Channel PSD Analysis and Optimal Antenna Weighting for High-Mobility Massive MIMO

Yinghao Ge,Weile Zhang,Feifei Gao,Shun Zhang,Xiaoli Ma

2018-01-01

Abstract:Communications in high-mobility environments have caught a lot of attentions recently. In this paper, fast time-varying channels for massive multiple-input multiple-output (MIMO) systems are addressed. We derive the exact channel power spectrum density (PSD) for the uplink from a high-speed railway (HSR) to a base station (BS) and propose to further reduce the channel time variation via beamforming network optimization. A large-scale uniform linear array (ULA) is equipped at the HSR to separate multiple Doppler shifts in angle domain through high-resolution transmit beamforming. Each branch comprises a dominant Doppler shift, which can be compensated to suppress the channel time variation, and we derive the channel PSD and the Doppler spread to assess the residual channel time variation. Interestingly, the channel PSD can be exactly expressed as the product of a pattern function and a beam-distortion function. The former reflects the impact of array aperture and is the converted radiation pattern of ULA, while the latter depends on the configuration of beamforming directions. Inspired by the PSD analysis, we introduce a common configurable amplitudes and phases (CCAP) parameter to optimize the beamforming network, by partly removing the constant modulus quantized phase constraints of matched filter (MF) beamformers. In this way, the residual Doppler shifts can be ulteriorly suppressed, further reducing the residual channel time variation. The optimal CCAP parameter minimizing the Doppler spread is derived in a closed form. Numerical results are provided to corroborate both the channel PSD analysis and the superiority of beamforming network optimization technique.