Paper 2014/966

Privacy-Preserving Data Publish-Subscribe Service on Cloud-based Platforms

Kan Yang, Xiaohua Jia, Kuan Zhang, and Xuemin (Sherman) Shen

Abstract

Data publish-subscribe service is an effective approach to share and filter data. Due to the huge volume and velocity of data generated daily, cloud systems are inevitably becoming the platform for data publication and subscription. However, the privacy becomes a challenging issue as the cloud server cannot be fully trusted by both data publishers and data subscribers. In this paper, we propose a privacy-preserving data publish-subscribe service for cloud-based platforms. Specifically, we first formulate the problem of privacy-preserving data publish-subscribe service by refining its security requirements on cloud-based platforms. Then, we propose a bi-policy attribute-based encryption (BP-ABE) scheme as the underlying technique that enables the encryptor to define access policies and the decryptor to define filtering policies. Based on BP-ABE, we also propose a \underline{P}rivacy-preserving \underline{D}ata \underline{P}ublish-\underline{S}ubscribe (PDPS) scheme on cloud-based platforms, which enables the cloud server to evaluate both subscription policy and access policy in a privacy-preserving way. The security analysis and performance evaluation show that the PDPS scheme is secure in standard model and efficient in practice.