多核嵌入式系统的节能优化调度算法仿真

发布时间：2018-04-27 12:27

本文选题：多核 + 嵌入式系统　；参考：《计算机仿真》2015年06期

【摘要】：在对多核嵌入式系统进行节能优化调度的过程中,不同嵌入式系统拥有不同的电压可变性能,也拥有离散的电压可变等级,传统的嵌入式节能优化调度方法在应用到多核系统中时,由于只能根据唯一标准的电压可变等级进行任务调度,无法高效实现多核嵌入式系统的节能调度。为了高效节能,提出一种多核嵌入式系统的动态分批节能优化调度算法,分析了多核嵌入式系统的系统模型以及功耗和能耗模型,依据上述两种模型,针对多核嵌入式系统处理器集与任务集,综合分析各种影响因素,建立目标函数,在每次进行当前局部调度时,依据分批规则,在待调度任务集中选取一批任务,针对每一项任务,在每个处理器上执行目标函数,将离散电压等级转换成非平衡分配问题,一次性为上述每一个任务安排一个处理器,或为每个处理器安排一项任务,从而实现多核嵌入式系统的节能优化调度。仿真结果表明,所提方法具有很好的节能效果及高效性。
[Abstract]:In the process of energy saving optimization scheduling for multi-core embedded systems, different embedded systems have different voltage variability and discrete voltage variable classes. When the traditional embedded energy-saving scheduling method is applied to the multi-core system, it can not efficiently realize the energy-saving scheduling of the multi-core embedded system because the task scheduling can only be carried out according to the only standard voltage variable level. In order to save energy efficiently, a dynamic batch energy-saving scheduling algorithm for multi-core embedded system is proposed. The system model, power consumption and energy consumption model of multi-core embedded system are analyzed. According to the processor set and task set of multi-core embedded system, this paper synthetically analyzes various influencing factors, establishes objective function, selects a batch of tasks in the set of tasks to be scheduled according to the batch rules each time the current local scheduling is carried out. For each task, execute the objective function on each processor, convert the discrete voltage level into an unbalanced assignment problem, assign a processor for each of these tasks at once, or assign a task to each processor, In order to achieve the multi-core embedded system energy-saving optimal scheduling. Simulation results show that the proposed method has a good energy-saving effect and high efficiency.
【作者单位】：华中科技大学武昌分校;
【分类号】：TP301.6;TP332

【参考文献】