HELP PUF Entropy Tuning With Delay Offsets and TV Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Hardware-Embedded Delay PUF (HELP) systems face limitations in generating unique, random, and reproducible bitstrings due to restricted challenge-response pairs and vulnerability to machine learning attacks, with limited entropy and reliability in resource-constrained environments.
Innovation Solution
The TVCOMP process enhances bitstring generation by compensating for temperature and voltage variations, applying linear transformations to path delays, and using population-based and chip-specific offsets to expand the challenge-response space and improve entropy, reliability, and uniqueness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional PUF architectures use precisely placed and routed arrays of identically designed components, then manufacturing precision is improved, but device complexity increases and area overhead increases
Solution Approach 1:
The patent replaces the mechanical/physical approach of precisely placing and routing identical components with an algorithmic/software-based approach. The HELP PUF uses standard cell placement and routing tools to generate paths, then applies digital signal processing and statistical analysis to extract unique identifiers, substituting physical precision requirements with computational methods
Solution Approach 2:
The patent changes the fundamental parameters of the PUF system by using delay measurements along logic paths as the source of uniqueness instead of physical dimensions or positions of components. By measuring propagation delays through combinational logic and applying statistical transformations, the system extracts unique identifiers without requiring precise physical placement
2Area of stationary object
If HELP PUF uses on-chip logic paths for delay measurements, then area overhead is reduced, but measurement precision deteriorates due to environmental variations
Solution Approach 1:
The patent implements feedback mechanisms where delay measurements are continuously collected under varying environmental conditions, and statistical models are updated to compensate for these variations. The system uses helper data and reconciliation protocols to correct measurement drift, ensuring consistent unique identifier extraction despite temperature and voltage changes
Solution Approach 2:
The patent prepares for environmental variations in advance by collecting delay measurements under multiple operating conditions during fabrication and calibration phases. Helper data and compensation parameters are pre-computed and stored, cushioning against the impact of future environmental variations during operation
3Device complexity
If the number of challenge-response pairs is limited to polynomial size, then device complexity is reduced, but security strength deteriorates due to vulnerability to exhaustive readout
Solution Approach 1:
The patent transitions from a weak PUF model with polynomial challenge space to a strong PUF model by introducing additional dimensions of complexity through machine learning resistance. This is achieved by using diverse path selections, multiple delay measurements per path, and statistical transformations that create an exponentially large effective challenge space while maintaining polynomial hardware complexity
4Device complexity
If resource-constrained versions of HELP restrict the number of paths to 2^20, then device complexity is reduced, but entropy and uniqueness deteriorate
Solution Approach 1:
The patent transforms the limited path delay measurements into high-entropy bitstrings by applying statistical transformations, diversity extraction, and randomness amplification techniques. By changing the parameter space from raw delay values to processed bitstring representations, the system achieves high entropy output from a limited number of physical paths
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
The Distribution Effect is proposed for the HELP PUF that is based on purposely introducing biases in the mean and range parameters of path delay distributions to enhance entropy. The biased distributions are then used in the bitstring construction process to introduce differences in the bit values associated with path delays that would normally remain fixed. Offsets are computed to fine tune a token's digitized path delays as a means of maximizing entropy and reproducibility in the generated bitstrings: a first population-based offset method computes median values using data from multiple tokens (i.e., the population) and a second chip-specific technique is proposed which fine tunes path delays using enrollment data from the authenticating token.