Gaussian mixture model for crowdsensing incentivization

Abstract

Participatory Sensing (PS) is a novel technology that involves people in the process of gathering sensing data (i.e. local temperature, pollution levels, etc.), most of the time the sensing process is carried out using embedded sensors in smartphones which implies an active participation of the users. However, even though, there has been a lot research in areas such as: new applications of PS, studies of PS energy consumption, etc., there have been very few works that tackle the problem of incentive the participation of the users in order to keep a minimum number of participant that guarantee quality of the service and none of them that address both reliable sensing and good coverage. In this project, we address the problem of user participation incentive in participatory sensing scheme under constraints of budget and coverage by the use of combination of techniques such as Game theory, unsupervised learning, and multi-objective optimization.