A Source-level Energy Optimization Framework for Mobile Applications

TL;DR

This paper introduces a source-level energy optimization framework for mobile applications, enabling developers to understand and reduce energy consumption through targeted code refactoring, with significant energy savings demonstrated.

Contribution

It presents the first high-level language source-level energy optimization framework guided by an energy model for mobile apps.

Findings

Achieved 6.4% to 50.2% CPU energy savings in case studies.

Framework helps developers identify energy-intensive code segments.

Lays foundation for automatic code optimization tools.

Abstract

Energy efficiency can have a significant influence on user experience of mobile devices such as smartphones and tablets. Although energy is consumed by hardware, software optimization plays an important role in saving energy, and thus software developers have to participate in the optimization process. The source code is the interface between the developer and hardware resources. In this paper, we propose an energy-optimization framework guided by a source code energy model that allows developers to be aware of energy usage induced by the code and to apply very targeted source-level refactoring strategies. The framework also lays a foundation for the code optimization by automatic tools. To the best of our knowledge, our work is the first that achieves this for a high-level language such as Java. In a case study, the experimental evaluation shows that our approach is able to save from 6.4% to 50.2% of the CPU energy consumption in various application scenarios.