• Title/Summary/Keyword: 푸시-메시 구조

Search Result 2, Processing Time 0.016 seconds

An Efficient Peer-to-Peer Streaming Scheme Based on a Push-Mesh Structure (푸시-메시 구조 기반의 효율적인 피어투피어 스트리밍 기법)

  • Kim, Jin-Sung;Kim, Dong-Il;Kim, Eun-Sam;Pae, Sung-Il
    • Journal of the Korea Society of Computer and Information
    • /
    • v.15 no.3
    • /
    • pp.81-89
    • /
    • 2010
  • The research on peer-to-peer streaming schemes has largely focused on tree-push and mesh-pull structures. However, the tree-push structure has a defect that the tree restructuring time is long, and the mesh-pull structure has long startup delay and lag time from source servers. In this paper, we propose a new peer-to-peer live streaming scheme based on a push-mesh structure that takes advantages of tree-push and mesh-pull structure simultaneously. This structure basically provides the mesh-pull mechanism for data transmission and utilizes peers with high network upload capacity. It also supports the push mechanism along with paths from a source server, super peers, and selected general peers. By NS-2 simulation experiments, we finally show that our proposed scheme can achieve shorter startup delay than the mesh-pull structure, similar lag time to tree-push structure and best playback continuity among the three schemes.

A Mobile P2P Streaming Architecture for Efficient Handover (효율적인 핸드오버를 위한 모바일 P2P 스트리밍 구조)

  • Kim, Sang-Jin;Kim, Eun-Sam;Hwang, Ho-Young
    • Journal of the Korea Society of Computer and Information
    • /
    • v.17 no.2
    • /
    • pp.95-103
    • /
    • 2012
  • In this paper, we propose an efficient mobile P2P streaming architecture to minimize playback jitters in P2P overlay networks based on wired/wireless networks even though mobile peers experience the handover. In this architecture, mobile peers receive data in a push manner to maximize the data receiving speed before and after the handover. In addition, they can maintain the buffering above a specific level through handover prediction and re-selection of neighbor peers. By simulation experiments, we show that our proposed architecture can improve the performance significantly when the handover occurs compared to the existing mesh structure.