Biometric Identity Verification Using Encrypted Score Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional identity verification methods, such as the Colmade process, are inefficient due to long execution times and require centralized steps with multiple entities, which can be a security concern and necessitate the use of multiple databases.
Innovation Solution
A method that calculates a score representing the distance between biometric test and reference data using encrypted data, applying decryption and masking treatments with secondary keys and masks, allowing for faster centralized processing and secure identity verification without plaintext calculation, and can be implemented with a single database.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the Colmade method performs score calculation and threshold comparison in the encrypted domain with centralized steps distributed across several entities, then security is improved, but execution time increases significantly
Solution Approach 1:
The patent divides the decryption process into multiple segments by introducing secondary decryption keys (sk1, sk2) that are derived from the primary decryption key. Each entity holds a segment of the decryption capability, and the score is decrypted through coordinated processing by multiple entities. This segmentation maintains security through distributed key management while optimizing execution time by enabling parallel processing of decryption operations.
Solution Approach 2:
The patent introduces an intermediary masking mechanism where a mask is applied to the score during the decryption process. This intermediary element allows the score to be processed in a secured manner without requiring full plaintext decryption at any single point, enabling faster processing while maintaining security guarantees. The mask acts as a mediator that facilitates efficient computation without compromising the encrypted domain security model.
2Reliability
If multiple entities perform parallel calculations with centralized recombination steps, then security is improved through distribution, but device complexity and coordination overhead increase
Solution Approach 1:
The patent merges the security benefits of distributed processing with the efficiency of centralized coordination by combining multiple decryption operations into a unified decryption framework. Instead of maintaining separate complex distributed systems, the invention combines the secondary decryption keys and masking operations into a cohesive process that achieves both security through key distribution and simplicity through unified processing logic.
Solution Approach 2:
The patent creates a universal decryption framework that can handle both centralized and distributed scenarios through the same mechanism. The secondary key derivation and masking process serves multiple functions: it provides security through key segmentation, enables parallel processing, and maintains compatibility with existing encrypted domain computation infrastructure. This multi-functional approach reduces overall system complexity by eliminating the need for separate specialized systems.
Data Source
Figure 1~2
Figure 3
AI summary
Method comprising: calculation (204) of a cipher (cs) of a score representing a distance between a test biometric data (x) and a reference data (yu); implementation of the following steps by a device of index i=1,2: processing of decryption and masking to the cipher (cs) of the score using a secondary decryption key ((sk)i) of index i and a secondary mask ((r)i) of index i, the processing producing a data (ŝ) representing the score in a decrypted form and masked by a primary mask (r) without having calculated the score in plain text, and generation (308) of a partial result (oi) of index i from the data (9) and a demasking data (ki) of index i; in which the partial results (o1; o2) allow the calculation of a result (o) indicating whether the biometric test data corresponds to the reference data.