Privacy-Preserving Range Query for High-Dimensional Uncertain Data in a Two-Party Scenario

Abstract

With the fast evolution of sensor technology, massive high-dimensional data are collected by various service providers. Other institutions also want to use the data for analysis and statistics. However, for the privacy and legal concerns, data owners should not directly share the data with others. In addition, some factors such as measurement limitations, noise, and network delays may result in uncertain data. Compared with processing certain data, managing and processing uncertain data is more challenging. In this paper, we propose a privacy-preserving range query scheme for high-dimensional uncertain data owned by the other party, in which range query problem can be solved by data owners without revealing their data and the query range is invisible except the query requestor. We utilize the Paillier encryption as the basic block of our scheme. The data owner utilizes a binary tree index to promote query, which combines pivot-mapping and Bloom filter. We analyze the security and evaluate the performance of the scheme with a synthetic dataset. The analysis and experimental results show that our scheme is secure and efficient.