Skip to main content
Ctrl+K
Aerial CUDA-Accelerated RAN - Home

Aerial CUDA-Accelerated RAN

  • PDF
Aerial CUDA-Accelerated RAN - Home

Aerial CUDA-Accelerated RAN

  • PDF

Table of Contents

  • Supported Systems
  • Release Notes
    • Software Manifest
    • Supported Features and Configurations
    • Multicell Capacity
    • Supported Test Vector Configurations
    • Limitations
    • Acknowledgements
  • Installation Guide
    • Installing Tools on Grace Hopper MGX System
    • Installing Tools on Dell R750
    • Installing and Upgrading cuBB
    • Aerial System Scripts
    • Aerial CUDA-Accelerated RAN Versioning in YAML Files
    • Troubleshooting
  • Quickstart Guide
    • Quickstart Overview
    • Generating TV and Launch Pattern Files
    • Running Aerial cuPHY
    • Running cuBB End-to-End
    • Running cuBB Performance tests
    • E2E gNodeB on MIG
  • cuBB
    • Getting Started
    • Features and Architecture
      • cuPHY Features Overview
      • Aerial CUDA-Accelerated RAN Features for 5G gNB
    • cuBB Integration Guide
      • NVIPC
        • NVIPC Overview
        • NVIPC Integration
      • FAPI Support
        • Standard FAPI Support
        • Vendor-Specific Extensions
        • FAPI Message Formats and Specifications
      • Run-time Configuration and Status
    • cuPHY Developer Guide
      • Overview
      • Components
      • Test MAC and RU Emulator Architecture Overview
      • 5G MATLAB Models for Testing and Validation
      • AI/ML Components for PUSCH Channel Estimation
      • References
    • OAM
      • Configuration
      • Operation
      • Logging
      • Metrics
    • cuMAC
      • Implementation Details
      • Examples
      • cuMAC API Reference
      • cuMAC-CP integration guide
        • cuMAC-CP API Procedures
        • cuMAC-CP API Messages
        • L2 integration notes
        • cuMAC-CP Tests
    • Glossary
  • Data Lake
  • pyAerial
    • Overview
    • Getting Started with pyAerial
    • Examples of Using pyAerial
      • Using pyAerial to run a PUSCH link simulation
      • Using pyAerial for LDPC encoding-decoding chain
      • Using pyAerial to run 5G sounding reference signal transmission and reception
      • Using pyAerial to run CSI-RS transmission and reception
      • Using pyAerial for data generation by simulation
      • LLRNet: Dataset generation
      • LLRNet: Model training and testing
      • Using pyAerial to evaluate a PUSCH neural receiver
      • Channel Estimation for Uplink Shared Channel (PUSCH) in PyAerial
      • Using pyAerial for channel estimation on Aerial Data Lake data
      • Using pyAerial for PUSCH decoding on Aerial Data Lake data
      • Using pyAerial for PUSCH decoding on Aerial Data Lake data
    • API Reference
      • Physical layer for 5G
        • Receiver algorithms
        • Configuration classes
        • PDSCH
        • PUSCH
        • LDPC 5G
        • CSI reference signals (CSI-RS)
        • Sounding reference signals (SRS)
        • Fading channel
        • Pipeline API definitions
      • Utilities
  • pyAerial

pyAerial#

pyAerial provides a Python API towards the 5G/6G signal processing functionality included in the cuPHY library.

  • Overview
    • Key Features
    • Release Notes
  • Getting Started with pyAerial
    • Pre-requisites
    • Installing pyAerial
    • Testing the installation
    • Running the example Jupyter notebooks
  • Examples of Using pyAerial
    • Running a PUSCH link simulation
    • LDPC encoding-decoding chain
    • Sounding reference signal transmission and reception
    • CSI reference signal transmission and reception
    • Dataset generation by simulation
    • Dataset generation for LLRNet
    • LLRNet model training
    • Neural receiver validation
    • Machine learning based channel estimation for 5G NR PUSCH
    • Channel estimation on transmissions captured using Data Lake
    • Decoding PUSCH transmissions captured using Data Lake
    • Decoding PUSCH transmissions from multiple radio units using Data Lake
  • API Reference
    • Physical layer for 5G
    • Utilities

previous

Data Lake

next

Overview

NVIDIA NVIDIA

Copyright © 2025, NVIDIA Corporation.

Last updated on Dec 10, 2025.