REVIEW 55 cited by
Sionna: An Open-Source Library for Next-Generation Physical Layer Research
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Sionna: An Open-Source Library for Next-Generation Physical Layer Research
read the original abstract
Sionna is a GPU-accelerated open-source library for link-level simulations based on TensorFlow. It enables the rapid prototyping of complex communication system architectures and provides native support for the integration of neural networks. Sionna implements a wide breadth of carefully tested state-of-the-art algorithms that can be used for benchmarking and end-to-end performance evaluation. This allows researchers to focus on their research, making it more impactful and reproducible, while saving time implementing components outside their area of expertise. This white paper provides a brief introduction to Sionna, explains its design principles and features, as well as future extensions, such as integrated ray tracing and custom CUDA kernels. We believe that Sionna is a valuable tool for research on next-generation communication systems, such as 6G, and we welcome contributions from our community.
Forward citations
Cited by 55 Pith papers
-
Quantum LDPC codes with design rate 1/5 and good performance below 1000 physical qubits
A new family of rate-1/5 quantum LDPC codes built from non-abelian group symmetries approaches teraquop-region memory error rates below 1000 physical qubits under an optimistic, extrapolated benchmark.
-
MAC-Gyver: Open, Programmable, Scheduling for AI-RAN 6G Systems
An open framework turns the OpenAirInterface MAC scheduler into a pluggable dApp pipeline and pairs it with a real-time, PHY-less, 90-user L2 emulator so the same scheduling policy runs in emulation and over the air.
-
AI-RAN on NPUs: Baseband Processing Without Baseband Chips
A complete OFDM transceiver runs end-to-end over the air on a commercial edge NPU by remapping baseband operators onto dense matrix and vector engines.
-
RadioRange: An Open-Source Digital Twin-based Ranging Simulator for UWB, Wi-Fi, and 5G
RadioRange is an open-source digital twin simulator for UWB, Wi-Fi, and 5G ranging that incorporates eleven hardware impairments and is validated on real measurements.
-
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...
-
Near-Field User Location Inference From Far-Field Power Measurements
Distributed far-field sensors can infer near-field user locations from received power signatures created by location-dependent leakage in beamfocusing, supported by a Bayesian CRLB and evaluated via grid-search and de...
-
Not All Symbols Are Equal: Importance-Aware Constellation Design for Semantic Communication
A framework combining VQ-VAE, task-relevance scoring, DRL subset selection, and a learned semantic-aware constellation claims near-100% semantic protection probability and proves standard Gray-coded constellations are...
-
Inductive Latent Context Persistence: Closing the Post-Handover Cold Start in 6G Radio Access Networks
ILCP transfers 128-byte compressed latent context during RAN handovers, achieving 0% ping-pong handovers and +5.1 pp average post-HO accuracy gains versus baselines on Vienna drive-test data.
-
Planar Gaussian Splatting with Bilinear Spatial Transformer for Wireless Radiance Field Reconstruction
BiSplat-WRF applies 2D planar Gaussians rendered on angular domains plus a bilinear spatial transformer to capture electromagnetic interactions, outperforming prior NeRF and GS methods on SSIM for wireless radiance fi...
-
The AI Telco Engineer: Toward Autonomous Discovery of Wireless Communications Algorithms
An LLM-powered agentic framework autonomously designs competitive and sometimes superior explainable algorithms for wireless PHY and MAC layer tasks.
-
Diffusion Inpainting MIMO-OFDM Channels with Limited Noisy Observations
A Conditional Diffusion Transformer recovers full MIMO-OFDM channels from sparse noisy pilots, delivering over 5 dB gain versus baselines even at 1/32 pilot density and completing inference in 10 steps.
-
Equivariant Multi-agent Reinforcement Learning for Multimodal Vehicle-to-Infrastructure Systems
A self-supervised multimodal alignment step plus equivariant GNN-based MARL yields over twofold sensing accuracy and 50% performance gains in decentralized V2I rate maximization.
