Anchor-CKM reconstructs channel knowledge maps from sparse irregular measurements by constructing a pilot-supported representation with partial convolutions followed by layout-conditioned Fourier refinement, achieving 0.79-1.33 dB RMSE reduction versus baselines on DeepMIMO scenarios.
hub
DeepMIMO: A Generic Deep Learning Dataset for Millimeter Wave and Massive MIMO Applications
22 Pith papers cite this work. Polarity classification is still indexing.
22
Pith papers citing it
abstract
Machine learning tools are finding interesting applications in millimeter wave (mmWave) and massive MIMO systems. This is mainly thanks to their powerful capabilities in learning unknown models and tackling hard optimization problems. To advance the machine learning research in mmWave/massive MIMO, however, there is a need for a common dataset. This dataset can be used to evaluate the developed algorithms, reproduce the results, set benchmarks, and compare the different solutions. In this work, we introduce the DeepMIMO dataset, which is a generic dataset for mmWave/massive MIMO channels. The DeepMIMO dataset generation framework has two important features. First, the DeepMIMO channels are constructed based on accurate ray-tracing data obtained from Remcom Wireless InSite. The DeepMIMO channels, therefore, capture the dependence on the environment geometry/materials and transmitter/receiver locations, which is essential for several machine learning applications. Second, the DeepMIMO dataset is generic/parameterized as the researcher can adjust a set of system and channel parameters to tailor the generated DeepMIMO dataset for the target machine learning application. The DeepMIMO dataset can then be completely defined by the (i) the adopted ray-tracing scenario and (ii) the set of parameters, which enables the accurate definition and reproduction of the dataset. In this paper, an example DeepMIMO dataset is described based on an outdoor ray-tracing scenario of 18 base stations and more than one million users. The paper also shows how this dataset can be used in an example deep learning application of mmWave beam prediction.
hub tools
citation-role summary
dataset 2
citation-polarity summary
roles
dataset 2polarities
use dataset 2representative citing papers
A cross-fused generative beamforming method decouples 2D UPA channel sensing and uses bidirectional cross-attention plus conditional normalizing flows to generate high-gain beam candidates, reporting up to 83.6% normalized gain improvement with 93.8% less overhead in DeepMIMO simulations.
WiSER introduces a transmitter-conditioned sparse 3D scene encoder queried by a ray-corridor decoder for radiomaps and a DETR-style set decoder for variable-cardinality CIR taps, trained on co-registered ScanNet++ and Sionna data.
PrismQuant achieves near rate-distortion optimality for Gaussian-mixture sources by losslessly transmitting the mixture component label at H(C)/n bits per dimension and applying component-matched KLT plus scalar quantization, with vanishing gap to the genie-aided bound.
SAT-CTS delivers the first finite-time regret bounds for combinatorial semi-bandits with satisficing objectives, bounding satisficing regret by a constant when the threshold is realizable and yielding O((log T)^2) standard regret otherwise.
A conditional diffusion model learns beam alignment priors to guide efficient top-k sweeps, achieving Hit@1 of 0.61 and improving over deterministic baselines by 180% on simulated data.
A flow-matching 1D diffusion transformer predicts angular radio maps from building geometry, achieving distributional fidelity, per-bin accuracy, beam selection, and Bayesian localization from a single zero-shot model across 99 environments.
A new site-specific model uses 3D geometry maps and recursive UTD diffraction calculations to predict urban radio path loss and time-varying Doppler more accurately than 3GPP models, with RMSE reductions of 7.1 dB in complex NLOS cases.
An ellipsoid-guided selective refinement algorithm improves radio-map fidelity in urban wireless digital twins by prioritizing refinement of a small subset of buildings using only low-fidelity models.
RFPrompt adapts the Large Wireless Model via deep prompt tokens to improve out-of-distribution robustness in modulation classification while training only a small number of parameters.
A discrete denoising diffusion model learns from probing histories to generate promising beam candidates, yielding better SNR, lower beam-miss probability, and reduced probe regret than baselines under tight probing budgets.
