Marginal Deep Architectures Based Massive MIMO Beam Forming and Predictions

Abstract

The increasing demands in 5G technology is fulfilled by MIMO systems which employ hundreds of antennas to serve the broadband terminals. Deep learning algorithms are powerful machine learning algorithms capable of advanced learning and predictions. Deep learning is largely employed in MIMO. This paper explores the usage of deep learning architectures in beam forming and predictions in MIMO. Specifically, an improvised version of
deep learning is the marginal deep architecture. This paper also briefs about the usage of marginal deep architecture based methods for beam forming and predictions. Such proposed models will greatly improve the efficiency and capability of beam forming in MIMO resulting in improved 5G services.