Share Sens: An Approach to Optimizing Energy Consumption of Continuous Mobile Sensing Workloads

Smartphones and a growing number of wearable devices are equipped with a variety of powerful sensors. This has led to increased interest in developing of applications across a wide variety of domains including health-care, entertainment, environmental monitoring and transportation, which use sensor feeds to offer services. However, most of these applications require continuous sensing, which places a heavy demand on the device's typically limited battery power. This problem is further amplified as multiple applications attempt to monitor multiple sensors simultaneously. In this paper, we present ShareSens, our approach to opportunistically merge independent sensing requirements of applications. We achieve this using sensing schedulers for sensors, which determine the lowest sensing rate which would satisfy all requests, and then use custom filters to send out only the needed data to each application. Sensing requests made through the ShareSens API (which we have implemented for Android) are forwarded to the relevant schedulers which determine the optimum sensing rates to satisfy all requests. The paper presents the design and implementation of ShareSens, as well as results from our experimental work on the power savings that can be achieved by using it.