Enhanced Physical Layer Authentication via Robust and Trustworthy Sensing

Angle-of-arrival (AoA) has been utilized as a reliable feature for physical layer authentication (PLA), as impersonation attacks require restrictive spatial and signal conditions. However, in configurations where the legitimate and adversary users are spatially close or aligned, AoA-based authentication may fail. This paper investigates enhanced sensing-based authentication using channel state information (CSI). We develop a phase-correction pipeline to sanitize CSI and propose antenna subarray aggregation across frequencies, drastically improving AoA estimation. Moreover, we show that by leveraging sensing parameters such as time-of-flight (ToF), received signal strength indicator (RSSI), impersonation attacks can be accurately revealed where AoA performs poorly. Our proposed scheme is validated using a real dataset collected on the Nokia campus in Stuttgart, Germany. Experimental results show that, in addition to the highly accurate AoA estimation, proximal impersonation attacks can be detected with an accuracy of 94.4-100%. These findings demonstrate that spatial and frequency-domain processing of CSI enables effective, hardware-friendly, robust PLA through attacks detection.