A hybrid evolutionary approach for heterogeneous multiprocessor scheduling

This paper considers the assignment of tasks with interdependencies in a heterogeneous multiprocessor environment where task execution time varies with task as well as the processing element processing it. The solution to this heterogeneous multiprocessor scheduling problem involves the optimization of complete task assignments and processing order within the assigned processors with minimum makespan, subject to the precedence constraint. To solve such a NP-hard combinatorial optimization problem, this paper presents a hybrid evolutionary algorithm that incorporates two local search heuristics that exploits the intrinsic structure of the solution as well as specialized genetic operators to encourage exploration of the search space. The effectiveness and contribution of the proposed features are validated on a set of benchmark problems characterized by different degrees of communication times, task and processor heterogeneities. Simulation results demonstrate the algorithm is capable of finding useful schedules on the set of new benchmark problems.