Secure Exponent Unification for Efficient Floating-Point MPC
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for uniforming the exponent parts of floating points in secure computation are inefficient while maintaining secrecy.
Innovation Solution
A system and method for uniformly exponent parts of floating points using a secret exponent part uniforming system comprising three or more apparatuses, which includes a maximum value calculation, difference calculation, mantissa part calculation, and output means to efficiently uniform exponent parts while maintaining confidentiality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional secure computation methods are used to uniform exponent parts, then confidentiality is maintained, but computation efficiency deteriorates
Solution Approach 1:
The patent divides the exponent uniforming process into distinct functional modules: maximum value calculation unit, difference calculation unit, and mantissa part calculation unit. Each unit handles a specific aspect of the exponent uniforming operation, allowing parallel processing and reducing overall computation time while maintaining secure multi-party computation protocols for confidentiality
Solution Approach 2:
The patent performs preliminary calculations of maximum exponent values and difference values before the actual exponent uniforming operation. By pre-computing these values in a secure manner, the system avoids repeated complex calculations during the main processing phase, significantly improving computation efficiency while preserving confidentiality through secure multi-party computation
2Loss of information
If secure multi-party computation protocols are implemented for exponent uniforming, then secrecy is preserved, but communication overhead increases
Solution Approach 1:
The patent combines multiple secure computation operations into unified calculation units. The maximum value calculation unit and difference calculation unit work together to simultaneously determine exponent uniforming parameters, reducing the number of separate communication rounds required in traditional secure multi-party computation protocols
Solution Approach 2:
Each participating party in the secure computation independently computes local values (maximum exponents, differences) based on their own secret-shared data. This self-service approach minimizes inter-party communication requirements, as parties only need to exchange minimal aggregated information rather than continuously sharing intermediate computation results
Data Source
Figure 1~2
Figure 3
Figure 4~5
AI summary
Provided is a secure computation technique for efficiently uniforming exponent parts of floating points. A secret exponent part uniforming system which, from a share ([[→a]]P, [[→ρ]]Q) of a floating point vector (→a=(a0,..., am-1), →ρ=(p0, ..., ρm-1)), calculates a share ([[→b]]P, [[→ρmax]]Q) of a floating point vector with uniformed exponent parts (→b=(b0,..., bm-1), →ρmax=(ρmax, ..., ρmax) (ρmax=max{ρ0, ..., ρm-1}), 2ρ_iai≒2ρ_maxbi is satisfied), comprises a mantissa part calculation means for calculating a share [[→b]]P by calculating a share [[bi]]P (bi=2-ρ_dif,iai) of the number bi from the i-th element of the share [[→a]]P and the i-th element of a share <<→ρdif>>Q converted by replicated secret sharing from a share [[→ρdif]]Q=[[→ρ]]Q-[[→ρmax]]Q.