Secure Multi-Party Computation Inner Product via Blinded Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Secure Multi Party Computation (SMPC) protocols face significant challenges in communication complexity, particularly when evaluating nontrivial functions involving large numbers of inputs and operations, which hinders their applicability in real-world scenarios.
Innovation Solution
The proposed method involves a computer-implemented method for performing multi-party computations using a network of data processors, where first and second dealer nodes compute transformed vectors and blinded vectors using discrete linear transforms, allowing computing nodes to calculate result shares without accessing private inputs, and subsequently reconstructing the inner product of the private input vectors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional SMPC protocols are used to compute inner products, then security is maintained, but communication complexity increases significantly
Solution Approach 1:
The patent applies preliminary action by pre-computing and distributing blinded vectors to computing nodes before the actual inner product computation. The dealer nodes generate blinded vectors by multiplying random vectors with transformation matrices and distribute these pre-processed vectors to computing nodes, eliminating the need for complex communication during the computation phase.
Solution Approach 2:
The patent introduces blinded vectors as an intermediary mechanism between the dealer nodes and computing nodes. These blinded vectors serve as mediators that carry encrypted information without revealing the actual private inputs, allowing computing nodes to perform computations on obscured data while maintaining security and reducing communication requirements.
2Productivity
If more computing nodes are added to increase computational capacity, then productivity improves, but device complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the computation into independent tasks distributed across multiple computing nodes. Each computing node receives a portion of the blinded vectors and independently computes its contribution to the inner product, then combines results. This segmentation allows linear scaling of computational capacity without proportionally increasing overall system complexity.
Solution Approach 2:
The patent implements universality by designing a standardized protocol where computing nodes perform identical operations on identical input formats. All computing nodes use the same algorithm to process blinded vectors and contribute to the final result, enabling easy addition of more nodes without requiring custom configurations or increasing operational complexity.
Data Source
AI summary
A secure multiparty computation method permits the computation of an inner product of a pair of secret vectors. The vectors are transformed and blinded using various blinding factors with the transforms of the vectors being according to a discrete linear transform for which Parseval's theorem holds. Shares of the transformed, blinded vectors are distributed to computing nodes which each calculate shares of a result without access to the secrets, and the result shares can be combined to generate the inner product of the original vectors.