TRA-PS: Accountable data Pub/Sub service with fast and fine-grained controllable subscription

Abstract

The Cloud-based data publish/subscribe (Pub/Sub) service presents a selective manner for publishers and subscribers to share and receive data, where confidentiality and authorized access to the data are the fundamental security properties. Therefore, the methodology of attribute-based keyword search (ABKS) was applied that recently considered bilateral access control between the publisher and subscriber. However, existing data Pub/Sub solutions fail to take into account the following features: (i) the efficient revocation of subscribers; (ii) the traceability and auditability of subscriptions; (iii) the excessive time cost of decryption procedure. Therefore, we propose TRA-PS, an accountable data Pub/Sub service that implements efficient revocation, traceability, and public auditability for subscriptions. Technically, we refine the secret key of subscriber in the classic ABKS schemes with associating its identity information and binding accountability module. In addition, we implement TRA-PS and evaluate its practical performance compared to state-of-the-art work in the real cloud environment. Besides achieving accountable property, TRA-PS significantly diminishes the decryption time cost (number of attributes/keywords is 25) by 91.8%, independent of the number of attributes/keywords.