Task Graph Scheduling on Multiprocessor System using Genetic Algorithm

Task Graph Scheduling is an important issue in the distribution of programs on the processors of a parallel system. Because task graph scheduling is an NPComplete problem, methods of random search are utilized for finding the nearly optimal schedule. Recently, Genetic algorithms have received much awareness as they are robust and guarantee for a good solution. In this paper a new genetic algorithm is proposed based on Object Migration Automaton and is evaluated in comparison with FCFS and MET scheduling. The proposed algorithm begins with an initial population of randomly generated chromosomes and after some stages, each chromosome maps to an automaton.