Scaling Logic Locking Schemes to Multi-module Hardware Designs

The involvement of third parties in the integrated circuit design and fabrication flow has introduced severe security concerns, including intellectual property piracy, reverse engineering and the insertion of malicious circuits known as hardware Trojans. Logic locking has emerged as a prominent technique to counter these security threats by protecting the integrity of integrated circuits through functional and structural obfuscation. In recent years, a great number of locking schemes has been introduced, thereby focusing on a variety of security objectives and the resiliency against different attacks. However, several major pitfalls can be identified in the existing proposals: (i) the focus on isolated and often small circuit components, (ii) the assumption of unrealistic attack models that enable powerful attacks on logic locking and (iii) the design of very specific locking schemes targeted towards achieving resilience against specific attacks. These observations strongly impair the practicality of logic locking. Therefore, in this paper we present a holistic framework for scaling logic locking schemes to common multi-module hardware designs, thereby showcasing an industry-ready pathway of applying logic locking in a realistic design flow. The framework represents an enhancement of the previously published Inter-Lock methodology, offering several algorithmic improvements as well as toolflow implementation details to facilitate the applicability of the framework to large multi-module designs. The framework is tested and evaluated on a real-life 64-bit RISC-V core.