Enhancing IoT Security: Optimizing PUF Responses through Pre-Processing Techniques

Abstract

In this paper, we propose and detail the implementation of pre-processing techniques—specifically truncation and uniformization—to enhance the performance of authentication processes utilizing Physical Unclonable Functions (PUFs) within the context of the Internet of Things (IoT). Traditional authentication methods are often critiqued for their reliance on static secret storage, presenting inherent security risks. Physical Unclonable Function (PUF) technology addresses this concern by dynamically generating keys, akin to a device's "biometric" signature, thereby offering a more secure alternative. However, despite the dynamic nature of PUF-generated secret keys, vulnerabilities to specific attacks persist. Prior research has not focused on optimizing the secret key generated by PUFs, resulting in a lack of additional security layers and maintaining the susceptibility to PUF-targeted attacks at a constant level. This study introduces a PUF-based IoT device framework that optimizes PUF responses, aiming to significantly improve the security performance of the system. This enhancement is evaluated through metrics such as decidability, the confusion matrix, and randomness value, presenting a comprehensive approach to reinforcing system security. The optimization of Physical Unclonable Function (PUF) responses, through methods such as truncation or bit uniforming, plays a critical role in enhancing the security of IoT devices. Our findings indicate that bit uniforming notably improves system security, evidenced by a significant increase in the decidability value from 0.73 (unoptimized) to 1.37. This improvement is further reflected in the confusion matrix values, with False Rejection Rate (FRR), False Acceptance Rate (FAR), True Rejection Rate (TRR), and True Acceptance Rate (TAR) showing marked improvements from 18.02%, 4.93%, 95.06%, and 81.97% in the unoptimized state to 3.04%, 0.98%, 99.02%, and 96.96%, respectively, post-optimization. The proposed pre-processing techniques show its effectiveness in the PUF authentication systems, as superior results are obtained.