From theory to reproducible, hands-on experimentation on realistic emulated mobile networks with Open Source
The BubbleRAN Teaching & Education Solution is a complete academic package built on MX-PDK, with optional MX-AI extensions for advanced AI-powered RAN teaching.
It combines a fully hosted and managed cloud laboratory, reproducible deployment blueprints, isolated student environments, ready-made courseware, instructor tools, training, maintenance, and technical support. Universities, graduate schools, and research institutions can teach and experiment with end-to-end 5G SA, Open RAN, O-RAN, RIC, SMO, xApps, rApps, cloud-native automation, data pipelines, and AI-RAN without the cost and operational burden of assembling a telecom laboratory from scratch.
Students learn by operating individual 4G/5G network entities (e.g. gNB or CU or DU) and observing protocols among them (e.g. F1, NGAP). From a browser or command-line interface, they can deploy, configure, observe, program, troubleshoot, and optimize an end-to-end mobile network in a controlled and reproducible environment. Students will also learn 3GPP standards, O-RAN specifications, and new trends in mobile communication systems.
Why universities adopt the full package
-
Launch faster
Start a new course or refresh an existing curriculum with validated materials, deployment blueprints, instructor solutions, and preconfigured laboratory exercises instead of spending months integrating a custom testbed. -
Lower cost and operational risk
Begin with a fully virtual cloud laboratory that requires no dedicated telecom hardware. BubbleRAN can host, operate, maintain, update, and support the platform while the institution retains control of course design and assessment. -
Deliver consistent learning outcomes
Isolated Kubernetes environments, versioned configurations, repeatable blueprints, and common observability tools give every team a comparable starting point and make laboratory results easier to reproduce, explain, and grade. -
Teach both concepts and practice
Students connect 3GPP architecture and protocols to real deployment, signaling, data-plane, orchestration, RIC, xApp/rApp, and AI-RAN workflows—building skills that transfer directly to research and industry. -
Grow from teaching into research
The same environment can evolve from emulated coursework to physical radios, industrial-grade components, advanced AI agents, network digital twins, internships, thesis projects, and shared research infrastructure.
The complete university package
| Component | Included Deliverables |
|---|---|
| Course Materials | Lecture notes, presentation slides, textbook, recommanded books, module outlines, and learning objectives |
| Laboratories | Laboratory instructions, instructor solutions, deployment blueprints, validation steps, troubleshooting guidance, and assessment rubrics |
| Student resources | Homework, exams, configuration files, logs, packet captures (pcaps), datasets, SDK and code samples |
| Cloud laboratory | End-to-end 5G SA, Core and RAN functions, CU/DU deployments, O-RAN RIC/SMO, Kubernetes, dashboards, CLI tools, and reproducible experiment environments |
| Instructor tools | Environment administration, cohort and resource management, monitoring, reset/recovery workflows, and reproducibility controls |
| Enablement | Instructor and administrator training, onboarding, deployment support, course tailoring, and teaching-assistant sessions |
| Operations & support | Maintenance, updates, technical support, high-availability options, and SLA coverage aligned with teaching schedules and critical lab windows |
Core platform capabilities
- Fully managed cloud hosting, deployment, and lifecycle management
- Terraform-based deployment on public or private cloud infrastructure
- Remote access through browser-based interfaces, dashboards, APIs, and CLI tools
- Individual or isolated environments for students, teams, and research groups
- High-availability deployment options
- End-to-end 5G Standalone architecture
- 5G Core and RAN components, including gNB, CU, and DU functions
- Open RAN and O-RAN architecture and interfaces
- Near-Real-Time RIC and Non-Real-Time RIC capabilities
- xApp and rApp development, deployment, and experimentation
- Introductory and advanced AI-RAN, including AI-for-RAN and AI-on-RAN concepts
- Kubernetes and cloud-native network functions
- Network automation, orchestration, monitoring, and observability
- Multi-vendor learning paths spanning open-source stacks such as OpenAirInterface, OCUDU, and Open5GS, with optional industrial-grade components for advanced interoperability exercises
- Reproducible laboratory blueprints and preconfigured exercises
- Instructor training, maintenance, updates, technical support, and SLA options
Deployment and delivery models
| Model | Hosting | Best suited for | Hardware requirement |
|---|---|---|---|
| Fully managed cloud | Hosted and operated by BubbleRAN | Rapid course launch; institutions without telecom infrastructure; predictable operations | No dedicated telecom hardware; students use standard computers with a modern browser or SSH client |
| Customer-managed public cloud | Deployed in the institution’s public (e.g. GKE, AZUR) cloud resources | Institutions manage infrastructure, identity, data, or policy control | Lab rooms to access the cloud resources |
| Customer-hosted private cloud | Integrated and deployed with an existing campus, national testbed, or shared facility | Long-term research programs, advanced laboratories, and physical-radio extensions | Depends on existing servers, radios, UEs, timing, transport, and test equipment |
A university can start with a managed public cloud subscription and later add physical radios, UEs, test equipment, industrial-grade 5G components, or existing research infrastructure without redesigning the teaching program.
