Auditable Data Sharing Model for Multi-Domain Privacy Information Retrieval
Boyi Wang, Peng Liu, Chunpei Li, Hao Huo, Yuhan Chen · 2024
In recent years, heightened concerns regarding personal privacy and trade secrets underscore the need to safeguard retrieved statements in data interactions, often containing easily compromised sensitive information. While Privacy Information Retrieval (PIR) techniques aim to address this, current applications primarily focus on protecting indexed resource locations within a single administrative domain. In multi-domain PIR, entities demand higher resource availability and trustworthiness, necessitating a balance between access record audit and retrieval keyword confidentiality. This paper addresses privacy information retrieval and resource authorization in multi-management domain scenarios, proposing a novel auditable blockchain-driven multi-domain privacy information retrieval model, BDMD-PIR. Leveraging a cloud-chain storage collaboration approach and blockchain ledgers, BDMD-PIR facilitates asset sharing across multiple data domains. Additionally, mainstream PIR schemes are adapted and optimized to alleviate communication overhead and computational burdens of resource-constrained entities. Rigorous security analysis and experimental simulations demonstrate BDMD-PIR’s resilience to various attacks and offer viable solutions to inherent security challenges in PIR and multi-data domain sharing.