MX-PDK CAMPUS

Multi-vendor 5G/6G Platform DevKit with Infrastructure at Campus Scale

MX-PDK CAMPUS

service thumb

🏷️ Request a quotation

Campus-scale platform with enterprise O-Cloud, site-scoped indoor and outdoor O-RUs, synchronized fronthaul, GPU acceleration, agentic AI-RAN, Network Digital Twin, energy management, and one year of updates and support

3GPP REL 16/17 O-RAN Agentic AI-RAN

MX-PDK CAMPUS gives operators, vendors, system integrators, enterprises, and large R&D programmes a complete, licensed, and validated multi-vendor 5G/6G O-RAN and AI-RAN platform at campus scale. It combines enterprise cloud infrastructure, indoor and outdoor radio coverage, programmable RAN control, agentic automation, GPU acceleration, and Network Digital Twin capabilities in one integrated environment.

It extends the full MX-PDK O-RAN and AI-RAN foundation with 3+ enterprise O-Cloud nodes, site-scoped O-RUs, on-premises GPU acceleration, the Observability, Automation, and Optimization Blueprints, MX-DT and its Scenario Kit, PDU-based energy visibility, and deployment services. It is designed for campus deployments, field validation, pre-production, and continuous optimization without changing the software foundation used across the MX-PDK range.

What makes MX-PDK CAMPUS different

  1. Scale: Move from a lab testbed to indoor and outdoor site coverage with enterprise compute, synchronized O-RAN fronthaul, commercial devices, and multi-cell configurations.
  2. O-RAN: Use the SMO/OAM, near-RT RIC, non-RT RIC, xApp SDK, rApp SDK, and CDK, with support for E2, A1, O1, R1, and O-RAN 7.2 fronthaul.
  3. AI-RAN: Run Observability, Automation, and Optimization Blueprints, build custom agents with the BAT Agent DevKit, and connect agents to network data and control through A2A and MCP.
  4. Network Digital Twin: Replicate live network scenarios, run what-if experiments, validate proposed changes, and use the MX-DT Scenario Kit for repeatable Digital Twin workflows.
  5. Infrastructure: Combine network, cloud, application, and per-node energy data, with PDU monitoring and control integrated through the SMO.

What can you do with MX-PDK CAMPUS?

  1. Deploy and validate campus-scale 5G networks using indoor and outdoor O-RAN radios, multiple cells, commercial devices, and end-to-end network slicing.
  2. Test multi-vendor O-RAN interoperability across the O-Cloud, O-CU/O-DU, O-RUs, RIC, SMO/OAM, fronthaul fabric, and commercial UE domains.
  3. Run agentic AI-RAN workflows for observability, automation, configuration, SLA assurance, slicing, mobility, and network optimization.
  4. Validate changes against a Network Digital Twin before applying them to the live network, supporting safer what-if analysis and closed-loop optimization.
  5. Experiment with advanced 5G/6G use cases, including TN/NTN, ISAC, sensing, mobility, V2X, industrial IoT, edge AI, and applications in the loop.
  6. Measure infrastructure and energy behavior across the network and O-Cloud, including per-node power visibility and control.
  7. Prepare field trials and pre-production deployments on the same software foundation used throughout the MX-PDK product family.

What arrives in the deployment?

MX-PDK CAMPUS hardware

A CAMPUS configuration is scoped to the site and programme. A typical reference deployment includes:

