Streaming Caching Technologies for Next generation Multimedia Internet

（1）视频流的分布式/分层缓存算法由于视频数据量大，传统的Web缓存机制不能直接应用于视频流。为了克服这个问题，研究了几种算法，例如将视频数据分割成小段，以及通过引入以分布式方式合作的分层定位的缓存服务器。研究结果在国际会议上发表，并与下一个主题合并，提交给IEICE Transactions（已接受）。(2)Web友好的流缓存算法流缓存技术应该避免当前广泛部署的Web缓存方法的性能下降。然后，我们提出了“网络友好”的流缓存算法，利用统计特性的Web对象和流媒体对象在一个统一的方式。(3)层次图像数据格式缓存算法的改进研究了层次图像数据格式缓存算法的改进。开发了基于偏好统计的替换/预取算法，并对Web访问模型进行了理论分析。研究结果在国际会议上发表，提交给IEICE Transactions（已接受），并对视频流缓存的上述两个主题做出了贡献。

We investigated technologies on video streaming and related issues as follows:(1) Distributed/hierarchical caching algorithms for video streamingClassical web caching mechanisms cannot be directly applied to video streaming due to huge size of video data. To overcome this problem, several algorithms were investigated, such as splitting video data into small segments and by introduction of hierarchically located cache servers that cooperate in a distributed manner. Results were presented at international conferences and, with merged with the next topic, submitted to IEICE Transactions (accepted).(2) Web-friendly streaming caching algorithmsStream caching technologies should avoid performance degradation of current web caching methods that are widely deployed. We then proposed "web-friendly" stream caching algorithms that exploited statistical characteristics of web objects and streaming objects in a unified manner.(3) Improvement of caching algorithms for hierarchical image formatImproved caching algorithms for hierarchically represented image data format were investigated. Development replacing/pre-fetching algorithm according to preference statistics and theoretical analysis of the web access model were carried out. Results were presented at international conferences, submitted to IEICE Transactions (accepted), and contributed to the above two topics on video stream caching.