PUF Key Reconstruction Using Helper Data for Noise Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Physical unclonable functions (PUFs) produce unpredictable and device-unique responses but are susceptible to measurement noise and environmental influences, which can compromise the reliability and security of cryptographic keys, necessitating post-processing to correct for noise and ensure key consistency.
Innovation Solution
The use of fuzzy extractors, which include an enrollment stage to derive and store helper data from PUF responses, and a reconstruction stage to reevaluate and correct these responses, ensuring identical key derivation while maintaining helper data as public and non-sensitive, thereby reducing information leakage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If PUF responses are used directly to generate cryptographic keys, then device uniqueness and unpredictability are achieved, but measurement noise and environmental influences compromise reliability and key consistency
Solution Approach 1:
The patent introduces helper data as an intermediary between the noisy PUF response and the cryptographic key. The helper data contains correction information that mediates the transformation from noisy PUF output to reliable, consistent cryptographic keys across multiple uses while maintaining security properties
Solution Approach 2:
The patent performs preliminary processing of PUF responses during an enrollment phase, where helper data is generated and stored before actual key usage. This preliminary action prepares the system to correct noise in future operations, ensuring reliability without requiring real-time complex processing
2Reliability
If post-processing is applied to correct PUF noise, then key reliability improves, but additional handling may introduce security risks and information leakage
Solution Approach 1:
The patent converts the potentially harmful helper data into a beneficial element by designing it to contain only public, non-sensitive information that aids in noise correction without revealing cryptographic secrets. The helper data structure itself becomes a security feature by separating public correction information from private key material
Solution Approach 2:
The patent applies different quality requirements to different parts of the system: the helper data is designed to be public and non-sensitive (lower security requirement), while the cryptographic key remains private and highly sensitive (higher security requirement). This local differentiation of security properties allows noise correction without compromising key security
Data Source
AI summary
Some embodiments are directed to an electronic cryptographic device arranged to determine a cryptographic key. The cryptographic device can include a physically unclonable function (PUF) arranged to produce a first noisy bit string during the enrollment phase and a second noisy bit string during the reconstruction phase, and a statistical unit arranged to execute a statistical test for verifying correct functioning of the physical unclonable function. The statistical test computes a statistical parameter for the physical unclonable function using helper data. The statistical test determines correct functioning if the statistical parameter satisfies a criterion of the statistical test.