HeterCSI: Channel-Adaptive Heterogeneous CSI Pretraining Framework for Generalized Wireless Foundation Models

• URL: http://arxiv.org/abs/2601.18200v1
• Date: Mon, 26 Jan 2026 06:35:48 GMT
• Title: HeterCSI: Channel-Adaptive Heterogeneous CSI Pretraining Framework for Generalized Wireless Foundation Models
• Authors: Chenyu Zhang, Xinchen Lyu, Chenshan Ren, Shuhan Liu, Qimei Cui, Xiaofeng Tao,
• Abstract summary: HeterCSI is a channel-adaptive pretraining framework that reconciles training efficiency with robust cross-scenario generalization.<n>HeterCSI achieves superior average performance over full-shot baselines.<n>Compared to the state-of-the-art benchmark WiFo, it reduces NMSE by 7.19 dB, 4.08 dB, and 5.27 dB for CSI reconstruction, time-domain, and frequency-domain prediction, respectively.
• Score: 24.285127409979342
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Wireless foundation models promise transformative capabilities for channel state information (CSI) processing across diverse 6G network applications, yet face fundamental challenges due to the inherent dual heterogeneity of CSI across both scale and scenario dimensions. However, current pretraining approaches either constrain inputs to fixed dimensions or isolate training by scale, limiting the generalization and scalability of wireless foundation models. In this paper, we propose HeterCSI, a channel-adaptive pretraining framework that reconciles training efficiency with robust cross-scenario generalization via a new understanding of gradient dynamics in heterogeneous CSI pretraining. Our key insight reveals that CSI scale heterogeneity primarily causes destructive gradient interference, while scenario diversity actually promotes constructive gradient alignment when properly managed. Specifically, we formulate heterogeneous CSI batch construction as a partitioning optimization problem that minimizes zero-padding overhead while preserving scenario diversity. To solve this, we develop a scale-aware adaptive batching strategy that aligns CSI samples of similar scales, and design a double-masking mechanism to isolate valid signals from padding artifacts. Extensive experiments on 12 datasets demonstrate that HeterCSI establishes a generalized foundation model without scenario-specific finetuning, achieving superior average performance over full-shot baselines. Compared to the state-of-the-art zero-shot benchmark WiFo, it reduces NMSE by 7.19 dB, 4.08 dB, and 5.27 dB for CSI reconstruction, time-domain, and frequency-domain prediction, respectively. The proposed HeterCSI framework also reduces training latency by 53% compared to existing approaches while improving generalization performance by 1.53 dB on average.

