Beam-Response Contrastive Learning for Transmitter-Side MIMO CSI Representation

2026/07/27 by Sehyun Ryu, Yumin Kim, Minjae Lee +2
Computer Science · Engineering · Mathematics · #cs.IT #cs.SY #eess.SY #math.IT

paper · pdf

Abstract

Self-supervised representation learning from unlabeled channel state information (CSI) can reduce labeling and adaptation overhead in learning-based multiple-input multiple-output (MIMO) systems. Existing CSI pretraining methods typically use reconstruction objectives or contrastive pairs from generic augmentations, which do not explicitly reflect transmission-relevant channel similarity. This paper proposes beam-response contrastive learning (BRCL), a self-supervised CSI pretraining framework based on the transmit-side Gram matrix. For a channel matrix H, R=H^HH determines the received power of any unit-norm transmit beam w as |Hw|₂²=w^HRw. BRCL maps each CSI sample to a beam-response profile and uses the induced soft similarity as a label-free relational target for contrastive pretraining. Combined with reconstruction learning, BRCL enforces both sample-level CSI recovery and beam-response-level consistency, yielding transferable CSI representations without task-specific labels or manual positive pairs. Experiments on diverse MIMO channel datasets show that BRCL improves label efficiency and outperforms autoencoder- and channel-charting-based pretraining across beam selection, user selection, and future beam selection tasks.

Citations

• SSNet: Flexible and robust channel extrapolation for fluid antenna systems enabled by an self-supervised learning framework
• WiFo-CF: Wireless Foundation Model for CSI Feedback
• Standards-Compliant DM-RS Allocation via Temporal Channel Prediction for Massive MIMO Systems
• A Multi-Task Foundation Model for Wireless Channel Representation Using Contrastive and Masked Autoencoder Learning
• 6G WavesFM: A Foundation Model for Sensing, Communication, and Localization
• A MIMO Wireless Channel Foundation Model via CIR-CSI Consistency
• Building 6G Radio Foundation Models with Transformer Architectures
• Radio Foundation Models: Pre-training Transformers for 5G-based Indoor Localization
• Learning Latent Wireless Dynamics from Channel State Information
• Autosen: improving automatic wifi human sensing through cross-modal autoencoder
• Wireless Channel Charting: Theory, Practice, and Applications
• Self-Supervised and Invariant Representations for Wireless Localization
• CRONOS: Colorization and Contrastive Learning for Device-Free NLoS Human Presence Detection using Wi-Fi CSI
• AutoFi: Towards Automatic WiFi Human Sensing via Geometric Self-Supervised Learning
• Masked Autoencoders Are Scalable Vision Learners
• BEiT: BERT Pre-Training of Image Transformers
• Efficient channel charting via phase-insensitive distance computation
• Triplet-Based Wireless Channel Charting: Architecture and Experiments
• A Simple Framework for Contrastive Learning of Visual Representations
• Momentum Contrast for Unsupervised Visual Representation Learning
• DeepMIMO: A Generic Deep Learning Dataset for Millimeter Wave and Massive MIMO Applications
• Channel Charting: Locating Users within the Radio Environment using Channel State Information
• Representation Learning with Contrastive Predictive Coding
• Unsupervised Representation Learning by Predicting Image Rotations
• Deep Learning for Massive MIMO CSI Feedback
• Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles
• Unsupervised Visual Representation Learning by Context Prediction
• Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
• Joint Spatial Division and Multiplexing
• Joint Spatial Division and Multiplexing—The Large-Scale Array Regime
• BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

Related

• A MIMO Wireless Channel Foundation Model via CIR-CSI Consistency
• Generative Decoding of Compressed CSI for MIMO Precoding Design
• Exploiting Partial FDD Reciprocity for Beam Based Pilot Precoding and CSI Feedback in Deep Learning
• MIMO Channel as a Neural Function: Implicit Neural Representations for Extreme CSI Compression in Massive MIMO Systems
• Self-supervised Contrastive Learning for 6G UM-MIMO THz Communications: Improving Robustness Under Imperfect CSI