An Improved Artificial Chemical Reaction Optimization Algorithm for Job Scheduling Problem in Grid Computing Environments

The non-deterministic polynomial-time-hard job scheduling problem can be regarded as the optimal assignment of a set of jobs to a set of computing nodes to minimize the completion time. Such problems can be efficiently addressed through a meta-heuristic optimization approach, such as the new artificial chemical reaction optimization method. This approach mimics a chemical reaction process in which reactants interact with one another to reach the minimum enthalpy (potential energy) state. Therefore, this study proposes a novel approach of artificial chemical reaction optimization for job scheduling (ACROAJS) in grid computing environments based on the recently proposed chemical reaction-inspired meta-heuristic. Software simulation results show that the proposed ACROAJS algorithm significantly improves job schedule quality (makespan) in grid computing environments compared with two existing solutions [genetic algorithm and heterogeneous earliest finish time algorithm] over a set of randomly generated graphs and over graphs for real-world problems with various characteristics. With this algorithm, makespan was reduced by approximately 5.06% on average.