摘要
针对现有基于推、拉以及推拉混合模式调度算法不能高效分发数据的同时保证节点数据接收质量,导致服务器负载重的问题,提出一种基于数据接收质量的自适应推拉调度算法。该算法根据节点缓存区数据被及时正确填充的情况动态调节推拉获取数据比例,当数据接收质量好时主要采用推方式获取数据,数据接收质量变差时则过渡到拉方式获取数据。仿真实验表明所提算法可以充分利用推拉方式各自优点加速数据在网络中传播的同时保证节点数据接收质量,降低服务器负载,提高系统可扩展性。
In order to solve the problem that existing push,pull and hybrid model cannot guarantee peer data receiving quality when distributing data among network and leads to heavy server pressure,a data receiving quality based adaptive Push-Pull scheduling algorithm is proposed.The algorithm adjusts data ratio received by peer between push and pull method dynamically according to condition that whether buffer is filled in time correctly.Push method is utilized when data receiving quality is good and pull method is utilized when data receiving quality turns bad.Simulation shows that the proposed algorithm can make full use of advantages of both push and pull to accelerate data spread among network as well as guarantee peer data receiving quality,which can reduce server pressure effectively and improve system scalability.
出处
《微计算机应用》
2011年第4期1-8,共8页
Microcomputer Applications
基金
十一五"国家科技支撑计划课题(2008BAH28B04)
国家自然科学基金青年基金项目(60903218F0208)
中国高技术研究发展计划(863)项目(2009AA01A339)
关键词
对等网络
流媒体
推拉调度
数据接收质量
自适应
Peer-to-Peer
Streaming
Push-Pull Scheduling
Data Receiving Quality
Adaptive