Searching Untrusted Clouds Meets Multiple Keys: Privacy-Preserving Spatio-Textual Top-k Query

Ningning Cui, Dong Wang, Yuliang Ma, Mengxiang Wang, Kang Qian, Jianxin Li

Abstract: Cloud platform has emerged as a promising choice for outsourcing query processing services. However, since the cloud server is untrusted, the outsourced data security has become the thorny concerns. To tackle this issue, a prevailing approach of almost existing studies resorts to encrypt database prior to outsourcing but shares identical public/private key pair among query users, which unfortunately raises up neglected security issues in practice, e.g., leaking the private key or query privacy. Therefore, in such a scenario, how to guarantee strong security/privacy (e.g., multiple keys) and the rich query functionality (e.g., both space and text) simultaneously is a quite challenging problem. In this paper, we study Privacy-Preserving Spatio-Textual Top-\(\underline{k}\) Query under Multiple Keys, called PT k MK. First, we propose a novel secure index SBlock\(^{inv}\) and secure protocols by adopting Distributed Two Trapdoors Public-Key Cryptosystem to provide multiple keys setting and data privacy guarantee. Further, to hide access patterns privacy, we propose an enhanced secure index EnSBlock\(^{inv}\) and corresponding secure protocols by leveraging bucket partition scheme. At last, security analysis and complexity analysis are theoretically demonstrated and the performance of our proposed approaches are empirically evaluated.