Theoretical study of randomly distributed gold nanoparticles as potential plasmonic physically unclonable functions

We theoretically study a random arrangement of cylindrical gold nanoparticles (NPs) deposited on a dielectric/gold substrate as a potential plasmonic physically unclonable function (PUF). We use the coupled dipole approximation (CDA) method in order to reduce the computation time and memory demand. Each NP is modeled as a single electric dipole through the polarizability tensor of a spheroidal-shaped NP, while the interparticle coupling is described by Green’s tensors. The concept of plasmonic PUF considered here consists of transforming the electric field maps obtained by the CDA into 2D-binary matrices using a key generation algorithm. A PUF must satisfy different criteria to be efficient, including uniqueness that is evaluated by the distributions of Hamming inter-distances. In this work, we propose to evaluate the uniqueness property of different PUFs characterized by substrates whose dielectric layer thickness is variable. We also test the ability to generate unique keys when the challenges to which our PUF concept is subjected change, namely, the wavelength, angle, and plane of incidence.