Physics-intrinsic wireless foundation model enforces wave equivariance to capture electromagnetic propagation and improve robustness to unseen environments.
OmniLoc presents a geometry-aware foundation model for anchor-free user equipment localization from wireless signals that claims to generalize across diverse indoor environments.
SimART is an open platform that unifies robotics, ray tracing, and wireless tools via ROS for reproducible multimodal simulation in 6G integrated sensing and communication.
Learning-based radio map construction is taxonomized as source-aware forward prediction versus source-agnostic inverse reconstruction, spanning five neural families, optics-inspired continuous fields, and a three-level physics integration framework.
MDFCE applies mixture-of-experts and self-attention to extrapolate sub-6 GHz CSI to mmWave CSI, showing better performance and efficiency than prior methods in simulations across antenna scales and SNR levels.
A pretrained wireless foundation model claims to unify zero-shot channel reconstruction and few-shot adaptation across 9 CSI tasks, outperforming task-specific supervised baselines.
A Transformer-based generative model builds an environment-aware channel knowledge base that is injected into JSCC encoders and decoders, achieving 10^{-3} level channel estimation error and outperforming benchmarks in semantic communication performance.
RAFC adaptively composes multi-level hidden features from wireless foundation models via a task-driven router, outperforming standard adaptation methods on four tasks with under 50K extra parameters.
A unified deep learning model predicts FR3 signal strength from FR1 data and sparse measurements to cut simulation and measurement costs in 6G networks.
Linear regression on ray-tracing data predicts 7 GHz outdoor channel coefficients with MAE 7.5155e-5 and RMSE 9.2861e-5, beating SVR and decision-tree regression.
citing papers explorer
-
Channel Knowledge Map Reconstruction From Sparse Measurements via Pilot-Anchored Layout-Conditioned Fourier Refinement
Anchor-CKM reconstructs channel knowledge maps from sparse irregular measurements by constructing a pilot-supported representation with partial convolutions followed by layout-conditioned Fourier refinement, achieving 0.79-1.33 dB RMSE reduction versus baselines on DeepMIMO scenarios.
-
Generative Site-Specific Beamforming for UPAs via Decoupled Channel Sensing
A cross-fused generative beamforming method decouples 2D UPA channel sensing and uses bidirectional cross-attention plus conditional normalizing flows to generate high-gain beam candidates, reporting up to 83.6% normalized gain improvement with 93.8% less overhead in DeepMIMO simulations.
-
WiSER: A Wireless Scene Encoder for Geometry-Grounded Multi-View Wireless Prediction
WiSER introduces a transmitter-conditioned sparse 3D scene encoder queried by a ray-corridor decoder for radiomaps and a DETR-style set decoder for variable-cardinality CIR taps, trained on co-registered ScanNet++ and Sionna data.
-
PrismQuant: Rate-Distortion-Optimal Vector Quantization for Gaussian-Mixture Sources
PrismQuant achieves near rate-distortion optimality for Gaussian-mixture sources by losslessly transmitting the mixture component label at H(C)/n bits per dimension and applying component-matched KLT plus scalar quantization, with vanishing gap to the genie-aided bound.
-
Multi-User mmWave Beam and Rate Adaptation via Combinatorial Satisficing Bandits
SAT-CTS delivers the first finite-time regret bounds for combinatorial semi-bandits with satisficing objectives, bounding satisficing regret by a constant when the threshold is realizable and yielding O((log T)^2) standard regret otherwise.
-
Diffusion-Based Generative Priors for Efficient Beam Alignment in Directional Networks
A conditional diffusion model learns beam alignment priors to guide efficient top-k sweeps, achieving Hit@1 of 0.61 and improving over deterministic baselines by 180% on simulated data.
-
RadioDiff-v2: Generative Angular Radio Maps for Multi-Beam Selection and Localization
A flow-matching 1D diffusion transformer predicts angular radio maps from building geometry, achieving distributional fidelity, per-bin accuracy, beam selection, and Bayesian localization from a single zero-shot model across 99 environments.
