Real-World Wireless Channel Data Augmentation Using Adversarial Autoencoder for DL-Based Massive Mimo CSI Feedback
Xin Liang, Zhenyu Liu, Lin Zhang · 2023
In massive multiple-input multiple-output (MIMO) systems, deep learning (DL)-based channel state information (CSI) feedback can provide high downlink throughput with limited feedback overhead. However, insufficient wireless channel data in the actual situation impairs the reconstruction accuracy of CSI feedback network. In this paper, we propose an adversarial autoencoder (AAE)-based wireless channel data augmentation network named CsiAAE to improve the CSI reconstruction accuracy of DL-based CSI feedback network trained with limited dataset. Experimental results under real-world wireless channel data show that compared with traditional data augmentation methods, CsiAAE can provide higher reconstruction accuracy for the DL-based CSI feedback network trained with insufficient dataset.