Abstract: Focused on the task scheduling of cloud computing issue, a Template-based Task Scheduling（TTS） strategy is proposed. TTS gives a full consideration to network communication cost and realizes tasks scheduling based on task assignment preprocessing. There are two steps in TTS. First of all, given a task set, the strategy calculates the template of assignment for each processer in a way of dividing subtasks in divisible task scheduling model. Then, based on the template, an effective algorithm is applied for task allocation to obtain an optimal result. Based on TTS strategy, this paper improves classic greedy algorithm to put forward Template-based Task Scheduling Greedy Algorithm（TTSGdA）. Compared with Min-min algorithm and genetic algorithm, TTSGdA can effectively reduce the completion time of task set.

Key words: cloud computing, task scheduling strategy, template, combination optimization, greedy algorithm

摘要： 针对云计算任务调度，提出了一种基于模板的任务调度（Template-based Task Scheduling，TTS）策略。该策略充分考虑了通信开销，在对任务分配进行预处理的基础上实现任务调度，主要分为两步：针对一个任务集合，采用可分任务调度求解子任务大小的方法，求出各个处理机应该分担的任务量模板;根据求出的模板，采用合理的调度算法对任务进行调度，从而得到较优的调度结果。在TTS策略下，对传统贪心算法加以改进，最终提出基于模板的任务调度贪心算法（Template-based Task Scheduling Greedy Algorithm，TTSGdA）。与Min-min算法和遗传算法的对比实验结果表明，TTSGdA能够有效减少任务集合完成时间。

关键词: 云计算, 任务调度, 模板, 组合优化, 贪心算法