NDNlive:命名数据网络下的视频直播系统

发布时间：2018-05-08 23:18

本文选题：命名数据网络 + 视频流　；参考：《计算机科学与探索》2017年07期

【摘要】：命名数据网络(named data networking,NDN)通过将IP网络中以地理位置驱动的信息交互方式转变成为以数据为中心的信息交互模式,为内容分发应用例如视频播放提供了更好的支持。通过利用NDN的命名机制与数据获取模式,设计并实现了基于NDN的视频直播系统(NDNlive),将实时捕捉的视频传输给多用户。与传统的定长切片技术不同,NDNlive将视频流按照应用数据单元(帧)进行切分与获取。同时对于音频、视频和元数据信息,依照其数据属性和生成模式采用不同的数据获取方法。由于帧获取流水线策略提供的灵活性,NDNlive可以容忍小的网络问题。NDNlive被部署在NDN全球测试平台中,实验结果表明,NDNlive可以在全球跨11个时区提供流畅和同步的视频直播流。
[Abstract]:Named data networking provides better support for content distribution applications such as video playback by transforming geographically driven information interactions in IP networks into data-centric information interaction modes. By using the naming mechanism and data acquisition mode of NDN, a live video broadcasting system based on NDN is designed and implemented to transmit real-time captured video to multiple users. Different from the traditional fixed-length slicing technique, NDNlive segments and acquires the video streams according to the application data units (frames). At the same time, for audio, video and metadata information, different data acquisition methods are adopted according to their data properties and generation patterns. Because of the flexibility provided by the frame acquisition pipeline policy, NDNlive can tolerate small network problems. NDNlive is deployed in the NDN global test platform. The experimental results show that NNlive can provide smooth and synchronized live video streaming across 11 time zones around the world.
【作者单位】：清华大学计算机科学与技术系;清华大学清华信息科学与技术国家实验室;加州大学洛杉矶分校计算机系;清华大学电子工程系;
【基金】：国家自然科学基金No.61373025 国家重点研发计划No.2016YFB1000303~~
【分类号】：TN948.6

【相似文献】