MOS PUF Authentication Using Trap Occupancy for Larger CRP Space
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing PUFs based on MOS devices are limited by small challenge-response pairs space, leading to reduced robustness and increased server load, making them inadequate for 'strong' authentication systems and inefficient in adapting to various application scenarios.
Innovation Solution
A PUF-based method that involves randomly selecting MOS devices, calculating time constants of traps, constructing a probabilistic model, and generating code keys based on trap occupancy probabilities to expand the challenge-response pairs space, reducing server data storage and improving efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing RTN PUFs are used to achieve high stability and robustness, then system reliability is improved, but the challenge-response pairs space remains small limiting applicability
Solution Approach 1:
The patent segments the trap states into multiple energy levels (first energy level, second energy level, etc.) within the gate oxide layer. Each energy level can be independently detected and used to generate challenge-response pairs. This segmentation transforms a single-bit response system into a multi-bit response system, expanding the CRPs space while maintaining the reliability benefits of trap-based PUFs.
Solution Approach 2:
The patent introduces a new dimension by detecting trap states at multiple energy levels rather than just presence/absence. By adding the energy level dimension to the trap state detection, the system generates significantly more challenge-response pairs from the same physical traps, thereby expanding applicability without compromising reliability.
2Reliability
If all CRPs are input to a database for key management, then authentication is achieved, but server load increases exponentially reducing working efficiency
Solution Approach 1:
The patent extracts only the essential authentication information (trap occupancy states at different energy levels) and stores it in a compact format in the database. By extracting only necessary data rather than storing all possible CRPs, the system maintains authentication security while dramatically reducing database size and server processing load.
Solution Approach 2:
Instead of storing all challenge-response pairs in the database, the patent inverts the approach by storing only the minimal required trap state information and generating CRPs on-demand during authentication. This inversion reduces database storage requirements from exponential to linear scale, improving server efficiency while maintaining security.
3Ease of manufacture
If MOS devices are used for PUF implementation, then low cost and small size are achieved, but threshold voltage offset during aging reduces robustness
Solution Approach 1:
The patent copies the intrinsic trap characteristics from the gate oxide layer to create a stable reference for authentication. By detecting trap states at multiple energy levels and using their relative relationships rather than absolute threshold voltages, the system creates a robust digital fingerprint that remains stable despite aging-induced threshold voltage shifts, maintaining both low cost and high robustness.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly increases the challenge-response pairs space, enhancing system reliability, robustness, and reducing server load, while uniquely identifying physical devices and automatically encrypting transmission keys, thus improving security and reducing design and manufacturing costs.
Implementation Method 1
a trap in a gate oxide layer of the MOS device captures or releases a carrier, thereby changing a threshold voltage of the MOS device
Implementation Method 2
time constants of noise signals are closely associated with activation energy of traps, and noise is also effective against electrical stresses. By virtue of strong stability of the noise, the use of the noise in the PUFs, i.e., random telegraph noise physical unclonable functions (RTN PUFs)
Data Source
AI summary
A physical unclonable function (PUF)-based method for enhancing system reliability is provided, including: requesting, by a client, data transmission with a server; randomly selecting, by the server, a plurality of metal oxide semiconductor (MOS) devices in an MOS array, and acquiring positional information of the plurality of MOS devices; calculating, by the server, a probabilistic PUF that the trap in each of the plurality of MOS devices is occupied by a carrier and constructing a probabilistic model; randomly generating, by the server, detection time according to the probabilistic model and sending the detection time and the positional information to the client; and determining, by the server, an occupancy probability of the trap in each of the plurality of MOS devices at the detection time according to the probabilistic model, and generating a theoretical code key.