A multi-task neural network recovers superimposed downlink CSI and uplink data sequences in FDD massive MIMO, improving CSI estimation over standalone SC while maintaining similar UL-US detection across varying SNR and PPC.
Deep learning-based CSI feed- back approach for time-varying massive MIMO channels
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.NI 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Deep Learning for CSI Feedback Based on Superimposed Coding
A multi-task neural network recovers superimposed downlink CSI and uplink data sequences in FDD massive MIMO, improving CSI estimation over standalone SC while maintaining similar UL-US detection across varying SNR and PPC.