Privacy assurance in mobile sensing networks: Go beyond trusted servers

Mobile devices are becoming the largest sensor network around the world. They could be used to collect a large amount of data with little effort and cost which is leading to a promising future for participatory sensing networks or urban sensing. However, privacy concerns of the mobile users are the major inhibitors hindering massive participation. This paper proposes a solution to user privacy preserving problem in a participatory sensing network. Each user is considered as a node in a social network and users are connected through friendship links which are represented as edges on the network. Typically, each user contributes to the participatory system by uploading his/her acquired data to a server. Instead of uploading data to the server directly, we devised a Hot-Potato-Privacy-Protection Algorithm (HP3) in which data is sent to one of the friends of the user and the friend will choose another friend to deliver the data to the next hop. Hopping goes on until some user-defined threshold is reached, then the last user uploads the data to the server. Friend selection is random and the number of hops is also random and independent. HP3 ensures that the probability that the server can make a successful attack on the data owner is no better than where n is the number of mobile users in the system. Therefore, HP3 protects location privacy as well as data ownership privacy of mobile users. We simulate our approach on some large scale social networks and report some findings in the paper. Experiments show that our system achieves privacy protection for each user against the server with tolerable communication overhead.