MambaCSP: Hybrid-Attention State Space Models for Hardware-Efficient Channel State Prediction
arXiv 2604.21957•3d4150dd2a6a6b87237ad00b7c0a1a15bfd740dac8d96cf146324504c85288dd
CSI predictionGamma-distributed constellationISACMIMOMambaCSPPCRBattentionbeamformingchannel-state-informationconstellation designdiffusiongenerative modelsgrouped patternintegrated sensing and communicationslow-latencynull-space flow matchingrange estimationsecrecy outage probability','wireless secrecy','reconfigurable-@semantic HARQsemantic communicationsemantic error correctionshort block codesstate-space modelstransformerswireless
Paper metadata
- arXiv ID
- 2604.21957
- Version
- Not specified by this published record
- Category
- Mathematics — Information Theory (math.IT)
The PDF link points to arxiv.org. Baitaphish does not expose a private stored PDF.
Evidence and limitations
- Source ID
- arxiv_math_it
- Record identifier
- 3d4150dd2a6a6b87237ad00b7c0a1a15bfd740dac8d96cf146324504c85288dd
- Enrichment time
- 2026-04-27T08:51:49Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.