Towards an energy-efficient framework for location-triggered mobile application

Numerous Location-Based Applications (LBAs) have been emerging in our smart phones over the past several years. Most of them are user-triggered LBAs which heavily depend on end users' manual operations and are easy to lose their accessibility and usability when fail to get instructions in time. In contrast, location-triggered applications usually bring better user experiences by performing actions automatically when reaching predefined locations. However, there are still some burdens of location-triggered mechanisms to be widely adopted. The most challenging one is the lack of a unified framework to define location-triggered data collection behaviors and corresponding system internal behaviors. In this paper, we propose a preliminary framework to manage location-specific actions associated with specific targets of interest. We explored both dynamic data collection scenarios and the mobile system internal reaction scenario. We also examine the battery draining problem caused by periodically duty-cycling power-intensive GPS sensor. A new application level energy-efficient location detection approach is proposed which intelligently duty-cycles GPS with other on-device sensors. The results of our real-world experiment demonstrate that mobile device could reduce the usage of powerintensive GPS location update operation by as much as 90% with this new approach in location-triggered application.