Delegated Private Set Intersection via Key Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for private set intersection, such as using cryptographic hash functions, are vulnerable to brute force attacks and require multiple interactions between participants, leading to inefficiencies and increased costs due to the need for dedicated servers.
Innovation Solution
A method that delegates the calculation of a keyed hash to a chain of delegate modules, where each module applies a partial key to generate a virtual private key, allowing for secure and efficient determination of data set intersections in a single round of interaction without revealing confidential information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If cryptographic hash functions are used for private set intersection, then data confidentiality is maintained, but the system becomes vulnerable to brute force attacks
Solution Approach 1:
The patent segments the cryptographic key into multiple shares distributed across different delegate modules. Each module holds only a portion of the key, making it impossible for any single module or external attacker to perform brute force attacks on the complete key. This segmentation resolves the contradiction by maintaining confidentiality while eliminating the brute force vulnerability.
Solution Approach 2:
The patent introduces delegate modules as intermediaries between the participants and the hash calculation process. These modules apply the segmented key shares to transform the input data, creating an additional layer of security that prevents direct brute force attacks on the original data or complete key. The intermediaries protect against harmful factors while maintaining the confidentiality function.
2Reliability
If multiple interactions are used between participants for private set intersection, then security is improved, but computational efficiency and cost decrease
Solution Approach 1:
The patent performs preliminary actions by pre-distributing key shares to delegate modules before the actual private set intersection computation. This preparation allows the main computation to proceed with fewer interactions, as the security infrastructure is already in place. The preliminary key distribution enables efficient computation while maintaining security guarantees.
Solution Approach 2:
The delegate modules autonomously apply their key shares to transform data without requiring continuous interaction between participants. Each module independently performs its cryptographic function, reducing the need for multiple back-and-forth interactions. This self-service approach maintains security while significantly improving computational efficiency and reducing costs.
3Device complexity
If a single key is used for hashing, then the process is simple, but the risk of key theft increases
Solution Approach 1:
The patent divides the single key into multiple key shares distributed across different delegate modules. This segmentation maintains relative simplicity in key management (each module manages only its share) while dramatically reducing key theft risk, as an attacker would need to compromise multiple independent modules to obtain the complete key.
Solution Approach 2:
The patent applies different key shares to different delegate modules, giving each module a unique local quality (its specific key share). This allows the system to maintain simple key management at the global level while creating localized security zones where each module's compromised state does not expose the entire key. The local quality differentiation reduces overall risk while preserving management simplicity.
Data Source
AI summary
Embodiments disclosed herein describe computing private set intersection (PSI) between various parties using delegation to other devices and in one round of interaction (request and response). The various parties involved and their associated computing devices are referred to herein as participants. The protocol is forward-secure and completely hides the data of participants from an eavesdropper. Because the protocol only uses a single round of interaction, it is more efficient and does not require each participant to have servers that remain online continuously.


