Digital Twin Aided Massive MIMO: CSI Compression and Feedback

TL;DR

Results show that a DL model trained solely on the digital twin data can achieve high performance when tested in a real-world deployment and requires orders of magnitude less real-world data to approach the same performance of the model trained completely on a real-world CSI dataset.

Abstract

Deep learning (DL) approaches have demonstrated high performance in compressing and reconstructing the channel state information (CSI) and reducing the CSI feedback overhead in massive MIMO systems. One key challenge, however, with the DL approaches is the demand for extensive training data. Collecting this real-world CSI data incurs significant overhead that hinders the DL approaches from scaling to a large number of communication sites. To address this challenge, we propose a novel direction that utilizes site-specific \textit{digital twins} to aid the training of DL models. The proposed digital twin approach generates site-specific synthetic CSI data from the EM 3D model and ray tracing, which can then be used to train the DL model without real-world data collection. To further improve the performance, we adopt online data selection to refine the DL model training with a small real-world CSI dataset. Results show that a DL model trained solely on the digital twin data can achieve high performance when tested in a real-world deployment. Further, leveraging domain adaptation techniques, the proposed approach requires orders of magnitude less real-world data to approach the same performance of the model trained completely on a real-world CSI dataset.

Figures (5)

• Figure 1: This figure shows the key idea of utilizing a digital twin to train the DL CSI model. A small amount of real-world data can then be used to compensate for the mismatch between the digital twin and real-world CSI data, and refine the DL model to achieve higher performance.
• Figure 2: This figure shows the geometry layout of the target scenario, which reflects a real-world section in Boston City. The BS is located in a vertical street, and the service area is annotated in blue.
• Figure 3: This figure shows the geometry layout of the baseline scenario, which adopts the O1 scenario from the DeepMIMO dataset deepmimo. We set "BS 6" as the BS, and activate rows 1150 to 1650 as the service area (annotated in blue).
• Figure 4: This figure shows the NMSE performance of the direct generalization and three model refinement approaches. All NMSE performance is evaluated on the target data unseen in the training and refining.
• Figure 5: This figure shows the empirical CDF of the normalized correlation between the target and digital CSI matrices.