Proximity Determination Using Key Matrix Modulus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing proximity determination technologies using masked numerical information are vulnerable to information leakage, as the same mask value is shared among devices, potentially revealing user information to others.
Innovation Solution
A method involving a key matrix generated with a random number vector added as a last column, where each device uses a unique random number to create a concealed vector, and a remainder vector is calculated using the key matrix as a modulus to determine proximity without revealing actual positions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a shared mask value is used to conceal numerical information, then the concealment process is simple and efficient, but information leakage occurs as the same mask can be reversed by third parties
Solution Approach 1:
The patent divides the single shared mask into multiple unique masks, with each device having its own private mask value. This segmentation prevents third parties from reversing the concealment operation, as each device uses a different mask that cannot be guessed or reversed by others. The proximity determination server collects concealed values using these unique masks and performs proximity determination without needing to know the actual mask values.
Solution Approach 2:
The patent introduces a proximity determination server as an intermediary that performs the proximity determination operation without directly accessing or knowing the actual numerical information or mask values. The server receives concealed values from multiple devices, performs mathematical operations on these concealed values to determine proximity, and returns results without revealing any underlying data, thus maintaining security while enabling the desired functionality.
2Reliability
If unique random numbers are used for each device to prevent information leakage, then security is improved, but the complexity of the system increases
Solution Approach 1:
Each device generates its own unique random number (mask) independently without requiring coordination or complex setup procedures. The devices autonomously create their own concealment values by combining their unique random numbers with the numerical information they wish to conceal. This self-service approach simplifies the overall system architecture while maintaining security, as each device handles its own concealment without needing complex inter-device communication or centralized mask management.
3Loss of information
If numerical information is concealed using masking, then user privacy is protected from third parties, but the ability to perform proximity determination accurately is reduced
Solution Approach 1:
The patent changes the mathematical parameters and operations used in proximity determination to work correctly with concealed values. Instead of computing distances using raw numerical information, the system performs operations on concealed values (adding concealed values, computing differences modulo a key matrix) to derive proximity relationships. This parameter change allows accurate proximity determination while maintaining the concealment of actual position data, as the mathematical operations preserve the relative relationships between values even when masked.
Data Source
AI summary
A proximity determination method includes: collecting a first concealed vector based on a linear combination of a row vector of a key matrix, first numerical information, and a first random number, the key matrix being generated by adding a random number vector as a last column to a matrix in which a threshold for determining a proximity range is a diagonal element; collecting a second concealed vector based on the linear combination, second numerical information, and a second random number; calculating a remainder vector, the remainder vector being a remainder obtained from a difference between the collected first concealed vector and the second concealed vector when the key matrix is used as a modulus; and determining whether the first numerical information and the second numerical information belong to the proximity range, based on the calculated remainder vector.