Student isolation, scalability, and reproducibility
A representative teaching configuration assigns one dedicated Kubernetes cluster to each team of two to three students. Indicatively, each cluster can host a complete 5G deployment with:
- 1-2 gNB or disaggregated CU/DU topology with a single core network
- 40-100 MHz, 2x2, FR1 radio configuration in emulation mode
- 1-4 UEs
- Approximately 16-24 vCPUs per group
The number of UEs, network topology, and bandwidth among others can be adapted to the course objectives (e.g. O-RAN with xApps and rApps) and available cloud resources. Environments may be allocated per student, per team, per laboratory, or per research project.
This model allows multiple teams to work at the same time without interfering with one another (subject to proper cloud resource allocation), while instructors retain a consistent reference environment for demonstrations, troubleshooting, and assessment.
Example Curriculum and Learning Path
The course portfolio follows a coherent progression from the Radio Access Network, to the 5G Core, and then to automation and intelligence. Each course can be delivered independently or combined into a complete 84-hour graduate program.
| Course | Primary focus | Indicative duration | Progression |
|---|---|---|---|
| MobiSys: Next Generation Mobile Communication Systems | 5G RAN architecture & landscape terrestrial and non-terrestrial networks (TN/NTN), MAC/PHY layer and protocols (CP and UP), 5G networking, Open RAN, O-RAN, RIC, xApps/rApps, cloud-native RAN, and introductory AI-RAN | 42 hours | Foundation |
| MobiCore: Next Generation Mobile Core Networks | 4G/5G Core, Service-Based Architecture, NAS, authentication, mobility, sessions, PFCP, GTP, and user plane | 21 hours | Core-network specialization |
| TelcoAI: Automation and Intelligence in Telecommunication Networks | Automation, data pipelines, O-RAN applications, AI-RAN, LLM-based agents, closed-loop control, and autonomous networks | 21 hours | Advanced intelligence and automation |
Note: While below we highligt three major courses, many other courses can be designed with particular focus.
Course 1 — MobSys: Next Generation Mobile Communication Systems
Indicative duration: approximately 42 hours
Organization: 7 lectures and 7 laboratory sessions
Key topics
- Evolution from LTE, virtualized RAN, and Cloud RAN to Open RAN, O-RAN, AI-RAN, and future 6G RAN
- 5G New Radio architecture, physical layer, protocol stack, scheduling, radio-resource management, and mobility management
- gNB architecture, CU/DU/RU functional splits, and the F1 interface
- O-RAN architecture and interfaces including E2, A1, O1, and R1
- Near-RT RIC, Non-RT RIC, xApps, and rApps
- Introduction to AI-RAN and AI-assisted radio-resource optimization
- Future evolution toward intelligent 6G radio networks
Representative labs
- Deploy a cloud-native 5G Open RAN using Kubernetes
- Configure gNB, CU, DU, and radio parameters
- Capture and analyze RAN protocols with Wireshark
- Deploy and connect a Near-RT RIC
- Run, configure, and modify xApps
- Observe network KPIs, logs, traces, and resource usage
- Run introductory AI-RAN optimization experiments
- Complete an end-to-end Open RAN deployment project
Learning outcomes
Students can explain 5G and O-RAN architectures and protocols, analyze radio signaling, deploy a cloud-native RAN, operate a Near-RT RIC, use basic xApps, and evaluate radio-network behavior using protocol and observability tools.
Course 2 — MobiCore: Next Generation Mobile Core Networks
Indicative duration: approximately 21 hours
Organization: 3 lectures and 4 laboratory sessions
Key topics
- 4G Core architecture and its point-to-point model
- 5G Service-Based Architecture, network functions, and virtualization
- NAS authentication, registration, mobility, and session-establishment procedures
- User-plane establishment and forwarding with PFCP and GTP
- Emerging 5G/6G functions such as NWDAF and CAPIF
Representative labs
- Deploy and configure an OpenAirInterface 5G Core on Kubernetes
- Register subscribers and establish PDU sessions
- Trace authentication, registration, session management, PFCP, and GTP procedures
- Analyze control-plane and user-plane traffic with Wireshark
Learning outcomes
Students can design a 5G Core architecture, deploy and configure a cloud-native Core, analyze signaling and user-plane procedures, and program or extend selected network functions.
