CAJ | 학술논문

传统的基于粒子群的云计算任务调度算法不能感知QoS用户偏好，任务调度活性不强，适应度低。对此提出一种融合QoS静默爬虫感知的云计算任务调度算法，定义活性因子进行活性表征，提高云计算任务调度中的活性，进而提高用户的满意率。设计了网络QoS静默爬虫算法，实现了对云任务的弹性抓取，静默爬虫数据在Javascript程序内部经过变量赋值、传递，字符编码和过滤，在向量空间模型中，进行多QoS因素的静默爬虫活性度表征，得到任务调度分配链路中的比例分值，有效反映出云计算任务调度的结构特征，提高了任务调度效率和活性成分。仿真结果表明，采用多QoS因素静默爬虫感知技术，能有效地反映不同类型任务的QoS偏好，提高了任务调度的活性，整个调度过程的时间开销影响不大，总执行时间最多也仅比传统PSO多了10%左右，但以较少的执行时间增加换取用户的满意率是值得的。展示了算法的优越性能和较好的应用价值。
전통적기우입자군적운계산임무조도산법불능감지QoS용호편호，임무조도활성불강，괄응도저。대차제출일충융합QoS정묵파충감지적운계산임무조도산법，정의활성인자진행활성표정，제고운계산임무조도중적활성，진이제고용호적만의솔。설계료망락QoS정묵파충산법，실현료대운임무적탄성조취，정묵파충수거재Javascript정서내부경과변량부치、전체，자부편마화과려，재향량공간모형중，진행다QoS인소적정묵파충활성도표정，득도임무조도분배련로중적비례분치，유효반영출운계산임무조도적결구특정，제고료임무조도효솔화활성성분。방진결과표명，채용다QoS인소정묵파충감지기술，능유효지반영불동류형임무적QoS편호，제고료임무조도적활성，정개조도과정적시간개소영향불대，총집행시간최다야부비전통PSO다료10%좌우，단이교소적집행시간증가환취용호적만의솔시치득적。전시료산법적우월성능화교호적응용개치。
The traditional particle swarm cloud computing task scheduling algorithm cannot perceive QoS user preference, task scheduling activity is not strong with low fitness. An improved cloud computing task scheduling activity description al?gorithm is proposed based on fusion of QoS silent crawler aware, activity factor is defined, the cloud task scheduling activity is improved. Then improve user's satisfaction rate, the network silencing of QoS crawler algorithm is designed. The elastic grab on the cloud task is realized, silent crawler data within the Javascript program is taken with assignment, transfer, char?acter encoding and filtering. In the vector space model, multi factor QoS silence crawler activity characterization is taken. The task scheduling and allocation proportion of value link are obtained. It can effectively reflect structure characteristics of task scheduling in cloud computing, improve the scheduling efficiency and active ingredients. The simulation results show that, the factors of QoS silent crawler sensing technology can effectively reflect the different QoS preference, the task scheduling activity is improved. The entire scheduling process time cost is reduced. The total execution time is about 10%than traditional PSO. The algorithm has superior performance and good application value.