CAJ | 학술논문

传统的多媒体数据流调度方法在云平台环境下，未考虑服务器信息调度的差异性，容易形成数据调度冲突，调度效率低。为了解决上述分析的问题，通过构建多媒体云计算下数据流调度平台，实现对多路大规模多媒体数据流的合理调度，利用多级分层结构将多媒体云计算下的多服务器大规模数据流调度过程划分成管理层面、控制层面以及数据层面，可确保将数据包均匀的分配给各个服务器，充分发挥云计算下的多服务器可并行处理的特性，增强数据流的调度质量。对调度平台的软件框架进行了详细的描述，依据该软件框架的多层次实现大规模数据流调度优化，分析了大规模数据流调度的具体实现过程，并给出大规模数据流调度平台软件代码的设计。实验结果表明，所提方法增强了多媒体云计算下的大规模数据流调度的并发性能，提高数据流调度质量。
전통적다매체수거류조도방법재운평태배경하，미고필복무기신식조도적차이성，용역형성수거조도충돌，조도효솔저。위료해결상술분석적문제，통과구건다매체운계산하수거류조도평태，실현대다로대규모다매체수거류적합리조도，이용다급분층결구장다매체운계산하적다복무기대규모수거류조도과정화분성관리층면、공제층면이급수거층면，가학보장수거포균균적분배급각개복무기，충분발휘운계산하적다복무기가병행처리적특성，증강수거류적조도질량。대조도평태적연건광가진행료상세적묘술，의거해연건광가적다층차실현대규모수거류조도우화，분석료대규모수거류조도적구체실현과정，병급출대규모수거류조도평태연건대마적설계。실험결과표명，소제방법증강료다매체운계산하적대규모수거류조도적병발성능，제고수거류조도질량。
The traditional multimedia data stream scheduling method does not consider the difference of the server informa?tion scheduling in the cloud environment,and is easy to form the data scheduling conflict,which may result in low scheduling efficiency. In order to solve the problem of the above analysis,by constructing the data stream scheduling platform under the condition of multimedia cloud computation to achieve rational scheduling of multi?channel large?scale multimedia data stream,the massive data stream scheduling process of multiple servers in multimedia cloud computation is divided into management level, control level and data level by using multi?stage stratified structure,so as to ensure that the data packet is evenly assigned to each server,give full play to the parallel processing features of multiple servers in cloud computation,and enhance the scheduling quality of data stream. The software framework of dispatching platform is described in detail. According to the multi?level feature of software framework,the scheduling optimization of large?scale data stream is realized. The specific implementation process of massive data streams scheduling is analyzed. The design of the software code for massive data streams scheduling platform is giv?en. The experimental results indicate that the proposed method has enhanced the concurrent performance of massive data flow scheduling under the condition of multimedia cloud computation and improved the scheduling quality of data stream.