Channel Modeling Aided Dataset Generation for AI-Enabled CSI Feedback
收藏资源简介:
本数据集由IEEE支持创建,专注于AI辅助的CSI反馈中的信道建模数据增强。数据集通过有限的现场信道数据提取主要随机参数,并利用这些参数更新TR 38.901模型参数,从而生成新的信道模型数据集。数据集包含800条样本,主要用于优化数据收集和增强,旨在减少数据收集开销和增强模型泛化能力。该数据集的应用领域主要是在大规模MIMO系统中,通过AI模型改进信道状态信息的反馈效率和准确性。
This dataset was developed with the support of IEEE, focusing on data augmentation for channel modeling in AI-assisted CSI feedback. The dataset extracts core stochastic parameters from limited on-site channel measurement data, and updates the parameters of the TR 38.901 model using these parameters to generate a new channel model dataset. Comprising 800 samples, this dataset is primarily used for optimizing data collection and augmentation, with the goal of reducing data collection overhead and enhancing model generalization capability. Its main application scenario is massive MIMO systems, where AI models are employed to improve the feedback efficiency and accuracy of channel state information.




