ZeKi: A Zero-Knowledge Dynamic Logic Locking Implementation with Resilience to Multiple Attacks

In recent years, hardware security has gained significant attention, particularly in the context of protecting Integrated Circuits (ICs) from reverse engineering (piracy), counterfeiting, and overproduction. Logic locking, also known as logic obfuscation, has emerged as a solution to safeguard the vast majority of general ICs. However, traditional logic locking methods, including some state-of-the-art approaches, are vulnerable to Satisfiability (SAT) attacks, structural attacks, and sensitization attacks. Additionally, there is a potential threat from internal IC designers who are well-versed in logic locking. To address these threats by dynamic way with less overhead, we proposes a comprehensive logic locking design and implementing framework called ZeKi: Dynamic Logic Locking (DLL) by ‘Zero-Knowledge’ Implementation, an adaptive approach that customises the implementation of diverse locking mechanisms with the lower overhead for each unique design. The implemented DLL provides defence against all attacks mehtioned above. To our best knowledge, $\mathbf{Z e K i}$ is the first logic locking strategy to implement ‘zero-knowledge’ feature which effectively safeguards the design against potential malicious attacks from the design team. ZeKi’s DLL matches SFLL-HD0 in SAT attack resistance, with logarithmically increasing DIPs; and a superior average Hamming difference of $\mathbf{8 9. 5}$ compared to zero in other schemes on structural attack resistance. It also offers five times the resistance of RLL against sensitization attacks. Zeki’s DLL incurs lowest average power and area overheads of $4.27 \%$ and $4.48 \%$, respectively.