Efficient Channel Prediction based on Gram-Square-Root Factorization using GMMs
This work proposes a Gaussian mixture model (GMM)-based prediction framework for MIMO-orthogonal frequency-division multiplexing channels under partial feedback using Gram-square-root factorization, and introduces an efficient parameter reduction technique that exploits structured covariance matrices, significantly low...