Scheduling tasks of a parallel program in two-processor systems with use of cellular automata

Abstract

Cellular automata (CA) are proposed to design scheduling algorithms for allocating tasks of a parallel program in multiprocessor systems. For this purpose a program graph is considered as CA which control a process of migration of tasks in a multiprocessor system. In the first phase of the algorithm, effective rules for CA are discovered by a genetic algorithm. In the second phase, for any initial allocation of tasks, CA-based scheduler is able to find an allocation which minimizes the total execution time of the program in a multiprocessor system.