Private Database Queries Using MPC and Updatable Set Intersection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cryptographic technologies in database management systems are outdated and focused on securing data in transit and at rest, failing to address modern data management needs, particularly in cloud environments where data is stored and managed by third parties, and do not efficiently support collaborative data access without compromising privacy.
Innovation Solution
Integrate modern cryptographic protocols, such as multi-party computation (MPC) directly into database systems, specifically through new operators in query languages like MQL, enabling secure data computations between parties while maintaining data privacy, and implement updatable private set intersection (UPSI) protocols for dynamic database environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If modern cryptographic protocols (MPC, PSI) are integrated into database systems, then data privacy and security are improved, but system complexity increases
Solution Approach 1:
The patent introduces a dedicated cryptographic operations module that acts as an intermediary between the database engine and the cryptographic protocols. This module handles the complex MPC and PSI operations, isolating the complexity from the core database system while maintaining data privacy guarantees. The module translates high-level privacy-preserving queries into low-level cryptographic operations.
Solution Approach 2:
The database system is segmented into distinct functional components: the core database engine, the cryptographic operations module, and the query processing layer. This segmentation allows each component to be optimized independently, with the cryptographic module handling privacy-preserving operations while the core engine maintains standard database functionality, thus managing system complexity through modular architecture.
2Adaptability or versatility
If updatable private set intersection protocols are implemented, then support for dynamic database environments is improved, but computational overhead increases
Solution Approach 1:
The system performs preliminary setup operations during database initialization, pre-computing cryptographic parameters and structures that will be needed for future UPSI operations. This includes pre-establishing encryption keys, hash functions, and other cryptographic infrastructure, thereby reducing the computational overhead during actual dynamic updates and intersections.
Solution Approach 2:
The cryptographic operations are structured to be performed periodically or in batches rather than continuously for every data change. The system efficiently handles updates by batching cryptographic operations and performing them at optimal intervals, reducing overall computational overhead while maintaining support for dynamic environments.
3Adaptability or versatility
If secure multi-party computation operations are added to query languages, then collaborative data access capability is improved, but query processing complexity increases
Solution Approach 1:
The query language is extended with universal cryptographic operators that can handle multiple types of secure operations (MPC, PSI, encrypted searches) through a unified syntax. This allows the same query language framework to support both traditional SQL operations and advanced privacy-preserving operations, improving collaborative data access capability while managing query processing complexity through a consistent interface.
Data Source
AI summary
Systems and methods are provided for implementing updatable private set intersection in distributed database architectures (e.g., the well-known MongoDB). The systems and methods can include a new specification language referred to for convenience as “MPPL,” for specifying general multi-party computation protocols. This architecture incorporates private set intersection, accessible via command line operators or other operators formatted according to a native query language. Other embodiments detail systems and methods for integrating multi-party database operations via new server node(s) added to a distributed database system (e.g., MongoDB cluster) that manage communication between parties holding private data.


