Compressive channel estimation for millimeter wave multiuser MIMO systems via pilot reuse

Abstract

The multi-user millimeter wave (mmWave) MIMO system has aroused great interest in recent years for future fifth generation (5G) cellular networks. We investigate the problem of uplink channel estimation in mmWave systems, which is challenging due to the strong pathloss and massive antennas deployment. Taking hardware constraints into consideration, we assume that hybrid analog and digital beamforming structures are employed by both the base station (BS) and the mobile stations (MSs). To reduce the pilot overhead, a pilot reuse scheme is proposed in our paper. Then we propose a compressed sensing (CS)-based method, exploiting the sparse nature of the mmWave channel, for pilot interference elimination and joint estimation of the channels from multiple users. Furthermore, sufficient conditions for estimation error are obtained to study the performance of the CS recovery. Simulation results illustrate that the proposed method is efficient to acquire the channel state information (CSI) with a substantial reduction of training overhead.