-
Telecom World Models: Unifying Digital Twins, Foundation Models, and Predictive Planning for 6G
Telecom World Models introduce a three-layer architecture for learned, action-conditioned, uncertainty-aware modeling of 6G network dynamics, combining digital twins and foundation models, with a network slicing proof...
-
Calibrating the Digital Twin Channel: Statistics-Consistent Sim-to-Lab Adaptation for W-Band Industrial OFDM Links
SC-SLA is a learned statistics-matching post-processor that makes Sionna ray-traced W-band CFRs match measured 92–95 GHz OFDM statistics, cutting the τrms KS gap from 0.86 to 0.050 at 95 GHz.
-
CFM-Bench: A Unified Multi-Domain, Multi-Task Benchmark for Channel Foundation Models
CFM-Bench provides six curated radio-channel datasets with leakage-resistant splits and six task groups for standardizing evaluation of channel foundation models.
-
Physics-Aware Conditional SetGAN for Spatially Consistent Multi-User TR 38.901 Channel Generation
A physics-aware SetGAN reproduces UMa/NLoS multi-user TR 38.901 channels with ~0.41 dB Wasserstein power error and <0.03 median spatial-profile deviation while delivering 3.45× generation-time speedup versus Sionna.
-
GenAI-Enhanced Digital Twins for Predictive Interference Management in Ultra-Dense Networks
cGAN-enhanced digital twins synthesize rare mmWave channels to drive worst-case ZF beamforming, yielding 5–8 dB SINR gains and 60–70% lower packet loss in Sionna simulations of ultra-dense indoor networks.
-
Semantic Communications in the THz Band
AWGN-trained DeepSC generalizes to THz fading channels with a pilot-free neural detector that outperforms MMSE equalization with perfect CSI.
-
CSI-CLIP++: A Scalable Channel Foundation Model for Wireless Communication via CIR-CSI Consistency
CSI-CLIP++ uses CSI-CIR contrastive alignment to pretrain a channel encoder that improves beam prediction by up to 19.31 percentage points and supports positioning on DeepMIMO data across environments.
-
Active Perception for Radio Map Reconstruction in Uncharted 3D Air-Ground Environments
3D-URAM decouples radio map reconstruction into Bayesian UNet recovery with uncertainty and transformer-based PPO waypoint selection, reporting over 50% error reduction in simulations and real 300x200x100m field tests.
-
RxGS: Receiver-Generalizable 3D Gaussian Splatting for Radio-Frequency Data Synthesis
RxGS delivers a single unified 3DGS model for RF synthesis that generalizes across seen and unseen receivers by freezing shared geometry and conditioning directional radiance on receiver position.
-
Resilience Characterization of AI-Native Wireless Receivers via Persistent Homology
Proposes Topological Resilience Index (TRI) via persistent homology to quantify resilience of deep learning OFDM receivers to channel shifts, claiming superior warning lead and BER reduction in simulations across ITU-...
-
Near-Field User Location Inference From Far-Field Power Measurements
Far-field power measurements enable inference of near-field user locations via location-dependent leakage, with a Bayesian CRLB and two estimators evaluated under LoS and multipath conditions.
-
Plan2Cleanse: Test-Time Backdoor Defense via Monte-Carlo Planning in Deep Reinforcement Learning
Plan2Cleanse frames RL backdoor detection as a Monte Carlo planning problem to achieve over 61 percentage point gains in trigger detection and improved win rates in competitive environments.
-
Inductive Latent Context Persistence: Closing the Post-Handover Cold Start in 6G Radio Access Networks
ILCP transfers 128-byte compressed latent context over the Xn interface to eliminate post-handover cold start in GNN-based handover decisions, reporting 0.0% ping-pong rate and +5.1 pp accuracy gain on Vienna drive-test data.
