Channel Estimation for Millimeter-Wave MIMO Communications With Lens Antenna Arrays

Millimeter-wave (mmWave) communication is a promising technology for future wireless communication systems. By utilizing the advanced lens antenna arrays, a mmWave lens multiple-input multiple-output (MIMO) system significantly reduces the number of radio-frequency (RF) chains required, and thus, achieves cost-effective implementation. In this paper, we propose a novel channel estimation scheme for the mmWave lens MIMO system with limited RF chains. The proposed scheme is applicable for the general mmWave MIMO communications, regardless of narrow-band or wide-band, time-division-duplexing or frequency-division-duplexing operations. By exploiting the energy-focusing property of lens arrays, jointly with the multipath sparsity of mmWave channels, the proposed channel estimation scheme requires only low-complexity two-way energy-based antenna selections and reduced-size MIMO channel estimation. Numerical results show that under limited RF chains and taking the channel training overhead and imperfect estimation into account, the proposed lens MIMO system achieves significantly higher throughput compared to the more complex benchmark system with the conventional uniform planar array and MIMO-orthogonal-frequency-division-multiplexing-based transmissions.