Task Scheduling Based on Multi-objective Genetic Algorithm in Cloud Computing

Abstract

The performance of task scheduling in cloud computing mainly involves the total completion time, the average completion time and resource load balancing. However, the existing research has failed to synthetically consider the three objectives. This paper proposed a multi-objective genetic algorithm to solve the multi-objective task scheduling problem in cloud computing. A complex large task was divided into multiple small tasks, and let the mapping from the small tasks to the resources as the encoding of chromosomes. In the selection phase, using weights of the three objectives to determine which fltness functions should be adopted. Crossover probability and mutation probability were designed to ensure that the diversity of population and speed up the convergence speed. Finally, simulation results verify the efiectiveness of the algorithm proposed in this paper.