Signal Detection Of MIMO System With One bit ADCs Based On Deep Learning

The massive multiple-input multiple-output (MIMO) with one-bit analog-to-digital converters (ADCs) is a promising technique for the new generation of green mobile communications, which effectively reduces the power consumption and complexity of the massive MIMO communication system. But the performance loss of the massive MIMO system using conventional detection algorithms is severe because of the one-bit ADCs. Since deep learning has great performance advantages in dealing with nonlinear problems, we propose a deep learning (DL)-based detection algorithm and design a new loss function. but there is still a certain gap between DL-based detector and ML detector, for this reason, the result of the DL-based detector is further optimized based on the idea of Hamming distance. The simulation results show that the DL-based detector and two-stage detector is better than many other algorithms, and the proposed loss function is better than MSELoss in this task.