Research On Resource Allocation Technology In Cloud Computing Environment

Via virtualization technology and pooling management of cloud resources,the users' tasks are distributed to the cloud system,and the users can obtain computing resources,storage resources,and other resources.Consequently,the users only focus on their core business,and don't care about the bottom implementation and maintenance.This leads to reducing the users' spending on computer resources.How to schedule and manage the cloud system resources is an important issue.With the analysis of cloud computing resource allocation strategies,this thesis investigates the resource allocation problems for load balancing and cloud computing elastic expansion.The contributions of this thesis are as following:Tailored for the cloud computing resource allocation problem with load balancing,an improved brain storming algorithm is proposed.In order to avoid local optimum solution of the brainstorming algorithm,the discussion mechanism is adopted.This mechanism performs the discussion among different groups after the internal update.Additionally,the discussion frequency is increased with the number of iterations adaptively.Using the CloudSim tool,the simulation results show that the proposed scheme has a good effect for the load balance.Tailored for the resource scheduling problem of cloud alliance,it is modeled into a cooperative game using stability and fairness principles.A cloud alliance resource scheduling scheme via improved greedy algorithm is proposed,considering users' QoS and average resource utilization for the resource providers.Simulation results demonstrate that the proposed algorithm improves the users' QoS and resource utilization efficiency significantly.