Integrating WiFi communication with vital signs sensing, particularly under the emerging 802.11bf standard, offers a promising pathway toward next-generation authentication systems, providing non-intrusive, continuous, and scalable security. However, current WiFi-based respiration authentication approaches are limited to single-user scenarios, whereas existing multi-user respiration monitoring technologies either lack identity awareness or require users to remain completely stationary. In this paper, we present MURAL-Fi, the first system to achieve multi-user respiration authentication using WiFi, addressing three critical requirements: multi-user support, identity-aware capability, and involuntary-motion tolerance. Unlike conventional methods that monitor the movement of the highest-energy point within the respiration frequency band of the WiFi spectrum, MURAL-Fi introduces a novel baseline region variation exploration technique. This technique identifies a spatial region of points corresponding to the chest cavities and extracts respiration signals by analyzing strength variations within these regions. Specifically, we iteratively employ baseline region identification coupled with interference suppression utilizing spatial spectral information derived from WiFi signals. This iterative refinement progressively enhances the precision of chest-cavity localization and robustly mitigates various interferences, enabling identity-aware respiration pattern extraction despite involuntary body movements. Subsequently, we perform multi-level feature analysis to derive user-specific respiratory characteristics and utilize a triplet loss-based Siamese Neural Network to authenticate users. Through extensive experiments with 24 participants across diverse environments, we demonstrate that MURAL-Fi achieves high authentication accuracy, maintaining superior performance across challenging scenarios, including resistance to mimic attacks.
Li et al. (Mon,) studied this question.