A machine learning resistant crp obfuscation method for strong puf

By combining ROPUF and ArbiterPUF circuits with an excitation and response obfuscation module, the problem of high hardware resource overhead of strong PUF circuits is solved, achieving resistance to machine learning attacks while reducing hardware overhead and attack prediction rate.

CN116522296BActive Publication Date: 2026-07-03HEFEI UNIV OF TECH
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HEFEI UNIV OF TECH
Filing Date
2023-04-06
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

Existing robust PUF circuits that resist machine learning have complex structures and high hardware resource consumption, making them unsuitable for effective application in IoT devices with limited resources.

Method used

By combining ROPUF circuits, ArbiterPUF circuits, and excitation-response obfuscation modules, the excitation-response correlation is reduced through cyclic shifting, XOR operations, and obfuscation logic units, thus achieving resistance to machine learning attacks.

Benefits of technology

It effectively defends against machine learning attacks, reduces stimulus-response correlation, reduces hardware overhead, improves hardware resource utilization, and lowers attack prediction rate.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116522296B_ABST
    Figure CN116522296B_ABST
Patent Text Reader

Abstract

This invention relates to the field of CRP obfuscation technology and discloses a CRP obfuscation method for resisting machine learning attacks with strong PUF. The system includes a ROPUF circuit, an ArbiterPUF circuit connected to the ROPUF circuit, and an excitation obfuscation module. The excitation obfuscation module is connected to a response obfuscation module connected to the ArbiterPUF circuit. The invention also includes a CRP obfuscation method for resisting machine learning attacks with strong PUF, comprising the following steps: S1: Configure the system, input the original excitation signal C to the ROPUF circuit to generate an output response R1, and cyclically shift the output response R1 N times to obtain N different output responses R1'. This invention can effectively resist machine learning attacks, and through the ArbiterPUF and obfuscation logic unit, it can reduce the correlation between excitation and response to a certain extent, making it impossible for machine learning attacks to obtain the original excitation-response pair, ultimately greatly reducing the attack prediction rate, while also having low hardware overhead.
Need to check novelty before this filing date? Find Prior Art