  1. 3+ enterprise 2U O-Cloud server nodes, fully installed with the BubbleRAN O-Cloud, optimized Kubernetes, device discovery, Cilium networking, Multus, eBPF observability, and persistent storage. Additional nodes can be added when more capacity is required.
  2. Indoor and outdoor O-RAN 7.2 Radio Units, with LITEON FlexFi and supported Benetel indoor radios for in-building coverage, plus Benetel outdoor radios such as the RAN650 for campus and open-air deployments.
  3. 1x FibroLAN Falcon-RX fronthaul and synchronization fabric, with PTP grandmaster, GPS and local synchronization, PTP switching, QoS, and VLAN support.
  4. On-premises NVIDIA GPU acceleration, sized to the project for AI-RAN, agent inference, edge AI, and Digital Twin workloads, with NVIDIA RTX PRO 6000 Blackwell class as a reference configuration.
  5. Power Distribution Unit, integrated with the BubbleRAN SMO through Redfish for per-node power monitoring and control across the O-Cloud.
  6. Commercial device fleet and 20-SIM pool, using Quectel modules, Pictel edge nodes, and commercial devices for multi-UE, mobility, slicing, and application testing.
  7. Required deployment accessories, including antennas, RF, Ethernet and optical cables, SFPs, power supplies, adapters, and site-specific connectivity accessories.

The hardware configuration can be extended with additional O-Cloud nodes, GPUs, indoor or outdoor O-RUs, SDRs, commercial UEs and CPEs, edge devices, and multi-site configurations.

MX-PDK CAMPUS software

Software Stack included with MX-PDK CAMPUS

MX-PDK CAMPUS support, software updates, and services

The MX-PDK software license includes one year of software updates and technical support. The reference CAMPUS scope also includes a one-year hardware warranty. Site survey, installation, integration, hands-on training, feature development, and developer-level support can be scoped to the deployment requirements.

Your day-one blueprint

MX-PDK CAMPUS in use: MX-UI with the optimization assistant, a multi-node O-Cloud, the Falcon-RX fronthaul and indoor plus outdoor Radio Units with a SIM pool

MX-PDK CAMPUS is delivered as a site-scoped platform with reusable deployment blueprints for multi-cell networking, slicing, agentic operations, and Digital Twin workflows. A reference blueprint combines a live campus network with end-to-end slices and the Optimization Blueprint, so proposed configuration changes can be evaluated against a synchronized Digital Twin before rollout.

bash$ brc install network 5g-campus.yaml       # deploy the multi-cell campus network
bash$ brc install slices campus-slices.yaml    # create end-to-end slices
bash$ brc install aifabric optimization.yaml   # deploy Digital Twin-validated optimization
bash$ brc observe network                      # observe network, infrastructure, and energy state
Element What gets deployed
Access Indoor and outdoor gNBs over O-RAN 7.2 fronthaul across the covered area
Radio Site-scoped indoor and outdoor O-RUs synchronized through the Falcon-RX PTP fabric
Core 5G core with multiple end-to-end slices and per-slice assurance workflows
Terminals Commercial device fleet with a 20-SIM pool for multi-UE, mobility, slice, and application testing
RIC and xApps Near-RT RIC with monitoring, traffic-steering, mobility, sensing, and slice-control xApps
Non-RT RIC and rApps Policy, intent, SLA, optimization, and orchestration workflows over A1 and R1
AIFabric Observability, Automation, and Optimization Blueprints, with agents connected to network data and control through A2A and MCP
Digital Twin MX-DT replica workflows used by the Optimization Blueprint to test proposed changes before rollout
Energy PDU-based per-node power monitoring and control integrated through the SMO

The platform keeps the live network as the operational source of truth, while the Digital Twin provides a controlled environment for validating selected changes. This lets automation progress from observation, to proposed action, to Digital Twin validation, and finally to rollout when the configured validation criteria are met.


From AI-RAN to campus-scale infrastructure

MX-PDK CAMPUS extends the AI-RAN testbed into an infrastructure platform designed for site coverage, field validation, and pre-production.

Dimension What CAMPUS adds
Coverage Site-scoped indoor and outdoor O-RUs for networks spanning buildings, yards, test tracks, and campus environments.
Enterprise compute 3+ enterprise 2U O-Cloud nodes, with additional capacity available for concurrent networks, slices, agents, and heavier workloads.
GPU acceleration On-premises GPU capacity for agentic AI-RAN, AI-on-RAN, edge applications, and Digital Twin workloads.
Digital Twin MX-DT and its Scenario Kit for replica-based experimentation, what-if analysis, and validation alongside the live network.
Optimization The third agentic reference blueprint, Optimization, using Digital Twin validation before selected configuration changes are rolled out.
Energy visibility PDU integration through the SMO for per-node power measurement and control.
Slicing at scale End-to-end slice provisioning and assurance across multi-cell and site deployments.
Deployment services Site survey, installation, integration, training, and support scoped to the deployment.

