科技通报
科技通報
과기통보
BULLETIN OF SCIENCE AND TECHNOLOGY
2015年
6期
154-156
,共3页
QoS开销%满意度%权衡%云计算
QoS開銷%滿意度%權衡%雲計算
QoS개소%만의도%권형%운계산
QoS overhead%satisfaction%balance%cloud computing
提出一种引入QoS开销适应度运算的云计算任务权衡调度算法，首先进行了支持多QoS因素任务调度问题描述与网格拓扑结构构建，进行云计算任务权衡调度对多用户QoS偏好的影响力数学度量，通过多QoS开销适应度运算的引入，根据计算资源的成本和数据传输时间，来确定分配任务的位置。为了适应云存储中的多QoS偏好，重新定义PSO的适应度函数，实现任务权衡调度算法的改进。通过仿真实验研究得出，采用该算法对云计算任务节点的聚类准确性较高，进行任务调度中的实时性好。通过多QoS偏好分析，引入QoS开销适应度运算，用户满意率有明显上升，适应度函数随不同类别任务变化，有效地反映不同类型任务的QoS偏好。展示了较好的云计算任务权衡调度性能。
提齣一種引入QoS開銷適應度運算的雲計算任務權衡調度算法，首先進行瞭支持多QoS因素任務調度問題描述與網格拓撲結構構建，進行雲計算任務權衡調度對多用戶QoS偏好的影響力數學度量，通過多QoS開銷適應度運算的引入，根據計算資源的成本和數據傳輸時間，來確定分配任務的位置。為瞭適應雲存儲中的多QoS偏好，重新定義PSO的適應度函數，實現任務權衡調度算法的改進。通過倣真實驗研究得齣，採用該算法對雲計算任務節點的聚類準確性較高，進行任務調度中的實時性好。通過多QoS偏好分析，引入QoS開銷適應度運算，用戶滿意率有明顯上升，適應度函數隨不同類彆任務變化，有效地反映不同類型任務的QoS偏好。展示瞭較好的雲計算任務權衡調度性能。
제출일충인입QoS개소괄응도운산적운계산임무권형조도산법，수선진행료지지다QoS인소임무조도문제묘술여망격탁복결구구건，진행운계산임무권형조도대다용호QoS편호적영향력수학도량，통과다QoS개소괄응도운산적인입，근거계산자원적성본화수거전수시간，래학정분배임무적위치。위료괄응운존저중적다QoS편호，중신정의PSO적괄응도함수，실현임무권형조도산법적개진。통과방진실험연구득출，채용해산법대운계산임무절점적취류준학성교고，진행임무조도중적실시성호。통과다QoS편호분석，인입QoS개소괄응도운산，용호만의솔유명현상승，괄응도함수수불동유별임무변화，유효지반영불동류형임무적QoS편호。전시료교호적운계산임무권형조도성능。
A new QoS overhead fitness computing cloud computing task scheduling algorithm first constructs a trade-off is proposed, support for multiple QoS factors of task scheduling problem description and mesh topology structure of cloud com?puting, task scheduling metrics tradeoff influence of mathematics to multiuser QoS preference, adaptation through multiple QoS overhead of the introduction of operations, according to the cost of computing resources and the data transfer time, to determine the allocation of the location for the task. In order to adapt to the cloud storage of multi QoS preference, redefin?ing the fitness of PSO, improved task scheduling algorithm to achieve balance. Through the experimental study of simula?tion, the algorithm can use this algorithm in the face of cloud clustering accuracy higher calculation task node, for real-time task scheduling of a good. Through the multi QoS preference analysis, introducing the QoS overhead fitness operations, cus?tomer satisfaction rate increased significantly, the fitness function with the change of different types of tasks, it can effective?ly reflect the preferences of different types of task QoS, it show good cloud computing task scheduling performance.
