EPDQ: Efficient and Privacy-Preserving Exact Distance Query on Encrypted Graphs
Xuemei Fu
Abstract
With the explosive growth of graph-structured data, graph databases have become a critical infrastructure for supporting large-scale and complex data analysis. Among various graph operations, shortest distance queries play a fundamental role in numerous applications, such as path planning, recommendation systems, and knowledge graphs. However, existing encrypted graph query methods still suffer from limitations in computational efficiency and system scalability, making it challenging to support efficient query processing over large-scale encrypted graph data. To address these challenges, this paper proposes a tensor-based shortest distance query scheme for encrypted graph databases. The proposed method integrates an encrypted 2-hop cover indexing framework with the Pruned Landmark Labeling (PLL) technique, thereby constructing an efficient and privacy-preserving indexing mechanism. Furthermore, a tensorized representation is introduced to uniformly model graph structures, which effectively reduces computational complexity while ensuring data privacy, and significantly improves the scalability of the system. Extensive experimental evaluations on large-scale graph datasets demonstrate that the proposed approach achieves superior scalability and lower computational costs compared with existing encrypted graph query methods. Moreover, it provides strong privacy protection guarantees, making it well suited for privacy-preserving graph query applications in cloud computing and distributed environments.