Spatio-Temporal Representation with Deep Neural Recurrent Network in MIMO CSI Feedback
This work addresses a domain-specific problem in wireless communication systems, offering an incremental improvement over existing deep learning methods for CSI feedback compression.
The paper tackles the challenge of compressing large amounts of channel state information (CSI) in massive MIMO systems for feedback transmission, proposing a deep learning-based approach that uses a recurrent neural network and depthwise separable convolution to improve recovery quality and accuracy, achieving remarkable robustness at low compression ratios.
In multiple-input multiple-output (MIMO) systems, it is crucial of utilizing the available channel state information (CSI) at the transmitter for precoding to improve the performance of frequency division duplex (FDD) networks. One of the mainchallenges is to compress a large amount of CSI in CSI feedback transmission in massive MIMO systems. In this paper, we propose a deep learning (DL)-based approach that uses a deep recurrent neural network (RNN) to learn temporal correlation and adopts depthwise separable convolution to shrink the model. The feature extraction module is also elaborately devised by studyingdecoupled spatio-temporal feature representations in different structures. Experimental results demonstrate that the proposed approach outperforms existing DL-based methods in terms of recovery quality and accuracy, which can also achieve remarkable robustness at low compression ratio (CR).