MoEformer: An Input-Adaptive Multi-Scale Mixture-of-Experts Transformer for Automatic Modulation Recognition
Sensors · 07.11.2026 · 26 (21) · 7045
DOI: 10.3390/s26217045 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 07.11.2026. Verfasst von Jiale Wang und 2 weiteren Autor:innen. Beteiligte Einrichtung: School of Information and Communication Engineering, Beijing Information Science and Technology University, Beijing 100101, China. Die Arbeit nennt 50 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Automatic modulation recognition (AMR) is essential for spectrum sensing and non-cooperative wireless monitoring. Although multi-scale temporal representations provide complementary information from received in-phase and quadrature (I/Q) signals, their fusion faces two challenges: mismatched feature sequence lengths and the inability of fixed fusion rules to adapt scale contributions to individual inputs. To address these challenges, we propose MoEformer, a multi-scale Mixture-of-Experts Transformer with input adaptability for AMR from raw I/Q sequences. Parallel experts extract scale-specific features from temporally resampled inputs. Frequency-domain feature alignment maps the expert features to a common sequence length, while a learned gating network assigns sample-dependent weights before Transformer-based temporal modeling and attention pooling. Experiments on RadioML2016.10a, RadioML2016.10b, and RadioML2018.01A achieve overall classification accuracies of 63.74%, 66.24%, and 64.22%, respectively, outperforming all compared baselines. For SNR from −20 to 0 dB, MoEformer exceeds the strongest baseline on each dataset by 1.07, 1.56, and 1.43 percentage points, respectively. Ablation studies support the contributions of adaptive expert weighting and frequency-domain feature alignment, while the additional inference overhead highlights the trade-off between recognition accuracy and computational efficiency.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Information and Communication Engineering, Beijing Information Science and Technology University, Beijing 100101, China.
- Artificial Intelligence Institute of China Electronics Technology Group Corporation, Beijing 100041, China.
ORCID-Profile: Jiale Wang, Wupeng Xie, Yaxin Mu
Zitieren
APA
Wang, J., Xie, W., & Mu, Y. (2026). MoEformer: An Input-Adaptive Multi-Scale Mixture-of-Experts Transformer for Automatic Modulation Recognition. Sensors, 26(21), 7045. https://doi.org/10.3390/s26217045
Vancouver
Wang J, Xie W, Mu Y. MoEformer: An Input-Adaptive Multi-Scale Mixture-of-Experts Transformer for Automatic Modulation Recognition. Sensors. 2026;26(21):7045. doi:10.3390/s26217045
BibTeX
@article{Wang2026,
author = {Wang, Jiale and Xie, Wupeng and Mu, Yaxin},
title = {MoEformer: An Input-Adaptive Multi-Scale Mixture-of-Experts Transformer for Automatic Modulation Recognition},
journal = {Sensors},
year = {2026},
volume = {26},
number = {21},
pages = {7045},
doi = {10.3390/s26217045}
}
EUVIMED