透過您的圖書館登入
IP:3.15.147.215
  • 學位論文

支援可適性視訊編碼之內容傳遞網路的設計

Design of Content Delivery Network with Scalable Video Coding Support

指導教授 : 吳和庭

摘要


內容傳遞網路(Content Delivery Network,簡稱CDN)是將檔案分散放置在不同的邊緣伺服器(Surrogate Server)中,以減輕原始伺服器(Origin Server)的負擔,並能更有效的提供使用者請求的內容,以增進整體服務品質。然而,由於資源以及環境的限制,系統可能無法完全滿足使用者的需求。因此需要一動態配置內容的機制,適時的將使用者需求的內容新增到邊緣伺服器中,或是將使用率低的內容從伺服器中刪除,以確保伺服器中存放的內容為真的被使用者所需要,同時也可避免不被需要的內容佔用系統資源。 本論文以支援可適性視訊編碼(Scalable Video Coding,簡稱SVC)的內容為背景,設計相對應的內容配置演算法,將使用者需要的內容自原始伺服器複製至邊緣伺服器中。同時,內容管理服務器(Content Manager)也會持續監控所有邊緣伺服器,移除使用率過低的內容。此外,考慮到當系統負載過重時,導致使用者必須等待的狀況,本論文設計了頻寬分配演算法,讓持有較多頻寬的使用者釋放部份頻寬,以減少使用者的平均等待時間。 本論文最後透過模擬的結果,分析在不同的伺服器頻寬、容量的限制下,演算法能夠替使用者所帶來的效益。

並列摘要


A content delivery network (CDN) contains a collection of surrogate servers that store files in multiple distributed nodes. These cooperating surrogate servers effectively mitigate the loading of the content origin server, increase performance for user content request services, and thus lead to an overall improvement in service quality. Due to limited resources and environment factors, such systems may be nevertheless subject to overwhelming demand from users that cannot be satisfied. A dynamic content allocation mechanism is needed to distribute requested content to surrogate servers for better utilization of resources across the CDN. This thesis studies the CDN resource allocation issues from a scalable video coding (SVC) application standpoint. A content allocation algorithm is designed to place user requested content from the origin server to surrogate servers. Content managers are deployed to monitor all surrogate server statuses and to remove cached content that experiences low hit rates. Furthermore, a bandwidth allocation algorithm is proposed to lower the response times during heavily loaded server operation. This mechanism reduces bandwidth allocated to incumbent users which occupy large bandwidth to decrease the average waiting time for a new user. Performance evaluation via simulations has bean carried out to reveal the advantage employing theses algorithms under various network bandwidth and capacity conditions.

參考文獻


[1] A. Vakali and G. Pallis, “Content Delivery Networks: Status and Trends,” IEEE Internet Computing, IEEE Computer Society, pp. 68-74, Nov.-Dec. 2003.
[4] K. Pathan, and R. Buyya, “A Taxonomy and Survey of Content Delivery Networks,” Tech Report, University of Melbourne, 2007.
[5] Y. Chen, L. Qiu, W. Chen, L. Nguyen, and R. H. Katz, “Efficient and adaptive web replication using content clustering,” IEEE Journal on Selected Areas in Communications, vol. 21, no. 6, pp. 979-994, Aug. 2003.
[6] G. Pallis and A. Vakali, “Insight and Perspectives for Content Delivery Networks,” Communications of the ACM, vol. 49, no. 1, ACM Press, NY, USA, pp. 101-106, Jan. 2006.
[8] The Coral Content Distribution Network, URL:http://www.coralcdn.org/ .

延伸閱讀