This special session paper discusses recent advances on photonic physical unclonable functions (PUFs), providing a broader overview of and motivation for photonic PUFs. We discuss their potential advantages, such as a higher entropy, larger complexity, and possibly better resilience against machine learning attacks. We also deal with some recent implementations based on linear and non-linear optics, alongside with their main advantages and limitations.

Integrated silicon photonics is revolutionizing the field of optical communications as hybrid photonic/electronic integrated circuits made from this semiconductor material can seamlessly process signals spanning both the optical and electrical domains. The enormous bandwidth afforded by optical signaling enables extremely high aggregate data-rate communication signals. Correspondingly, signal processing devices must employ new mechanisms to keep up with the ever increasing need for greater capacity in communications. To this end, parametric nonlinear optical interactions in silicon-based materials provide a route toward processing and handling signal bandwidths well beyond a THz. These integrated silicon-based photonic structures also benefit from high optical confinement, large nonlinear optical responses, and long interaction lengths, making them highly suitable for power-efficient parametric nonlinear optical interactions. In particular, the parametric nonlinear optical interaction of four-wave mixing (FWM) in silicon-based photonic devices has enabled a wealth of devices for next-generation information handling applications including communications, all-optical computing, metrology, and security, which will be highlighted in this review.