BDEA: blockchain-based secure deduplication for low-entropy data supporting public auditing
Longxia Huang, Mao Xue, Chengzhi Ge · The Computer Journal · 2025
Abstract In cloud storage, deduplication techniques are essential for improving storage capacity. Low-entropy data, due to its predictable structure, pose security risks and challenges for convergent encryption, which can compromise semantic security and increase the likelihood of information leakage. To address these issues, robust auditing mechanisms tailored for low-entropy data are essential. Although double-layer encryption strategies have been suggested to enhance security, they often introduce significant overhead and do not ensure ownership verification or auditing for low-entropy data. To address these limitations, we propose BDEA, a blockchain-based secure deduplication scheme specifically designed for low-entropy data that improves security while enabling public auditing capabilities. BDEA employs advanced technologies such as random keys, Bloom filters, and chameleon hash functions to facilitate secure and efficient deduplication processes tailored for low-entropy datasets. The integration of blockchain technology with Rivest-Shamir-Adleman (RSA) signatures enhances the security and transparency of data integrity audits in cloud environments. Security analysis confirms the robustness of the BDEA scheme, while experimental results demonstrate its effectiveness. In Ubuntu 20.04, this scheme achieves approximately an 80% reduction in deduplication processes and a 70% decrease in audit verification time when applied to datasets ranging from 2 to 16 MB at intervals of 2 MB.