Electric Demand Response Management for Distributed Large-Scale Internet Data Centers

This paper evaluates the electric demand response (DR) management for distributed large-scale Internet data centers (IDCs) via the stochastic optimization approach. The electric DR of IDCs refers to the capability of optimally shifting cloud service tasks among distributed IDCs. Thus, the energy con...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on smart grid 2014-03, Vol.5 (2), p.651-661
Main Authors: Chen, Zhi, Wu, Lei, Li, Zuyi
Format: Article
Language:English
Subjects:
Cloud computing
Distributed databases
Distributed IDCs
Electricity
Energy consumption
environment
Linear programming
Optimization
Power demand
price-based demand response management
Servers
stochastic optimization
Stochastic processes
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper evaluates the electric demand response (DR) management for distributed large-scale Internet data centers (IDCs) via the stochastic optimization approach. The electric DR of IDCs refers to the capability of optimally shifting cloud service tasks among distributed IDCs. Thus, the energy consumption reduction at certain IDC locations could be considered as the DR provision capacity in day-ahead DR programs. Cloud service tasks of IDCs include processing, storage, and computing tasks, which are further categorized into interruptible and non-interruptible tasks. The proposed model determines the optimal hourly DR capabilities of individual IDCs while considering uncertain coming cloud service tasks to individual IDCs. The major contribution of this paper is to rigorously formulate the DR capability of IDCs as changes in the electricity consumption when shifting cloud service tasks among distributed IDCs in different time zones, while considering the energy consumption for providing IT service, cooling, shifting cloud service tasks, environmental impacts, and uncertain coming tasks. The proposed model would enhance the financial situation and improve the environmental impacts of distributed IDCs by participating in day-ahead DR programs. The stochastic optimization adopts scenario-based approach via the Monte Carlo (MC) simulation for minimizing the total electricity cost, which is the expected electricity payment minus the revenue from the DR provision. The proposed model is formulated as a mixed-integer linear programming (MILP) problem and solved by state-of-the-art MILP solvers. Numerical results show the effectiveness of the proposed approach for solving the optimal electric DR management problem for distributed large-scale IDCs.
ISSN:1949-3053
1949-3061
DOI:10.1109/TSG.2013.2267397