Resource Management in the Cloud Computing Using a Method Based on Ant Colony Optimization

Abstract

Regarding large amount of tasks and resource constraints, resource management is a significant challenge in cloud computing. There are several reasons, including heterogeneous and dynamic properties of resources and requests in cloud computing environments, which have caused it to be emerged as an NP-complete problem. Also, it is very difficult to manually allocate tasks to cloud computing resources. As a result, an appropriate method is needed to allocate cloud resources to requested tasks in order to increase exploiting cloud resources. So far, many methods and meta-heuristic algorithms such as Round Robin, Max-Min, GA and PSO is used as a solution for this problem. It is only reducing time or costs that is provided in most of presented methods. But in this article, scheduling algorithm aims to process and run users’ tasks at the least possible time and cost. Also, the maximum efficiency of resources is taken into account so that all tasks would be distributed evenly among the available resources in order to reduce the completion time of all tasks and to increase resource efficiency. Ant colony optimization algorithm is used to improve the management of resources in the cloud environment. Also, the gravitational algorithm with local search is used to prevent algorithm from fast converging to local optimal and to increase its dynamic ability. Simulation and comparison of the results of proposed strategy with the results of genetic algorithm and ant colony algorithm indicates that the proposed approach could satisfy users and also use resources well