¡¡Chinese Journal of Computers   Full Text
  TitleResearch on Key Technologies of Large-Scale Streaming Media
  AuthorsYIN Hao1) LIN Chuang1) WEN Hao1) CHEN Zhi-Jia1) WU Da-Peng2)
  Address1)(Department of Computer Science and Technology, Tsinghua University, Beijing 100084)
2)(Department of Electrical & Computer Engineering, University of Florida, Florida 32611-6130 USA)
  Year2008
  IssueNo.5(755¡ª774)
  Abstract &
  Background 		Abstract Large-scale online streaming video application has gradually become the most important and potential technology of multimedia applications. Since Internet is heterogeneous and there is no Quality of Service (QoS) and security guarantee, large scale online streaming video application over the Internet faces many challenges. This paper, following the development of video sourcing coding and network transmission technology, comprehensively and deeply presents the scalable encoding and error-resilient encoding, and the development of multicast and P2P technology. Specially a new streaming architecture is proposed to satisfy the requirement of QoS, heterogeneity,and security guarantee in practical application. Some detailed analysis and comparisons are conducted in this paper. At last, future research direction is proposed.
Keywords streaming video; QoS; video sourcing coding; peer to peer
Background This work is supported by the Key Project Foundation of Technology and Development Program of Beijing Municipal Commission of Education (No.KZ2005100005006), and the National Natural Science Foundation of China (Nos.90407017, 60403035), and the National High Technology Development Program(863 Program) of China (No.2006AA01Z205).
As an important solution providing QoS for users, packet scheduling has attracted many attentions. IETF has published many RFCs and drafts in this filed and a lot of top conferences and journals have proposed novel scheduling algorithms. However, most of existing packet scheduling algorithms do not guarantee fairness and priority at the same time. Thus some researchers improved them according to the proportion fairness principle, but regretfully their scheduling algorithms have high computational complexity. In this scenario, the authors propose a new packet scheduling algorithm realizing the proportion fairness principle, named PFWRR. PFWRR has low computation load because it renews each queue¡¯s weight only when the system is overload and the average packet arrival rate changes.
The authors¡¯ projects focus on the key technologies in wireless/mobile network including QoS, security and modeling and so on. They have developed a prototype system of wireless access point with QoS, and have published some relevant papers.