A mathematical model and NSGA-II algorithm for bi-objective grid scheduling problem with quality of service satisfaction

Abstract

Computational grids consist of the innovative technologies of the new era, which seek to accelerate performance through distributing tasks on computing resources. A grid system makes it feasible to run great computing operations through the connected processors. In this article, a bi-objective problem of grid scheduling based on quality of service concept is discussed. The first objective is to increase the profit earned from customers and the second, to increase the utilisation of computational resources. A mathematical programming model is proposed for the problem and a meta-heuristic NSGA-II algorithm is designed and customised for the problem. In the numerical analysis, by drawing Pareto diagrams and analysing the sensitivity thereof, the efficiency of the proposed methods and the effect of different parameters of the problem on both the methods are assessed. According to the results, the proposed NSGA-II algorithm is highly efficient in terms of solution quality and run time.