Homomorphic Encryption Noise Management via Intermediate Decryption

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current homomorphic encryption methods face challenges in maintaining confidentiality and preventing noise amplification during high calculation depth operations, leading to potential decryption failures and security vulnerabilities when executing programs on encrypted data.

Innovation Solution

A method that involves homomorphic evaluation of program instructions on encrypted data, where the result is masked with a random sequence, decrypted, and then reencrypted, using a modulo 2 summation operation in the ciphertext space to maintain confidentiality and control noise propagation, potentially utilizing a coprocessor for these operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If homomorphic operations are performed on encrypted data with high calculation depth, then computational functionality is improved, but noise level increases causing decryption failures

Engineering Contradiction:

Improvecomputational functionalityVSAvoiddecryption success

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The program execution is divided into multiple stages, with decryption operations inserted at intermediate points to reset noise levels. This segmentation allows high calculation depth to be achieved while preventing noise accumulation from causing decryption failures.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Decryption operations are performed periodically at strategic points during program execution to refresh the encrypted data and reduce noise levels, enabling sustained high calculation depth without decryption failures.

Inventive Principle:

Principle #19Periodic action

2Reliability

If decryption operations are performed frequently to reduce noise, then decryption reliability is improved, but computational efficiency deteriorates

Engineering Contradiction:

Improvedecryption successVSAvoidcomputational efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system pre-calculates and stores intermediate decryption points in the program analysis, identifying optimal locations where decryption should occur to balance noise reduction with minimal disruption to computational efficiency.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system monitors noise levels during execution and dynamically determines when decryption is necessary, using feedback from noise threshold checks to trigger decryption operations only when needed, thereby optimizing the balance between reliability and efficiency.

Inventive Principle:

Principle #23Feedback

3Speed

If unencrypted data is exposed during processing, then processing speed is improved, but security is compromised

Engineering Contradiction:

Improveprocessing speedVSAvoidsecurity vulnerability

Core Design Contradiction:

SpeedVSObject-affected harmful factors

Solution Approach 1:

The system uses encrypted intermediate representations and homomorphic encryption as intermediaries to enable processing without exposing unencrypted data. This allows security to be maintained while still performing necessary computational operations.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system replaces traditional decryption-encryption mechanics with homomorphic encryption operations that allow computation directly on encrypted data, eliminating the need to expose unencrypted data during processing while maintaining security.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS10439798B2Method for confidential execution of a program operating on data encrypted by a homomorphic encryption

Publication Date: 2019.10.08 COMMISSARIAT A LENERGIE ATOMIQUE ET AUX ENERGIES ALTERNATIVES

AI summary

A method of executing a program operating on data encrypted by a homomorphic encryption. Execution of a program instruction includes the homomorphic evaluation of an associated function in the ciphertext space, homomorphic masking of the result of the evaluation with a previously encrypted random sequence, decryption of the evaluation result thus masked followed by a new encryption and then homomorphic unmasking in the ciphertext space. The result of execution of the instruction does not appear in plain text at any time during execution of the instruction.