-
A Geometry Map-Based Site-Specific Propagation Channel Model for Urban Scenarios
A new site-specific model uses 3D geometry maps and recursive UTD diffraction calculations to predict urban radio path loss and time-varying Doppler more accurately than 3GPP models, with RMSE reductions of 7.1 dB in complex NLOS cases.
-
Fidelity Where it Matters: Site-Specific Nonuniform Refinement for Wireless Digital Twins
An ellipsoid-guided selective refinement algorithm improves radio-map fidelity in urban wireless digital twins by prioritizing refinement of a small subset of buildings using only low-fidelity models.
-
RFPrompt: Prompt-Based Expert Adaptation of the Large Wireless Model for Modulation Classification
RFPrompt adapts the Large Wireless Model via deep prompt tokens to improve out-of-distribution robustness in modulation classification while training only a small number of parameters.
-
Discrete Diffusion for Codebook-Based Beam Candidate Generation
A discrete denoising diffusion model learns from probing histories to generate promising beam candidates, yielding better SNR, lower beam-miss probability, and reduced probe regret than baselines under tight probing budgets.
-
Physics Equivariance for Robust Generalization in Wireless Foundation Model
Physics-intrinsic wireless foundation model enforces wave equivariance to capture electromagnetic propagation and improve robustness to unseen environments.
-
OmniLoc: A Geometry-Aware Foundation Model for Anchor-Free UE Localization Across Diverse Indoor Environments
OmniLoc presents a geometry-aware foundation model for anchor-free user equipment localization from wireless signals that claims to generalize across diverse indoor environments.
-
SimART: A Unified and Open Real-world Multimodal Simulation Platform for 6G Integrated Sensing and Communication
SimART is an open platform that unifies robotics, ray tracing, and wireless tools via ROS for reproducible multimodal simulation in 6G integrated sensing and communication.
-
A Tutorial on Learning-Based Radio Map Construction: Data, Paradigms, and Physics-Awareness
Learning-based radio map construction is taxonomized as source-aware forward prediction versus source-agnostic inverse reconstruction, spanning five neural families, optics-inspired continuous fields, and a three-level physics integration framework.
-
Deep Learning-Based Channel Extrapolation for Dual-Band Massive MIMO Systems
MDFCE applies mixture-of-experts and self-attention to extrapolate sub-6 GHz CSI to mmWave CSI, showing better performance and efficiency than prior methods in simulations across antenna scales and SNR levels.
-
WiFo-2: a generalist foundation model unifies heterogeneous wireless system design
A pretrained wireless foundation model claims to unify zero-shot channel reconstruction and few-shot adaptation across 9 CSI tasks, outperforming task-specific supervised baselines.
-
Generative Channel Knowledge Base With Environmental Information for Joint Source-Channel Coding in Semantic Communications
A Transformer-based generative model builds an environment-aware channel knowledge base that is injected into JSCC encoders and decoders, achieving 10^{-3} level channel estimation error and outperforming benchmarks in semantic communication performance.
-
A Unified Adaptive Feature Composition Framework for Multi-Task Generalization in Wireless Foundation Models
RAFC adaptively composes multi-level hidden features from wireless foundation models via a task-driven router, outperforming standard adaptation methods on four tasks with under 50K extra parameters.
-
CommUNext: Deep Learning-Based Cross-Band and Multi-Directional Signal Prediction
A unified deep learning model predicts FR3 signal strength from FR1 data and sparse measurements to cut simulation and measurement costs in 6G networks.
-
Data driven approach for Outdoor Channel Prediction in 5G and Beyond
Linear regression on ray-tracing data predicts 7 GHz outdoor channel coefficients with MAE 7.5155e-5 and RMSE 9.2861e-5, beating SVR and decision-tree regression.
- WiWorld-RealData: A Real-World Multi-Modal Dataset for 6G Wireless World Models