Grouped PUF Signatures Using Majority Voting for Stable Authentication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Physically unclonable functions (PUFs) in electronic devices often exhibit instability over time, leading to undesired errors due to thermal and electronic variability, which affects their reproducibility and reliability in authentication processes.
Innovation Solution
Implementing a system with multiple PUF elements grouped into subsets, where the PUF signature is generated based on majority values from these groups, and using a mask to store which elements belong to the subset, improving the repeatability of the PUF response by leveraging the collective stability of group bits.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If individual PUF elements are used to generate PUF signatures, then the PUF provides unique identification capability, but the PUF response exhibits instability over time due to thermal and electronic variability
Solution Approach 1:
The PUF array is divided into multiple groups, where each group contains multiple PUF elements. Instead of using individual PUF elements directly, the system segments them into groups and uses majority voting within each group to determine the group bit value. This segmentation approach reduces the impact of thermal and electronic variability on individual elements, thereby improving response stability and reproducibility.
Solution Approach 2:
Multiple PUF elements within each group are merged through majority voting to produce a single group bit. By combining the responses of multiple elements (e.g., 3 out of 5 elements), the system leverages the collective stability of the group, reducing the effect of random variations in individual elements and improving overall PUF response reliability.
2Measurement precision
If multiple PUF elements are grouped into subsets with majority voting, then the reproducibility of PUF response is improved, but the device complexity increases due to additional grouping logic and mask storage
Solution Approach 1:
The system performs preliminary action by pre-determining the optimal grouping configuration and storing it in a mask during manufacturing or initialization. This mask captures the stable subset of PUF elements for each group, eliminating the need for complex real-time evaluation of all possible element combinations during operation. The grouping logic is simplified because the mask pre-identifies which elements to vote on, reducing runtime complexity while maintaining high reproducibility.
3Reliability
If a mask is used to store subset information for PUF groups, then the error rate is reduced by selecting stable elements, but the loss of information increases due to storing additional mask data
Solution Approach 1:
The system extracts only the essential stability information into a compact mask, storing only the configuration data needed to identify stable PUF element subsets. Rather than storing complete PUF responses or extensive characterization data, the mask contains minimal information (element selection bits) that enables error reduction. This extraction approach achieves high authentication accuracy while minimizing information storage requirements.
Data Source
AI summary
A physically unclonable function (PUF) includes a plurality of PUF elements to generate an N-bit PUF signature. For each bit in the N-bit PUF signature, a PUF group of K number of individual PUF elements indicating a single-bit PUF value is used to generate a group bit. The group bits are more repeatable than the individual PUF elements. The value K may be selected such that (K+1)/2 is an odd number.