Energy-Efficient Data Acquisition in Mobile Crowdsensing Systems

Abstract

Mobile Crowdsensing (MCS) is one of the most promising paradigms for monitoring phenomena in urban environments. The success of a MCS campaign relies on large participation of citizens, who may be reluctant in joining a campaign due to sensing and reporting costs they sustain. Hence, it is fundamental to propose efficient data collection frameworks (DCFs). In the first stages of our work, we proposed an energy-efficient DCF that aims to minimize energy consumption while maximizing the utility of contributed data. Then, we developed an Android application and proposed a methodology to compare several DCFs, performing energy-and network-related measures with Power Monitor and Wireshark. Currently, we are investigating collaborative data delivery as a more efficient solution than the individual one. The key idea is to form groups of users and elect a responsible for aggregated data delivery. To this end, it is crucial to analyze device to device (D2D) communications and propose efficient policies for group formation and owner election. To evaluate the performance in realistic urban environments we exploit CrowdSenSim, which runs large-scale simulations in citywide scenarios.