In this paper, we propose a model-driven generative adversarial network (GAN)-based channel modeling framework to autonomously learn the reflected channel distribution, without complex theoretical ...