## Cryptology ePrint Archive: Report 2017/318

Key-Aggregate Searchable Encryption with Constant-Size Trapdoors for Fine-Grained Access Control in the Cloud

Abstract: Fine-grained access control, especially in shared data environments such as the cloud, is a common scenario. Suppose a data owner encrypts and stores a collection of $N$ documents in the cloud, and wishes to grant access to a subset $\mathcal{S}$ of these to a data user. The data user must therefore be able to perform search operations only on the subset $\mathcal{S}$ of documents, and nothing else. For example, the user may want to store the documents in separate folders pertaining to work, family, personal etc., and selectively grant search access to one or more folders among different users. This is a commonly encountered in document sharing environments such as Dropbox and Google Drive. Most existing searchable encryption schemes today assume that a user has search permissions over the entire document collection, and are hence not designed explicitly to address such a controlled-search scenario. The only previous work in this direction by Cui et al. possesses inherent security weaknesses that can be exploited to trivially break the privacy of their system. \noindent In this paper, we provide a novel, efficient and secure solution to this problem by introducing a new public-key searchable encryption primitive referred to as the key-aggregate searchable encryption~(KASE). KASE is secure under well-defined cryptographic assumptions, and requires optimal network communication between the server and the data owners/data users. KASE is the first public key searchable encryption scheme to achieve performance and security at par with the most efficient symmetric key searchable encryption constructions. Additionally, while all existing schemes produce trapdoors that grow linearly with the size of the document collection, KASE produces constant-size trapdoors that are independent of the document collection size. KASE is also efficient and scalable with respect to data updates, making it ideal for use in dynamic cloud-based search applications.

Category / Keywords: cryptographic protocols / Searchable Encryption, Access Control, Data Privacy, Trapdoor Privacy

Date: received 10 Apr 2017, withdrawn 19 Apr 2017

Contact author: sikhar patranabis at iitkgp ac in

Available format(s): (-- withdrawn --)

Short URL: ia.cr/2017/318

[ Cryptology ePrint archive ]