Homomorphic Factorization Encryption for Secure Cloud Data Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Businesses using cloud computing face security vulnerabilities as they lack control over their data, making them susceptible to attacks and legal issues when using cloud services for data analysis, as existing homomorphic encryption methods require decryption by the cloud provider and have limitations in scalability and communication costs.

Innovation Solution

A novel encryption scheme using a private encryption key, a public computation key, and a decryption key allows arithmetic operations on encrypted data without decrypting it, enabling machine learning algorithms to run on encrypted data, ensuring data security and reducing communication costs by allowing efficient arithmetic operations on large datasets.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Power

If cloud computing is used for data analysis, then computational power and availability are improved, but security control and vulnerability to attacks worsen

Engineering Contradiction:
Improvecomputational powerVSAvoidsecurity control
Core Design Contradiction:
PowerVSReliability

Solution Approach 1:

Homomorphic encryption acts as an intermediary mechanism that enables cloud computing operations on encrypted data without requiring decryption. The encryption scheme allows the cloud provider to perform computations on ciphertexts and return encrypted results, which the client can then decrypt. This intermediary layer maintains security control while enabling cloud computational power.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If existing homomorphic encryption methods are used, then data encryption is maintained, but decryption by cloud provider is required which limits scalability

Engineering Contradiction:
Improvedata encryptionVSAvoidscalability
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The encryption scheme segments the decryption capability from the computation process. Only the client holds the decryption key, while the cloud provider performs computations on encrypted data without decryption capability. This segmentation enables scalable cloud-based homomorphic encryption operations while maintaining data encryption throughout the process.

Inventive Principle:
Principle #1Segmentation

3Reliability

If data is encrypted and uploaded to cloud, then data security is improved, but communication costs and operational complexity worsen

Engineering Contradiction:
Improvedata securityVSAvoidcommunication costs
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The homomorphic encryption scheme enables self-service operations where the encrypted data itself can be processed without requiring decryption or special handling protocols. The cloud provider can directly operate on ciphertexts using homomorphic operations, eliminating the need for complex secure communication protocols and reducing communication overhead between client and cloud service.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3639127B1Homomorphic factorization encryption
Publication Date: 2021.09.01 MICROSOFT TECHNOLOGY LICENSING LLC
  • EP3639127B1 patent drawingFigure 1
  • EP3639127B1 patent drawingFigure 2
  • EP3639127B1 patent drawingFigure 3

AI summary

Systems, methods, and computer-executable instructions for secure data analysis using encrypted data. An encryption key and a decryption key are created. The security of encryption using the encryption key and the decryption key are based upon factoring. A computation key is created based upon the encryption key. Data is encrypted using the encryption key. The encrypted data and the computation key are provided to a remote system. The remote system is requested to perform data analysis on the encrypted data. An encrypted result of the data analysis is received from the remote system. The encrypted result of the data analysis is decrypted with the decryption key.