GATS 1.0: a novel GA-based scheduling algorithm for task scheduling on heterogeneous processor nets

We present a novel GA-based scheduling algorithm for heterogeneous processor networks that succeeds in generating task schedules with completion times that are 7% and 10.1% shorter than those produced by two of the best existing scheduling algorithms for heterogeneous networks of processors: HEFT [3] and DLS [2]. The new algorithm (GATS 1.0) achieves these results by employing an innovative genotype to phenotype encoding scheme and matching crossover and mutation operators. In addition, GATS 1.0 uses a simple fitness evaluation function and a small population, which makes it efficient (relative to classic GA implementations), as well as effective.