Szczegóły publikacji

Opis bibliograficzny

Automated orchestration of virtualized deployments of mobile core networks / Piotr BORYŁO, Wojciech Stasiak, Piotr Piwowar, Nashid Shahriar // Computer Networks ; ISSN 1389-1286. — 2024 — vol. 255 art. no. 110888, s. 1-20. — Bibliogr. s. 19, Abstr. — Publikacja dostępna online od: 2024-11-15

Autorzy (4)

Słowa kluczowe

network deployment automationnetwork softwarization5G core networkcontrol plane performanceproof-of-concept testbed

Dane bibliometryczne

ID BaDAP157198
Data dodania do BaDAP2024-12-12
Tekst źródłowyURL
DOI10.1016/j.comnet.2024.110888
Rok publikacji2024
Typ publikacjiartykuł w czasopiśmie
Otwarty dostęptak
Czasopismo/seriaComputer Networks

Abstract

Given the ongoing process of softwarization and virtualization of communications networks, including 5G and beyond 5G (B5G), automation has become an inevitable part of network deployment. Ensuring a robust, efficient, and zero-touch deployment process is expected to drive the advancement of B5G and 6G technologies. Automation will become increasingly critical because of increasing complexity and requirements on dynamic provisioning of these networks. A well-designed automation process can streamline the deployment of private mobile networks for Industry 4.0 or on-demand networks for mass events, ensuring efficient video signal transmission and reducing the need for physical wiring. This paper presents a Proof-of-Concept testbed tailored for the automated orchestration of virtualized deployment of a 5G mobile core network, along with insights into potential challenges and solutions. Furthermore, we compare two open-source core network implementations in terms of feasibility for virtualization, automation and performance of user plane (delay, throughput) and control plane (user equipment registration time and PDU session establishment time). We evaluate the performance of automated deployment of 5G core network locally in the testbed and based on the results we propose enhancements that reduce deployment time by 40% on average for one of the implementations. We also present the process and necessary tools to automatically orchestrate deployment of a virtualized 5G core network in a cloud infrastructure. Finally, we performed preliminary performance assessments of the automation process in the proposed cloud-based deployment.