Sparse Mobile Crowdsensing: Components and Frameworks

Mobile crowdsensing (MCS) allows crowdsourcing of data sensed through mobile phones. It provides numerous applications, varying from detecting potholes to monitoring pollution levels and temperature in a given area. However, a major challenge that MCS faces is the high cost of sensing which includes incentive costs for participants and energy costs for computation. Thus, a new paradigm of sparse mobile crowdsensing has been introduced where only a small number of values are sensed, and the rest of the values are inferred using an inference algorithm. Recent research indicates that the accuracy of the inferred values depends on the inference algorithm as well as the subset of cells selected by the MCS platform for sensing. This paper discusses how the shortcomings of traditional MCS led to the discovery of a new research area, sparse MCS. Further, the paper describes the components of an MCS framework and provides a comprehensive review of various state-of-the-art frameworks, together with potential future research areas.