The paper introduces a finite-rate deep-learning-based channel state information (CSI) feedback method for massive MIMO systems. It leverages vector quantization to provide a finite-bit representation of the latent vector, reducing computational complexity. The proposed method improves CSI reconstruction performance while minimizing feedback overhead. By separating the shape and gain of the latent vector, the approach significantly reduces computational complexity compared to conventional methods. Multi-rate codebook design enhances performance under varying feedback overheads.
To Another Language
from source content
arxiv.org
Djupare frågor