Research On Cloud Resource Scheduling Based On Ant Colony Optimization

Cloud computing is both a service model and a computing model.As a service model: users have certain requirements on the quality of their services.Either to obtain the most efficient service in the shortest time or to obtain the smoothest service at the lowest cost,etc.Both short time and low cost are users' requirements for cloud computing service quality,which should be met as much as possible.As a computing model: virtualization technology encapsulates a large number of heterogeneous computing,storage and network resources for users to access and use.When cloud users increase on a large scale,the heterogeneity of cloud server resources and network nodes,as well as the diversity of user needs,may cause the cloud system load imbalance,which in turn affects the overall performance of the cloud system.How to schedule cloud resources reasonably and efficiently is the basis for meeting user service quality requirements and ensuring the overall performance of the cloud system.This paper starts from the related optimization goals of cloud users and cloud systems,takes the multi-objective optimization of time and cost as the service quality requirements of cloud users,and reduces the load balance difference as the performance guarantee standard of the cloud system.At the same time,a cloud resource scheduling model is built,and for the second-level scheduling mode of cloud resources,a cloud resource scheduling algorithm based on the improved ant colony algorithm(Improved Ant Colony Optimization,IACO)is proposed.There are two improvements: on the one hand,the time cost constraint function is created to redefine the pheromone;on the other hand,the load adaptive function is introduced to calculate the ant transition probability.The purpose of the improvement is to realize the multi-objective optimization of time and cost and balance the load of the cloud system respectively.In addition,the improved ant colony algorithm(IACO)proposed in this paper has many parameters and it is difficult to determine the optimal combination of parameters.This paper uses Taguchi-Grey Relational Analysis method to complete parameter optimization.According to the proposed method,this article simulates cloud resource scheduling on the Cloud Sim and implements two sets of experiments:the first set of experiments evaluates and analyzes the optimized parameters,the second set of experiments takes the methods mentioned in this article as the main body to compare and analyze other related algorithms.The first set of experimental results show that when the cost is constant,the optimized parameter combination can make the IACO algorithm complete the scheduling task more efficiently.The second set of experiments shows that compared with the other two ant colony algorithms,the IACO algorithm has the lowest scheduling time and cost,and has more advantages in terms of load balancing.In general: the method proposed in this article can improve the user's service quality requirements while improving the load balance of the cloud system.