以使用者為導向的智慧型手機能源管理，在平衡能源消耗和手機效能時，同時也考慮使用者體驗。相關研究近年備受關注。由於目前智慧型手機主要透過螢幕與使用者互動，因此相關研究多以幀速率 (frame rate，單位為幀/秒 (FPS)) 做為使用者體驗的指標。而手機能源管理最常採用的方法則為動態電壓/頻率調整 (dynamic voltage frequency scaling, DVFS)，藉由動態調整CPU/GPU的頻率來因應系統的工作負載，讓FPS維持在使用者滿意的水平。然而，前人的研究主要集中在遊戲或視頻應用程式。這類程式需要不斷地繪製和更新幀緩衝器 (frame buffer)，因此令使用者滿意的FPS值固定而且容易掌握。另一方面，智慧型手機上的許多應用程式是以頁面為基礎與使用者互動，其畫面的更新取決於使用者的互動事件，例如社交應用程式臉書，讓使用者滿意的FPS值是動態變動的。更進一步，透過實際使用者的滿意度調查，我們發現若系統丟掉若干比例的幀並不影響使用者對手機顯示的體驗，手機省電機制可以做更多。這便為以使用者為導向的智慧型手機能源管理開了一條新路。根據這樣的實測結果，我們提出了一個新的使用者體驗指標，稱為畫幀率(frame drawn rate, FDR) 來量化使用者體驗。依據此一新指標，我們建立了一套新的CPU-GPU DVFS機制，並且將其實作在智慧型手機上。我們徵求使用者來使用和評估我們建立的DVFS調整機制，並且與手機現有的DVFS調整機制比較。根據受邀者使用手機的滿意度回饋，發現我們的DVFS調整機制在使用者體驗方面比現有方法表現一樣好甚至更好，並且能達到平均17.2％、最高20.58%的省電。

User-centric power management for smartphones, which takes user’s quality of experience (QoE) into account while trading off power consumption and performance, has received much attention recently. Most existing works focus on gaming and video streaming applications and use the frame rate, measured in frame per second (FPS), as the metric to quantify user’s QoE. The key idea is to adjust the CPU/GPU frequency just enough to maintain the frame rate at a user satisfactory level. What are less studied are applications that draw and update the frame buffer aperiodically in response to user inputs, such as social applications and web browsers. These applications are normally page based. Although the above frequency scheduling policy can still work, our experiments with real users show that a more aggressive frequency governor may be developed for these applications that purposely drops a fraction of frames and the users’ perceived visual quality is not affected. Based on this observation, we propose in this thesis a new QoE metric called the frame drawn ratio (FDR). A new CPU/GPU frequency governor using FDR is then developed to determine the optimal CPU/GPU frequency setting. Real users are asked to use the smartphones with the new governor, and their use experiences and power consumption of the phones are evaluated. The experimental results show that the proposed governor can save in average 17.2% power consumption against the default governor, while maintaining the same or even better QoE rating.

1 Introduction...........................1
2 Methodology...........................7
2.1 FDRmetric.................................... 7
2.2 Relationship between FDR and user’sQoE................... 9
2.3 FDR governor................................... 11
2.3.1 Achieving FDRgood ............................ 12
2.3.2 Maintaining FDRgood........................... 13
3 Evaluation ...........................16
3.1 ExperimentSetup................................. 16
3.2 EvaluationResult................................. 18
3.2.1 Correlation between FDR and the performance . . . . . . . . . . . . 18
3.2.2 Correlation between FDR and user’sQoE . . . . . . . . . . . . . . . 20
3.2.3 Power consumption comparison with different governor . . . . . . . . 22
3.2.4 User satisfaction with different governor . . . . . . . . . . . . . . . . 24
3.2.5 FDR governor overhead......................... 27
4 Discussion and limitation...........................28
5 Conclusion...........................30

