Meta-heuristics based approach for workflow scheduling in cloud computing: A survey

Abstract

The Cloud computing is an emerging distributed systems which follows a “pay-as-you-use” model. It is a new type of shared infrastructure able to offer several resources through the Internet. There is large number of users using the services over the cloud, which generating large volume of data. The scheduling of dependent tasks is a NP-complete problem and has become as one of the most challenging problems in cloud environment. There is a need of specifying a sequence of execution of these tasks to satisfy the user requirements in terms of QoS parameters such as cost, execution time, etc. The workflow scheduling is considered to be difficult, when it becomes a multi-objective optimization problem. In this paper, we presented a comprehensive description of the existing approaches based on meta-heuristics for workflow scheduling. On the basis of the related works, we found the Genetic algorithm as the best method for scheduling. A GA searches the problem space globally and therefore, scholars have investigated combining GAs with other meta-heuristic methods to resolve the local search problem. We feel that there is a scope of using hybrid meta-heuristics approach that combines Artificial Bee Colony algorithm and Genetic Algorithm (ABC-GA) for scheduling workflows in Cloud computing. Cross-over and mutation operators of GA can be embedded into ABC to improve scheduling strategy.