报告人：Gan Zheng

时间：2025年1月18日 11:00-12:00

地点：A8厚德楼-110

嘉宾简介

      Gan Zheng，现任英国拉夫堡大学客座教授，IEEE Fellow，被斯坦福大学列入世界前 2% 的顶尖科学家。他的研究方向包括智能无线通信、智能超表面、无人机通信、以及边缘计算等。Gan Zheng教授曾获2013年IEEE Signal Processing Letters最佳论文奖、2015年IEEE GLOBECOM最佳论文奖、以及2018年IEEE通信学会绿色通信与计算技术委员会颁发的最佳论文奖。目前，他是IEEE Wireless Communications Letters和IEEE Transactions on Communications 的副编辑。

报告简介

The multiuser MISO downlink transmission with channel uncertainties presents significant challenges in optimization, especially when addressing robustness against Gaussian channel uncertainties. Traditional convex-based optimization methods, while tractable, suffer from high computational complexity, rank relaxation, and conservative solutions. In this talk, I will introduce  an unsupervised deep learning-based approach that incorporates the sampling of channel uncertainties in the training process to optimize the probabilistic system performance. Our method incorporates unsupervised learning, sampling channel uncertainties during training to enhance system performance. We propose a model-driven learning framework, with a new beamforming structure and trainable parameters, coupled with a graph neural network for efficient parameter inference. I will demonstrate how this approach is applied to the minimum rate quantile maximization problem subject to outage and total power constraints and solve power minimization problems with probabilistic rate constraints, achieving superior performance over state-of-the-art methods.