Application behavior analysis in resource consumption for mobile devices

The understanding of applications is a must to design proper resource scheduling policies and a cost-effective design of a mobile device. Different from the past work on user behavior studies or user-oriented resource allocation, we are interested in the "behavior" analysis of applications in resource consumption for mobile devices. In particular, we propose a way to select indices to capture the characteristics of a mobile device and then a way to visualize its resource consumption. Example streaming applications, namely YouTube and Vimeo, are taken as a case study to show the "behaviors" of the streaming applications in consuming resources and their difference. We then show how to use the same methodology to explore the differences in resource consumption for some functionalities of some popular social-network applications, namely Facebook and Twitter. The study provides not only insights to the resource needs of applications but also potential directions in software optimization, such as energy efficiency and data caching.