Task Scheduling With Genetic Approach and Task Duplication Technique

The basic and prime job in parallel multiprocessor environment is accurately mapping of tasks and set the scheduling of different tasks on different processors to minimize the makespan. Total runtime is time taken time for all jobs with the individual runtime of tasks and their communication cost among tasks. In parallel multiprocessor environment, an optimal scheduling of parallel tasks having precedence relationship on such system is same as NPcomplete problem. In this paper, the scheduling hindrance is bringing out the optimal mapping of tasks and their efficiently possible execution stream on multiprocessor system configuration. Numbers of alternative solutions and heuristics techniques are proposed to solve such type of problem. By applying the GA with task duplication technique enhances the efficiency of the task scheduling in parallel multiprocessor environment. Here also by comparing this new approach with some existing deterministic scheduling techniques for minimizing inter processor traffic communication will speed up the scheduling.