Cloud-side Collaborative Privacy Protection Based on Differential Privacy

Cloud-side Collaborative Privacy Protection Based on Differential Privacy

With the rapid development of 5G and the Internet of Things, edge computing is playing an increasingly important role in real-world applications. But at the same time there is the risk of leaking user privacy information. Therefore, how to ensure users’ personalized privacy requirements has become one of the hot issues in cloud-side collaborative computing scenarios. This paper focuses on the problems of the mean estimation and histogram estimation algorithms in the differential privacy protection model. The traditional personalized local differential privacy based on data derivation does not consider the influence of coding on the estimation error of histogram, and the data derivation algorithm has high algorithm complexity. This paper solves the above problems in the cloud-side collaboration scenario, and its main work is as follows: Established a distributed and personalized local differential privacy protection model for the privacy protection scenario of cloud-side collaboration. Under the premise of meeting the personalized privacy requirements of data at different edge nodes, the use of optimized unary coding reduces the mean square error of histogram estimation. Proposed an optimized personalized privacy data derivation algorithm based on optimized unary encoding. And confirmed the algorithm complexity of the data derivation algorithm is greatly reduced.