 This paper proposes a novel approach to optimize the waveform design of MIMO radars operating in multi-path environments. The authors leveraged the power of deep learning to construct a deep residual network which converts the CM constraint into a phase sequence optimization problem. They then use Adam as an optimizer to train the network, resulting in improved SNR values compared to existing methods. This article was authored by Zixiang Zheng, Yu Zhang, Xiang Yupeng, and others.