verify the effectiveness of the proposed algorithm and demonstrate that we can tradeoff between the average cost and average delay. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1109/ICCA.2018.8444191 | 2018 IEEE 14th International Conference on Control and Automation (ICCA) |
Keywords | Field | DocType |
Data center operation,battery,stochastic network optimization,Lyapunov optimization,worst case delay | Mathematical optimization,Workload,Control theory,Queue,Average cost,Lyapunov optimization,Distributed algorithm,Energy supply,Engineering,Data center,Optimization problem | Conference |
ISSN | ISBN | Citations |
1948-3449 | 978-1-5386-6090-4 | 0 |
PageRank | References | Authors |
0.34 | 6 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
jun sun | 1 | 17 | 15.65 |
Zaiyue Yang | 2 | 188 | 15.59 |
Zhenwei Guo | 3 | 2 | 2.39 |
Qinmin Yang | 4 | 295 | 23.86 |