Template-Based Genetic Algorithm for QoS-Aware Task Scheduling in Cloud Computing

Abstract

Task scheduling is ones of the most important issues in cloud computing environment, which directly affects the overall performance of the cloud platform. QoS-aware Task scheduling in cloud computing is NP-hard problem. There is no efficient method to solve it, and most of current task scheduling algorithms bias total task completion time than single task completion time. This paper proposes a template-based genetic algorithm (TBGA) with users' QoS constraints for task scheduling. Firstly, according to processors' CPU, bandwidth and etc, the algorithm calculates the maximal size of tasks that should be allocated to each processors, which is called template, secondly, according to the template, the algorithm combines tasks into multiple subsets and finally allocate the subset of tasks to the corresponding processors by using genetic algorithm. Simulation experiments in CloudSim are given and the results show that the algorithm TBGA can obtain minimal makespan for total task. Compared with other task scheduling algorithms, TBGA is better than them in performance of task scheduling.