A Novel Approach to Optimize Response-Based Cryptography for Secure Data Transmission

In today’s technological landscape, where devices like robots, automotive vehicles, IoT devices, and machinery rely heavily on data for training and functionality. Therefore, storing sensitive information such as passwords and cryptographic keys poses considerable risks. Physical Unclonable Functions (PUFs) have the potential to mitigate some of these risks; however, integrating PUFs into cryptographic systems presents challenges due to their susceptibility to environmental conditions, making it difficult to consistently recover reliable keys. Response-Based Cryptography (RBC) has been proposed as a server-based search engine to address the computational challenges posed by error correction codes, offering a solution for effectively managing discrepancies in IoT devices with limited resources. The issue of burst errors in binary key streams poses significant security risks, requiring effective mitigation strategies like RBC with fragmentation. This paper proposes a solution involving a fixed-error in sub-response to address burst errors, against attackers attempting to replicate the PUF through modeling attacks. The paper also presents an analysis demonstrating a significant increase in computational complexity for attackers using an interleaver in the proposed method. Statistical analysis are introduced to quantify the rates of successful key recovery and latency data for keys of varying lengths. The method is capable to handle up to 35% bit error rates in the PUF-generated keys within milliseconds. In the rare cases of failure to recover the keys, repeating the challenge-response-pair mechanism of the PUF proves to be a faster and effective alternative than persisting in a lengthy search.