Multi-Cell Multi-Group Massive MIMO Multicasting: An Asymptotic Analysis

This paper considers a multi-cell multi-group massive MIMO multicasting system. It is assumed the system has L cells, where each cell has N groups of K single antenna users. The users in each group are interested to the same message and they form a multicasting group. In this paper first, we present a precoding scheme that asymptotically removes the interference and noise and provide infinite signal to interference plus noise ratio, when the channel state information is available at the base station. Second, we propose a channel state information acquisition scheme using uplink training and channel reciprocity based on time division duplexing operation. The main difference of the proposed method with the common methods in the literature is that it decreases the number of required orthogonal training sequences for channel state information acquisition from NK to N. Third, we elaborate the effect of pilot contamination on the proposed training and precoding and present the asymptotic performance of multi-cell multi- group massive MIMO Multicasting systems under pilot contamination. Also we show for these systems even in the asymptote the signal to interference plus noise is bounded by a known upper bound. Finally we revise the proposed multicasting scheme such that we can achieve an infinite signal to interference plus noise ratio in the asymptote even when accounting for channel state information acquisition and under pilot contamination.