A 0.26% BER, 10<sup>28</sup> Challenge-Response Machine-Learning Resistant Strong-PUF in 14nm CMOS Featuring Stability-Aware Adversarial Challenge Selection
Vikram Suresh, Raghavan Kumar, Mark Anders, Himanshu Kaul, Vivek De, Sanu Mathew
Abstract
A 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">28</sup> challenge-response strong-PUF in 14nm CMOS, demonstrates machine learning (ML) attack resistance across 6-million training samples. The 2-stage non-linear cascaded PUF array with adversarial challenge selection limits ML attack accuracy to ~50%. The configurable cross-coupled inverter-based entropy source with stability-aware challenge pruning enables 9.8× higher array density and 0.26% peak BER across 650-850mV and 0-100°C.
Topics & Concepts
CMOSSelection (genetic algorithm)Adversarial systemStability (learning theory)InverterComputer scienceEntropy (arrow of time)PruningArtificial intelligenceMachine learningEngineeringElectronic engineeringElectrical engineeringPhysicsBiologyBotanyQuantum mechanicsVoltagePhysical Unclonable Functions (PUFs) and Hardware SecurityIntegrated Circuits and Semiconductor Failure AnalysisAdvancements in Semiconductor Devices and Circuit Design