Coexistence of Direct and Relayed Transmission Users in Multi-Cell Massive MIMO Systems

In this paper, we investigate the uplink performance of a multi-cell massive MIMO relay network, where the base stations (BSs) and the relay stations (RSs) are equipped with large scale antennas and conventional number of antennas, respectively. In each cell, a portion of users communicate with the BS directly while others are aided by the RSs, which are called direct transmission users (DTUs) and relayed transmission users (RTUs), respectively. The two types of users and RSs of all cells share the same time-frequency resource and thus interfere with each other. After describing the system model, we investigate linear minimum mean-squared error channel estimation for the DTUs and RTUs and derive tractable expressions of channel estimation accuracy. With maximal ratio combining and zero-forcing detections based on the estimated channel state information (CSI), we derive the closed-form lower bound and approximation of the achievable rates for DTUs and RTUs to reveal their mutual impact, including intra- and inter-cell interference and pilot contamination. Then, we propose the power scaling law, which proves to be a simple and energy-efficient way to enable the coexistence of DTUs and RTUs in all cells. Furthermore, an optimal power allocation (OPA) scheme is designed based on the proposed power scaling law to maximize the system sum rate. Numerical results verify the accuracy of our theoretical analysis and the effectiveness of the OPA scheme to guarantee performance improvement and fairness simultaneously.