Enabling privacy-preserving non-interactive computation for Hamming distance

Abstract

Hamming distance is a measure of the similarity between two strings of the same length. Privacy-preserving Hamming distance computation allows data users to obtain the Hamming distance between their data without disclosing their respective private data. The existing privacy-preserving protocols for Hamming distance computation require multiple rounds of online interactive computation between two data users. To address this issue, we propose a new Privacy-preserving non-interactive Hamming Distance Computation (PHDC) protocol. Different from previous works, we adopt the strategy of secure outsourcing to avoid the online interactive computation between data users. We move the Hamming distance computation from the user side to the cloud side. The cloud server is responsible for the Hamming distance computation under privacy protection. To preserve data privacy, we propose a novel blinding technique for user data. Data users initially blind their data using homomorphic encryption and randomization techniques. The edge server is responsible for data aggregation and further data blindness. In this way, data users only need to outsource their encrypted data to the edge server, and there is no online interactive computation between data users. With the assistance of the edge server and the cloud server, the privacy-preserving Hamming distance computation is achieved. The security analysis demonstrates that the protocol guarantees the data privacy under the semi-honest adversarial model. The theoretical analysis and experimental results illustrate the efficiency of the proposed protocol.