The same blueprints, xApps, rApps, agents, APIs, datasets, and operational workflows used on MX-PDK O-RAN and MX-PDK AI-RAN carry forward into CAMPUS. Compare the MX-PDK range.


What the Network Digital Twin adds

MX-DT extends the live platform with replica-based experimentation. It observes the physical network, creates scoped replicas, and provides reusable workflows for scenario testing, validation, and AI-driven analysis.

MX-DT also enables the Optimization Blueprint. The configuration planner can propose a change, evaluate it through the Digital Twin backend, and proceed only when the configured validation criteria are satisfied. This provides a controlled path for testing optimization decisions before they reach the live network.

Component What it does
DigitalTwin Operator Cloud-native Digital Twin orchestrator built from Perceptors and Replicators.
Perceptor Observes the physical network, or another replica, and collects the state needed to track how it evolves.
Replicator Uses the collected state to create one or more replicas of the physical network, synchronized when required.
Scenario Kit Designs and deploys repeatable test scenarios, with reusable Digital Twin workflows integrated with the BAT toolkit.
Capability MX-DT on MX-PDK CAMPUS
Replica spawn latency 10 to 120 s
Sandboxing Full 5G sandboxing with emulated UEs in the loop
Parallel replicas Parallel replica sets with selective scoping
Time control Synchronize or freeze each replica independently, live or time-locked, with time-aware state throughout
App catalog Configuration manager, traffic mirroring, channel models, synthetic data generation
Interfaces CLI and UI
Current 5G replication scope OpenAirInterface, contact BubbleRAN for other 5G stacks

This makes what-if analysis, AI/ML validation, and optimization practical alongside a live deployment, with the replica providing a controlled environment for evaluating selected changes first.


Who it is for

  1. Operators, vendors, and industrial R&D teams: A field-oriented platform for validating O-RAN, AI-RAN, automation, and optimization before broader deployment.
  2. Enterprises and industrial sites: A multi-vendor private 5G platform for campus, plant, logistics, test-track, and operational environments.
  3. National labs and large research programmes: A shared infrastructure for multi-cell, slicing, AI-RAN, Digital Twin, TN/NTN, and advanced 5G/6G experimentation.
  4. AI-RAN and Digital Twin teams: An environment combining live radios, GPU capacity, agentic workflows, and synchronized network replicas.
  5. System integrators: A configurable platform for integrating, validating, and demonstrating end-to-end solutions on top of the MX-PDK foundation.

Scale and 5G/6G capabilities

MX-PDK CAMPUS combines the complete MX-PDK foundation with 3+ enterprise O-Cloud nodes, site-scoped indoor and outdoor radio infrastructure, GPU acceleration, full agentic AI-RAN, Network Digital Twin and Scenario Kit, energy visibility, and deployment services. The exact hardware and radio scope is configured per site and programme.

