analiza aiqbench
This research investigates the robustness of hybrid CNN-Transformer models for automatic modulation classification in real-world free-space scenarios, specifically addressing the gap between synthetic training data and physical link impairments like path loss and antenna misalignment. By employing a curriculum fine-tuning approach at 4 GHz, the study demonstrates that sequential adaptation to varying distances and alignment errors can achieve high accuracy while honestly quantifying the confounding effects of its experimental design on cumulative link performance.
Sub-6 GHz Over-the-Air AMC via Curriculum Fine-Tuned CNN-Transformers
arXiv:2609.07726v1 Announce Type: cross 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.
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.
Accepted and presented at IDAP 2026. 5 pages, 8 figures
Signal Processing (eess.SP); Machine Learning (cs.LG)
arXiv:2609.07726 [eess.SP]
(or arXiv:2609.07726v1 [eess.SP] for this version)
https://doi.org/10.48550/arXiv.2609.07726 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 280 cuvinte
Subiecte conexe
Small Language Models with Hugging Face transformers Library + smolLM3
KDnuggets · 7 aug. 2026
Iris-3B: Going Beyond the Latent with Pixel-Space Diffusion Training, Conversion and Fine-Tuning
arXiv · 8 oct. 2026
Deep recurrent models are less robustly CoT-monitorable than normal CoT models in a toy setting
LessWrong · 18 sept. 2026
[AINews] Muse Spark 1.3 matches GPT-5.6-Sol, confirming Meta Superintelligence as the newest Frontier Lab, >90% discount for training
Latent Space · 3 sept. 2026