R&D: Encrypted Data Reduction, Removing Redundancy from Encrypted Data in Outsourced Storage

ACM Transactions on Storage has published an article written by Jia Zhao, Zuoru Yang, Department of Computer Science and Engineering, The Chinese University of Hong Kong, Shatin, Hong Kong, Jingwei Li, Center for Cyber Security, University of Electronic Science and Technology of China, Chengdu, China, and Patrick P. C. Lee, Department of Computer Science and Engineering, The Chinese University of Hong Kong, Shatin, Hong Kong.

Abstract: “Storage savings and data confidentiality are two primary goals for outsourced storage. However, encryption by design destroys the content redundancy within plaintext data, so there exist design tensions when combining encryption with data reduction techniques (i.e., deduplication, delta compression, and local compression). We present EDRStore, an outsourced storage system that realizes encrypted data reduction to achieve both storage savings and data confidentiality. EDRStore’s core idea is a careful design of the encryption and data reduction workflows. It proposes new key generation and encryption schemes to preserve the content similarity of encrypted data for deduplication and delta compression. It further proposes selective local compression based on content similarity, so as to achieve storage savings of encrypted data from both delta compression and local compression. Evaluation on real-world datasets shows that EDRStore achieves higher storage savings than existing encrypted storage approaches and incurs moderate performance overhead compared with plaintext storage.“