Managing Self-Interference Cancellation and IoT Connectivity in Full-Duplex Massive MIMO Networks

In this paper, we consider the sum-rate maximization of a single cell full-duplex (FD) massive MIMO network. The network has a single FD base station (BS) with multiple half-duplex (HD) downlink (DL) users, uplink (UL) users, and Internet-of-Things (IoT) DL users. The BS utilizes zero-forcing (ZF) precoding and a ZF receiving. In the ZF precoding, in the DL transmission, the precoder cancels the multi-user interference between the DL transmissions and the IoT DL transmission. Additionally, the ZF precoder will help in reducing the self-interference (SI) on some of the BS receive (RX) antennas. We start by analyzing the asymptotic sum-rate behavior when the number of antennas grows to infinity. Accordingly, we derive lower bounds for the DL, UL, and IoT-DL rates. Afterwards, we study the trade-off between SI cancellation via ZF transmit-beamforming and IoT connectivity, i.e., we try to find the best utilization of the BS transmit (TX) antennas. Accordingly, a rate-gain maximization problem is formulated to maximize the gain from the massive TX antennas. The problem is proved to be a convex optimization problem that can be solved efficiently. Finally, in the numerical results, we validate the derived bounds as well as the proposed rate-gain maximization solution.