Paper 2019/900

Multi-client Secure Encrypted Search Using Searching Adversarial Networks

Kai Chen, Zhongrui Lin, Jian Wan, Lei Xu, and Chungen Xu.


With the rapid development of cloud computing, searchable encryption for multiple data owners model (multi-owner model) draws much attention as it enables data users to perform searches on encrypted cloud data outsourced by multiple data owners. However, there are still some issues yet to be solved nowadays, such as precise query, fast query, dimension disaster and flexible system dynamic maintenance. To target these issues, this paper proposes a secure and efficient multi-keyword ranked search over encrypted cloud data for multi-owner model based on searching adversarial networks (MRSM\_SAN). Specifically, we exploit searching adversarial networks to achieve optimal pseudo-keyword filling, and obtains the optimal game equilibrium for query precision and privacy protection strength. In order to achieve fast query, maximum likelihood search balanced tree is proposed, which brings the query complexity closer to $O(\log N)$. we reduce data dimension with fast index clustering, and enable low-overhead system maintenance based on balanced index forest. In addition, attribute based encryption is used to achieve more secure and convenient key management as well as authorized access control. Compared with previous work, our solution maintains query precision above 95\% while ensuring adequate privacy protection, significantly improving search efficiency, enabling more flexible system dynamic maintenance, and reducing the overhead on computation and storage.

Available format(s)
-- withdrawn --
Publication info
Published elsewhere. Minor revision. The 18th International Conference on Cryptology and Network Security (CANS 2019)
Searchable EncryptionMulti-keyword Ranked SearchMulti-owner ModelSearching Adversarial NetworksMaximum Likelihood Search Balanced TreeBalanced Index Forest
Contact author(s)
kaichen @ njust edu cn
2019-08-22: withdrawn
2019-08-08: received
See all versions
Short URL
Creative Commons Attribution
Note: In order to protect the privacy of readers, does not use cookies or embedded third party content.