-
Optimal Transmitter Placement in Realistic Urban Environments
A submodular optimization algorithm called IA-SPA with realistic ray-tracing on urban 3D maps achieves approximately 2x mean data rate and 2-8x edge rate gains over existing base station placements.
-
Optimal Transmitter Placement in Realistic Urban Environments
A new algorithm for optimal transmitter placement in realistic urban environments using submodular optimization achieves approximately 2x higher mean data rates and 2-8x higher edge rates compared to existing deployme...
-
Diffusion-OAMP for Joint Image Compression and Wireless Transmission
Diffusion-OAMP combines a pre-trained diffusion model with the OAMP algorithm under an SNR-matching rule to enable training-free reconstruction of compressed images transmitted over noisy wireless channels.
-
Joint Clustering and Prediction of the Quality of Service in Vehicular Cellular Networks
A joint clustering and prediction method for QoS distributions in vehicular cellular networks reduces mean absolute error by 9-27% compared to local or global baselines by adapting clusters to network changes.
-
Measuring Robustness of Speech Recognition from MEG Signals Under Distribution Shift
An OTA-MIMO differentially private bidding algorithm for energy sharing converges in expectation near the generalized Nash equilibrium and can cut required artificial privacy noise by exploiting channel noise.
-
Environment-Aware Channel Inference via Cross-Modal Flow: From Multimodal Sensing to Wireless Channels
Multimodal sensing data (image, LiDAR, GPS) can be mapped to full MIMO channel matrices via a cross-modal flow matching model, achieving sub-10 dB NMSE in a simulated dynamic intersection without pilots.
-
Communicating Properties of Quantum States over Classical Noisy Channels
STT-UEP transmits classical shadow data with unequal error protection, enabling a decoder to estimate M local Pauli observables using O(9^w log(M)/epsilon^2) state copies, with basis bits coded more strongly than outc...
-
Synthetic Counterfactual Labels for Efficient Conformal Counterfactual Inference
SP-CCI augments conformal calibration sets with synthetic counterfactual labels and uses RCPS with PPI debiasing to achieve tighter prediction intervals while preserving marginal coverage guarantees.
-
Deep-OFDM: Neural Modulation for High Mobility
A CNN modulator jointly trained with a neural receiver spreads information across local time-frequency neighborhoods in OFDM, breaking QAM rotational symmetry to support sparse or zero pilots under high Doppler.
-
The S-ICDF Dataset: Sionna-Simulated Dynamic Interference Characterization and Direction Finding
S-ICDF is a public Sionna-simulated multi-antenna IQ dataset of 102 indoor GNSS interference configurations with classical and ML direction-finding baselines.
-
Large Language Model-Enhanced Multi-hop Parallel Image Semantic Communication
An LLM-guided residual-compensation link reduces distortion accumulation in multi-hop wireless image transmission with modest extra bandwidth.
-
Multi-Modal Environment-Aware Beam Management for Massive MIMO: A Geometry-Driven Virtual Base Station Framework
A geometry-driven virtual base station framework uses multi-modal LiDAR data for coarse channel reconstruction and dual-agent DRL to reduce beam training overhead and improve coordinated selection in massive MIMO.
-
Low-Overhead Receiver Design for Data-Dependent Superimposed Training via Deep Learning
A low-overhead DDST framework with mix transmission and ViT-based receiver shows performance gains in low-to-medium SNR under time-varying channels.
-
GENESIS: Harnessing AI Agents for Autonomous 6G RAN Synthesis, Research, and Testing
The paper presents GENESIS, an agentic AI framework for autonomous 6G RAN synthesis, research, and testing that converts intents into over-the-air validated solutions via composable primitives and a knowledge layer.
-
Against the Monolithic Wireless World Model: Why NextG Needs Composable and Agentic Intelligence
Wireless data lacks the self-contained tokenized substrate of text, so monolithic wireless world models are unsuitable for 6G; composable agentic systems using specialized components and explicit interfaces are the re...
