基於效能剖析之動態應用程式轉移之方法探討-以Android系統為例

A Profile-Driven Dynamic Application Offloading Scheme for Android Systems

陳勇偉(Yong-Wei Chen)

指導教授 ： 洪士灝

國立臺灣大學/電機資訊學院/資訊工程學研究所/碩士(2012年)

https://doi.org/10.6342/NTU.2012.01956

摘要

近年來, 由於處理器、記憶體、儲存裝置、網路與應用程式的進步,使得智慧型行動裝置日趨普及,如智慧型手機及平板電腦。然而,因為行動裝置有限的資源與電量,造成目前的智慧型行動裝置無法完全處理重量級的工作及滿足複雜應用程式的要求。若可將這些工作轉移到雲端,將有可能克服這些限制。在這篇論文中,我們藉由改進之前的雲端服務與引進flow-based programming,使此系統有更多的靈活性且使開發者易於在此平台上開發程式。此外,我們也將探討如何妥善運用雲端資源與如何決定轉移決策,提供一套動態轉移的方法。

關鍵字

雲端運算 ； 普及運算 ； 協同應用 ； 效能分析 ； 能源分析

並列摘要

Smart mobile devices, such as smartphones and tablets, are gaining strong demands recently with the advances in processors, memories, storages, communication networks, and software applications. However, due to their limited form factors and battery capacities, the current generation of smart mobile devices cannot yet fulfill the requirements of sophisticated applications. Leveraging a cloud service to offload the application workload can potentially overcome these limitations. In the thesis, based on previous works, a novel scheme is proposed on mobile-cloud computing by extending the programming model on Android smartphones with a flow-based programming paradigm to provide more flexibility for application migration. For properly provisioning the resources and making smart decisions based on the performance gain, a profile-based policy manager and profiling service are added into our framework to enable a dynamic offload scheme.

並列關鍵字

cloud computing ； pervasive computing ； collaborative application ； performance analysis ； energy-consumption analysis ； process offloading

參考文獻

execution,” in Proceedings of the 12th conference on Hot topics in operating systems,

Strategy Through 2015.” [Online]. Available: http://www.gartner.com/it/page.jsp?id=

plications on the cloud: Framework and issues,” in Computers; Mathematics with Appli-

[4] J. P. Morrison, “Flow-based programming, 2nd edition: A new approach to application

development,” 2012.