Homomorphic Encryption Module for Dynamic Data Security
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fully homomorphic encryption (FHE) systems face challenges in maintaining the security and longevity of encrypted control signals in cyber-physical systems, as noise increases with computations and the secret key's vulnerability to attackers threatens security.
Innovation Solution
A method involving a homomorphic encryption module and parallel computing modules that perform fully homomorphic encryption on dynamic data, allowing updates to encrypted state variables without bootstrapping, using catch-up vectors to reduce the number of matrix multiplications and extend the lifespan of ciphertext.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If fully homomorphic encryption is applied to control signals in cyber-physical systems, then security against cyber-physical attacks is improved, but noise increases as the number of computations increases
Solution Approach 1:
The patent applies bootstrapping in advance to refresh the ciphertext before noise accumulates to critical levels. By periodically performing bootstrapping operations on the encrypted state variables, the system prevents noise from degrading the ciphertext to the point where decryption becomes impossible, thus maintaining both security and information integrity over extended computation periods.
Solution Approach 2:
The patent changes the parameter of ciphertext noise level through bootstrapping operations. By transforming the ciphertext and recalibrating its noise characteristics, the system resets the noise parameter to acceptable levels, allowing continued secure computation without permanent degradation of signal quality.
2Loss of information
If bootstrapping is performed to reduce noise in encrypted data, then noise is reduced, but the lifespan of encrypted data is reduced due to computational overhead
Solution Approach 1:
The patent segments the computation process into multiple parallel computing modules that operate simultaneously. Different modules handle different aspects of the control computation, with some dedicated to bootstrapping operations and others to state variable updates. This segmentation allows bootstrapping to be performed concurrently rather than sequentially, reducing the overall time penalty and extending the effective lifespan of encrypted data.
Solution Approach 2:
The patent implements periodic bootstrapping at strategically determined intervals rather than continuously. By monitoring noise levels and performing bootstrapping only when necessary, the system minimizes the computational overhead while maintaining data integrity. This periodic approach extends the lifespan of encrypted data by reducing the frequency of expensive bootstrapping operations.
3Ease of operation
If secret key is stored inside the controller for decryption, then decryption capability is improved, but security is threatened by potential key theft
Solution Approach 1:
The patent extracts the secret key from the controller entirely, storing it only in secure hardware modules dedicated to encryption/decryption operations. The controller itself operates only on ciphertext without ever possessing the decryption key, eliminating the vulnerability to key theft while maintaining full decryption capability through secure enclaves or trusted execution environments that hold the key in isolated memory.
Solution Approach 2:
The patent introduces secure hardware modules as intermediaries between the controller and the decryption process. These dedicated security modules hold the secret key in protected memory and perform decryption operations only when properly authenticated, acting as a mediator that prevents direct access to the key by the main controller while still enabling necessary decryption functionality.
Data Source
AI summary
The present disclosure provides a computer-implemented method for processing dynamic data by dynamic data processing device. The device comprises a homomorphic encryption module and a plurality of computing modules running in parallel. The method comprises carrying out, by the homomorphic encryption module, fully homomorphic encryption to dynamic data received from an object which generates the dynamic data; updating, by the computing module which is not in bootstrapping, the encrypted state variable; and carrying out, by the computing module which completes bootstrapping, the first update to the encrypted state variable. The first update to the encrypted state variable after completion of bootstrapping is carried out by x(t+Nboot)→AN<sub2>boot</sub2>x(t)+Σj=0N<sub2>boot</sub2>−1AN<sub2>boot</sub2>−1−jB(r(t+j)−y(t+j)).