Energy Efficiency Analysis And Optimization For Mobile Platforms

<p>The introduction of mobile devices changed the landscape of computing. Gradually, these devices are replacing traditional personal computer (PCs) to become the devices of choice for entertainment, connectivity, and productivity. There are currently at least 45.5 million people in the United States who own a mobile device, and that number is expected to increase to 1.5 billion by 2015.</p> <p>Users of mobile devices expect and mandate that their mobile devices have maximized performance while consuming minimal possible power. However, due to the battery size constraints, the amount of energy stored in these devices is limited and is only growing by 5% annually. As a result, we focused in this dissertation on energy efficiency analysis and optimization for mobile platforms. We specifically developed SoftPowerMon, a tool that can power profile Android platforms in order to expose the power consumption behavior of the CPU. We also performed an extensive set of case studies in order to determine energy inefficiencies of mobile applications. Through our case studies, we were able to propose optimization techniques in order to increase the energy efficiency of mobile devices and proposed guidelines for energy-efficient application development. In addition, we developed BatteryExtender, an adaptive user-guided tool for power management of mobile devices. The tool enables users to extend battery life on demand for a specific duration until a particular task is completed. Moreover, we examined the power consumption of System-on-Chips (SoCs) and observed the impact on the energy efficiency in the event of offloading tasks from the CPU to the specialized custom engines. Based on our case studies, we were able to demonstrate that current software-based power profiling techniques for SoCs can have an error rate close to 12%, which needs to be addressed in order to be able to optimize the energy consumption of the SoC. Finally, we summarize our contributions and outline possible direction for future research in this field.</p>

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Open Access Dissertation

Discipline thesis:degree_discipline

Computer Science

Year dc:date.available

2014

Author dc:creator

• Metri, Grace

Contributors dc:contributor

• Monica Brockmeyer
• Weisong Shi

• Energy Efficiency
• Energy Efficient Programming
• Mobile Computing
• Mobile device
• Computer Sciences
• Oil, Gas, and Energy

Repository record dc:identifier

https://digitalcommons.wayne.edu/oa_dissertations/970

OAI identifier oai:identifier

oai:digitalcommons.wayne.edu:oa_dissertations-1969