使用熱門廣播法及支援點對點傳輸之影音內容傳遞網路; Content Delivery Network with Hot-video broadcasting and Peer-to-peer Approach

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/8759

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/8759

Title:	使用熱門廣播法及支援點對點傳輸之影音內容傳遞網路;Content Delivery Network with Hot-video broadcasting and Peer-to-peer Approach
Authors:	劉之揚;Chih-Yang Liu
Contributors:	資訊工程研究所
Keywords:	點對點傳輸網路;影音多媒體串流;快取伺服器;內容傳遞網路;熱門廣播法;分散式雜湊表;Distributed hash;Peer-to-peer network;Media streaming;Cache;Content delivery network;Hot-video broadcasting
Date:	2003-06-30
Issue Date:	2009-09-22 11:34:19 (UTC+8)
Publisher:	國立中央大學圖書館
Abstract:	隨著網路頻寬的成長，對於影音多媒體串流的需求也隨之增加，隨選視訊系統 (VOD) 的發展也被認為是下一波網際網路應用的主流。由於影音多媒體串流的資料量大、頻寬需求高、持續時間長，因此對於播放伺服器的負擔也較重，使得系統規模受到限制。為解決此問題，常見的做法為在網路中建置影片快取伺服器，藉由快取部份或全部的影片 (Content Delivery Network, CDN)，來分擔影片伺服器的負載、降低頻寬的使用量；另一種方式為利用內容傳輸網路，將影片預先放置在各地的內容快取伺服器中，將使用者的請求導向至離其最近的內容快取伺服器，使其可以快速的取得要看的影片；另外也有人提出使用點對點傳輸網路的方法來做影音多媒體的串流，因為點對點傳輸網路 (Peer-to-peer Network) 中沒有集中式的伺服器，資料是由使用者間彼此分享，所以不會有伺服器過載的情況，並且系統可以有較佳的擴充性及容錯能力。本文提出了一結合熱門廣播法及點對點傳輸的影音內容傳遞網路。系統使用兩層式的雜湊表，將所有的影片及使用者映射到系統中的影片伺服器及內容快取伺服器，並且經由影片名稱來快速的找到影片伺服器，經由使用者的來源位置快速的找到距離其最近的內容快取伺服器。內容快取伺服器負責接受使用者的請求，向影片伺服器抓回影片區段，將其存在快取中，並且使用熱門廣播法傳送給使用者，使用者在接收過影片後，利用動態更新將自已的位置記錄在雜湊表中，讓後來進入系統的使用者可以找到已經看過影片的使用者，而向其請求影片區段。最後，我們實作出整套系統，並且實際在網路上播放以測試其效能。 As the growth of bandwidth of Internet, the demand of media streaming is increasing as well. The development of video on demand system is considered as next main application on the internet. Because media streaming uses more bandwidth and last longer than conventional web page, the loading on video server is much heavier. Thus the system’s scalability is limited. To solve these problems, the common approach is using video cache to share load of original server and reduce bandwidth usage. Another approach is using Content Delivery Network, by deploy content replica over internet, user request is redirect to closest replica thus effectively access the video. Peer-to-peer network is also using in media streaming. There is no centralized server in p2p network, and data is shared among users. So there isn’t server overloading problem and system will have better scalability and ability on fault-tolerating We purposed a content delivery network with hot-video broadcasting and peer-to-peer approach. By using two-level hash table, we can map all videos and clients to video servers and content replicas. Time for finding video server is minimized by hashing the name of video and through hashing client’s address we can quickly locate the closest content replica. Content replica is responsible for client’s request, fetching video segments from video server and cache at disk. Then using hot-video broadcast video segments to client. Once client had received video segments, it registers its location on the hash table thus clients arrive later can retrieve video from it. Finally we implement the system and test it on real network.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Size	Format

社群 sharing

Loading...