Secure multiparty aggregation with differential privacy
Slawomir Goryczka, Li Ping Xiong, Vaidy S Sunderam · 2013
This paper considers the problem of secure data aggregation in a distributed setting while preserving differential privacy for the aggregated data. In particular, we focus on the secure sum aggregation. Security is guaranteed by secure multiparty computation protocols using well known security schemes: Shamir's secret sharing, perturbation-based, and various encryption schemes. Differential privacy of the final result is achieved by distributed Laplace perturbation mechanism (DLPA). Partial random noise is generated by all participants, which draw random variables from Gamma or Gaussian distributions, such that the aggregated noise follows Laplace distribution to satisfy differential privacy. We also introduce a new efficient distributed noise generation scheme with partial noise drawn from Laplace distributions.