|
|
 |
| |
|
|
|
Reinforcement Learning Based Energy Consolidation Model for Efficient Cloud Computing System |
|
|
|
PP: 67-77 |
|
|
|
doi:10.18576/amis/170109
|
|
|
|
Author(s) |
|
|
|
B. Prabha,
J. Thangakumar,
K. Ramesh,
|
|
|
|
Abstract |
|
|
| Cloud computing offers significant deployment services to the end-users by using various network & software resources which consumes enormous energy. Energy plays a key factor in the ubiquitous cloud system because it involves financial investment for the hardware infrastructure and also contributes for service quality. There are numerous models available the consumption of cloud data center energy. In the prevailing cloud models certain count of resources were experimented which are not providing significant energy consumption due to wide set of applications actively running in the data center. One of the classifications of Machine Learning model is Reinforcement Learning (RL) model which gives maximum support to minimizing the application energy through learning model and reward points. The Data Center (DC) in the cloud consists of a host with multiple Virtual Machines (VMs), so the energy consolidation technique is needed for minimizing energy. RL based Multi agent model is proposed for minimizing the energy in VM level to DC level by implementing three agent’s levels namely VM agent, Host Agent and DC agent. The reward points are analyzed based on these agent’s responses with corresponding resource level parameters. Based on the resource action, the policy determines in finding energy efficient resources and terminate the i
|
|
|
|
|
|
|
|
