Analysis Of Task Scheduling Algorithms And Virtual Resource Optimization In Cloud Computing

Cloud Computing has become a new vehicle for delivering resources such as computing and storage to customers on demand based on per-use pricing model.The effect has been a shift to outsourcing and mobile cloud computing such as iCloud storage,and so on.The role of server virtualization in cloud computing allows multiple instances of an operating system and associated application to run on single physical machine.The amount of resources allocated to these instances and the storage that they use can all be managed anywhere at any time through web interfaces.In cloud computing system,task scheduling and virtual resource optimization are NP-hard optimization problems,this brings new challenges to the cloud computing service providers and cloud computing researchers to find out how to use cloud computing resources efficiently and get the maximum profits for cloud users' side and for cloud computing service providers' side.Our main work and our innovations were summarized as follows:(1)In order to solve the task scheduling problem in cloud computing,this work proposed a modified Particle Swarm Optimization(MPSO)to optimize task scheduling and cloud resources.The simulation results prove that our algorithm could quickly and dynamically optimize virtual resources with a reduced total time at a reduced cost.(2)In order to solve the resource optimization problem in cloud computing environment;this work proposed a Multi-Objective Resource Scheduling Algorithm with the objective of balancing the workload between nodes within clusters in order to reduce the waiting time and response time.The algorithm detects the system status and makes a decision.If all nodes are in busy state,the submitted tasks should remain in queue until receiving the notification to proceed or to migrate to the available node.(3)A cloud platform task scheduling and resource optimization algorithm(EGA-TS)is proposed based on the evolutionary algorithm.In that proposed algorithm,selection,crossover,mutation operators were investigated.The simulation results showed that EGA-TS is optimal compared it with five heuristic methods(HEFT-T,HEFT-B,Basic GA,CPOP,and MOSCOA).The results showed that the proposed methods improve the performance of execution time and communication costs.This work is divided into 6 chapters.Chapter 1 gives a short introduction of cloud computing,motivation of the research,objectives,challenges,methodology,and research problems formulation.Chapter 2 gives the review on the existing cloud virtualization resource management and Task Scheduling Optimization Algorithms in Cloud Computing.In chapter 3,we proposed the system architecture for automatic migration process based on Hadoop YARN environment,and design an algorithm of Multi-Objective Resource Scheduling(MORS)to maximize resource utilization.In chapter 4,we proposed a Modified Particle Swarm Optimization(MPSO)algorithm to solve task scheduling and virtual resource utilization problems.We conducted a performance comparison approach based on most critical objective functions of task scheduling problems which are execution time and computation cost of tasks in cloud computing resources.In chapter 5,we proposed Task Scheduling and Cloud Resources Optimization method based on Evolutionary Algorithms.Chapter 6 concludes the work and enumerates some future direction.