• DocumentCode
    3237727
  • Title

    A two-way available bandwidth estimation scheme for multimedia streaming networks adopting Scalable Video Coding

  • Author

    Li, Mingfu ; Chang, Chia-Rong

  • Author_Institution
    Dept. of Electr. Eng., Chang Gung Univ., Taoyuan
  • fYear
    2009
  • fDate
    March 30 2009-April 1 2009
  • Firstpage
    1
  • Lastpage
    6
  • Abstract
    Multimedia streaming applications have been increasing significantly due to the rapid growth of broadband access networks. Among the latest real-time streaming technologies, the scalable video coding (SVC) which is a layered coding scheme attracts more attention. The SVC encodes an original video into several streaming layers of different bit rates simultaneously. Thus, clients can subscribe a part of the coded layers of SVC video according to their available bandwidth, which should be measured by a bandwidth estimation methodology. Additionally, in SVC clients can augment or abate the number of subscribed streaming layers to fit the time-varying bandwidth during viewing programs. In this paper1, we exploit the concepts of Traceroute and ICMP timestamps to design a two-way available bandwidth estimation mechanism which is never presented in literature. The proposed scheme can infer the available bandwidth of both uplink and downlink paths, and is feasible to the paths with multiple tight links. Hence, it is specially applicable to the multimedia streaming applications adopting SVC. Numerical results show that the presented bandwidth estimation approach can achieve good accuracy.
  • Keywords
    video streaming; multimedia streaming networks; scalable video coding; time-varying bandwidth; two-way available bandwidth estimation scheme; Bandwidth; Bit rate; Delay estimation; Downlink; Explosives; Network servers; Optical fiber subscriber loops; Static VAr compensators; Streaming media; Video coding;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Sarnoff Symposium, 2009. SARNOFF '09. IEEE
  • Conference_Location
    Princeton, NJ
  • Print_ISBN
    978-1-4244-3381-0
  • Electronic_ISBN
    978-1-4244-3382-7
  • Type

    conf

  • DOI
    10.1109/SARNOF.2009.4850352
  • Filename
    4850352