Note
RIA Hub is now available at riahub.ai . Public repos can be explored, but access to the application and its tools are only throgh the private beta.
Radio Inference Applications Hub (RIA Hub) by Qoherent is an AI development platform for software-defined radios, designed for radio signal processing engineers and researchers in wireless technologies. RIA accelerates intelligent radio development by significantly reducing integration between stages, allowing a sharper focus on innovation. The platform streamlines dataset creation, algorithm design, and model deployment, enhancing productivity and reducing project timelines. Built on a foundation of open-source accessibility, a comprehensive IP library, and automation capabilities, RIA drives the creation of intelligent radios, unlocking solutions in an increasingly congested and challenged wireless spectrum.
End-to-end workflow for deploying machine learning into software-defined radios
- Automated and procedural signal generation in Python, GNU Radio and MATLAB, as well as SDR testbed control for over-the-air emulation.
- Plugins for popular signal processing and deep learning back ends.
- Automated design, training, and optimizations of high performance models.
- Automated testing, integration, and deployment into SDR-enabled systems, including OpenRAN-ready 5G solutions.
- Tools for sharing datasets, models, and inference applications with colleagues and the broader community.
Comprehensive reference library
- Labelled datasets, pre-trained models, and optimized inference applications for radio.
- Reference designs for radio ML solutions.
- Example workflows and reusable actions for common tasks.
- Open Source and community libraries.
The development workflow you love, tailored for radio
- RIA Hub is built on Gitea, an open-source DevOps platform.
- Write workflows in GitHub Actions-flavoured YAML.
- Store and version datasets and models directly in repositories using Git LFS.
- Manage containerized inference applications on the RIA Hub Container Registry.