Homomorphic Encryption for Secure Multi-Party Statistical Computation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing secure computation methods fail to ensure the privacy and anonymity of private data shared among multiple parties, particularly in scenarios where data from competing enterprises or peer groups is exchanged, as they lack effective mechanisms to protect against unauthorized access and maintain confidentiality.
Innovation Solution
The implementation of a public key encryption system that enables homomorphic encryption and blind computation, allowing for secure data processing and exchange among multiple parties while ensuring privacy and anonymity, using a protocol that reduces the number of rounds required for data exchange and is independent of the number of parties or values exchanged.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing secure computation methods are used to process private data among multiple parties, then data processing can be performed, but privacy and anonymity of the private data cannot be ensured against unauthorized access
Solution Approach 1:
The patent introduces homomorphic encryption as an intermediary mechanism that enables computation on encrypted data without decryption. The encryption system acts as a mediator between the need for data processing and the requirement for privacy protection, allowing the platform provider to compute statistical quantities from encrypted private values while maintaining confidentiality against unauthorized access
Solution Approach 2:
The patent transforms the security parameter from traditional encryption-decryption models to homomorphic encryption models. By changing the cryptographic parameter state from decrypted to encrypted-computable form, the system enables processing while maintaining privacy. The blind computation technique further changes parameters by introducing random blinding factors that hide the actual values during statistical computation
2Reliability
If traditional encryption methods are used for secure data exchange, then confidentiality can be maintained, but the number of exchange rounds increases and efficiency decreases
Solution Approach 1:
The patent applies preliminary action by pre-establishing homomorphic encryption keys and protocols before data exchange begins. The system prepares the cryptographic infrastructure in advance, including generating public and private keys for homomorphic encryption, so that subsequent data exchange can proceed efficiently without requiring multiple rounds of key establishment and decryption operations
Solution Approach 2:
The patent extracts the decryption step from the computation process. Instead of requiring data to be decrypted before processing, the system extracts and eliminates the decryption operation entirely by using homomorphic properties that allow computation on encrypted data. This removes a critical bottleneck that would otherwise require multiple exchange rounds between parties
3Productivity
If private data from multiple parties is collected for statistical computation, then useful statistical quantities can be generated, but the privacy of individual party data is compromised
Solution Approach 1:
The patent changes the parameter state of private data from plaintext to homomorphically encrypted form throughout the computation process. This parameter transformation allows statistical quantities to be computed from encrypted values, and the results remain encrypted until the final stage where only the statistical output (not individual inputs) is revealed, thus preventing privacy loss while maintaining computational capability
Solution Approach 2:
Homomorphic encryption serves as an intermediary that enables statistical computation without direct access to private data. The encryption system mediates between the need to aggregate data from multiple parties and the requirement to protect individual privacy, allowing the platform provider to compute statistics like averages and sums while the encrypted nature of the data prevents unauthorized access to individual party information
Data Source
AI summary
An embodiment may include a system having a communication unit and a processing unit. The communication unit may be configured to receive an encrypted private value of a party, the encrypted private value being generated from a private value with a public-key encryption system and a public key, to send an encrypted blinded result to the party, and to receive a blinded result generated from the encrypted blinded result. The processing unit may be configured to compute a result of a function, the function having as input the private value, to blind the result of the function to generate the encrypted blinded result, and to compute the result by unblinding the blinded result.