Lightweight PCGAE-Net: Parallel CrossGate Attention and Bottleneck AutoEncoder for Efficient 5G Channel Prediction

Paper demonstrates that IEEE 802.11 MAC address randomization used during network discovery can be defeated by ML-based fingerprinting. The authors show a passive eavesdropper can extract unencrypted probe-frame fields (hardware/spec hints) and combine them with temporal features such as inter-probe-frame arrival time (IFAT) and physical-layer reception characteristics to reliably re-identify and track devices despite randomized MACs. The work evaluates multiple eavesdropping scenarios and configurations, highlighting a practical privacy risk for users and deployment-level implications for Wi‑