WhiteNet: Robust Identification of Overlapping IEEE 802.11 Signals Across Unseen Channels

2026-08-06

AI Summary

This work addresses the significant performance degradation of existing deep learning models in recognizing overlapping IEEE 802.11 signals when training and deployment channel conditions differ. To mitigate this issue, the authors propose a physics-informed spectral whitening method combined with a synthetic overlapping signal generator based on realistic channel models for pretraining. This approach effectively suppresses frequency-selective fading and substantially enhances model generalization under unseen channel conditions. By reducing reliance on multi-transmitter real-world measurements, the method markedly narrows the accuracy gap caused by channel mismatch on public datasets. Moreover, it achieves a model size only 1/7.7 that of the current state-of-the-art while enabling knowledge distillation to lightweight edge-device models, thereby unifying high robustness with low computational complexity.

Abstract

Deep learning (DL) classifiers trained on I/Q samples achieve high accuracy for IEEE 802.11 protocol identification of overlapping signals, but their performance degrades sharply when channel conditions at deployment differ from those encountered during training. We present WhiteNet, a framework that addresses the problem of channel variability in I/Q samples. The central idea is spectral whitening, a physics-grounded preprocessing step that suppresses frequency-selective fading while preserving protocol-discriminative features. To reduce dependence on costly multi-transmitter over-the-air captures for training, we complement it with a synthetic overlap mixer featuring a physically accurate per-transmitter channel and shared-receiver signal chain for pre-training without extensive field data collection. On public over-the-air IEEE 802.11 data, WhiteNet closes a substantial portion of the accuracy gap caused by unseen channel conditions while using 7.7 times fewer parameters than the prior state of the art, and optionally distills to compact variants for coarse spectrum awareness on power-constrained edge devices.

Problem

• channel variability
• IEEE 802.11
• overlapping signals
• I/Q samples
• protocol identification

Innovation

• spectral whitening
• channel robustness
• synthetic overlap mixer
• I/Q sample classification
• model distillation