作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程

• 体系结构与软件技术 • 上一篇    下一篇

基于资源预测的智能终端资源缓存算法

徐 超1,2,曾学文1,郭志川1   

  1. (1. 中国科学院声学研究所国家网络新媒体工程技术研究中心,北京100190; 2. 中国科学院大学,北京100049)
  • 收稿日期:2014-04-01 出版日期:2015-03-15 发布日期:2015-03-13
  • 作者简介:徐 超(1986 - ),男,博士研究生,主研方向:嵌入式系统,多媒体技术;曾学文,研究员、博士生导师;郭志川,副研究员。
  • 基金资助:
    国家科技支撑计划基金资助项目“电视商务综合体新业态运营支撑系统开发”(2012BAH73F01);中国科学院先导专项课题 基金资助项目“智能电视平台与服务支撑环境研制”(XDA06040501)。

Smart Terminal Resource Cache Algorithm Based on Resource Prediction

XU Chao 1,2,ZENG Xuewen 1,GUO Zhichuan 1   

  1. (1. National Network New Media Engineering Research Center,Institute of Acoustics,Chinese Academy of Sciences,Beijing 100190,China; 2. University of Chinese Academy of Sciences,Beijing 100049,China)
  • Received:2014-04-01 Online:2015-03-15 Published:2015-03-13

摘要: 针对智能电视终端应用间资源竞争导致的系统性能下降问题,基于资源消耗预测,提出一种智能终端资源缓存算法。根据系统记录的各应用程序的资源消耗统计数据,应用Markov 模型预测下一时间段可能出现的资源瓶颈和应用的资源状态,利用应用的资源状态动态调整应用权重,并以最小化应用切换时间为目标,将资源缓存问题转化为多维多选择背包问题,采用轻量级的启发式算法求解资源缓存问题。仿真实验结果表明,在智能终端中该算法对于资源消耗的预测精确度比其他算法提高5. 4% ,而应用响应时间缩短约45% 。

关键词: 智能电视终端, 资源预测, Markov 模型, 资源缓存算法, 多维多选择背包问题, 启发式算法

Abstract: In order to solve the problem of the performance degradation caused by the resource competition among the applications in smart TV, this paper presents a resource usage prediction based resource cache algorithm. The applications’ resource consumption data is recorded and the resource state and resource bottleneck of the next time interval are predicted by Markov model. The weight of each application is adjusted dynamically and the resource cache problem is converted to multidimensional multiple-choice knapsack problem to minimize the switch time of the application. A lightweight heuristic solution algorithm with lower time complicity is presented. Simulation results show that the precision of the resource usage prediction of the algorithm is superior to others by about 5. 4% ,and the switch time of the application is reduced by about 45% .

Key words: smart TV terminal, resource prediction, Markov model, resource cache algorithm, multidimensional multiple-choice knapsack problem, heuristic algorithm

中图分类号: