Beyond Encryption: A Holistic Approach to Privacy-Preserving Query Processing in Modern Database Systems

Beyond Encryption: A Holistic Approach to Privacy-Preserving Query Processing in Modern Database Systems

Nishant Kumar Rathi,

Associate Professor,

Department of Computer Applications,

Shri Ram College, Muzaffarnagar, Uttar Pradesh

Neetu Singh,

Assistant Professor,

Department of Computer Applications,

Shri Ram College, Muzaffarnagar, Uttar Pradesh

Abstract

The exponential growth of data-centric applications in cloud and distributed environments has intensified the demand for robust privacy-preserving mechanisms in database systems. While encryption techniques—such as homomorphic encryption and secure multiparty computation—offer foundational security, they often incur significant computational overhead and fail to address broader privacy concerns such as inference attacks, access pattern leakage, and insider threats. This study presents a comprehensive, layered framework for privacy-preserving query processing that extends beyond traditional encryption paradigms. Integrating fine-grained access control, differential privacy, secure hardware enclaves, and privacy-aware query rewriting, the proposed architecture balances query expressiveness, performance, and privacy guarantees. A prototype implementation on PostgreSQL was evaluated using standard workloads (TPC-H and synthetic sensitive datasets) to assess system latency, accuracy trade-offs, and privacy leakage. Results indicate a substantial reduction in leakage exposure with minimal performance degradation, demonstrating the framework’s practicality for real-world deployment. This research contributes to the evolving discourse on database privacy by advocating a shift from encryption-centric approaches to holistic privacy engineering, paving the way for secure, trustworthy, and regulation-compliant data systems.

Keywords: Privacy-preserving query processing, differential privacy, encrypted databases, trusted execution environments, data confidentiality, query optimization, homomorphic encryption, secure multi-party computation, access control, privacy-aware data management.