Task Scheduling problem in distributed systems considering communication cost and precedence by population-based ACO

With regard to the fact of the rapid growth of distributed systems and their large spectrum of usage of proposing and representing controlling solutions and optimization of task execution procedures is one of the most important issues. Task scheduling in distributed systems has determining role in improving efficiency in applications such as communication, routing, production plans and project management. The most important issues of good schedule are minimizing makespan and average of waiting time. However, the recent and previous effort usually focused on minimizing makespan. This article presents and analyze a new method based on Ant Colony Optimization (ACO) algorithm with considerations to precedence and communication cost for task scheduling problem. In the mentioned method in addition to optimization of finish time, average of waiting time and number of needed processors are also optimized. In this method, by using of a new heuristic list, an algorithm based on ant colony is proposed. The results obtained in comparison with the latest similar models of random search algorithms, proves the higher efficiency of algorithm.