Bidirectional location privacy protection scheme for epidemiological investigation based on OT

Abstract

Epidemiological investigation is a crucial method to deal with epidemics (e.g., COVID-19), and the location data in the epidemiological investigation is widely used to determine close contacts. Through the analysis of the location data of the respondents and the diagnosis point data, it can effectively determine the close contacts and then promote the prevention and control of the epidemic. Due to the sensitivity of location data, the direct release and epidemiological investigation of location data will cause social panic and seriously threaten to personal privacy. Therefore, two aspects need to be considered in epidemiological investigation: the location privacy of responder and the location privacy of the diagnosis point. Focusing on the above two privacy requirements, we propose a bidirectional location privacy protection scheme based on oblivious transfer. Firstly, we introduce the Hilbert curve into Centers for Disease Control and Prevention (CDCP) to encode the position of the diagnosis point. Then the Bloom filter is used to perturb the encode of diagnosis point to ensure the location privacy of CDCP. Secondly, 1-out-of-n oblivious transfer protocol is carried on between the CDCP and the responder to achieve screening of close contacts. At the same time, the bidirectional privacy protection of the location between CDCP and respondents has been ensured. Security analysis and performance analysis prove that the overall scheme is secure, feasible, and efficient.