Inherit Differential Privacy in Distributed Setting: Multiparty Randomized Function Computation
Genqiang Wu, Yeping He, Jingzheng Wu, Xianyao Xia · 2016
How to achieve differential privacy in the distributed setting, where the dataset is distributed among the istrustful parties, is an important problem. We consider in what condition can a protocol inherit the differential privacy property of a function it computes. The heart of the problem is the secure multiparty computation of randomized function. A notion obliviousness is introduced, which captures the key security problems when computing a randomized function from a deterministic one in the distributed setting. By this observation, a sufficient and necessary condition about securely computing a randomized function from a deterministic one is given. The above result can not only be used to determine whether a protocol computing differentially private function is secure, but also be used to construct a secure one. Then we prove that the differential privacy property of a function can be inherited by the protocol computing it if the protocol securely computes it. A composition theorem of differentially private protocols is also presented. Finally, we construct protocols of Gaussian mechanism and Laplace mechanism, which inherit the differential privacy property.