5G/6G radio and network
Capability MX-PDK CAMPUS
5G stack OpenAirInterface (R2.4.0), OCUDU (R24.06), and Open5GS, with 3GPP Release 17 support, including OCUDU UE and OAI UE soft terminals, with supported LITEON All-in-One small cells
Deployment modes Standalone (SA) and Non-Standalone (NSA), monolithic gNB, CU/DU split, and CU-CP/CU-UP split
TN/NTN Terrestrial networking plus 5G NTN experimentation via OCUDU, including non-terrestrial emulation, with over-the-air NTN available as an option
Spectrum FR1 with the selected indoor and outdoor O-RUs, plus all TDD FR1 and limited FR2 support on the software platform, subject to the radio configuration
Bandwidth Up to 100 MHz per cell and bandwidth part
Modulation 256QAM in downlink and uplink
MIMO Up to 8x8 downlink and 2x2 uplink, subject to the selected radio configuration
Mobility Xn, NG, and inter-DU handover across deployed cells
Interfaces E2, F1, NG, Xn, plus O-RAN 7.2a and supported fronthaul workflows
O-Cloud 3+ enterprise 2U nodes with GPU, Kubernetes (Kubeadm), Cilium CNI with Multus, Harbor registry, Rook, and Ceph persistent storage
Radio units Site-scoped indoor and outdoor O-RAN 7.2 O-RUs, including supported LITEON and Benetel configurations
Fronthaul O-RAN 7.2 FHI with PTP synchronization through the FibroLAN Falcon-RX grandmaster, with GPS and local synchronization
Commercial UE fleet Commercial devices, Quectel modules, Pictel edge nodes, and a 20-SIM pool in the reference scope
Concurrent networks Multiple networks and slices in parallel, sized to the deployment
Over-the-air 5G Site-scoped and included in the deployment configuration
On-premises GPU Included and sized to project requirements
O-RAN, AI-RAN and Digital Twin
Capability MX-PDK CAMPUS
Near-RT RIC 300 µs to 1 ms control loop; E2AP v3.0 and A1AP v4.04
Non-RT RIC Included, with R1AP v8.0 or Kubernetes CRDs, plus A1AP v4.04
E2 service models KPM v3.0, RC v1.03, CCC v3.01, LLC v1.0, plus BubbleRAN Traffic Control, Slice Control, and L2 Statistics
xApp capabilities Performance measurement, traffic steering, handover control, RAN reconfiguration, slicing, interference detection, and sensing from SRS I/Q samples
rApp capabilities Intent-driven automation, QoS/QoE optimization, SLA assurance, slice provisioning, and AI/ML deployment workflows
Agentic AI-RAN AIFabric controller and CRD, Observability, Automation, and Optimization Blueprints, A2A and MCP, BAT Agent DevKit, AI-for-RAN, and AI-on-RAN
Agent decision loop Approximately 1 s to 1 min for agentic workflows, alongside the faster near-RT and non-RT RIC control loops
AI agent catalog Supervisor, SMO Agent, RIC Agent, API Agent, Observability Planner, Cluster Agent, Logs Agent, VM Agent, and reusable catalog artifacts
Model endpoints Local or on-premises Ollama, NVIDIA NIM and NeMo Agent Toolkit, remote model APIs, or OpenAI-compatible endpoints
Network Digital Twin MX-DT included, with DigitalTwin Operator, Perceptor, Replicator, Scenario Kit, parallel replicas, and independent synchronization or freeze control
Digital Twin-validated optimization Included through the Optimization Blueprint and MX-DT backend
ISAC and sensing LLC (Low Layer Control) service model over E2, providing access to SRS I/Q samples for sensing, positioning, interference detection, and object detection workflows
Development kits xApp SDK (C, C++, Python), rApp SDK (Python), BAT Agent DevKit, CDK, and MX-DT Scenario Kit, plus the TelcoFabric reusable artifact catalog
Automation, observability, and operations
Capability MX-PDK CAMPUS
SMO/OAM operations Day 0 in 1 to 5 s, Day 1 in 1 to 30 s, Day 2+ in 1 to 75 s
Northbound Kubernetes CRDs (YAML), REST API, O1 interface, Redfish for PDU integration, BubbleRAN CLI (brc)
Lifecycle scope Resource discovery, NF onboarding, service design, scheduling, deployment, configuration, reconfiguration, testing, and upgrades
Observability Multi-source data lake with RAN statistics, logs, traces, infrastructure usage, energy consumption, and Grafana dashboards
Agentic observability Included Observability Blueprint with Planner, cluster, logs, and metrics agents
Agentic automation Included Automation Blueprint with Supervisor, RIC, SMO, and API agents
Agentic optimization Included Optimization Blueprint using the Network Digital Twin for what-if validation
Energy visibility and control PDU integration through the SMO for per-node power measurement and control
Security RBAC, network isolation, signed rootless artifacts, SBOM, runtime network security, and process security
Services Site survey, installation, integration, hands-on training, feature development, and developer-level support, scoped per deployment

Ready to scope your deployment?

