A novel discrete grey wolf optimizer for scientific work ow scheduling in heterogeneous cloud computing platforms

Abstract

There are several scientific work flow applications which need a vast amount of processing. Therefore, cloud offerings are made to give them a sense of economy. Work flow scheduling has drastic impact on gaining the desired Quality of Service (QoS). The main objective of work flow scheduling is to minimize the makespan. This scheduling is formulated into a discrete optimization problem, which is NP-hard. This paper presents a novel Discrete Grey Wolf Optimizer (D-GWO) for scientific work flow scheduling problems in heterogeneous cloud computing platforms with the aim of minimizing makespan. Although traditional Grey Wolf Optimizer (GWO) has great achievements with continuous optimization problems, a clear gap exists in utilizing GWO for combinatorial discrete optimization problems given that the continuous changes in search space during the course of discrete optimization lead to inefficient or meaningless solutions. To this end, the proposed algorithm is customized to optimize the discrete work flow scheduling problem by leveraging some new binary operators and Walking Around approaches to balancing between exploration and exploitation in a discrete search space. Scientific unstructured work flows were investigated in different circumstances to prove the effectiveness of the proposed D-GWO. The simulation results witnessed the superiority of the proposed DGWO to other state-of-the-arts in terms of scheduling assessment metrics.