Scheduling Tasks with Non-negligible Intertask Communication onto Multiprocessors by using Genetic Algorithms

The paper deals with genetic algorithms for scheduling the tasks with non-negligible intertask communication. Three genetic algorithms for scheduling with the primary goal to minimise finishing time are reported. The basic genetic algorithm includes the reproduction, crossover and mutation operators. Its improved version has additional cloning operator that allows duplicated scheduling. The third algorithm is adaptive. The experiments describing influence of genetic operators’ probabilities, population size and number of generations on the resulting schedules, comparison of algorithms and the results obtained for different task granulation are discussed.