SCPIR-V: An Optimized CPIR-V Algorithm for Privacy Protection Nearest Neighbor Query

摘要

With the Privacy issues drawing more and more concerns, privacy protection techniques based on Computational Private Information Retrieval (CPIR) allow a user to retrieve data from a service provider without revealing the users query information. For large-scale applications, there exists a gap between privacy protection techniques and its feasibility. In this paper, we propose an optimized CPIR-V based algorithm SCPIR-V for privacy protection nearest neighbor query which reduces the computational cost and communication cost efficiently. Inclusion relation among the candidate data sets of nearest neighbor points are utilized to compress the matrix where data are stored, and new data structures and query algorithms are designed. Compared to the existing work, the computation cost is reduced by 2-5 times and the communication cost by nearly 2 times.