Biometric Authentication Using Inner Product Encryption and Rejection Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biometric authentication methods are hindered by noise variability and have slow encryption speeds with long ciphertext lengths, making them inefficient for secure and fast authentication.
Innovation Solution
A computer-implemented method using an authentication server that processes vectors of secret information by calculating inner products and determining authentication success based on predetermined values, with the option to encrypt and transmit vectors, employing a matrix Q and error vectors to manage noise and security.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional encryption methods such as Fuzzy extractor and Inner product encryption are used for encrypting biometric information, then security is improved, but authentication speed becomes slow and ciphertext length becomes long
Solution Approach 1:
The patent changes the mathematical parameters of the encryption scheme by using Rejection Sampling technique to generate ciphertexts with specific properties. The method transforms the biometric authentication problem into a computational problem where the ciphertext length and computation time are controlled by parameter choices in the rejection sampling process, achieving both security and efficiency
Solution Approach 2:
The patent extracts the essential authentication function from complex encryption operations. By using inner product encryption with carefully constructed matrices and vectors, the method separates the encryption overhead from the actual authentication computation, allowing fast verification while maintaining security through the matrix structure
2Reliability
If conventional encryption methods are used for biometric information, then security is improved, but ciphertext length becomes long
Solution Approach 1:
The patent controls ciphertext length by changing the dimensional parameters of the matrix Q and vector operations. The rejection sampling process generates ciphertexts with optimized length properties, and the inner product computation uses vectors of controlled dimensionality to achieve secure authentication with minimal ciphertext overhead
3Productivity
If biometric information is processed without encryption, then authentication speed is improved, but security and risk of leakage worsen
Solution Approach 1:
The patent introduces encrypted vectors and matrices as intermediaries between the biometric data and the authentication decision. The inner product computation operates on encrypted representations, allowing the authentication server to verify identity without accessing or storing the raw biometric information, thus maintaining both speed and security
Data Source
AI summary
The computer-implemented method for authenticating secret information according to an aspect of the present disclosure, comprises receiving, by an authentication server, Q{right arrow over (X)} from a terminal for registering secret information; storing, by the authentication server, the received Q{right arrow over (X)}; receiving, by the authentication server, a vector {right arrow over (Z)} from a terminal for requesting authentication of secret information; calculating, by the authentication server, the inner product of Q{right arrow over (X)} and {right arrow over (Z)}; calculating, by the authentication server, ½(n−the inner product); and determining, by the authentication server, that the authentication is successful if ½(n−the inner product) is within a predetermined value and that the authentication fails otherwise.