雲端遊戲系統效能評估:利用使用者滿意度提供及解剖式分析

Evaluation of Cloud Gaming Systems’ Performance with QoE Provisioning and Anatomical Analysis

張裕鈞(Yu-Chun Chang)

指導教授 ： 雷欽隆

國立臺灣大學/電機資訊學院/電機工程學研究所/博士(2012年)

https://doi.org/10.6342/NTU.2012.00977

摘要

雲端遊戲是一種現代化的服務，它結合了雲端計算及線上遊戲的概念。這項服務可從遠端伺服器提供給玩家一整個遊戲體驗。由於高普及率的寬頻網路存取以及雲端計算技術的利用，玩家不需依賴高等的硬體和複雜的軟體程序，但可利用精簡型終端機配備，如輕量級電腦，有機上盒的電視，甚至移動裝置來玩遊戲。玩家也不用帶專屬的遊戲機，便可以隨時隨地玩他自己最喜愛的遊戲。鑑於這些潛在的優勢，對於遊戲開發商以及消費者來說，雲端遊戲一直被視為一個潛在的範式，其可改變電腦遊戲的提供和遊玩方式。然而，這也產生了新的服務品質的挑戰，例如哪個雲端遊戲系統提供最好的服務品質，哪個設計元素可構成一個良好的雲端遊戲系統。

我們這次的研究主要分為兩部分。在第一部分中，我們的目標是研究線上遊戲對於網路的健壯性以及了解其設計方案，例如航位推測法以及時間同步機制應該如何實施。在第二部分中，我們試著利用解剖式分析去評估雲端遊戲系統的效能。我們的目標是想要提出一種新的方法，來判斷哪個雲端遊戲系統可提供最好的服務品質。

我們研究的貢獻有三個方面：1）我們提出一個配對比較以及選擇機率模組來量化線上遊戲對於不同網路狀況的體驗質量。有了這樣的方法，我們能夠判斷遊戲的網絡設計和功能，如航位推測法如何設計，以及時間同步機制如何實施。2）我們提出了一種方法來量化透過精簡型終端機軟體玩遊戲的效能，即使這些軟體是封閉式的。並指出該軟體在客戶端部分中哪一個元素對於遊戲效能會有重要的影響。 3）我們提出了一個針對封閉性雲端遊戲系統的效能評估方法。利用這個方法，我們可以從三個層面: 流量特性，響應性，以及串流媒體的品質來衡量雲端遊戲系統的效能。

關鍵字

體驗質量 ； 精簡型終端機 ； 雲端遊戲系統 ； 雲端計算

並列摘要

Cloud gaming is a modern kind of service, which combines the concepts of cloud computing and online gaming. This service gives the entire game experience to players from a remote data center. Because of the high penetration rate of broadband Internet access and the use of cloud computing technology, players no longer rely on the high-level hardware and complex software, but are able to use thin-client devices, such as a lightweight PC, a TV with a set-top box, or even a mobile device. Players can play their favorite games anytime, anywhere without bringing their own game consoles. Given these potential advantages for both game developers and consumers, cloud gaming has been seen as a potential paradigm that would change the way computer games are delivered and played. However, it also raises new challenges of service quality, such as which cloud gaming system delivers the best quality of service and which design elements constitute a good cloud gaming system.

Our work is classified into two distinct parts. In the first part, we aim to investigate onine games' network robustness and find out design alternatives, such as how dead reckoning algorithms and time synchronization mechanisms should be implemented. In the second part, we strike for analyzing anatomically system-level performance of cloud gaming systems. Our goal is to define a novel methodology to identify which cloud gaming system delivers the best quality of service.

Our contributions are three-fold: 1) We proposed to use paired comparisons and probabilistic choice models to quantify online games' QoE under various network situations. With such a methodology, we were able to make decisions regarding game's network design and functionalities, such as how a dead reckoning algorithm should be designed and where the time synchronization mechanism should be implemented. 2) We proposed a methodology for quantifying the performance of thin-clients on gaming, even for thin-clients which are close-sourced, and pointed out which element at the client side has critical effect to gaming performance. 3) We proposed a novel methodology for evaluating the performance of proprietary and closed cloud gaming systems. With our methodology, the performance of a cloud gaming system can be evaluated in three aspects: traffic characteristics, responsiveness, and streaming quality.

並列關鍵字

quality of experience ； thin-client ； cloud gaming system ； cloud gaming

參考文獻

[21] G. Armitage. An experimental estimation of latency sensitivity in multiplayer Quake 3. In The 11th IEEE International Conference on Networks, pages 137–141, 2003.

[22] G. Armitage and L. Stewart. Limitations of using real-world, public servers to estimate jitter tolerance of first person shooter games. In Proceedings of ACM SIGCHI ACE 2004 Conference, pages 257–262, 2004.

[23] N. E. Baughman and B. N. Levine. Cheat-proof playout for centralized and distributed online games. In Proceedings of IEEE INFOCOM 2001, Anchorage, AK, April 2001.

[24] T. Beigbeder, R. Coughlan, C. Lusher, J. Plunkett, E. Agu, and M. Claypool. The effects of loss and latency on user performance in Unreal Tournament 2003. In Proceedings of NetGames’04, pages 144–151. ACM Press, 2004.

[25] R. A. Bradley and M. E. Terry. Rank analysis of incomplete block designs: I. the method of paired comparisons. Biometrika, 39(3/4):324–345, 1952.