Physical Layer Authentication for Distributed RIS (DRIS) Enabled VLC Systems

The introduction of Reflective Intelligent Surfaces (RIS) brings significant advancements in communication systems, such as increased capacity, communication secrecy, and novel physical layer-based authentication schemes. Besides Radio Frequency (RF) communications, RISs also benefit Visible Light Communications (VLC). Indeed, recent results showed that it is possible to increase the communication secrecy of VLC systems by leveraging RISs and their reconfiguration capabilities. However, no solution exists to authenticate a transmitter at the physical layer in VLC systems. Despite the existence of RIS-based Physical Layer Authentication (PLA) schemes in the RF domain, the geometrical behaviour of VLC channels renders these solutions not trivially portable to VLC systems. This paper proposes the first physical layer-based authentication scheme for VLC systems. The legitimate transmitter leverages a time-slotted communication to send a certain number of pre-agreed challenges to the receiver. Although an attacker might be able to replicate some of these challenges, the probability of correctly guessing all of them is very low. As an enabling component of our scheme, we propose the novel Distributed RIS (DRIS) concept, i.e., a RIS whose Reflecting Elements (RE) are spread over a wider area than traditional RISs. Thanks to DRIS, we increase the spatial diversity of the VLC channel model available at the transmitter, breaking hence the limits imposed by the symmetries of the widely accepted geometrical VLC channel. We validate our scheme via numerical simulations and compare our results with those obtained with a similar scheme implemented with a conventional RIS. We show that thanks to Distributed Reflective Intelligent Surface (DRIS), we achieve 10–
³
probability of false alarm and 10–
¹
probability of misdetection with and Signal-to-Noise Ratio (SNR) of 10dB, while classical Reflective Intelligent Surface (RIS) achieve 0.7 probability of false alarm and 0.5 probability of misdetection for the same SNR value. DRIS represents a significant improvement for physical layer-based authentication schemes in VLC, paving the way for further research on the subject.