all AI news
Data-Driven Deep Learning Based Hybrid Beamforming for Aerial Massive MIMO-OFDM Systems with Implicit CSI. (arXiv:2201.06778v1 [eess.SP] CROSS LISTED)
cs.LG updates on arXiv.org arxiv.org
In an aerial hybrid massive multiple-input multiple-output (MIMO) and
orthogonal frequency division multiplexing (OFDM) system, how to design a
spectral-efficient broadband multi-user hybrid beamforming with a limited pilot
and feedback overhead is challenging. To this end, by modeling the key
transmission modules as an end-to-end (E2E) neural network, this paper proposes
a data-driven deep learning (DL)-based unified hybrid beamforming framework for
both the time division duplex (TDD) and frequency division duplex (FDD) systems
with implicit channel state information (CSI). For …
arxiv data data-driven deep learning hybrid learning systems