基于动态双向优先级的任务分配与调度算法  

Task allocation and scheduling algorithm based on dynamic dual-directional priority

作　　者：龚跃[1] 张真真[1] 黄小珂[1] 刘建军 

机构地区：[1]长春理工大学计算机科学技术学院,长春130022 [2]长春创力信息技术有限公司开发部,长春130022

出　　处：《计算机应用》2009年第4期1131-1134,共4页journal of Computer Applications

基　　金：国家科技部中小企业创新基金资助项目(05C26212200378)

摘　　要：提出了一种基于动态双向优先级的任务分配与调度算法,称作动态双向优先级(DDDP)算法。该算法综合考虑了实时任务的优先级和子机的优先级,构造了动态双向优先级任务分配模型,实现了数据传输中主机/子机模式的任务动态分配与调度。在模拟实验中,通过使用正常负载和过载情况下的典型数据对算法进行仿真研究表明,这种算法比单纯考虑截止期的EDF算法在性能方面有明显的改进和提高。A task allocation and scheduling algorithm called Dynamic Dual-Directional Priority （DDDP） was presented. This algorithm considered the priority of real-time task and the sub synthetically, and constructed a dynamic dual-directional priority task allocation model, realized the task allocation and scheduling of the master/sub model in data transmission. In the simulation experiments with some typical data of various parameters under normal workload and overload situation, the DDDP algorithm has improved the performance of scheduling obviously compared with a classical Earliest Deadline First （EDF） algorithm which only considers the deadlines of real-time tasks.

关 键 词：任务调度 动态双向 优先级 主机/子机模式 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]