Massive MIMO Based on Compressed Sensing Adaptive Clustering Algorithm

Abstract

Joint Spatial Division and Multiplexing (JSDM) is a recently proposed scheme to realize massive MIMO for FDD systems. This scheme is carried out using two steps: partitioning MS into groups in terms of covariances of each group; harnessing a two-stage beamforming: a precoder to mitigate inter-group interference by spatial correlation and a mutli-user precoder to counterbalance intra-group interference. Dual polarized massive MIMO aims to stave off the spatial constraint of 5G massive mimo. Diverging from pervious contributions which equally distribute MS into groups, this paper focuses mainly on an adaptive MS clustering algorithm for massive mimo JSDM scheme to optimize system performance. Simulation results show that the system outdoes the former JSDM FDD system.