Hardware Hacking Space: "Machine Learning vs. Hardware Integrity Protection: Who will prevail?"

Date:

2021

Language:

English

Abstract:

As the designated root of trust, hardware is undoubtedly the most critical layer to security in modern electronic systems. However, due to the involvement of third parties in the integrated circuit design and fabrication flow, protecting the hardware against malicious modification, i.e., hardware Trojans, has become ever more complex. This challenge has triggered great interest in the academic and industrial sector to explore novel hardware design for trust methodologies. In particular, logic locking—a hardware obfuscation technique—has evolved as a prominent method to safeguard hardware designs throughout the integrated circuit supply chain. However, the widespread proliferation of machine learning has recently introduced a new pathway to evaluating and breaking locking schemes, thereby uncovering fundamental security vulnerabilities.

In this talk, the speaker will first take a look at the concept of hardware integrity protection w.r.t the mitigation of malicious design modifications known as hardware Trojans. Second, he will analyze the impact of data-driven attacks on modern hardware obfuscation policies.
Finally, the speaker will discuss future steps in designing machine-learning-resilient locking schemes.

Download:

BibTeX