Scheduling Tasks on Heterogeneous Multi-Core Processors Based on Modified Ant Colony Optimization
Zhulin Li, Cuirong Wang, Haiyan Lv, Xin Song · International Journal of Control and Automation · 2014
Traditional techniques of single CPU no longer fit the trend of big data processing and parallel computing environment.Accordingly, heterogeneous multi-core processors bring opportunities.However, task scheduling strategies on heterogeneous multi-core processors are not as mature as the hardware techniques.To this end, in this paper, we study on the task scheduling algorithm on heterogeneous multi-core processors.Specifically, we assume there exists dependence between tasks, and due to the different computing capabilities of processors, the workload on each processor should be considered as well.Therefore, we formulate the problem as to optimize the makespan and load balance of the whole task execution process, and propose a modified Ant Colony Optimization (ACO) algorithm to find the optimal solution.Experiments are also conducted for evaluation.