The excessive overhead of millimeter-wave beam measurement induces a significant challenge for massive MIMO systems. To reduce the measurement overhead and improve the accuracy of beam selection, DL is employed to predict the signal quality of all beam pairs by measuring a few selected beam pairs in this chapter. The proposed beam prediction network is mainly composed of input module, residual module, and MLP module, therefore denoted as ReMBP net. Moreover, to improve the generalization ability of the proposed ReMBP net for different communication environments, the input module with a flexible input size of signal quality data is designed.