CAJ | 학술논문

为了在云计算任务调度过程中保证云设备效率的同时提高资源利用率，提出了一种基于优先级和动态电压频率调节的调度算法。首先，为调度算法定义了问题模型以规范异构服务器的性能；然后，利用优先级为任务提供可行的组合或调度；最后，利用动态电压频率调节为服务器提供适当的电压和频率供应，并且向虚拟机管理器发送分配结果。利用可扩展的模拟工具CloudSim进行实验评估了本文方法的能耗和调度时间，结果表明，本文方法的执行时间与MMF-DFVS方法相当，而能耗比MMF-DFVS降低了5％-25％。
위료재운계산임무조도과정중보증운설비효솔적동시제고자원이용솔，제출료일충기우우선급화동태전압빈솔조절적조도산법。수선，위조도산법정의료문제모형이규범이구복무기적성능；연후，이용우선급위임무제공가행적조합혹조도；최후，이용동태전압빈솔조절위복무기제공괄당적전압화빈솔공응，병차향허의궤관리기발송분배결과。이용가확전적모의공구CloudSim진행실험평고료본문방법적능모화조도시간，결과표명，본문방법적집행시간여MMF-DFVS방법상당，이능모비MMF-DFVS강저료5％-25％。
In order to guarantee efficiency of cloud equipment and improve resource utilization in thejob scheduling process of cloud computing, a scheduling algorithm based on priority and dynamic voltage frequency adjustment is pro_posed in this paper. Firstly, problem model is defined to regulate the performance of the heterogeneous server for the scheduling algorithm. Then, the priority is used to provide feasible combination or schedule for tasks. Finally, dynamic voltage frequency regulation is used for the server to provide the proper voltage and frequency supply, and it will send al_location result to the virtual machine manager. Extensible simulation tool CloudSim is used to do experiments and evalu_ate energy consumption and operation time of proposed method. Experimental results show that proposed method has exe_cution time with MMF-DFVS, but its energy consumption has reduced by 5%-25% than MMF-DFVS.