QoS guaranteed schedule for large-scale MIMO systems with pilot reuse

The large-scale multiple-input multiple-output (MIMO) system achieves high spectral efficiency, and its performance is ultimately limited by pilot contamination. Since the channel energy usually concentrates in a significant angle region, some mean-square-error (MSE)-based schedule methods have been proposed to mitigate the contamination, but these methods cannot guarantee the quality of service (QoS) for each served terminal. In order to overcome this deficiency, we derive an asymptotic achievable rate for each served terminal under the spatially correlated channel. Further, we analyze how the achievable rate is affected by the degree of channel overlap in the angle region. Based on the analysis result, we design a low-complexity schedule algorithm to guarantee the QoS and find the optimal number of served terminals. Numerical results show that the proposed schedule algorithm outperforms the MSE-based one in terms of QoS and fairness, and can maximize the number of served terminals.