Course 3 — TelcoAI: Automation and Intelligence in Telecommunication Networks
Indicative duration: approximately 21 hours
Organization: 3 lectures and 4 laboratory sessions
Key topics
- 5G RAN and Core as programmable cloud-native systems
- End-to-end network slicing and service-level objectives
- Network automation, orchestration, and intelligence fundamentals
- Data collection, observability, analytics, and closed-loop control
- AI-RAN Services and Applications (AI-FOR-RAN, AI-ON-RAN, and AI-AND-RAN)
- Autonomous networks with multi-agent workflows
Representative labs
- Two labs on 5G deployment and lifecycle automation using Kubernetes
- One lab on data collection, processing, and observability
- Two labs on AI-RAN intelligence for SLA optimization and RAN slicing
Learning outcomes
Students can design networks for automation, automate network-function lifecycle management, build data pipelines for analytics and control, develop O-RAN intelligence applications, and assess the opportunities and guardrails of autonomous and LLM-based telecom agents.
Textbook
A dedicated textbook, entitled “From 5G to 6G: A Practical Guide from Radio to Agentic Networks”, is provided to support the students during the different courses.
Practical Guide.
Rather than surveying the landscape from a distance, we go deep into the details that actually matter: how the physical layer multiplexes users in time and frequency, how the MAC scheduler arbitrates between competing flows, how a UE attaches to a 5G Standalone core, how the Core Network routes a session from the gNB to the internet.
From 5G to 6G.
5G was designed to be highly configurable and is getting evolved to be programmable and intelligence. First, Open RAN disaggregates the Radio Access Network into standardised, interoperable components and introduces the RAN Intelligent Controller (RIC), which exposes real-time network state to programmable applications - xApps (near-real-time control) and rApps (non-real-time optimisation). Second, AI RAN embeds machine learning directly into the protocol stack, automating resource allocation, beam management, and anomaly detection at timescales no human operator can match. Third, agentic RAN, targeted for 3GPP Release 22 and deployment around 2030, is designed so that autonomous agents perceive network state, reason about objectives, and act, closing the loop from observation to control without human intervention.
Laboratory portfolio
The final laboratory set is tailored to the institution, program level, semester duration, and available infrastructure.
| Domain | Indicative number of labs | Examples |
|---|---|---|
| 5G RAN and Core | 5–10 | Deployment, configuration, protocol analysis, mobility, slicing, QoS, and performance |
| RIC | 2–4 | Near-RT RIC deployment, E2 integration, subscription, policy, and control workflows |
| xApps | 5–15 | KPM monitoring, RAN control, slicing, mobility, data collection, and optimization |
| rApps | 5–10 | Non-RT optimization, policy generation, orchestration, SLA assurance, and data workflows |
| AI-RAN | 5–10 | Dataset collection, model integration, anomaly analysis, SLA optimization, and AI-assisted RAN control |
Laboratories generate tangible evidence for learning and assessment, including deployment manifests, configuration files, CLI output, packet captures, logs, traces, time-series KPIs, dashboards, datasets, source code, and written analysis.
Assessment and evidence-based grading
A recommended course model combines:
- Laboratory work: 50% attendance, execution, reports, and practical analysis
- Final examination: 50% architecture, protocols, troubleshooting, lab work and interpretation of network evidence
The package can also support homework, open-PCAP assessments, practical demonstrations, team projects, oral defenses, and capstone deployments. Built-in observability and isolated environments make grading more consistent because instructors can review both the student’s answer and the network evidence behind it.
Optional advanced extensions
-
MX-AI for advanced AI-RAN teaching
Add customizable agentic workflows, multi-agent orchestration, observability and configuration agents, AI-for-RAN and AI-on-RAN scenarios, and development with the BubbleRAN Agentic Toolkit. -
Physical-radio and research-infrastructure extension
Extend the virtual laboratory with SDRs, O-RUs, indoor/outdoor small cells, commercial UEs, timing infrastructure, GPU resources, or existing campus testbeds for over-the-air research and advanced student projects.
Why BubbleRAN is the right long-term academic partner
Building a modern mobile-network course is not only a software decision. It requires curriculum design, reproducible infrastructure, instructor enablement, lifecycle maintenance, student support, and a credible path from introductory teaching to advanced research. Such Teaching & Education Solution shall also be reliable and validated with students with proven learning outcomes.
