Teaching learning-based optimisation for job scheduling in computational grids.

Grid computing is a framework that enables the sharing, selection and aggregation of geographically distributed resources dynamically to meet the current and growing computational demands. Job scheduling is a key issue of grid computing and its algorithm has a direct effect on the performance of the whole system. Because of distributed heterogeneous nature of resources, the job scheduling in computational grid is an NP-complete problem. Thus, the use of meta-heuristic is more appropriate option in obtaining optimal results. In this paper, a recently developed optimisation algorithm known as teaching learning-based optimisation (TLBO) is proposed to solve job scheduling problem in computational Grid system with minimisation of makespan, processing cost and job failure rate, and maximisation of resource utilisation criteria. In order to measure the efficacy of proposed TLBO, genetic algorithm (GA), particle swarm optimisation (PSO), firefly algorithm (FA) and differential evolution (DE) are considered for comparison. The comparative results exhibit that the proposed TLBO technique outperforms other algorithms.