Large System Performance and Distributed Scheme of Downlink Beamforming in F-RANs With Distributed Antennas

This paper addresses the performance of downlink beamforming in fog radio access networks (F-RANs) with distributed antennas. The wireless channel is modeled as Rician fading. We analyze the asymptotic behavior of the signal-to-interference-plus-noise ratio (SINR) and the data rate of maximum ratio transmission (MRT) and regularized zero-forcing (RZF) beamformers in the large system limit, where the numbers of antennas and users go to infinity while keeping a fixed ratio. In some practical F-RAN environment, it is difficult to perform centralized processing and share channel state information (CSI) globally, so distributed beamforming schemes are of interest. Therefore, we propose the local zero-forcing (ZF) beamformer that relies only on local CSI. The simulation results show that the asymptotic results of MRT and RZF are accurate for finite system size, and the local ZF beamformer provides a tradeoff between MRT and RZF in terms of CSI sharing complexity and interference control ability.