A Stochastic Performance Analysis Framework for DNA-Based Authentication Protocols

Abstract

Synthetic DNA is gaining significant attention for its potential across various fields, both in biology and as an information medium. One particular emerging field of application is product tagging, driven by the need for enhanced security, precise traceability, and reliable authenticity verification. The field of DNA-based authentication protocols is still in its early stages. In this work we propose a framework for analyzing and comparing the performance of such protocols. We introduce precise definitions and present analytical results. We evaluate two representative protocols inspired by research published in recent literature. The first protocol employs the whole genome of a cell line with mutations at $n$ specific locations and the second uses Chemical Unclonable Functions (CUFs), inducing unique patterns based on primer pair challenges. We model the interaction between three entities: the manufacturer ( $\mathcal {M}$ ), the customer ( $\mathcal {C}$ ), and an adversary ( $\mathcal {A}$ ). After receiving the product, $\mathcal {C}\ $ obtains verification challenges from $\mathcal {M}$ , applies them to the DNA embedded in the product, and sends the resulting pattern back to $\mathcal {M}\ $ for verification. Meanwhile, $\mathcal {A}\ $ attempts to counterfeit products by imitating the tagging signature based on observed results. We propose a performance indicator that quantifies $\mathcal {A}$ ’s success probability as a function of the number of observed product instances and on the signature complexity. Finally, we discuss key challenges and propose optimization approaches to strengthen protocol resilience.