摘要：Cloud computing is a type of parallel and distributed systems to supply, use and delivery of IT services via internet. By growing demand and cloud computing infrastructures benefits, different types of computing can be performed in a cloud environment. Consequently, quickly move to the cloud is on the rise. To improve the response time and high cloud user satisfaction guarantee, task scheduling on virtual machines should be perform optimally. Cloud task scheduling is an NP-hard optimization problem, so using of nondeterministic and meta-heuristic algorithms to solve it in a reasonable time is crucial. In this paper, Quality of service aware task scheduling algorithm in cloud computing are presented that allocate virtual machines optimally to entry tasks. The proposed algorithm based on frog leaping evolutionary approach, is named shuffled frog leaping based task Scheduling algorithm (SLF-TSA). The proposed algorithm, sort the tasks by their running deadline, after specifying the number of available virtual machines, map task sets to the virtual machines by their running deadline with shuffled frog leaping algorithm. Therefore, tasks by lower running deadline time has higher priority to allocation. As a result, the rate of user satisfaction will increase. The proposed algorithm compared with existing task scheduling algorithms in different scenarios, and the results are presented. Simulation results with Cloudsim framework display that the proposed algorithm has better results than previous algorithms in the makespan time, average response time and load imbalance degree parameters.
