{"title":"Bottom-up trie structure for P2P live streaming","authors":"Boyuan Zhang, Changcheng Huang, James Yan","doi":"10.1109/ICC.2012.6364344","DOIUrl":null,"url":null,"abstract":"By simultaneously providing live video and audio contents to millions of users around the world, peer-to-peer live video streaming (P2P LVS) has become one of the most popular Internet applications in recent years. However, current P2P LVS software has problems such as non-smooth playback and long start-up delay for end users. To address these issues, we design a P2P-based multi-bit Trie structure, called Bottom-Up Trie (BU-Trie), for distributing P2P live contents. Different from other approaches, BU-Trie is a Trie formed and built inversely from leaf nodes (or child nodes) back to the root node (or parent node). This architecture consists of two phases: a diffusion phase and a swarming phase. The main design goal of the diffusion phase is to group the local peers together by discovering physical locations of peers, and design the paths for fast distributing live streams from the source node to end users. The objective of the swarming phase is to find an optimal way for exchanging the video stream chunks within a local group. We propose an algorithm called Most Popular Chunk First (MPCF) and apply it for the swarming phase for efficient chunk exchange. Performance evaluation of the proposed BU-Trie shows that, when compared to other approaches, the sequential throughput of video chunks is increased. The inter-domain traffic, the traffic between different Internet service providers (ISPs), is reduced as well. Such a reduction would benefit carriers economically.","PeriodicalId":331080,"journal":{"name":"2012 IEEE International Conference on Communications (ICC)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2012-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2012 IEEE International Conference on Communications (ICC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICC.2012.6364344","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
By simultaneously providing live video and audio contents to millions of users around the world, peer-to-peer live video streaming (P2P LVS) has become one of the most popular Internet applications in recent years. However, current P2P LVS software has problems such as non-smooth playback and long start-up delay for end users. To address these issues, we design a P2P-based multi-bit Trie structure, called Bottom-Up Trie (BU-Trie), for distributing P2P live contents. Different from other approaches, BU-Trie is a Trie formed and built inversely from leaf nodes (or child nodes) back to the root node (or parent node). This architecture consists of two phases: a diffusion phase and a swarming phase. The main design goal of the diffusion phase is to group the local peers together by discovering physical locations of peers, and design the paths for fast distributing live streams from the source node to end users. The objective of the swarming phase is to find an optimal way for exchanging the video stream chunks within a local group. We propose an algorithm called Most Popular Chunk First (MPCF) and apply it for the swarming phase for efficient chunk exchange. Performance evaluation of the proposed BU-Trie shows that, when compared to other approaches, the sequential throughput of video chunks is increased. The inter-domain traffic, the traffic between different Internet service providers (ISPs), is reduced as well. Such a reduction would benefit carriers economically.
P2P视频流(peer-to-peer live video streaming, P2P LVS)通过向全球数百万用户同时提供实时视频和音频内容,已成为近年来最受欢迎的互联网应用之一。然而,目前的P2P LVS软件对终端用户来说存在播放不流畅、启动延迟长等问题。为了解决这些问题,我们设计了一个基于P2P的多比特树结构,称为自底向上树(BU-Trie),用于分发P2P实时内容。与其他方法不同的是,BU-Trie是从叶节点(或子节点)反向构建到根节点(或父节点)的Trie。该体系结构由两个阶段组成:扩散阶段和群集阶段。扩散阶段的主要设计目标是通过发现节点的物理位置将本地节点分组在一起,并设计从源节点到最终用户快速分发实时流的路径。蜂群阶段的目标是找到在本地组内交换视频流块的最佳方法。我们提出了一种称为最受欢迎的块优先(MPCF)算法,并将其应用于群集阶段,以实现高效的块交换。性能评估表明,与其他方法相比,所提出的BU-Trie方法提高了视频块的顺序吞吐量。域间流量,即不同互联网服务提供商(isp)之间的流量,也减少了。这样的削减将使航空公司在经济上受益。