A self-supervised transformer pretrained on DeepMIMO channels produces embeddings that outperform raw channels on two downstream wireless tasks under data-limited settings.
LWM: A Pre-trained Wireless Foundation Model for Universal Feature Extraction,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IT 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Large Wireless Model (LWM): A Foundation Model for Wireless Channels
A self-supervised transformer pretrained on DeepMIMO channels produces embeddings that outperform raw channels on two downstream wireless tasks under data-limited settings.