Adapp: an adaptive network selection framework for smartphone applications

Possibility to access different cellular data network services from the same mobile platform is quite high in near future [kill1]. This gives smartphone applications an opportunity to choose amongst multiple services based on their functionality and user-desired QoE. In this paper we propose a framework that predicts the service, which suits an application best and at the same time provides desired QoE, while saving energy and dollar costs. We also develop a prototype system, Adapp, that trains itself online with user feedbacks and its prediction accuracy improves over use. We demonstrate with the help of rigorous experiments, how different users have varying service preferences for the same application, while the same user can have different preferences across applications. Experimental results validate the adaptive nature of the system. We have also analyzed Adapp's accuracy in selecting the most appropriate service and the errors show an appreciable receding trend over time.