Machine Learning vs. Hardware Integrity Protection - Dominik Sisejkovic - Ekoparty 2021: Hardware Hacking Space
Ekoparty Security Conference via YouTube
Overview
Explore the critical intersection of machine learning and hardware integrity protection in this 32-minute conference talk from Ekoparty 2021's Hardware Hacking Space. Delve into the challenges of safeguarding hardware against malicious modifications, particularly hardware Trojans, in the context of third-party involvement in integrated circuit design and fabrication. Examine logic locking as a prominent hardware obfuscation technique and its vulnerabilities to machine learning-based attacks. Gain insights into the concept of hardware integrity protection, analyze the impact of data-driven attacks on modern hardware obfuscation policies, and discuss future steps in designing machine-learning-resilient locking schemes. Learn from speaker Dominik Šišejković, an accomplished researcher and award-winning expert in hardware security and logic locking, as he shares his expertise on this crucial topic in cybersecurity.
Syllabus
Intro
Motivation
untrusted IC supply chain
Hardware trojans
Design dependent trojans
Reverse engineering
Logic locking
Attack model
Schemes
Exploitation
Snapshot
Oracle Guided Attacks
Preventing ML Attacks
Taught by
Ekoparty Security Conference