Reliable Beamspace Channel Estimation for Millimeter-Wave Massive MIMO Systems with Lens Antenna Array

摘要

Millimeter-wave massive MIMO with lens antenna array can considerably reducethe number of required radio-frequency (RF) chains by beam selection. However,beam selection requires the base station to acquire the accurate information ofbeamspace channel. This is a challenging task, as the size of beamspace channelis large while the number of RF chains is limited. In this paper, weinvestigate the beamspace channel estimation problem in mmWave massive MIMOsystems with lens antenna array. Specifically, we first design an adaptiveselecting network for mmWave massive MIMO systems with lens antenna array, andbased on this network, we further formulate the beamspace channel estimationproblem as a sparse signal recovery problem. Then, by fully utilizing thestructural characteristics of mmWave beamspace channel, we propose a supportdetection (SD)-based channel estimation scheme with reliable performance andlow pilot overhead. Finally, the performance and complexity analyses areprovided to prove that the proposed SD-based channel estimation scheme canestimate the support of sparse beamspace channel with comparable or higheraccuracy than conventional schemes. Simulation results verify that the proposedSD-based channel estimation scheme outperforms conventional schemes and enjoyssatisfying accuracy, even in the low SNR region as the structuralcharacteristics of beamspace channel can be exploited.