计算机应用与软件2016，Vol.33Issue(8)：179-182,4.DOI:10.3969/j.issn.1000-386x.2016.08.040
基于带宽预测的流媒体超级节点选择算法
STREAMING MEDIA SUPER NODES SELECTION ALGORITHM BASED ON BANDWIDTH PREDICTION
摘要
Abstract
The selection of super group peer (SGP)is an important factor affecting the system fluency and playing quality of P2P streaming media.According to the characteristics of P2P system,we improved the traditional random selection algorithm,and proposed the nodes selec-tion algorithm,namely ELM-SGP,which uses ELMextreme learning machine.Through estimating the node bandwidth and the CPU real-time loading at the next time moment,the comprehensive availability of super nodes can be assessed.The ordinary nodes select SGP according to its comprehensive strength of availability,this effectively avoids the blindness and randomness in random selection algorithm,and also enables the system to provide users with a stable and reliable high-quality service.It is proved by the experiment that relative to traditional random se-lection algorithm,ELM-SGP algorithm’s system throughput is increased by 7.74%,and its playback delay is reduced by 44.4%,the broad-cast quality remains stable at 90% or higher.关键词
ELM/超级节点/P2P 流媒体/带宽Key words
ELM/Super node/P2P streaming media/Bandwidth分类
信息技术与安全科学引用本文复制引用
魏赟,韩少恒..基于带宽预测的流媒体超级节点选择算法[J].计算机应用与软件,2016,33(8):179-182,4.基金项目
国家自然科学基金项目（61170277）；上海市教委科研创新基金项目（12YZ094）。 （）
