SAACO: A Self Adaptive Ant Colony Optimization in Cloud Computing

摘要

The cloud environment is a heterogeneous, dynamic and complex environment. The characteristic of Ant Colony Optimization (ACO), such as robustness and self adaptability, can just match the cloud environment. ACO is an algorithm that imitates the ants foraging, and it has a good application in the problems that want to find the optimal solution. The task scheduling in cloud computing is also the problem that want to find the optimal solution actually. In this paper, a self adaptive ant colony optimization (SAACO) is proposed. For the drawback of PACO we proposed before, such as the parameters' selection and the pheromone's update, in SAACO, we use particle swarm optimization (PSO) to make the parameters of ACO to be self adaptive. And we also improve the calculation and update of the pheromone. The results show that SAACO has a better performance than PACO both in makespan and load balance.