Perturbation for Privacy-Preserving Participatory Sensing on Mobile

Participatory sensing applications collect data from participants to construct statistical information of environment or phenomenon, using their mobile phone. Mobile phone is closely related to participant's daily life, therefore the invasion of privacy in participatory sensing would have dire consequences. In this research, we study privacy-preserving participatory sensing technique which is the perturbation using negative surveys and limited negative surveys on mobile to promote use of participatory sensing in healthcare, investigation, and other useful applications. When participants report the data in negative surveys, their mobile phones automatically select a value from the set complement of the sensed data value at random. In other words, we can construct public statistics without knowing the personal information of citizens. Additionally, our research extends negative surveys to limited negative surveys, which have the capable of change, according to the feature of data, especially the number of categories. We combine negative survey and limited negative survey, because it is difficult to construct valuable databases when the categories of perturbed is large size on mobile phone. We also present the evaluation of these schemes on the view point of privacy and utility of data sets in central collection server.