WiID: WiFi-based Identity Recognition Using In-Air Handwriting Gestures

WiFi-based identity recognition has garnered significant attention in recent years owing to its contact-free operation, low deployment cost, and non-intrusive nature. This paper presents a WiFi-based identity recognition system founded on a concentric circle model. The proposed system employs a single transmitter–receiver pair and models wireless signal propagation through a carefully designed spatial configuration, enabling robust user identity recognition. Specifically, signal amplitudes are first extracted from the WiFi Channel State Information (CSI). Based on these amplitudes, a multidimensional feature vector is constructed, incorporating the raw amplitude, amplitude variance, signal energy, and amplitude first-order difference, to comprehensively characterize the temporal dynamics of user interactions. The resulting feature tensor is then fed into a two-dimensional convolutional neural network (2D-CNN) for end-to-end learning, enabling automatic user identity classification. Experimental results indicate that the proposed method achieves robust identity recognition performance across various test gestures.