Fine-Grained Energy Modeling for the Source Code of a Mobile Application (1510.04165v2)

Abstract: Energy efficiency has a significant influence on user experience of battery-driven devices such as smartphones and tablets. The goal of an energy model for source code is to lay a foundation for the application of energy-saving techniques during software development. The challenge is to relate hardware energy consumption to high-level application code, considering the complex run-time context and software stack. Traditional techniques build the energy model by mapping a hardware energy model onto software constructs; this approach faces obstacles when the software stack consists of a number of abstract layers. Another approach that has been followed is to utilize hardware or operating system features to estimate software energy information at a coarse level of granularity such as blocks, methods or even applications. In this paper, we explain how to construct a fine-grained energy model for the source code, which is based on "energy operations" identified directly from the source code and able to provide more valuable information for code optimization. We apply the approach to a class of applications based on a game-engine, and explain the wider applicability of the method.