MX-PDK CAMPUS is quoted per opportunity because the variables that matter are specific to the site and programme, including coverage area, spectrum, number of cells, indoor and outdoor radio requirements, O-Cloud capacity, GPU capacity, Digital Twin scope, services, and support model.

📧 contact@bubbleran.com


Frequently Asked Questions

1️⃣ How is MX-PDK CAMPUS different from MX-PDK AI-RAN? MX-PDK AI-RAN is designed for multi-cell O-RAN and AI-RAN R&D with three O-Cloud nodes, two live O-RUs, agentic observability and automation, and the BAT Agent DevKit. MX-PDK CAMPUS extends that foundation with enterprise 2U O-Cloud infrastructure, site-scoped indoor and outdoor radio coverage, on-premises GPU acceleration, the included Network Digital Twin and Scenario Kit, the Optimization Blueprint, PDU-based energy management, and deployment services. It is designed for campus deployment, field validation, and pre-production.
2️⃣ Why is MX-PDK CAMPUS quotation-based? The platform is configured around the deployment. Coverage area, spectrum, number and type of radios, O-Cloud capacity, GPU requirements, Digital Twin scope, device fleet, installation, and support all depend on the site and programme. BubbleRAN therefore scopes the configuration and quotation to the actual deployment requirements.
3️⃣ Can MX-PDK CAMPUS cover outdoor sites or an entire campus? Yes. Indoor and outdoor O-RAN radio configurations are supported, including Benetel outdoor radios such as the RAN650 class. Coverage, radio count, synchronization, fronthaul, installation, and site survey are scoped to the deployment.
4️⃣ Is GPU acceleration included? Yes, on-premises GPU capacity is part of the CAMPUS scope and is sized to the project requirements. NVIDIA RTX PRO 6000 Blackwell class is a reference configuration for AI-RAN, agent inference, edge AI, and Digital Twin workloads. Other supported GPU configurations can be scoped when required.
5️⃣ What does the Network Digital Twin add? MX-DT adds replica-based experimentation alongside the live network. It uses the DigitalTwin Operator, Perceptors, Replicators, and Scenario Kit to create scoped network replicas for what-if analysis, scenario testing, validation, and AI/ML workflows. It also provides the Digital Twin backend used by the Optimization Blueprint to evaluate selected changes before rollout.
6️⃣ Which 5G stack does MX-DT replicate? The current MX-DT replication scope supports OpenAirInterface deployments. Contact BubbleRAN when the Digital Twin must replicate a different 5G stack.
7️⃣ Does CAMPUS include all three agentic AI-RAN blueprints? Yes. MX-PDK CAMPUS includes the Observability, Automation, and Optimization Blueprints. The Optimization Blueprint is the additional capability at this tier because CAMPUS includes the Network Digital Twin backend required for what-if validation before selected configuration changes are applied.
8️⃣ Is the software foundation the same across the MX-PDK range? Yes. MX-PDK CAMPUS uses the same O-Cloud, SMO/OAM, RIC, CLI, blueprint, SDK, and reusable artifact foundation as the broader MX-PDK range. Blueprints, xApps, rApps, agents, APIs, datasets, and operational workflows developed on the lower tiers carry forward as you scale to CAMPUS infrastructure.
9️⃣ What support and services are available? The MX-PDK software license includes one year of software updates and technical support, and the reference CAMPUS scope includes a one-year hardware warranty. Site survey, installation, integration, hands-on training, feature development, and developer-level support can be included in the project scope.

🔟 Ask your questions

Every network is unique and so is the path to the right solution!

Contact US