ACO-MTS: A new approach for multiprocessor task scheduling based on ant colony optimization

In this paper, a new proposed approach named ACO-MTS to multiprocessor task scheduling based on ant colony optimization is introduced. Optimized task scheduling is one of the most important problems in parallel and distributed systems. Task scheduling in multiprocessor architectures is NP-hard so that finding the best possible solution is generally impossible. Ant colony optimization is a metaheuristic approach inspired from social behavior of real ants. It is a multi-agent approach, in which agents (artificial ants) try to find the shortest path for solving the given problem using an indirect communication. The proposed ACO-MTS is evaluated in comparison with not only the traditional heuristics but also the genetic algorithm. Finally, it outperforms the heuristic approaches, and had identical performance with genetic algorithm. However, the genetic algorithm examines too more solutions to achieve the best scheduling compared to ACO-MTS. Presented results demonstrate that the proposed approach is so successful in multiprocessor task scheduling.