[1] Ravishankar Rao, Sarma Vrudhula, and Daler N Rakhmatov, “Battery modeling for energy aware system design”, Computer, vol. 36, no. 12, pp. 77–87, 2003.
[2] Aaron Carroll, Gernot Heiser, et al., “An analysis of power consumption in a smart- phone.”, in USENIX annual technical conference. Boston, MA, 2010, vol. 14, pp. 21–21.
[3] Anuj Pathania, Qing Jiao, Alok Prakash, and Tulika Mitra, “Integrated cpu-gpu power management for 3d mobile games”, in Proceedings of the 51st Annual Design Automa- tion Conference. ACM, 2014, pp. 1–6.
[4] Alex Shye, Benjamin Scholbrock, and Gokhan Memik, “Into the wild: studying real user activity patterns to guide power optimizations for mobile architectures”, in Mi- croarchitecture, 2009. MICRO-42. 42nd Annual IEEE/ACM International Symposium on. IEEE, 2009, pp. 168–178.
[5] Alex Shye, Benjamin Scholbrock, Gokhan Memik, and Peter A Dinda, “Characteriz- ing and modeling user activity on smartphones: summary”, in ACM SIGMETRICS Performance Evaluation Review. ACM, 2010, vol. 38, pp. 375–376.
[6] Yu-Ming Chang, Pi-Cheng Hsiu, Yuan-Hao Chang, and Che-Wei Chang, “A resource- driven dvfs scheme for smart handheld devices”, ACM Transactions on Embedded Com- puting Systems (TECS), vol. 13, no. 3, pp. 53, 2013.
[7] Venkatesh Pallipadi and Alexey Starikovskiy, “The ondemand governor”, in Proceedings of the Linux Symposium. sn, 2006, vol. 2, pp. 215–230.
[8] Lei Yang, Robert P Dick, Gokhan Memik, and Peter Dinda, “Happe: Human and application-driven frequency scaling for processor power efficiency”, IEEE transactions on mobile computing, vol. 12, no. 8, pp. 1546–1557, 2013.
[9] Sascha Bischoff, Andreas Hansson, and Bashir M Al-Hashimi, “Applying of quality of experience to system optimisation”, in Power and Timing Modeling, Optimization and Simulation (PATMOS), 2013 23rd International Workshop on. IEEE, 2013, pp. 91–98.
[10] Wei-Ming Chen, Sheng-Wei Cheng, Pi-Cheng Hsiu, and Tei-Wei Kuo, “A user-centric cpu-gpu governing framework for 3d games on mobile devices”, in Proceedings of the IEEE/ACM International Conference on Computer-Aided Design. IEEE Press, 2015, pp. 224–231.
[11] Youngho Ahn and Ki-Seok Chung, “User-centric power management for embedded cpus using cpu bandwidth control”, IEEE Transactions on Mobile Computing, vol. 15, no. 9, pp. 2388–2397, 2016.
[12] Yuhao Zhu, Matthew Halpern, and Vijay Janapa Reddi, “Event-based scheduling for energy-efficient qos (eqos) in mobile web applications”, in High Performance Computer Architecture (HPCA), 2015 IEEE 21st International Symposium on. IEEE, 2015, pp. 137–149.
[13] Wook Song, Nosub Sung, Byung-Gon Chun, and Jihong Kim, “Reducing energy con- sumption of smartphones using user-perceived response time analysis”, in Proceedings of the 15th Workshop on Mobile Computing Systems and Applications. ACM, 2014, p. 20.
[14] Mark Claypool, Kajal Claypool, and Feissal Damaa, “The effects of frame rate and resolution on users playing first person shooter games”, in Electronic Imaging 2006. International Society for Optics and Photonics, 2006, pp. 607101–607101.
[15] Benjamin F Janzen and Robert J Teather, “Is 60 fps better than 30?: the impact of frame rate and latency on moving target selection”, in Proceedings of the extended abstracts of the 32nd annual ACM conference on Human factors in computing systems. ACM, 2014, pp. 1477–1482.
[16] Daecheol You and Ki-Seok Chung, “Quality of service-aware dynamic voltage and frequency scaling for embedded gpus”, IEEE Computer Architecture Letters, vol. 14, no. 1, pp. 66–69, 2015.
[17] “https://source.android.com/devices/graphics/implement-vsync/”.
[18] “https://developer.android.com/studio/profile/dev-options-rendering.html/”.
[19] “https://developer.android.com/studio/command-line/adb.html/”.
[20] “https://play.google.com/store/apps/details?id=com.facebook.katana/”.
[21] “https://play.google.com/store/apps/details?id=com.instagram.android/”.
[22] Katarzyna Wac, Selim Ickin, Jin-Hyuk Hong, Lucjan Janowski, Markus Fiedler, and Anind K Dey, “Studying the experience of mobile applications used in different contexts of daily life”, in Proceedings of the first ACM SIGCOMM workshop on Measurements up the stack. ACM, 2011, pp. 7–12.
[23] Ulrich Reiter, Kjell Brunnstro ̈m, Katrien De Moor, Mohamed-Chaker Larabi, Manuela Pereira, Antonio Pinheiro, Junyong You, and Andrej Zgank, “Factors influencing quality of experience”, in Quality of experience, pp. 55–72. Springer, 2014.
[24] Yu Bai and Priya Vaidya, “Memory characterization to analyze and predict multime- dia performance and power in embedded systems”, in Acoustics, Speech and Signal Processing, 2009. ICASSP 2009. IEEE International Conference on. IEEE, 2009, pp. 1321–1324.
[25] “https://browser.primatelabs.com/”.
[26] “http://www.cs.waikato.ac.nz/ml/weka/”.
[27] Jurn-Gyu Park, Chen-Ying Hsieh, Nikil Dutt, and Sung-Soo Lim, “Quality-aware mo- bile graphics workload characterization for energy-efficient dvfs design”, in EmbeddedSystems for Real-time Multimedia (ESTIMedia), 2014 IEEE 12th Symposium on. IEEE, 2014, pp. 70–79.
[28] Benedikt Dietrich and Samarjit Chakraborty, “Lightweight graphics instrumentation for game state-specific power management in android.”, Multimedia Systems, vol. 20, no. 5, 2014.
[29] Po-Kai Chuang, Ya-Shu Chen, and Po-Hao Huang, “An adaptive on-line cpu-gpu gov- ernor for games on mobile devices”, in Design Automation Conference (ASP-DAC), 2017 22nd Asia and South Pacific. IEEE, 2017, pp. 653–658.
[30] “https://android.googlesource.com/kernel/common/+/a7827a2a60218b25f222b54f77ed38f57aebe08b/ freq/governors.txt”.
[31] “https://www.msoon.com/”.
[32] “https://developer.android.com/training/testing/ui-testing/uiautomator- testing.html/”.
[33] Arindam Mallik, Jack Cosgrove, Robert P Dick, Gokhan Memik, and Peter Dinda, “Picsel: Measuring user-perceived performance to control dynamic frequency scaling”, in ACM SIGARCH Computer Architecture News. ACM, 2008, vol. 36, pp. 70–79.