Int J Performability Eng ›› 2019, Vol. 15 ›› Issue (8): 2208-2216.doi: 10.23940/ijpe.19.08.p21.22082216

Previous Articles     Next Articles

Cloud-OM Patching: A Novel Video Stream Scheduling Scheme based on Hybrid Cloud-Overlay Architecture

Guangqian Kong*, Xun Duan, and Yun Wu   

  1. School of Computer Science, Technology, Guizhou University, Guiyang, 550025, China
  • Submitted on ;
  • Contact: * E-mail address: gq_kong@163.com
  • About author:Guangqian Kong received his Ph.D. from Guizhou University in 2009. He is currently an associate professor, graduate supervisor, and member of the China Computer Society. His research interests include computer networks, cloud computing, software defined networks, and deep learning and its applications. Xun Duan received his Ph.D. from Guizhou University in 2007. He is currently an associate professor and graduate supervisor. His research interests include distributed computing and cloud computing and its applications. Yun Wu received his Ph.D. from Guizhou University in 2009. He is currently an associate professor, graduate supervisor, and member of the China Computer Society. His research interests include distributed computing, game theory, recommender system, and big data and its applications.
  • Supported by:
    This research is supported by the Natural Science Foundation of China (No. 61741124) and the Science and Technology Plan Project of Guizhou Province (Qiankehe Platform Talent) (No. 5781).

Abstract: Patching is an effective multicast video stream scheduling technology that provides "real" VoD services. However, patching streams cannot be shared by other users, resulting in insufficient server scalability. In addition, IP multicast cannot be deployed on the Internet in a large scale; thus, the application of multicast stream scheduling technology is limited. Based on the above problems, this paper first proposes a hybrid cloud-overlay multicast architecture based on overlay network and cloud computing. It consists of three parts: cloud layer, overlay layer, and monitoring layer. Then, based on this architecture, a Cloud-OM patching stream scheduling algorithm is proposed, which combines patching stream sharing and multicast tree level time difference cache sharing technology. The experimental results show that Cloud-OM patching can effectively reduce the server bandwidth requirements and has better performance than standard patching, double patching, and batched patching, especially for popular videos.

Key words: overlay multicast, cloud computing, video-on-demand, video stream scheduling