Static Task Graph Scheduling in Real Time Homogenous Multiprocessor Systems Using Learning Automata

摘要

Multiprocessor system have widely application in parallel computation. one of the application is using them in real time systems. a job in multiprocessor real time systems divide in the set of the tasks with the relation between them in order that one task can be execute only when it's parents executed. by difficulty in scheduling the task graph and it's complexity, many effort accomplish for finding the best optimized solution. In this paper, we tried to have the balance between the processor and also reduce relation between the processor and importance of them we tried to improved the speed of getting the response. In most of the activity and experiment, the run time of the scheduling algorithm ignored. finally, the result of maintaining this solution show that we can have the appropriate schedule in acceptable time. also in this paper, at the end, we compared the proposal algorithm with the other famous scheduling algorithm.