Sciences & Société
Soutenance de thèse : Mohamed Anis FEKIH
Low-cost Wireless Sensor Networks in Participatory Air Quality Monitoring
Doctorant : Mohamed Anis FEKIH
Laboratoire INSA : CITI
Ecole doctorale : ED512 : Informatique et mathématiques de Lyon
Mobile crowdsensing sensing is an emerging and promising paradigm that has attracted much attention in recent years, especially for environmental monitoring. Coupled with the power of low-cost wireless sensor networks (WSN), it leverages population density to collect extensive data in many applications, such as air pollution monitoring. In fact, air pollution is one of the main problems that still suffers from a lack of characterization due to the limitations of traditional assessment methods employed in terms of cost, network size, and flexibility. Mobile crowdsensing and WSN aim at filling this gap by enabling large-scale deployments in order to improve the local knowledge on the phenomenon.
In this thesis, we consider the air quality monitoring application using mobile crowdsensing, while focusing on three parts : 1) the design of low-cost participatory air quality monitoring systems; 2) the analysis of dense low-cost WSN data and their benefit in generating high-resolution air pollution maps; 3) the selection of the participants' paths in order to improve the characterization, while considering the constraints of travel distance and sensor errors. Through this work, we aim to show the potential of using low-cost WSN and participatory sensing in air quality monitoring. In this vein, we conduct substantial experimental work on the design of a participatory air quality monitoring system from scratch. We provide engineering guidelines regarding the design of low-cost participatory environmental monitoring platforms. Moreover, we propose a general framework that allows the comparison of different strategies based on numerical simulations and an adequate estimation of the simulation error as well as the sensing error. We also explore the impact of the sensing rate on the energy consumption and the characterization error. Furthermore, we propose two new algorithms for route selection in participatory sensing. Our approaches take into account the participants' constraints and the characteristics of low-cost WSN.
Amphithéâtre Claude Chappe, Département TC (Bâtiment Hedy Lamarr) - Villeurbanne