UML-based Modeling and Analysis of 5G Service Orchestration
| Autor: | Ashalatha Kunnappilly, Cristina Seceleanu, Peter Backeman |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Service (systems architecture)
Computer science Quality of service Distributed computing 020206 networking & telecommunications 02 engineering and technology Service-level agreement Unified Modeling Language 0202 electrical engineering electronic engineering information engineering Orchestration (computing) Routing (electronic design automation) computer Virtual network computer.programming_language Object Constraint Language |
| Zdroj: | APSEC |
| DOI: | 10.1109/apsec51365.2020.00021 |
| Popis: | The fifth generation of cellular wireless technol- ogy, 5G, bears the promise to transform the future network connectivity by providing seamless, low-latency and reliable interconnections between devices. In this paper, we focus on modeling and analyzing 5G service orchestration that deals with virtual network function placement, resource assignment and traffic routing, which are the building blocks of generating network slices catering to various application requirements. In order to ensure that a particular network slice works as stated by the application's service level agreement, it is essential that the constituent virtual network functions are placed in proper hosts, allocated adequate resources in terms of processing power, memory, bandwidth, and routed such that the constraints of the hosts and the network are met. This is a complex problem to solve if one considers the diverse set of requirements of 5G services. We tackle this problem by proposing a UML-based modeling and analysis framework, called UML5G Service Orchestration Profile, which allows one to describe 5G network slices and service orchestration via a specialized profile, and analyze as-sociated quality-of-service requirements by checking constraints expressed in Object Constraint Language. Our framework allows a designer to model any candidate orchestration scheme for 5G networks and verify if the network function placement, resource assignment, and routing guarantee the application's quality-of-service requirements, at design time. We evaluate the framework on a prototype implementation of an orchestration algorithm that generates a multitude of allocation configurations that we automatically check against requirements formalized in Object Constraint Language. Our contribution facilitates modeling and design-time evaluation of network slicing and service orchestration schemes in 5G-based solutions. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|