Participant Comfort Adaptation in Dependable Mobile Crowdsensing Services

Abstract

Mobile Crowdsensing (MCS) is a ubiquitous sensing concept under the Internet of Things (IoT) ecosystem where builtin sensors in smart mobile devices are utilized as users join in sensing campaigns launched by the crowdsensing platform. The pervasive and non-dedicated nature of the sensing instruments in MCS raises the trustworthiness issue. On the other hand, due to granting access to the hardware on their devices, user comfort –which is directly related to the information revealed or the type of sensor activation by user– is also another barrier against wide adoption of MCS in the IoT Era. In this article, we present an adaptive mechanism to manage user comfort in an adaptive manner while ensuring the trustworthiness of the crowdsensed data through auction based reputation maintenance at the MCS platform. The proposed mechanism allows the users to adaptively switch their sensory allocation that are made available to the MCS platform based on historical tracking of the changes in their utility. Through simulations, we show that adaptive management of sensory selection in the auction-based MCS campaign can result in up to >3% increase in user comfort and up to >2% improvement in platform utility when compared to the fixed configuration of sensory arrays based on constant comfort levels used in user recruitment.