Related papers

• CSI-4CAST: A Hybrid Deep Learning Model for CSI Prediction with Comprehensive Robustness and Generalization Testing [44.045995554758385]
  This paper introduces CSI-4CAST, a hybrid deep learning architecture that integrates 4 key components, i.e., Convolutional neural network residuals, Adaptive correction layers, ShuffleNet blocks, and Transformers.<n>The dataset spans multiple channel models, a wide range of delay spreads and user velocities, and diverse noise types and intensity degrees.<n> Experimental results show that CSI-4CAST achieves superior prediction accuracy with substantially lower computational cost.
  arXiv  Detail & Related papers  (2025-10-14T21:19:52Z)
• Green Learning for STAR-RIS mmWave Systems with Implicit CSI [53.03358325565645]
  Green learning (GL)-based precoding framework is proposed for simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS)-aided millimeter-wave (mmWave) broadcasting systems.<n>Motivated by the emphasis on environmental sustainability in future 6G networks, this work adopts a transmission framework for scenarios where multiple users share identical information, improving spectral efficiency and reducing redundant transmissions and power consumption.
  arXiv  Detail & Related papers  (2025-09-08T15:56:06Z)
• Distributed Gossip-GAN for Low-overhead CSI Feedback Training in FDD mMIMO-OFDM Systems [65.23921727688749]
  We propose a novel gossiping generative adversarial network (Gossip-GAN)-aided CSI feedback training framework.<n>Gossip-GAN enables the CSI feedback training with low-overhead while preserving users' privacy.
  arXiv  Detail & Related papers  (2025-08-31T07:46:16Z)
• Standards-Compliant DM-RS Allocation via Temporal Channel Prediction for Massive MIMO Systems [4.251030047034567]
  We introduce the concept of channel prediction-based reference signal allocation (CPRS)<n>CPRS jointly optimize channel prediction and DM-RS allocation to improve data throughput without requiring CSI feedback.<n>We show up to 36.60% throughput improvement over benchmark strategies.
  arXiv  Detail & Related papers  (2025-07-15T07:56:37Z)
• RGE-GS: Reward-Guided Expansive Driving Scene Reconstruction via Diffusion Priors [54.81109375939306]
  RGE-GS is a novel expansive reconstruction framework that synergizes diffusion-based generation with reward-guided Gaussian integration.<n>We propose a reward network that learns to identify and prioritize consistently generated patterns prior to reconstruction phases.<n>During the reconstruction process, we devise a differentiated training strategy that automatically adjust Gaussian optimization progress according to scene converge metrics.
  arXiv  Detail & Related papers  (2025-06-28T08:02:54Z)
• Channel Fingerprint Construction for Massive MIMO: A Deep Conditional Generative Approach [65.47969413708344]
  We introduce the concept of CF twins and design a conditional generative diffusion model (CGDM)<n>We employ a variational inference technique to derive the evidence lower bound (ELBO) for the log-marginal distribution of the observed fine-grained CF conditioned on the coarse-grained CF.<n>We show that the proposed approach exhibits significant improvement in reconstruction performance compared to the baselines.
  arXiv  Detail & Related papers  (2025-05-12T01:36:06Z)
• Lessons from Deploying Learning-based CSI Localization on a Large-Scale ISAC Platform [18.4186280784439]
  We explore the deployment of a large-scale CSI-based localization system involving over 400 Access Points (APs) in a real-world building under the Integrated Sensing and Communication (ISAC) paradigm.<n>We propose a novel CSI-based learning framework for WiFi localization, tailored for large-scale ISAC deployments on the server side.
  arXiv  Detail & Related papers  (2025-04-24T01:16:40Z)
• CSI-BERT2: A BERT-inspired Framework for Efficient CSI Prediction and Classification in Wireless Communication and Sensing [19.12026243010111]
  We propose a unified framework named CSI-BERT2 for CSI prediction and classification tasks.<n>The framework adapts BERT to capture the complex relationships among CSI sequences through a bidirectional self-attention mechanism.<n>Extensive experiments on both real-world collected and simulated datasets demonstrate that CSI-BERT2 achieves state-of-the-art performance across all tasks.
  arXiv  Detail & Related papers  (2024-12-09T06:44:04Z)
• GAQAT: gradient-adaptive quantization-aware training for domain generalization [54.31450550793485]
  We propose a novel Gradient-Adaptive Quantization-Aware Training (GAQAT) framework for DG.<n>Our approach begins by identifying the scale-gradient conflict problem in low-precision quantization.<n>Extensive experiments validate the effectiveness of the proposed GAQAT framework.
  arXiv  Detail & Related papers  (2024-12-07T06:07:21Z)
• Model Inversion Attacks Through Target-Specific Conditional Diffusion Models [54.69008212790426]
  Model inversion attacks (MIAs) aim to reconstruct private images from a target classifier's training set, thereby raising privacy concerns in AI applications. Previous GAN-based MIAs tend to suffer from inferior generative fidelity due to GAN's inherent flaws and biased optimization within latent space. We propose Diffusion-based Model Inversion (Diff-MI) attacks to alleviate these issues.
  arXiv  Detail & Related papers  (2024-07-16T06:38:49Z)
• Physics-Inspired Deep Learning Anti-Aliasing Framework in Efficient Channel State Feedback [25.68689988641748]
  This work introduces a new CSI upsampling framework at the gNB as a post-processing solution to address the gaps caused by undersampling. We also develop a learning-based method that integrates the proposed algorithm with the Iterative Shrinkage-Thresholding Algorithm Net (ISTA-Net) architecture. Our numerical results show that both our rule-based and deep learning methods significantly outperform traditional techniques and current state-of-the-art approaches in terms of performance.
  arXiv  Detail & Related papers  (2024-03-12T23:40:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.