Database Query Verification Using Commitments and GPU Proofs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for verifying database query results in outsourced databases are inefficient and add significant computation overhead, limiting scalability, especially in decentralized systems where untrusted nodes can tamper with data.
Innovation Solution
A method using commitments and graphics processing units (GPUs) to verify query results by generating and storing commitments, allowing verification without direct access to the data, and leveraging query plans to generate and verify proofs efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If current verification methods are used to verify database query results, then verification capability is provided, but computation overhead increases significantly and scalability is limited
Solution Approach 1:
The verification process is segmented into distinct phases: commitment generation phase (performed once when data is inserted) and verification phase (performed when queries are executed). This segmentation allows the heavy computational work to be done upfront and stored as commitments, making subsequent verification operations much lighter and more scalable.
Solution Approach 2:
Commitments are generated in advance when data is inserted into the database, before any verification is needed. These pre-computed commitments are stored alongside the data and used later to efficiently verify query results without repeating the heavy computational work.
2Reliability
If verification mechanisms are implemented in decentralized databases, then trust in query results is improved, but system complexity increases
Solution Approach 1:
Cryptographic commitments act as an intermediary between the stored data and the verification process. Instead of requiring complex trust mechanisms between database nodes and clients, the commitments serve as a mathematical mediator that enables verification through simple algebraic checks, significantly reducing system complexity.
Solution Approach 2:
The patent replaces complex mechanical trust mechanisms (requiring multiple trusted nodes, consensus protocols, or manual verification) with a mathematical/cryptographic system based on polynomial commitments. This substitution transforms a complex distributed trust problem into a simpler mathematical verification problem.
3Reliability
If traditional verification approaches are used, then query results can be checked, but computation time and overhead increase
Solution Approach 1:
Verification-related computations (generating commitments to data and intermediate results) are performed in advance during data insertion and query execution, respectively. The results are stored and reused, eliminating the need to re-compute these values during verification, thus reducing computation time significantly.
Solution Approach 2:
Instead of storing and verifying the actual data values, the system stores and verifies cryptographic commitments (hashes) of the data. These commitments are compact copies that preserve the essential verification properties while requiring minimal storage and computation space.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
Methods, non-transitory computer readable media, and query verification apparatuses are disclosed that receive data, store the data into a table of a database, and receive a query that is associated with the database table. A query plan and a query result are generated for the query, one or more partial proofs are generated from one or more commitments, and an overall proof is generated from the one or more partial proofs. Each of the one or more partial proofs is associated with at least one node of the query plan in some examples, which can include a directed acyclic graph. The overall proof is returned along with the query result in response to the query to facilitate verification of the query result. One or more GPUs are configured to generate the one or more commitments in some examples to thereby accelerate the verification process and improve database scalability.