分组交换网络语音回放控制及性能分析  

Playback Control and its Performance Analysis of Voice Traffic over Packet-Switch Network

作　　者：谭献海[1] 温碧丽[1] 金炜东[1] 

机构地区：[1]西南交通大学信息科学与技术学院,成都610031

出　　处：《计算机科学》2007年第7期113-115,119,共4页Computer Science

基　　金：西南交通大学科学研究基金项目(2005A03);国家自然科学资金项目(No60572143)

摘　　要：语音通信中的回放控制是保证语音平滑性的关键。本文深入研究分组交换网中的语音回放控制及其性能分析问题。在总结语音通信系统的特点和端到端时延的组成及其计算方法的基础上,提出了一个简单可行的语音业务回放控制模型,采用双令牌桶对语音流量进行整形。基于最新的网络演算理论,推导出了给定端到端时延、语音到达曲线和网络服务曲线条件下的语音回放时延、需要分配的速率和缓冲区长度的计算公式。最后通过应用实例分析验证了本文的分析结论。Playback control is the key guarantee scheme of voice srnoothing. In this paper, the problems of voice playback control and its performance analysis are studied. Based on the summary of the cbaracteristic of voice communication and its end-toend components, a simple and effective voice playback control model is proposed, which exploiting double token buckets as voice traffic shaper. Based on the novel network calculus theory, the formulations of playback delay, desired rate and buffer size are derived, given the end-to-end delay, the voice arrival curve and the network service curve Finally, the validity and effectiveness of the analytic results of the paper is verified with an application example.

关 键 词：分组交换网络 语音业务 回放 性能分析 网络演算 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]