Secure Spatial Query Scheme with Access Pattern Obfuscation

Abstract

The existing secure spatial query methods allow authorized users to generate spatial query tokens that are executed in ciphertext mode by the service provider. However, since only the correct query results that match the query tokens are returned, the service provider can observe clearly which encrypted data is accessed and returned to the client when a query is launched. It results in the leakage of data access pattern. Existing studies have shown that the attacker can use the leaked access patterns to recover query information with some prior knowledge. In this paper, we propose a secure spatial range query scheme with access pattern obfuscation to protect the location data from attackers. This scheme not only prevents access pattern leakage but also ensures the feasibility of spatial range query on encrypted data, which makes up for the deficiencies of existing privacy protection methods. Extensive experiments are conducted on real-world datasets, and experimental results demonstrate that performing range query on the obfuscated dataset can still obtain acceptable query accuracy. Besides, we find both privacy budget and query area are proportional to query accuracy. Moreover, we prove our scheme is more suitable for circular range queries than rectangular ones. In addition, the choosing of privacy budget is studied to enable datasets with different sizes to achieve similar obfuscation performance.