提出了适于异构环境独立任务调度的可调节动态调度算法(AS算法)。该算法以任务与处理机的执行时间和完成时间作为参数共同构造任务调度顺序的衡量值,其中二者所占的比重能进行适当调整。AS算法克服了Min-min算法单纯追求局部最优的局限性,更适合异构环境。实验结果表明AS算法可以有效地降低调度跨度,其性能比Min-min算法有所提高。
This paper presents a new algorithm called adjustable scheduling algorithm (AS algorithm) suitable for independent job scheduling on heterogeneous computing environment. The algorithm adopts a mechanism of the measuring value that considers the completion time and the execution time between tasks and processors as the parameters together, and it adjusts the proportions of both to achieve a better performance. AS algorithm overcomes the limita- tion of Min-min algorithm which pursues only local optimization, so it is more suitable for heterogeneous computing environment. The experimental results show that AS algorithm reduces the scheduling span effeetively and its performance is better than Min-min algorithm.