Resource Allocation for Massive MIMO Systems with Pilot Contamination in Spectrum Sharing Networks

In this paper, we investigate the resource allocation problem for massive multiple-input-multiple-output (mMIMO) systems with pilot contamination in spectrum sharing networks. Firstly, we propose a dynamic pilot sharing scheme for pilot transmission phase, where secondary users (SUs) are allowed to use pilots for channel estimation only when there are temporarily unused primary pilots. Based on the above scheme, we formulate the optimal power allocation problem to maximize the downlink sum rate of the SN under the limited total transmit power. Meanwhile, to guarantee the quality of service (QoS) of PUs, we define the minimum signal-interference-plus-noise-ratio (SINR) that PUs can tolerate as the constraint. Since the formed optimization problem is nonconvex, we transform the original problem into a convex optimization problem and propose an iterative update algorithm to obtain the optimal power allocation. Next, we theoretically analyze the impact of the number of the secondary base station (SBS) antennas or the primary base station (PBS) antennas on the downlink rate of the SN and PN. Finally, simulation results are conducted to show the downlink rate of the SN under different parameters.