Fine-Grained Energy Modeling for the Source Code of a Mobile Application

The goal of an energy model for source code is to lay a foundation for the application of energy-aware programming techniques. State of the art solutions are based on source-line energy information. In this paper, we present an approach to constructing a fine-grained energy model which is able to provide operation-related information that is more valuable for guiding code-optimization than source-line information. The modeling is enabled by a set of novel and practical techniques such as source-level operation identification, block-varied execution-case design and measurement variability control. Using the model we observed several counter-intuitive effects, e.g., in a common game scenario, control flow operations consume around 38% of the total CPU energy use, while arithmetic operations consume only 6%. Our model is being integrated into a source-level energy-optimization approach, which we briefly describe and the paper includes a case study to illustrate how the model guides energy optimization.