RA-LWLM uses retrieval from per-scene databases and in-context learning with a frozen foundation model to achieve cross-scene wireless localization without retraining.
CSI2Vec: Towards a universal CSI feature representation for positioning and channel charting,
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
eess.SP 3years
2026 3verdicts
UNVERDICTED 3representative citing papers
CSI-CLIP++ uses CSI-CIR contrastive alignment to pretrain a channel encoder that improves beam prediction by up to 19.31 percentage points and supports positioning on DeepMIMO data across environments.
LatentWave applies JEPA pretraining to wireless data for more transferable representations than masked reconstruction baselines, with evaluations on RF classification, 5G positioning, beam prediction, and LoS/NLoS tasks.
citing papers explorer
-
RA-LWLM: Retrieval-Augmented In-Context Localization with Wireless Foundation Models
RA-LWLM uses retrieval from per-scene databases and in-context learning with a frozen foundation model to achieve cross-scene wireless localization without retraining.
-
CSI-CLIP++: A Scalable Channel Foundation Model for Wireless Communication via CIR-CSI Consistency
CSI-CLIP++ uses CSI-CIR contrastive alignment to pretrain a channel encoder that improves beam prediction by up to 19.31 percentage points and supports positioning on DeepMIMO data across environments.
-
LatentWave: JEPA Pretraining for Wireless Foundation Models
LatentWave applies JEPA pretraining to wireless data for more transferable representations than masked reconstruction baselines, with evaluations on RF classification, 5G positioning, beam prediction, and LoS/NLoS tasks.