使用可信任的快取及網路感知來加速行動設備瀏覽網頁;TCM: Accelerating Mobile Page Loads Using Trusted Cache and Network-Aware Compaction

NCUIR > College of Electrical Engineering & Computer Science > Executive Master of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/81152

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/81152

Title:	使用可信任的快取及網路感知來加速行動設備瀏覽網頁;TCM: Accelerating Mobile Page Loads Using Trusted Cache and Network-Aware Compaction
Authors:	周裕傑;Chou, Yu-Chieh
Contributors:	資訊工程學系在職專班
Keywords:	網路感知;Server Push;Server Hints;Inline;Cache;TPM;Network-Aware;Server Push;Server Hints;Inline;Cache;TPM
Date:	2019-07-31
Issue Date:	2019-09-03 15:37:19 (UTC+8)
Publisher:	國立中央大學
Abstract:	現在行動設備瀏覽網頁的流量已經正式超過桌上型電腦，然而在行動設備上瀏覽網頁的速度緩慢，讓使用者感到沮喪，並損害了網站提供商的收入。我們提出了一種可信任的快取模組框架(Trusted Cache Module; TCM)，能夠在標準的商業化瀏覽器上，根據用戶端的網路條件來動態的壓縮影像資源，並使用 HTTP/2 的伺服器推送(Server Push)機制，將壓縮後的影像資源檔案推送給瀏覽器。我們的網站伺服器會使用資源嵌入(Inline)機制，儘量的減少用戶端請求資源的數量，並且能保有資源嵌入後的可快取性，同時，我們針對已被快取的資源檔案，使用透過硬體保護的安全驗證機制，提供了比標準商業化瀏覽器更安全的資料完整性驗證。我們在實際的商業化新聞網站和實體的行動設備上的實驗結果表明，TCM 將平均的網頁載入時間減少了 26%、能源消耗減少了 15%、頻寬成本減少了 18%。;Mobile browsing now generates more HTTP traffic than desktop browsing, however, the existing slowness of the web on mobile devices frustrates users and hurts the revenue of website providers. We propose a Trusted Cache Module(TCM) framework that dynamically compresses image resources based on the client′s network conditions on a standard browser and uses the HTTP/2 push, the compress image resource file is pushed to the browser. Our web server to inline the page’s JavaScript and CSS files, and to minimize the number of HTTP requests needed to fetch the constituent resources, and can provide cacheability for inline resources. We use hardware-protected security authentication mechanism for cached resource files, and provides more secure data integrity verification than standard browsers. Experiments with real news pages and real phones show that TCM reduces median page load time by 26%, energy expenditure by 15%, and bandwidth costs by 18%.
Appears in Collections:	[Executive Master of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	260	View/Open

社群 sharing

Loading...