Deep learning based user scheduling for massive MIMO downlink system

In this paper, we investigate a user scheduling algorithm for massive multiple-input multiple-output (MIMO) systems over more general correlated Rician fading channels. To achieve low latency and high throughput, a new user scheduling algorithm based on deep learning (DL) is proposed, which exploits only statistical channel state information. The proposed scheduling network is trained to grasp the mapping from the statistical signal and interference pattern to the user scheduling decision through supervised learning. It can predict the optimal scheduling scheme from statistical CSI without iterative calculation after offline training. Simulation results demonstrate the superior performance of the proposed algorithm in terms of calculation time, and it achieves almost the same throughput as the optimal scheduling algorithm which is obtained through exhaustive search. Furthermore, with the normalization of the input data, the proposed scheduling network is robust to the change of the channel environment and the number of transmit antennas.