Research On Low-power Task Scheduling Strategy Based On Mobile Agent In Cloud Environment

With the rapid development of cloud computing, how to effectively reduce the enormousenergy consumption from large-scale cloud data center has become the focus that the academicand business communities is paying close attention to. The energy consumption issue of cloudcomputing is very urgent, this thesis is conducted a series of exploration and development inthese areas, mainly includes:1. This thesis introduced the basic concepts and characteristics of cloud computing andmobile agent. According to the specific conditions of the cloud environment and the advantagesof mobile agent technology in a distributed system, this thesis put forward a task schedulingmodel based on multi-agent in a cloud computing environment.2. This thesis improved the original migration mechanism by adding an Agent MigrationCache Manager (AMCM) in agency following the FIPA standards. In different situations, theAMCM will put the minimum information set into cache according to the corresponding rules. Inthis way the minimum information set will be sent to the remote agency directly withoutre-packaging when the same agent requests to migrate again. Finally, this thesis proved that thenew agent migration caching mechanism can effectively reduce the delay of migration and savethe computing resources by experiments.3. This thesis introduced the winner tree as a model which made the data nodes as the leafnodes of the tree and selected the final winner on the purpose of reducing energy consumption.This thesis fully considered the complexity of large-scale cloud data center, and defined the taskcomparing coefficients to make task scheduling strategy more reasonable while the cloud datacenter handled multiple tasks at the same time. Finally, this thesis proved the method in thisarticle can effectively improve the node utilization of the cloud data center, and reduce theoverall pothis thesisr consumption of the cloud data center by experiments.