Quantifying Security and Overheads for Obfuscation of Integrated Circuits

Logic obfuscation techniques are used to deter intellectual property piracy, reverse engineering, and counterfeiting threats in the manufacturing of integrated circuits (IC). The security of these obfuscation algorithms has been, however, compromised by Boolean satisfiability (SAT) based attacks. SAT attacks can reveal the deobfuscation key in seconds, rendering the IC design vulnerable to reverse engineering. The ever-changing landscape of attacks and defenses are typically vetted on small benchmark circuits where security is measured in terms of the time required to recover the encryption key from the obfuscated circuit. This paper introduces a uniform security metric for evaluating the existing obfuscation methods. The benchmark circuits are synthesized after each obfuscation method to determine the overhead in terms of area, power, and timing, including the impact of logic obfuscation algorithms on practical circuits with reasonable gate count (>100K gates). A thorough evaluation is conducted to determine the contributing factors toward attack resiliency time such as gate count, logic depth, and (area, power, timing) overhead before recommending the best obfuscation method for a specific circuit. Keywords—logic encryption; obfuscation; reverse engineering