Abstract

Differential evolution (DE) has recently emerged as simple and efficient algorithm for global optimization over continuous spaces. DE shares many features of the classical genetic algorithms (GA). But it is much easier to implement than GA and applies a kind of differential mutation operator on parent chromosomes to generate the offspring. Grid computing aims to allow unified access to data, computing power, sensors and other resources through a single virtual laboratory. Scheduling is a key problem in emergent computational systems, such as grid and P2P, in order to benefit from the large computing capacity of such systems. In this paper, we present differential evolution algorithm based on schedulers for efficiently allocating jobs to resources in a grid computing system. Several variations for DE are examined in order to identify which works best for the grid scheduling problem.

Key words: Grid computing, differential evolution, crossover probability, scale factor, population.