基於網格運算且具備工作負載分享之數位所有權管理平台

Abstract

分散式運算（Distributed Computing）允許應用程式透過網路，在不同地方的電腦上執行，將使得運算任務不必集中在某台電腦，而是以功能任務分派給各地電腦運算，再傳回結果與回覆任務。隨著系統架構的演進，相較於早期的主從式(Client-Server) 架構，同儕運算(Peer-to-Peer)或網格運算(Grid Computing)架構的分散式系統應用成為主流。 本研究提出一個基於網格運算下且具有工作負載分享之數位所有權管理(GC-DRM)平台並符合Grid Portal標準，使用Globus Toolkit 4與Condor6.9.2來作為平台的基礎，並利用Web 2.0 based介面來提供使用者選擇服務、遞交、控制、管理工作與即時監控進度，目前提供六項服務：1、浮水印嵌入與偵測2、影像攪亂3、可視浮水印嵌入4、影像竄改驗證與還原5、浮水印攻擊測試6、自訂工作。然而為了使得平台具有QOS的動態資源協調能力，來降低整體工作執行時間，在平台中，透過一個工作評估服務，利用電腦效能指標，來預測工作執行時間與降低所傳送的Overhead並進行工作合併，使平台保有穩定的運算能力。 在本研究中，此方法運用在分析數位浮水印強度的濾波器組（Filter Bank）選擇，以及平台服務所產生出來的大量工作，用以測試執行效果，其結果和未使用工作評估服務，在整體執行時間上，本研究的工作評估服務皆能加以改善。

Distributed Computing allows applications running on different locations computers to interact with each other across the network. It can dispatch task to remote computers and return results to finish mission without focus on one computer. As evolution of distributed computing environment are progressing, P2P and Grid Computing are gaining more attention and dominated than centric computing. In this research, we proposed a loading balancing digital rights management platform based on Grid Computing so-called GC-DRM and compliance of Grid Portal standards by using porlet. The platform integration of Globus Toolkit 4 and Condor 6.9.2 to represent it and use web 2.0 to construct the web-based user interface for providing job submission, control, management, monitor and DRM services. It provides six category of services, including watermark embedded and extraction, image scrambling, visible watermark embedded, image tamper-proof and recovery, watermark algorithms attack testing, user custom task. However, for increasing the platform computing capability and decreasing job transmission overhead, we show a module it called job evaluation services by using computer performance to predict job execution time and according predict result to merge jobs into one. Finally, the module has been applied to analyze the performance of digital watermark by filter bank selection. As the result of this study, the performance can be improved in the aspect of speedup, efficiency and process time.