Design of a Privacy-Preserving Algorithm for Peer-to-Peer Network Based on Differential Privacy

Abstract

In the peer-to-peer network (P2P), the private information of individual users faces the risks of being tracked, identified or leaked. This paper attempts to develop a security technique that fully protects the privacy of P2P users. Firstly, the node data were collected and analyzed from the P2P. Then, a privacy-preserving algorithm was proposed for the P2P based on the differential privacy model. The proposed algorithm adds noise to the degree distribution of the nodes, and solves the high sensitivity under the Laplace mechanism, which arises from the unique structure of the P2P. Finally, the proposed algorithm was verified through experiments. The results show that our algorithm protected the privacy of individual data in the storage and dissemination process, controlled the sensitivity to a low level through noise addition, and allocated the privacy budget rationally. Thus, the proposed algorithm is available and reliable for P2P operations like credit verification, transmission and storage, and enjoys excellent effectiveness and robustness.