A new AMC architecture with cross-channel self-attention and feature-preserving denoising achieves 3-14% higher accuracy than benchmarks at low-to-medium SNRs on the RML2018.01a dataset.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Cross-Validated Cross-Channel Self-Attention and Denoising for Automatic Modulation Classification
A new AMC architecture with cross-channel self-attention and feature-preserving denoising achieves 3-14% higher accuracy than benchmarks at low-to-medium SNRs on the RML2018.01a dataset.