Research On Multi-objective QoS Task Scheduling Algorithm Based On Discrete Particle Swarm In Cloud Environment

With the rapid development and popularization of information technology and the Internet,cloud computing has been developed rapidly and widely.Cloud computing,as a new business computing model,virtualizes a series of dynamically scalable computing resources for users to provide efficient and reliable application services.Cloud computing task scheduling is the core problem in the field of cloud computing research,the purpose is to get the best task scheduling strategy,that is,the user submitted to the task assigned to the cloud computing resources on the implementation of the best scheduling.The commercial characteristics of cloud computing determine the user-centric service requirements,and with the continuous development of cloud services,the rapid increase in the number of users,a large number of user tasks and their diverse needs of the service,making the cloud computing task scheduling problem more complex The In order to meet the different quality of service(QoS),it is necessary to consider the performance of the system not only to consider the performance of the system,but also to consider the user's service quality requirements.Aiming at the existing problem of task scheduling algorithm with the goal of user service quality,this paper proposes a task scheduling algorithm which satisfies the service quality requirement of users,and carries on the three goals from the reliability,time and cost of service quality Task scheduling research,the main research work includes the following two aspects:(1)A task scheduling model based on multi-objective QoS constraint is established.Combined with the current cloud computing system framework,the establishment of the task scheduling model and define the relevant mathematical model.This paper describes the multiobjective QoS constraint processing method used in this paper.For the three QoS target requirements of reliability,time and cost,the multi-objective optimization problem is transformed into a single objective optimization problem with constraints the time and cost are constrained by the deadline and the budget,and the reliability is used as the final goal of the task scheduling when the constraint is adopted.Finally,according to the task scheduling goal of this paper,the only solution with the highest reliability is obtained,that is,the task with the highest reliability is obtained under the time and cost constraints.(2)A task scheduling algorithm based on multi-objective QoS constraint is proposed.In this paper,a multi-objective QoS task scheduling algorithm based on discrete particle swarm optimization(QoS-DPSO)is proposed.This algorithm combines the three QoS target requirements of task deadline,scheduling budget and reliability,and combines the task scheduling model and the discrete particle swarm algorithm.The algorithm first defines the QoS of the task and the mathematical modeling.The deadline and the scheduling budget constraint redefine the search space of DPSO,and redefine the fitness function of DPSO according to the reliability.Finally,the optimal task scheduling scheme is searched according to the fitness value.The experimental results show that the QoS-DPSO algorithm has better reliability than other algorithms,and it has better advantages in other performance indicators while satisfying the scheduling deadline.