Secure Authentication Using Distributed LSH and Secret Hash Tables
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing secure authentication methods using Locality Sensitive Hashing (LSH) do not adequately protect user information and related feature amounts, as they either fail to keep information completely secret or are computationally inefficient, particularly in one-to-N authentication scenarios.
Innovation Solution
A secure authentication method and system that employs distributed Locality Sensitive Hashing (LSH) values and secret distribution methods to keep user information and feature amounts secure, using a unidirectional function to process inputs while maintaining secrecy, and a secret hash table to compute similarities and derive authentication results without revealing sensitive information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If LSH values are computed multiple times while keeping input data secret, then user information and feature amounts are protected, but computational amount increases and processing becomes inefficient
Solution Approach 1:
The patent segments the authentication process into two phases: a registration phase where LSH values are computed and stored in a hash table, and an authentication phase where pre-computed LSH values are retrieved and compared. This segmentation eliminates the need to compute LSH values multiple times during authentication, resolving the contradiction between maintaining secrecy and improving processing speed.
Solution Approach 2:
The patent performs the computationally intensive LSH value computation in advance during the registration phase, before actual authentication occurs. By preparing and storing the LSH values in a hash table during registration, the system avoids repeated computation during authentication, thus maintaining both security and efficiency.
2Ease of operation
If LSH values are used for authentication, then user authentication is enabled, but user information and feature amounts are not kept completely secret and may be guessed
Solution Approach 1:
The patent introduces LSH values as an intermediary representation that bridges the original user information and the authentication process. Instead of storing or transmitting raw user information, the system uses LSH values as an intermediate form that enables authentication while obscuring the original data, thus maintaining both functionality and secrecy.
Solution Approach 2:
The patent creates a transformed copy of the user information in the form of LSH values, which retains the necessary authentication characteristics while losing the ability to infer the original user information. This copying approach allows authentication to proceed without exposing the original sensitive data.
3Measurement precision
If feature amounts are used for authentication, then authentication accuracy is improved, but attacks can infer original user information from feature amounts
Solution Approach 1:
The patent converts the potential harm of feature amount leakage into a benefit by using LSH values as a protective transformation. The LSH values maintain the authentication accuracy needed for secure verification while simultaneously preventing inference attacks, as the transformed values do not reveal information about the original user data.
Data Source
AI summary
A secure authentication method includes: deriving a distributed LSH value using secret LSH, taking a first distributed feature amount which is a feature amount of user information distributed through a secret distribution method and encrypted LSH parameters as inputs; deriving a distributed hash value using a secret unidirectional function, taking the distributed LSH value and a distributed key as inputs; decoding the hash value by reversing distribution of the distributed hash value; selecting, from a secret hash table storing sets of a hash value as an index and a distributed feature amount as a data string, a set including a hash value matching the decoded hash value; computing, in secret, similarity between the distributed feature amount in the set and the first distributed feature amount; deriving, in secret, a user authentication result based on the similarity computed; and outputting the derived authentication result.