高级检索
    汪学舜, 余少华, 徐 宁, 戴锦友. 适用于流媒体传输的动态门限调度算法[J]. 计算机研究与发展, 2011, 48(1): 110-117.
    引用本文: 汪学舜, 余少华, 徐 宁, 戴锦友. 适用于流媒体传输的动态门限调度算法[J]. 计算机研究与发展, 2011, 48(1): 110-117.
    Wang Xueshun, Yu Shaohua, Xu Ning, Dai Jinyou. A Dynamic Threshold Schedule Algorithm for Media Traffic[J]. Journal of Computer Research and Development, 2011, 48(1): 110-117.
    Citation: Wang Xueshun, Yu Shaohua, Xu Ning, Dai Jinyou. A Dynamic Threshold Schedule Algorithm for Media Traffic[J]. Journal of Computer Research and Development, 2011, 48(1): 110-117.

    适用于流媒体传输的动态门限调度算法

    A Dynamic Threshold Schedule Algorithm for Media Traffic

    • 摘要: 为保证网络流媒体传输质量,在流媒体传输网络中需要对交换设备实施有效的缓存管理策略.结合流媒体数据对时延和丢包率敏感的特点,提出了一种基于E-Model传输级别的动态门限控制算法(EDTA),将交换设备的缓存管理分为全局门限控制和队列门限控制两部分.全局门限控制通过判断不同的传输场景进行门限控制,保证不同传输场景下的传输性能.队列门限控制则基于E-Model传输级别参数进行门限的划分和调整,保证流媒体的传输质量.实验结果表明该算法能有效改善流媒体的传输性能,对交换设备的缓存资源进行有效控制,并提高网络流媒体传输质量.

       

      Abstract: Real-time multimedia applications are getting more and more popular in the last several years with internet capability keeping improved, in order to guarantee the transmission quality of medium traffic, effective strategies for buffer management are applied in exchanging devices in flow medium network, such as switches and routers. Considering the characteristics of packet delay and loss sensitivity of medium data, this paper proposes a novel dynamic threshold algorithm at media traffic level for shared buffer packet switches, known as dual-dynamic threshold control based on E-model(EDTA). EDTA divides buffer management of switch into two parts: the global threshold strategy and the queue threshold strategy. The global threshold strategy controls the shared buffer by a whole threshold through judging different traffic scenarios to ensure good transmission performance in all situations. The queue threshold strategy decides and adjusts output queue threshold based on E-model traffic levels, which transforms several medium service quality effecting parameters into users’ psychological factors, thus guarantees the quality of flow medium data. Experimental results show that our proposed algorithm (EDTA) can improve the traffic performance for flow media more effectively, achieve better quality of network flow medium transmission, comparing with those algorithms using traditional control. And it is suitable to be used in any network condition and can improve the utilization of the network resources.

       

    /

    返回文章
    返回