Data Verification andPrivacy-respecting User Remuneration in Mobile Crowd Sensing

The broad capabilities of current mobile devices have paved the way forMobile Crowd Sensing (MCS) applications. The success of this emergingparadigm strongly depends on the quality of received data which, in turn, iscontingent to mass user participation; the broader the participation, the moreuseful these systems become. This can be achieved if users are gratified fortheir contributions while being provided with strong guarantees for the securityand the privacy of their sensitive information. But this very openness is adouble-edge sword: any of the participants can be adversarial and pollute thecollected data in an attempt to degrade the MCS system output and, overall,its usefulness. Filtering out faulty reports is challenging, with practically noprior knowledge on the participants trustworthiness, dynamically changingphenomena, and possibly large numbers of compromised devices. This workpresents a holistic framework that can assess user-submitted data and siftmalicious contributions while offering adequate incentives to motivate usersto submit better quality data. With a rigorous assessment of our systemâAZssecurity and privacy protection complemented by a detailed experimentalevaluation, we demonstrate its accuracy, practicality and scalability. Overall,our framework is a comprehensive solution that significantly extends thestate-of-the-art and can catalyze the deployment of MCS applications.QC 20160129