Hard Multi-Tenancy Kubernetes approaches in a local 5G deployment: Testing and evaluation of the available solutions

Loading...
Thumbnail Image

URL

Journal Title

Journal ISSN

Volume Title

Sähkötekniikan korkeakoulu | Master's thesis

Date

2022-08-22

Department

Major/Subject

Cloud and Network Infrastructures

Mcode

ELEC3059

Degree programme

Master's Programme in ICT Innovation

Language

en

Pages

85 + 9

Series

Abstract

With its capabilities, the fifth mobile network generation (5G) provides new opportunities and applications, which were not possible to exploit previously. In addition to novel technologies, also infrastructure management has been influenced by 5G. Micro-Operator business model was identified by University of Oulu and Nokia Oulu in 2017. They observed that TelCo stakeholders actors are no longer based only on a small group of Mobile Network Operators, but new actors are becoming relevant. Specifically, Micro-Operators are involved in the deployment of scalable services within local 5G deployments. In most cases, micro-operators deploy their containerized application by exploiting Kubernetes Orchestrator capabilities. At the same time, they share a pool of resources, which usually are provided by the infrastructure owner. Since current standards require strict data privacy and security mechanisms, hard multi-tenancy should be enforced in the design phase. Many solutions are available within the Kubernetes environment. Thus, it would be interesting to analyse how each approach affects network performance, which are very critical metrics in 5G requirements. In the context of this thesis, test and evaluation strategy consisted of deploying a set of applications with every identified hard multi-tenancy approach. The choice of the applications was based on the 5G application classes: extreme multi broadband, massive scale communication, ultra-reliable low latency communication. Once the deployment configurations were defined, data of specific critical application network performance metrics was collected for each of the approaches. By looking at the retrieved data, it is possible to observe that network performance is mainly affected by the application protocol and the corresponding transport protocol. Nevertheless, it is not possible to perform design choices only by looking at network performance. In conclusion, further investigation on other design characteristics may be required

Description

Supervisor

Manner, Jukka

Thesis advisor

Zizka, Jan

Keywords

5G, kubernetes, multi-tenancy, small cells, micro-operators, edge computing

Other note

Citation