-
Active Sensing with Meta-Reinforcement Learning for Emitter Localization from RF Observations
A recurrent deep-RL agent (DQN/PPO) actively senses RF with a 2x2 patch antenna and reaches 80.1% success localizing GNSS interferers in Sionna multipath simulations.
-
Active Sensing with Meta-Reinforcement Learning for Emitter Localization from RF Observations
A meta-reinforcement learning agent achieves 80.1% success in localizing RF emitters by sequentially sensing the environment with a 2x2 patch antenna in Sionna ray-tracing simulations.
-
Contextual Wireless Video Semantic Communication in MIMO-OFDM Systems
M-CVST aligns video context with MIMO subcarriers via a correlation map and applies recursive time-correlated sampling to boost semantic video performance over multi-path channels.
-
Advancing Network Digital Twin Framework for Generating Realistic Datasets
An open NDT integrating controllable vehicular mobility, Sionna RT, and ns-3 produces heterogeneous cross-layer urban wireless datasets for ML QoS research.
-
Mechanism and Communication Co-Design for Differentially Private Energy Sharing
A differentially private equilibrium-seeking algorithm for OTA MIMO-based energy sharing protects prosumers' private data while converging to near-optimal solutions with quantified accuracy loss.
-
Fronthaul-Efficient Distributed Cooperative 3D Positioning with Quantized Latent CSI Embeddings
An edge-cloud positioning network with quantized learned CSI embeddings achieves 0.48 m mean 3D error using 6.25% of the lossless fronthaul payload in a simulated urban 5G scenario.
-
On-board AI-based Channel Estimation for LEO NTNs
A compact convolution block (MDELAN) for LEO NTN uplink channel estimation reduces FLOPs roughly 29x versus LMMSE in QuaDRiGa simulations while improving MSE.
-
Towards a Joint Task-Oriented and Generative Semantic Communication Framework for 6G Networks
Proposes a dual task-oriented and generative semantic comms framework using scene graphs, ST-GNNs, and diffusion decoders over 3GPP vehicular channels, claiming 99.1% compression and improved FID scores.
-
ARIADNE: AI-RAN Informed Link Adaptation in Digital Twin Network Environments
ARIADNE uses online RL with SIONNA for MCS selection in digital twins, claiming up to 20% spectral efficiency gains over SOTA and divergence from OLLA behavior.
-
Automated Heuristic Design for Network Operations
Automated heuristic design is applied to network operations with an implementation for 5G decoding that produces LDPC heuristics on par with state-of-the-art production solutions.
-
NF-TrackLLM: Joint Prediction of UAV Trajectory and Near-Field Beam for LAE XL-MIMO Systems
NF-TrackLLM is a multi-modal GPT-2-based framework that first predicts UAV trajectories then uses them as priors for near-field beam prediction in XL-MIMO systems.
-
Against the Monolithic Wireless World Model: Why NextG Needs Composable and Agentic Intelligence
Argues that wireless data's configuration dependence and lack of self-containment make monolithic foundation models unsuitable for AI-native 6G, favoring instead composable agentic architectures.
-
Advancing Network Digital Twin Framework for Generating Realistic Datasets
An open integration of vehicular mobility, Sionna ray tracing, and ns-3 produces realistic cross-layer datasets for vehicular wireless networks and is released with code and example data.
-
User-Centric Stream Sensing for Grant-Free Access: Deep Learning with Covariance Differencing
Covariance differencing plus a deep-learning classifier lets a user device estimate the number of newly activated streams in overloaded grant-free access.
-
When Does a Neural Receiver Help? Calibration-Drift Benchmarking and Detect-and-Rollback for 5G/6G NR
Neural receivers outperform MMSE in-distribution but their robustness to calibration drift is poorly characterized; the work provides benchmarking and a detect-and-rollback approach for 5G/6G NR.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.