ALMOST: Adversarial Learning to Mitigate Oracle-less ML Attacks via Synthesis Tuning

TL;DR

ALMOST is a framework that uses adversarial learning and synthesis tuning to reduce the effectiveness of oracle-less ML attacks on logic locking schemes without compromising design optimization.

Contribution

It introduces a security-aware synthesis method employing adversarial models and simulated annealing to mitigate ML attacks on logic locking.

Findings

Attack accuracy drops to around 50% on ALMOST-synthesized circuits.

The framework maintains design optimization while enhancing security.

Experiments on ISCAS benchmarks validate effectiveness.

Abstract

Oracle-less machine learning (ML) attacks have broken various logic locking schemes. Regular synthesis, which is tailored for area-power-delay optimization, yields netlists where key-gate localities are vulnerable to learning. Thus, we call for security-aware logic synthesis. We propose ALMOST, a framework for adversarial learning to mitigate oracle-less ML attacks via synthesis tuning. ALMOST uses a simulated-annealing-based synthesis recipe generator, employing adversarially trained models that can predict state-of-the-art attacks' accuracies over wide ranges of recipes and key-gate localities. Experiments on ISCAS benchmarks confirm the attacks' accuracies drops to around 50\% for ALMOST-synthesized circuits, all while not undermining design optimization.