BubbleRAN combines these elements in one package:
-
Educational impact
Courses are organized around clear learning outcomes and hands-on evidence, helping students understand the relationship between architecture, protocols, operations, data, and intelligence. -
Time saved for faculty and teaching assistants
Ready-made courseware, instructor solutions, validated blueprints, and managed operations reduce the time spent assembling, debugging, resetting, and maintaining laboratories. -
Lower entry barrier
Universities can start without dedicated telecom hardware, scale according to cohort demand, and add physical equipment or advanced product modules only when academically justified. -
Inclusive and collaborative delivery
Guided, isolated, and reproducible environments support students with different backgrounds, encourage teamwork, and reduce the risk that infrastructure problems prevent students from completing the intended learning activity. -
Continuity from teaching to research and employability
Students use the same cloud-native, 3GPP standards, O-RAN specifications, observability, automation, and AI-RAN concepts encountered in advanced research and telecom engineering roles. -
Long-term enablement and support
BubbleRAN can provide instructor and administrator training, course tailoring, environment deployment, updates, maintenance, technical assistance, and SLA options aligned with the academic calendar.
FAQs
1️⃣ Can BubbleRAN provide a fully hosted and managed cloud service?
Yes. BubbleRAN can provide a turnkey environment hosted and operated for teaching, professional training, and research. The university can focus on course delivery while BubbleRAN manages deployment, lifecycle, monitoring, maintenance, and agreed support coverage.
2️⃣ Do we need RF hardware?
No. The program can begin in fully virtual emulation mode with real 5G and O-RAN network functions. Students only need standard computers, network access, and a browser or SSH client. Physical radios, UEs, and test equipment can be added later.
3️⃣ How many students can use the platform concurrently?
The platform scales with the available cloud resources. Sizing depends on cohort size, number of simultaneous laboratories, isolation model, topology, and course schedule. BubbleRAN provides a tailored capacity and cost estimate.
4️⃣ How are student environments isolated?
A typical model allocates one Kubernetes cluster to each team of two to three students. Environments may also be assigned per student, laboratory, or project, depending on the required isolation and available resources.
5️⃣ Can multiple teams work at the same time without interference?
Yes. Teams work in isolated environments and can deploy network functions, modify configurations, run traffic, collect data, and execute O-RAN activities without affecting other groups.
6️⃣ What teaching materials are included?
The package can include lecture notes, slides, textbook references, lab instructions and instructor solutions, homework, quizzes, examinations, packet captures, logs, configurations, datasets, code examples, grading guidance, and project templates.
7️⃣ Can we teach O-RAN, RIC, xApps, rApps, and AI-RAN—not just basic 5G?
Yes. The platform includes Near-RT and Non-RT RIC capabilities, O-RAN interfaces, xApp/rApp development workflows, data collection, observability, and AI-RAN exercises. Optional MX-AI extends the program with customizable multi-agent and LLM-assisted telecom workflows.
8️⃣ How are laboratory results assessed?
Students submit network artifacts and analysis such as deployment files, CLI output, PCAPs, logs, KPIs, dashboards, datasets, and code. This supports evidence-based grading rather than relying only on screenshots or theoretical answers.
9️⃣ Why not assemble separate open-source components ourselves?
A do-it-yourself stack can be valuable for research, but maintaining version compatibility, student isolation, reproducibility, upgrades, reset workflows, observability, and support across a full course creates significant hidden effort. BubbleRAN packages open technologies into a validated, repeatable teaching environment while retaining programmability and extensibility.
1️⃣0️⃣ Can the platform integrate with LMS, SSO, and university APIs?
Integration options are available and are reviewed against the university’s learning-management system, identity provider, access-control model, networking, and API requirements.
1️⃣1️⃣ Are annual, multi-year, support, and SLA options available?
Yes. Annual and multi-year options can be tailored to the selected hosting model, course portfolio, cohort size, support window, and service scope. SLA coverage can be aligned with the institution’s time zone, teaching schedule, examinations, and critical laboratory periods.
1️⃣2️⃣ Is a trial or proof of concept available?
A qualified trial or proof of concept can be arranged with an agreed scope, cohort, schedule, and success criteria.
1️⃣3️⃣ Where can instructors and developers find documentation and samples?
- Open documentation is available here.
- xApp SDKs and samples are available here.
- rApp and agent SDKs and samples are available here.
Request a tailored course, cloud-lab sizing, or demonstration
Build your university package
The final package is tailored to your program level, course calendar, cohort size, hosting preference, isolation model, learning outcomes, assessment approach, and research roadmap.
BubbleRAN can provide a sample syllabus, cloud-laboratory sizing, instructor-enablement plan, deployment schedule, and commercial proposal.