Energy-efficient crowdsensing at mass events

OData support
Supervisor:
Dr. Simon Vilmos
Department of Networked Systems and Services

To ensure safety at large scale events, describing the movement of a crowd is essential, which may be performed by knowing the location of participants. The level of development and penetration of our everyday communications technology allows for participants to be equipped with smartphones which are connected to mobile data networks and which incorporate many different sensors, and therefore able to measure and transmit location data. The thesis delivers a method that aims for localizing pedestrians participating at a music festival through their smartphones. While considering the desired accuracy of the use case, it takes energy efficiency into account to prevent batteries from draining too early. Besides acquiring fix points by GPS-based localization, the proposed solution uses the built-in ambient light sensor, the accelerometer, and the magnetometer to detect user activity and to estimate distance and heading, and keeps random error within required limits. The validation of the solution is performed by evaluating sensor data acquired from real measurements, which proved expectations and the efficiency of the algorithm.

Downloads

Please sign in to download the files of this thesis.