基于访问特征分析的流媒体存储系统节能方法研究

项目名称： 基于访问特征分析的流媒体存储系统节能方法研究

项目编号： No.61202115

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 计算机科学学科

项目作者： 柴云鹏

作者单位： 中国人民大学

项目金额： 24万元

中文摘要： 流媒体存储系统的高能耗问题逐渐成为制约其发展的一项重要挑战，虽然针对通用存储系统节能方法的研究已经成为研究热点，但这些方法很难适应流媒体应用的多种特征，流媒体存储系统的节能问题还有待于进一步深入研究。申请者在前期的调研和实验分析中发现，在流媒体环境中达到优异节能效果的关键，一是降低节能方法的数据迁移开销，二是充分利用流媒体用户访问规律的潜在信息。因此本项目拟基于大量流媒体用户访问记录，抽取和分析与节能密切相关的流媒体应用的用户访问特征，并建立准确的流媒体存储系统的能耗模型；然后在此基础上，提出低数据迁移开销的动态节能数据分布方法，以及基于流媒体用户访问规律的静态节能数据分布方法；并建立原型系统进行验证。总的来说，本项目能够为流媒体存储系统的节能减排提供理论与实验依据，促进流媒体应用的进一步发展，同时也有助于环境保护，具有重要的研究意义。

中文关键词： 节能；能耗；存储；流媒体；特征分析

英文摘要： The high energy consumption has gradually been a serious challenge for streaming media storage systems. Although saving energy in general storage systems appears as a hot research topic in recent years, the energy conservation problem of streaming media storage systems needs further in-depth study. Based on the surveys and preliminary experiments, we have found that the keys of efficiently saving energy in streaming media environments lie in two factors: The first one is lowering the data migration cost of energy saving methods; the second is fully ulitilizing some potential information of streaming media user accessing. Therefore, this project aims to extract and analyze the user accessing characteristics of streaming media applications, which is highly related to energy conservation, on the basis of a huge number of collected streaming media user traces. Then based on establishing accurate energy saving model of streaming media storage systems, the dynamic energy-efficient data layout method with low data migration cost and the static energy-efficient data layout algorithm based on streaming media user accessing patterns will be proposed in the project. Finally, we will build a prototype system to test and verify the energy saving algorithms we proposed. In a word, this project has important significance, beca

英文关键词： Energy Conservation；Power Consumption；Storage；Streaming Media；Characteristics Analysis