面向任务需求与服务能力的遥感任务调度算法

摘要

In allusion to the challenges on the low efficiency and load imbalance in cluster system,a new scheduling algorithm has been proposed based on the match of mission requirement and service capability.This algorithm focuses not only on the single factor of production nodes,but also introduces the mission requirement of remote sensing product simultaneously.By normalizing the mission requirement and service capacity of production nodes based on the characteristics of remote sensing image and the specialty of production process,the mission scheduling has been established from the match of the normalized values of the two factors.In comparison to conventional scheduling algorithms such as the methods of Round Robin,Use Rate of Product etc.,this proposed algorithm is much more advantageous in terms of increasing the production efficiency of remote sensing products,decreasing the imbalance degree and hopefully improving the serving capacity of remote sensing products for society.%针对目前遥感产品生产集群效率较低、集群负载不均衡的问题,提出了一种基于任务需求与服务能力匹配的遥感产品生产任务调度算法.算法不再单一考虑生产节点一方面因素,同时引入了遥感产品本身的任务需求因素;依据遥感影像的特点和遥感产品生产流程的特殊性,将任务需求与生产节点服务能力分别赋予一定的值,通过将2个因素的值相匹配,作为任务调度的依据.该算法与传统的轮转法、使用率乘积法等调度算法相比,具有调度效率高、集群不均衡度小的优点,提高了遥感产品的生产效率和遥感产品服务社会的能力.