Fraunhofer FOKUS
Testbed
The Fraunhofer FOKUS Testbed is a multi-vendor facility designed to integrate and test SA (Standalone) 5G Core and RAN (Radio Access Network). It is managed by Fraunhofer FOKUS in Berlin, Germany. It showcases various proof-of-concept scenarios incorporating multiple RAN vendors, diverse user equipment, and different backhaul technologies.
Infrastructure
The FOKUS Testbed is deployed with the Open5GCore toolkit built by FOKUS for research and testbed research deployments. The testbed is an all-in-one testbed that is shared for different projects and is used for developing B5G and towards 6G technologies.
In the 6G-PATH project, an additional edge node has been added to the testbed that will be deployed on-site in the Charitè Hospital for processing data on the edge with the required e-health software. The edge node is deployed with Proxmox virtualisation as a hypervisor for deploying software workloads.
The radios in the FOKUS Testbed operate on FR1. They are configured in band N78 with a 100MHz bandwidth, and the edge node is configured with a 20MHz bandwidth.
The FOKUS Testbed will use multiple features from the Open5GCore toolkit for use case deployments and external third-party testbed integrations. As the Open5GCore toolkit is highly configurable according to the requirements, the deployment configuration of the core network will be different. In addition to basic 5G core functionality, the testbed will use location services, Northbound NEF (Network Exposure Function) API integration and slicing for user rerouting.
In addition to the Open5GCore, the testbed uses open-source 5G Core networks for various tests and open-source RAN solutions with SDR boards.
The edge node will be a portable setup as the use case has movement of the node involved. The node is all-in-one machine packed with radio software, 5G core and the edge application. The edge node can operate without any backhaul network and serve the e-health apps to the devices connected locally in the network.
Services
Virtualisation, Management and Orchestration
The FOKUS Testbed is primarily based on the VMware ESXi cluster with additional Proxmox nodes for edge and small deployments. The network comprises Cisco switches and routers with 100G fibre backhaul connectivity between the radio and the core network.
In the first phase, the testbed does not use any special orchestration tool for deploying workloads. In standard deployment, workloads are deployed manually or via an Ansible script.
In the second phase, the testbed will be using the Closed Loop Coordinator from Onesource in order to deploy dynamically services, like the Wound Analytics, depending on the momentarily needs. Location Services for real-time positioning, e.g. of medical personnel that are currently on the road to attend to patients, for dynamic allocation of services, like caring for patients that need emergency care. A dashboard adds the necessary usability to this functionality.
For management, the interfaces from respective vendors are being used. There is no central management platform for the testbed. As most of the requirements are for custom builds, deployment using standard templates is not possible.
Monitoring
Monitoring for Open5GCore is mainly based on the command line and logs for status and debugging. The testbed runs multiple core network instances for different projects and purposes. An example of command-line access is represented in Figure 4.
In addition to the command line, the 5G components from Open5GCore are exporting metrics via a Prometheus exporter. All components export metrics from the platform base, which include memory usage for every 5G process and component-specific data. This data is plotted in Grafana with a pre-built dashboard. The core network components are managed from the command line. However, the devices in the testbed are managed by DeviceFarmer (Figure 5). This tool allows managing all the phones connected to it remotely via a web-UI, making it easier to conduct experiments with real devices remotely without needing to be present in the labs/datacentre.
Security
The infrastructure is secured by maintaining a hard separation between lab and regular networks. In addition to granular control of external access via Wireguard, the tenant separates access to third-party integrators and projects.
The Open5GCore is built for research as requirements to view the messages in transit, hence it does not perform TLS/OAuth between components. However, if needed, TLS could be enabled at the HTTP layer.
Remote Access
Access to the testbed is provided via Wireguard VPN server based on requirements.
Key Features and Assets
VMware and Proxmox, are hypervisor platforms for deploying containers and virtual machines with the required software. The control for the hypervisors is strictly limited to admins. Management of both platforms is done manually to create resources on the testbed.
Prometheus and Grafana, are the monitoring stack with node-exporter and Open5GCore exporter. This helps monitor the status of components and the overall system.
