A Robust Reconfigurable Intelligent Surface-Aided Physical Layer Authentication Scheme Under Confidentiality Constraints

Abstract

In this article, a robust reconfigurable intelligent surface (RIS)-aided physical layer authentication (PLA) scheme is proposed. The designed scheme leverages carrier frequency offset (CFO) as a hardware-based attribute for authentication and RIS’s reflective elements tunability to ensure an enhanced received signal-to-noise ratio at the authenticator in a wireless mesh network, resulting in improved authentication performance. Under the presence of a malicious eavesdropper in the network, and by leveraging semidefinite programming, the RIS phase-shifting optimization problem is solved, where a near-optimal RIS configuration maximizing the authentication performance (mitigating impersonation attacks) while fulfilling a minimal secrecy capacity (SC) value with respect to the eavesdropping attack is obtained. The obtained results show an enhanced authentication performance by the increase in the RIS size or the CFO difference between the pair of legitimate and illegitimate transmitters. Furthermore, it is shown that the proposed scheme achieves an authentication-confidentiality performance balance where the proposed scheme outperforms the benchmark one, maximizing the SC, in terms of authentication performance at the cost of a certain SC loss.