Spatiotemporal Location Privacy Preservation in 5G-Enabled Sparse Mobile Crowdsensing

 With the increasing popularity of 5G communications, smart cities have become one of the inevitable trends in the development of modern cities, and smart city services are the foundation of 5G smart cities. Sparse mobile crowdsensing (SparseMCS), as a new and informative urban service model, has attracted the attention of many researchers. Generally, the data required for a sensing task often has a high spatial and temporal correlation, which means that the data uploaded by users need to carry their location information, which may cause serious location privacy issues. The existing location privacy protection mechanism usually only pays attention to the location information of the user’s travel and ignores that people’s daily travel often has a fixed pattern. The attacker can use long-term observation and prior knowledge to infer the victim’s travel mode and analyze its location information. To achieve efficient, robust, and private data sensing, we built a SparseMCS framework with the following three elements: (1) We train the data adjustment model offline on the server-side and solve the position mapping matrix; (2) Design a noise-sensitive data reasoning algorithm improves the accuracy of data; (3) Combining differences and spatiotemporal location privacy to protect the user’s location information and travel mode. Experiments based on real datasets prove that our 5G-supported sparse mobile crowdsensing framework provides more comprehensive and effective location privacy protection.