NBQ-SE: an efficient and fine-grained searchable encryption scheme supporting non-monotonic boolean queries in cloud storage
摘要
The rapid evolution of high-performance computing (HPC) and big data ecosystems has intensified the conflict between massive data outsourcing requirements and privacy protection. While searchable encryption technology can effectively balance this conflict, existing schemes still exhibit significant limitations in retrieval expressiveness, security, and efficiency–particularly in supporting complex Boolean queries involving the NOT operator. To address these challenges, this paper proposes NBQ-SE, a novel searchable encryption scheme supporting non-monotonic Boolean queries. The scheme models Boolean formulas as circuit structures and integrates secret sharing with reconstruction labels, enabling full support for arbitrary Boolean expressions. It implements a public-key authentication mechanism to verify both data uploaders and query requesters during retrieval, effectively preventing keyword guessing attacks. Efficiency analysis shows NBQ-SE requires only a fixed number of pairing operations for monotonic queries while maintaining high performance for non-monotonic retrievals. Moreover, the proposed document-centric index structure inherently supports parallel processing, enabling scalable deployment in distributed supercomputing environments. Security proofs demonstrate the scheme’s resistance to keyword guessing attacks and chosen-plaintext attacks. Theoretical and experimental evaluations confirm its superiority over existing schemes in expressiveness, security, and efficiency.