Random Partition Region for Location Privacy Protection on Edge Computing

Edge computing, as a new computing paradigm, appears to solve the record generated by a great deal of edge devices that cannot be processed by traditional cloud computing. The real-time location big data of LBS (Location-Based Services) are transferred from the cloud server to the edge nodes, which further improves the service capacity and promotes the further development of LBS. However, based on the existing partition method, dynamic location data privacy cannot be well protected. In this paper, based on the differential privacy theory, we propose RPS (Random Partition and Selection) algorithm which can protect user’s location information. We achieve this by designing a dynamic partition mechanism and noise adding algorithm based on the number of users. We repartition the region in each snapshot and give a new noise probability. Experiment results show that our proposed method has the significant progress in the loss of service quality than the existing partition release, such as UG (Uniform Grid) and AG (Adaptive Grid) mechanism.