Update
Lab back online after the holiday — verified . Sign-in, GitLab, and the forum are reachable. GPU compute starts on demand. SSH key provisioning remains unavailable; see live service status.
O-RAN Alliance · nGRG · Open Research Lab

OpenLab// O-RAN is open and intelligent from day 1

O-RAN was designed open and intelligent from day 1. It encompasses AI for Open RAN, Open RAN for AI, and compute-and-communications integrated RAN. OpenLab is where researchers worldwide can build and test all three, on a neutral and reproducible base everyone can trust.

vendor-independent · multi-stack · multi-architecture OAI primary baseline srsRAN · FlexRIC · Open5GS NVIDIA GPU + Intel / AMD x86 software L1–L3 · no special hardware
Mission

Give Open RAN for AI research a common base — not a scoreboard.

OpenLab exists so that researchers anywhere can run real Open RAN experiments on a stable platform under O-RAN nGRG governance. We deliberately do not rank vendor platforms against each other. What we provide is a neutral stage: open stacks and vendor tooling run in the open, and results other groups can re-run for themselves.

Open baseline

OpenAirInterface anchors the lab as the primary research stack. It is fully open under LGPL/CeCILL, reproducible in an academic setting, and belongs to no vendor.

A neutral stage for vendors

NVIDIA and Intel each contribute Open RAN for AI enablement on equal footing. They compete by making research easier. The lab publishes no ranking of them.

Reproducible by default

Every stack is source-built and scripted, from protocol code to channel model. Experiments ship as dataset, model, and code, not as claims.

Flagship Direction

Open and intelligent from day 1

One open, disaggregated, GPU-accelerated platform, serving at once as the system AI improves and the substrate AI is served by. The three threads below run in parallel; none of them waits on the others.

Thread · AI for Open RAN

Learning inside the radio

Replace or augment RAN functions with learned components: neural receivers, ML-driven scheduling, beam management delivered as xApps over E2.

SionnaPyTorchFlexRICE2SM-KPM / RC
Thread · Open RAN for AI

Serving the AI workload

Co-locate edge AI inference with vRAN workloads on the same accelerators, then measure what happens to isolation, latency, and scheduling under contention.

CUDAOAI gNBOpen5GSGPU partitioning
Thread · Compute–communications integrated RAN

Where the two compete

Treat AI and RAN as competing tenants of the same GPUs, and build benchmarks for the energy-versus-throughput decisions that follow.

5G-LENAns-3energy xApptelemetry
Platform

A full O-RAN stack, open top to bottom.

Control, radio, core, and an AI/channel plane: all software, all source-built, running end-to-end without SDR or over-the-air hardware. Everything below is deployed and verified today.

Layer · ControlO-RAN / RIC
FlexRICnear-RT RIC xAppsKPM · RC · MAC · TC E2 service modelsKPM v3 · RC · E2AP v3
Layer · RAN L1–L3gNB & UE
OpenAirInterfacegNB + nrUE · RFsimbaseline srsRAN ProjectgNB · ZMQ srsRAN 4GsrsUE / eNB / EPC UERANSIMgNB / UE
Layer · Core5G Core & EPC
Open5GS5G SA + EPC · 20 NFs Subscriber DBMongoDB
Plane · AI & ChannelPHY / ML / SysSim
Sionnalink-level PHY + ray tracing PyTorchCUDA · GPU 5G-LENAns-3 · NR system-level MATLAB 5Glink-level cross-check
Layer · FabricCompute
Intel Xeon 6 SoCGNR-D · vRAN Boost · 8×25 GbE Intel Xeon 6+ NVIDIA RTX 5090 AMD Threadripper+ NVIDIA RTX 5090 NVIDIA CUDAGPU acceleration

The lab runs on both x86 vendors, Intel and AMD, each paired with an NVIDIA GPU, plus a dedicated Intel Xeon 6 SoC (Granite Rapids-D) DU-class server — 20 cores with built-in vRAN Boost and eight 25 GbE ports — now online as its own L1 home, away from the shared multi-tenant host. That is what backs the neutral-stage claim. All radio is still software-simulated; nothing here has yet touched an antenna.

Roadmap

From software testbed to live radio to new silicon.

Phase 0 Complete
2026 Q2
Infrastructure, OAI deployment, and the shared CI and registry that make every later result reproducible.
Phase 1 Current
2026 Q3
The triad joined up: OAI on NVIDIA GPU and Intel CPU, plus the first Open RAN for AI proofs of concept.
Phase 2 Planned
2026 Q4
With the dedicated Xeon 6 SoC DU server (vRAN Boost, 25 GbE fronthaul ports) already online, next is the first live RF via O-RU, and multi-scenario benchmarks anyone can repeat.
Phase 3 Planned
2027 H1
ARM, RISC-V, and accelerator cards join the platform, alongside the first major nGRG research output.
What comes out

Everything the lab produces is meant to be re-run.

Access

Open application, members first.

  1. Send a short proposalWhat you want to investigate, which parts of the stack you need, and roughly how much compute and for how long.
  2. Scoping conversationWe confirm the experiment fits the platform as it stands today, and agree what your published output will be.
  3. Get your environmentAn isolated tenancy, with the stack pre-built and your repository and CI already wired up.
  4. PublishResults land as an nGRG report or paper, with dataset, model, and code released alongside.

Current status

The platform is in Phase 1 build-out. Core infrastructure is back online after the holiday; GPU compute starts on demand. Multi-tenant onboarding is being staged as capacity comes online. Check live service status for current availability and known issues.

O-RAN Alliance member organisations are prioritised. Academic groups outside the Alliance are welcome to enquire.

Host Tongji University Governance O-RAN nGRG