使用者介面延遲測量與過度繪製檢測於手機應用程式效能測試

UI latency Measurement and Overdraw Detection in Performance Testing of Mobile Applications

王文樺(Wen-Hua Wang)

指導教授 ： 王凡

國立臺灣大學/電機資訊學院/電機工程學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202103996

摘要

隨著行動裝置的普及，使用手機App已經成為現代人日常生活的一部份，App的市場也日益競爭，其中對效能的要求也更加重要，不好的效能會嚴重的影響使用者體驗，本篇論文針對下列兩個效能測試的議題，各提出一個新的解決方案：
1.使用者介面延遲測量：以往的開發者假設要測量應用程式中各個動作的延遲，需要進行程式的插裝，也就是在目的程式中加入額外指令來收集需要的資訊，但這種測量的方式會花費開發者很大的心力，而測量出的時間也不能完全代表使用者在真實的情境下所花費的時間，本篇論文提出了一種基於影像且非侵入式的黑箱測試方法來測量手機App的使用者介面延遲。
2.過度繪製檢測：當應用程式中有重複背景的設計時，儘管使用者最終只會看到最上層的顏色，但其實這種設計會消耗GPU的資源，嚴重時可能會導致掉幀以至於降低使用者體驗，本篇論文提出了一種自動化的過度繪製檢測技術，並產生測試報告告知開發者多餘的重複背景出現的區域。

關鍵字

自動化測試 ； 黑箱測試 ； 介面延遲 ； 過度繪製

並列摘要

With the popularization of mobile devices, mobile applications have become a part of people’s daily lives, and the application market is more and more competitive. Therefore, the performance of an application is essential. Poor performance will seriously affect the user experience. This thesis mainly discusses the following two issues of performance testing:
1.UI latency measurement:
In the past, if developers wanted to measure the UI latency of each action in the application, they need to do program instrumentation. However, this measurement method would cost developers much effort, and the measured time cannot completely represent the time spent by the user in an actual situation. This thesis proposes an image-based and non-intrusive black box testing method to measure the UI latency of mobile applications.
2.GUI Overdraw detection:
When an application contains a design of overlapped background, although users will only see the uppermost color finally, this design will consume lots of GPU resources. In severe cases, it may cause jank and lead to poor user experience. This thesis proposes an automated overdraw detection method and generates a test report to inform the developer of the redundant background area.

並列關鍵字

automated testing ； black-box testing ； UI latency ； overdraw

參考文獻

S. Yang, D. Yan, and A. Routev, “Testing for poor responsiveness in Android applications,” In Proc. International Workshop on the Engineering of Mobile-Enabled Systems. MOBS '2013. 10-20.

M. Linares-Vásquez, C. Vendome, Q. Luo, and D. Poshyvanyk, “How developers detect and fix performance bottlenecks in android apps,” in ICSME’15, 2015, pp. 352–361.

Y. Gao, Y. Luo, D. Chen, H. Huang, W. Dong, M. Xia, X. Liu, and J. Bu, “Every pixel counts: Fine-grained UI rendering analysis for mobile applications,” in IEEE Conference on Computer Communications. IEEE, 2017, pp. 1–9

R. Zhou, S. Shao, W. Li, and L. Zhou, “How to define the user’s tolerance of response time in using mobile applications,” in 2016 IEEE International Conference on Industrial Engineering and Engineering Management (IEEM), Dec. 2016, pp. 281–285.

V. N. Inukollu, D. D. Keshamoni, T. Kang, and M. Inukollu. “Factors Influencing Quality of Mobile Apps:Role of Mobile App Development Life Cycle.” ArXiv e-prints, Oct. 2014.