Dr Huber Flores - Regularity-Based Optimization of the Energy Footprint of Mobile Sensing Applications (venia legendi)

Klipi teostus: Ahti Saar 22.03.2019 1300 vaatamist Arvutiteadus

Abstract: Mobile and wearable sensing is emerging as a fundamental field for understanding users, their behaviours, activities, and even personal states. While the sensing capabilities of mobile devices are continually evolving, with latest devices incorporating anything from accelerometers to location tracking technologies to even air quality monitoring capabilities and (near) medical grade heart monitoring, optimally taking advantage of the increasing sensing capabilities of smart devices is becoming a very complex problem. As devices are battery constrained, sensing needs to minimize its impact on the overall performance of the device. However, at the same time, care should be taken to ensure all relevant information about the monitored phenomena is being captured. In this talk, we analyze the impact that sensing has in the performance of the device. We demonstrate that significant benefits in performance can be achieved by optimizing the energy-information trade-off in mobile sensing applications through adaptive duty cycling.