Paper 2023/720
MUSES: Efficient Multi-User Searchable Encrypted Database
Abstract
Searchable encrypted systems enable privacy-preserving keyword search on encrypted data. Symmetric Searchable Encryption (SSE) achieves high security (e.g., forward privacy) and efficiency (i.e., sublinear search), but it only supports single-user. Public Key Searchable Encryption (PEKS) supports multi-user settings, however, it suffers from inherent security limitations such as being vulnerable to keyword-guessing attacks and the lack of forward privacy. Recent work has combined SSE and PEKS to achieve the best of both worlds: support multi-user settings, provide forward privacy while having sublinear complexity. However, despite their elegant design, the existing hybrid scheme inherits some of the security limitations of the underlying paradigms (e.g., patterns leakage, keyword-guessing) and might not be suitable for certain applications due to costly public-key operations (e.g., bilinear pairing). In this paper, we propose MUSES, a new multi-user encrypted search scheme that addresses the limitations in the existing hybrid design, while offering user efficiency. Specifically, MUSES permits multi-user functionalities (reader/writer separation, permission revocation), prevents keyword-guessing attacks, protects search/result patterns, achieves forward/backward privacy, and features minimal user overhead. In MUSES, we demonstrate a unique incorporation of various state-of-the-art distributed cryptographic protocols including Distributed Point Function, Distributed PRF, and Secret-Shared Shuffle. We also introduce a new oblivious shuffle protocol for the general 𝐿-party setting with dishonest majority, which can be of independent interest. Our experimental results indicated that the keyword search in our scheme is two orders of magnitude faster with 13× lower user bandwidth overhead than the state-of-the-art.
Metadata
- Available format(s)
- Publication info
- Preprint.
- Keywords
- privacy-enhancing technologiesencrypted searchdata privacy
- Contact author(s)
-
tungle @ vt edu
behnia @ usf edu
jorge guajardomerchan @ us bosch com
thanghoang @ vt edu - History
- 2023-05-22: approved
- 2023-05-18: received
- See all versions
- Short URL
- https://ia.cr/2023/720
- License
-
CC BY-NC
BibTeX
@misc{cryptoeprint:2023/720, author = {Tung Le and Rouzbeh Behnia and Jorge Guajardo and Thang Hoang}, title = {MUSES: Efficient Multi-User Searchable Encrypted Database}, howpublished = {Cryptology ePrint Archive, Paper 2023/720}, year = {2023}, note = {\url{https://eprint.iacr.org/2023/720}}, url = {https://eprint.iacr.org/2023/720} }