Modulation classification model adapts to real air signals with fine tuning
Sub-6 GHz Over-the-Air AMC via Curriculum Fine-Tuned CNN-Transformers
Machine Learning
Summary
Models that identify how radio signals are modulated often work well in perfect lab setups but struggle with real-world conditions like signal loss and antenna misalignment. The researchers tested a hybrid CNN-Transformer model trained on neat data at 915 MHz, then fine-tuned it using real over-the-air signals at 4 GHz. They adjusted the model gradually using signals received at different distances and antenna alignments, achieving about 92-94% accuracy. Their detailed analysis shows how the model handles real signal imperfections but also honestly discusses the limits of their approach.
Automatic modulation classificationCNN-TransformerCurriculum fine-tuningOver-the-air signalsSub-6 GHzAntenna alignmentRF theoryPath lossSignal modulationMachine learning
Authors
Nurettin Safak, Muhammet Sefa Demirel, Alperen Marasli, Taha Eren Atmaca, Durdu Can Yerdeyatar, Ozgun Ersoy
Abstract
Automatic modulation classification (AMC) models are frequently trained and validated on synthetic or channel-cabled data, leaving open the question of how they behave once path loss and antenna pointing error are introduced by a genuine free-space link. We report a curriculum fine-tuning study of a hybrid CNN-Transformer AMC model. The general-purpose, all-32-class dataset underlying the model was built entirely at 915 MHz, on a controlled, clock/PPS-synchronized MIMO-expansion-cable link (not spatial-multiplexing transmission); all subsequent free-space, real-hardware experimentation - the sequential fine-tuning curriculum, matched-distance evaluation, and every reported over-the-air accuracy figure - was carried out at 4 GHz, across five directional-antenna distances (25, 35, 50, 70, 75 cm) under fixed TX/RX gain. Three distances used near-ideal antenna alignment (~99%) and two used a deliberately introduced partial misalignment (~85%), fine-tuned last in the curriculum. We report matched-distance test accuracy (91.8-93.7% across all five 4 GHz conditions) and confusion-matrix analysis grounded in RF theory, and report honestly where our sequential fine-tuning order confounds cumulative link adaptation with antenna alignment, rather than overstating what the data can support.