Energy Diagnosis of Android Applications: A Thematic Taxonomy and Survey

Abstract

The abnormal energy consumption of Android applications is a significant problem faced by developers and users. In recent years, researchers have invested their efforts to develop energy diagnosis tools that pinpoint and fix the energy bugs from source code automatically. These tools use traditional software engineering methods such as program analysis, refactoring, software repair, and bug localization to diagnose energy inefficiencies. Existing surveys focus only on energy measurement techniques and profiling tools and do not consider automated energy diagnosis tools. Therefore, this article organizes state of the art by surveying 25 relevant studies on Android applications' automatic energy diagnosis. Further, this survey presents a systematic thematic taxonomy of existing approaches from a software engineering perspective. The taxonomy presented in this article would serve as a body of knowledge and help researchers and developers to understand the state of the field better. The future research directions discussed in this article might help prospective researchers to identify suitable topics to improve the current research work in this field. © 2020 ACM.