This paper presents the design and evaluation of a network slicing implementation in a simulated 5G Standalone (SA) mobile network deployed as a nomadic edge node, where “nomadic” refers to the physical portability and ease of redeployment of a self-contained, containerized 5G testbed suitable for university teaching and experimentation. The platform integrates Open5GS, UERANSIM, Kamailio, and Prometheus/Grafana to emulate a sliced 5G core and access network supporting differentiated service requirements typical of heterogeneous traffic classes and latency-sensitive applications. Slice provisioning is fully configurable, and Docker-based resource constraints are applied to enforce Quality of Service (QoS) differentiation. Performance was assessed through bandwidth and traffic-quality measurements, demonstrating measurable improvements in packet loss and jitter for high-priority slices, with corresponding degradation for lower-priority slices. Although the laboratory environment limits replication of distributed real-world deployments, the results confirm the effectiveness of network slicing for traffic isolation and service prioritization in 5G SA systems. These findings highlight the practical boundaries of container-based slicing enforcement in a single-host nomadic 5G SA node, and inform the design of future multi-host deployments.
Elena-Ramona Modroiu, Jianwei Cheng, Damian Atlaß et al.· International Conference on...· 0 citations
Sixth generation wireless mobile networks face many challenges. With the integration of non-terrestrial networks, wireless backhauls, multi-access edge computing, nomadic and non-public networks, the traditional assumption of a static and predictable control plane infrastructure has to be abandoned. The control plane network functions will no longer be deployed in a central location but geographically spread across the infrastructure. Different types of backhaul connections and deployment modes will mean less reliable and predictable control plane connections. This means that the core network has to be designed with these problems in mind. Running the 5G service-based architecture over a best-effort IP network is not sufficient in this regard. To address these new challenges, we propose the organic control plane. With an improved core network architecture and a robust control plane fabric, it is able to handle the heterogeneous and complex network infrastructure of the future, without compromising on performance. We implemented an organic 6G control plane, by combining our control plane fabric KIRA and our core network Open6GCore. Our evaluation using docker and Containernet shows that there is no significant performance impact, while providing improved dependability, scalability and flexibility.
Fabian Eichhorn, M. Corici, Thomas Magedanz et al.· IEEE Conference on Network S...· 0 citations
