Publication Type

Conference Proceeding Article

Publication Date



Attribute-based encryption (ABE) enables an access control mechanism over encrypted data by specifying access policies over attributes associated with private keys or ciphertexts, which is a promising solution to protect data privacy in cloud storage services. As an encryption system that involves many data users whose attributes might change over time, it is essential to provide a mechanism to selectively revoke data users’ attributes in an ABE system. However, most of the previous revokable ABE schemes consider how to disable revoked data users to access (newly) encrypted data in the system, and there are few of them that can be used to revoke one or more attributes of a data user while keeping this user active in the system. Due to this observation, in this paper, we focus on designing ABE schemes supporting selective revocation, i.e., a data user’s attributes can be selectively revoked, which we call ABE with granular revocation (ABE-GR). Our idea is to utilize the key separation technique, such that for any data user, key elements corresponding to his/her attributes are generated separately but are linkable to each other. To begin with, we give a basic ABE-GR scheme to accomplish selective revocation using the binary tree data structure. Then, to further improve the efficiency, we present a server-aided ABE-GR scheme, where an untrusted server is introduced to the system to mitigate data users’ workloads during the key update phase. Both of the ABE-GR constructions are formally proved to be secure under our defined security model.


granular revocation, ABE, efficiency, cloud storage


Information Security

Research Areas



Proceedings of the 12th EAI International Conference on Security and Privacy in Communication Networks (SecureComm): Guangzhou, October 10-12, 2016

First Page


Last Page




City or Country

Guangzhou, China

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.