摘要
提出了一种计算缓存区大小的改进方案.根据RTCP包中反馈回的丢包率,动态地改变音频数据的发送策略,有区分地采取差错控制手段,提高了网络带宽的利用率.在此基础上,设计了一个以RTP协议为基础的音频会议系统的模型,并在Linux平台上用多线程的方法实现了各个模块.模拟实验表明,该系统能自适应地调整缓存区大小和数据包的发送策略,在实时性及带宽资源的利用上都达到了较理想的效果.
Put forward a improved method of calculating the size of the buffer.Changing the packet sending tactic according to the packet loss rate fed back in RTCP and adopting mistake control method improve the using rate of the bandwidth.On the basis of it,design a set of Audio Meeting based on RTP,and realize various modules with multithread on Linux.The final experiment shows that the system can auto adjust the size of the buffer and the sending method of packets,and it has reached a ideal effect in several aspects such as the quality of real time and the the using rate of the bandwidth.
出处
《武汉大学学报（理学版）》
CAS
CSCD
北大核心
2004年第1期109-112,共4页
Journal of Wuhan University:Natural Science Edition
基金
教育部"新世纪网络课程建设工程"基金资助项目(144100123)
