Downlink Precoding for Multiple Users in FDD Massive MIMO Without CSI Feedback

Massive MIMO can provide downlink access to multiple user equipments (UEs) through appropriate precoding or beamforming. To obtain precoding matrices for users, channel state information at the transmitter (CSI-T) is usually mandatory, requiring downlink training and CSI feedback at least in the frequency division duplex mode. However, such training is typically considered impractical because of the considerable amount of pilot signals and feedback overhead. In this paper, we propose downlink precoding methods that do not require UEs to generate feedback CSI for massive MIMO systems with uniform linear arrays. By recognizing the similarity between uplink and downlink channels, the base station is assumed to have partial knowledge on downlink channels (more specifically, the angles of departure of the major propagation paths of each user). Using such partial channel knowledge, we propose two precoding design methods based on robust beamforming and the design of a spatial-domain optimum finite impulse response filter. The simulation results demonstrate that the proposed method achieves a sum rate near that of a feedback-based precoding method with ideal CSI-T. In contrast to an alternative method based on beamspace division, the numerical results also display the performance advantage of the proposed method.