Vector Quantized-Aided XL-MIMO CSI Feedback with Channel Adaptive Transmission

TL;DR

This work tackles the high CSI feedback overhead in 6G XL-MIMO under near-field spherical wave propagation by introducing VQ-DJSCC-F, a digital joint source-channel coding framework that exploits polar-delay domain sparsity and discrete latent representations. It fuses Transformer- and CNN-based encoders with vector quantization, a codebook updated via EMA, and an entropy regularization to prevent codeword collapse, augmented by an SNR-driven attention mechanism for robust end-to-end transmission. Key contributions include energy-focused dimensionality reduction, a dual-backbone VQ-VAE design, EMA- and entropy-based codebook optimization, and SNR-aware channel adaptation, enabling stable performance across dynamic channels. Empirical results show superior CSI reconstruction accuracy and lower feedback overhead compared to SSCC baselines, highlighting its practical potential for robust, low-overhead DL-enabled CSI feedback in 6G XL-MIMO systems.

Abstract

Efficient channel state information (CSI) feedback is critical for 6G extremely large-scale multiple-input multiple-output (XL-MIMO) systems to mitigate channel interference. However, the massive antenna scale imposes a severe burden on feedback overhead. Meanwhile, existing quantized feedback methods face dual challenges of limited quantization precision and insufficient channel robustness when compressing high-dimensional channel features into discrete symbols. To reduce these gaps, guided by the deep joint source-channel coding (DJSCC) framework, we propose a vector quantized (VQ)-aided scheme for CSI feedback in XL-MIMO systems considering the near-field effect, named VQ-DJSCC-F. Firstly, taking advantage of the sparsity of near-field channels in the polar-delay domain, we extract energy-concentrated features to reduce dimensionality. Then, we simultaneously design the Transformer and CNN (convolutional neural network) architectures as the backbones to hierarchically extract CSI features, followed by VQ modules projecting features into a discrete latent space. The entropy loss regularization in synergy with an exponential moving average (EMA) update strategy is introduced to maximize quantization precision. Furthermore, we develop an attention mechanism-driven channel adaptation module to mitigate the impact of wireless channel fading on the transmission of index sequences. Simulation results demonstrate that the proposed scheme achieves superior CSI reconstruction accuracy with lower feedback overheads under varying channel conditions.

Figures (4)

• Figure 1: The end-to-end framework of the proposed VQ-DJSCC-F.
• Figure 2: The Transformer-based architecture of the VQ-DJSCC-F.
• Figure 3: Channel adaptive performance of VQ-DJSCC-F.
• Figure 4: Performance of VQ-DJSCC-F under different methods.