The testbed has two backhaul options that could be used for tests. Fibre and Starlink connectivity provide one TN and one NTN connection for edge nodes based on the connectivity or required reliability of the backhaul.
Kubernetes is used for deploying e-health software on edge nodes with high availability and scalability, provided by the Kubernetes platform.
Testbed Extensions from Open Call #1
Slicefinity’s Trusted Application Function to enable Network APIs and Programmable Network extension leverages the northbound APIs defined in the 5G Core network component NEF to perform actions on the core network. The AF (Application Function) will interact with the core network over the secured Wireguard connectivity. The Open5GCore is providing currently 2 APIs viz. Traffic Influence and QoS management via NEF.
Traffic Influence API
This API allows changing PSA UPF for a given user. This is done by updating policies in PCF and SMF. The whole process of changing the UPF is transparent to the user, hence it is an important aspect for controlling data routing of users. The data could be routed to an edge node or to the central node simply by triggering the policies from the core.
QoS Management API
With this API, QoS of a specific user can be changed on demand. The process is similar as that of traffic influence API, the policies are updated in PCF and SMF by the NEF. The new QoS rules to be pushed could be GBR (Guaranteed Bit Rate) or non-GBR based on the type of application used.
The Slicefinity’s AF is exposing Camara Service APIs and other APIs via its northbound interface to be consumed by the consumers. These northbound APIs are taking care of authentication and authorization for the access of untrusted Afs. Slicefinity’s AF is deployed as a trusted AF in the 5G Core giving it access to multiple core network components.
Network Slice Booking API
This API is used for booking on-demand slices for users. Based on the latency or guarantees needed for the operations, the user can request a specific slice from the network operator and once the operator configures this slice, the user will be moved to this slice and will get the required guarantees.
Trials and Pilots of the supported Use Cases - Common Components
Both 6G-PATH usecases Health-UC1 regarding chronic wound management and Health-UC2 regarding elderly monitoring employ the FOKUS testbed. The trials and piloting activities are mainly located in Berlin. For supporting these usecases, set of key components were developed and are currently running on the same testbed as the core network.
- Admin & Care Portal – Modular Web/Mobile App, used for patient management, additional modules for each usecases, e.g. wound documentation, reha training & wearables
- Patient Data Storage Server – based on the HL7 FHIR (Fast Healthcare Interoperability Resources) standard and together with its database, it stores the patient records related to the nurse activities.
- Data Access Control Server – Provides access control service enabling access for the patient and the medical personnel to the Patient Data Storage Server using role-based policies and patient-specific information (e.g., caring doctor)
Targeted Use Cases
- All verticals are accepted, but Health vertical are preferred.
- All use case topics are accepted.
Published Papers
- Ancuta, Corici, A., Lahmann, N., Corici, M., Hoquel-Hans, M., Kuntz, S., Häusler, B., Deter, A., Grohnert, A. & Zope, H. (2024). 5G Wound Management System for Remote Areas. In: proceedings of the 4th IEEE International Conference on ICT Solutions for eHealth, ICT4eHealth, ESIEE Paris, France, 26-29 June 2024. doi.org/10.1109/ISCC61673.2024.10733682
- Ancuta Corici A. et al. (2024). Employing 5G Mobile Networks in Keeping Elderly Healthy and Active. In: Proceedings of the IEEE International Conference on E-health Networking, Application & Services (IEEE Healthcom 2024), Nara, Japan, 18–20 November 2024. doi.org/10.1109/HealthCom60970.2024.10880760
- Corici A., Hocquel-Hans M., Chaudhary K., Kuntz S., Corici M., Häusler B., Deter A., Grohnert A. & Lahmann N. (2025). Implementation Design and Validation Results of a Chronic Wound Management System Powered by Beyond 5G Private and Nomadic Networking. In: Proceedings of the 21st International Conference on Artificial Intelligence Applications and Innovations (AIAI 2025), Cyprus University of Technology, Limassol, Cyprus, 26-29 June 2025. AIAI 2025 Workshops, IFIP AICT 753, pp. 181-193, 2025. Springer Nature Switzerland AG 2025. doi.org/10.1007/978-3-031-97317-8_14