Homomorphic Encrypted Message Processing With Exponent-Mantissa Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing homomorphic encrypted messages use a fixed-point method, which makes it difficult to manage large or small data simultaneously due to limited precision, especially when data ranges are wide.
Innovation Solution
The method involves checking the exponential and mantissa data for numerical data, performing homomorphic encryption on the mantissa data, and merging the encrypted mantissa and exponential data to generate a homomorphic encrypted message in a floating-point type.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed-point method is used for homomorphic encrypted messages, then encryption simplicity is maintained, but precision consistency across wide data ranges deteriorates
Solution Approach 1:
The patent changes the data representation parameter from fixed-point to floating-point format. This allows the system to maintain encryption simplicity while achieving precision consistency across wide data ranges by using exponent and mantissa fields that can represent both large and small values with uniform precision characteristics.
Solution Approach 2:
The patent segments the numerical data into two distinct parts: exponent data and mantissa data. This segmentation allows independent handling of the magnitude (exponent) and precision (mantissa) components, enabling the system to maintain consistent precision across different data ranges while keeping the encryption process manageable.
2Device complexity
If fixed-point method is used, then data structure simplicity is maintained, but ability to manage large and small data simultaneously deteriorates
Solution Approach 1:
The patent changes the data structure from fixed-point to floating-point representation, introducing exponent and mantissa fields. This parameter change enables the system to manage both large and small data values simultaneously while maintaining reasonable structural organization through standardized floating-point format.
Solution Approach 2:
The patent adds an exponent dimension to the data structure, transforming the single-dimensional fixed-point format into a two-dimensional floating-point format with separate exponent and mantissa components. This dimensional change enables efficient management of wide data ranges.
3Productivity
If homomorphic encryption is applied to mantissa data only, then encryption efficiency is improved, but complete data security deteriorates
Solution Approach 1:
The patent applies homomorphic encryption selectively to only the mantissa data rather than encrypting the entire floating-point number. This partial action approach improves encryption efficiency by reducing computational overhead while maintaining adequate security by protecting the precision-critical mantissa portion, accepting that the exponent remains unencrypted.
Data Source
AI summary
Disclosed is a method of processing an encrypted message. The method of processing an encrypted message includes checking exponential data and mantissa data for numerical data, performing homomorphic encryption on the mantissa data, and merging the encrypted mantissa data and the exponential data to generate a homomorphic encrypted message.