摘要
针对智能电视终端面临的运行时资源冲突和过载控制问题,提出一种在线自适应多资源调度方法(OARSA)。该方法根据资源特点和应用优先级对独占型资源进行动态优先级抢占式调度,避免了资源冲突;采用反馈控制算法对共享型资源进行Qo S调度,直到实现Qo S比例公平性,并采用Qo S比例压缩法消除资源过载。实验结果表明,OARSA具备过载控制能力,仅以系统总体Qo S下降4.5%为代价,解决了资源冲突,并提高了系统重负载下的响应性能,且算法耗时比现有方法降低约30%,更加适用于智能电视终端的资源调度。
To solve the problem of resource contention and overload control during the run-time resource scheduling of the smart TV terminal, this paper proposed an online adaptive resource scheduling algorithm（OARSA）. According to the property of resources of the system, the method adopted the dynamic priority preemptive scheduling strategy of exclusive resources to a- void the resource contention. Then it allocated the shared resources by the feedback QoS scheduling algorithm until the appli- cations＇ QoS levels were proportional fair, and controlled the resource overload by the QoS proportional compression. The ex- periment result shows that OARSA improves the system response characteristic under heavy task load and prevents the resource contention at the cost of the 4.5% decrement of overall system QoS, and the time consumption of OARSA decreases about 30% compared with the existing methods, which is more applicable for the resource scheduling in the smart TV terminal.
出处
《计算机应用研究》
CSCD
北大核心
2015年第3期697-700,共4页
Application Research of Computers
基金
国家科技支撑计划资助项目(2012BAH73F01)
中国科学院先导专项课题(XDA06040501)
关键词
智能电视终端
资源调度
优先级调度
反馈控制
smart TV terminal
resource scheduling
priority scheduling
feedback control
