Strengths and weaknesses of genetic list scheduling for heterogeneous systems

Abstract

List scheduling combined with genetic algorithms has already been shown to be a powerful approach (Grajcar, 1999,Yu-Kwong Kwok and Ahmad, 1996). We investigate the problems associated with using a list scheduler for a heterogeneous system. Furthermore, we present cases, for which most list schedulers fail to find the optimum. We propose and experimentally evaluate novel ideas avoiding it.