CBL - Campus del Baix Llobregat

Projecte llegit

Títol: Design and development of a framework to standardize the configuration of 5G radios and cores


Estudiants que han llegit aquest projecte:


Director/a: GARCÍA LOZANO, MARIO

Departament: TSC

Títol: Design and development of a framework to standardize the configuration of 5G radios and cores

Data inici oferta: 15-12-2021     Data finalització oferta: 15-07-2022



Estudis d'assignació del projecte:
    DG ENG SISTE/TELEMÀT
Tipus: Individual
 
Lloc de realització: Fora UPC    
 
        Supervisor/a extern: Akshay Jain
        Institució/Empresa: Neutroon Technologies S.L.
        Titulació del Director/a: Doctorate in Network Engineering
 
Paraules clau:
5G, private networks, radio configuration, mobile network standardization, RAN telemetry
 
Descripció del contingut i pla d'activitats:
This project aims to create a single framework for standardizing the configuration of the radio or core from any vendor
providing 5G mobile network services.

Currently, if we want to work with one radio vendor it is necessary to study the parameters that they have, the data
structure that we need to change the configuration and learn about the method / standard that they use to configure the
radio (API, TR069'). Such a method can create delays in the deployment of 5G networks. With this project we want to create
a unified framework, which will standardize the structure of the data to be sent to the radio in order to configure it
whilst being agnostic of the vendor. The final user will not need to learn about each vendor as well as the operational
standards for communicating with the radio. The idea is to know only how this framework works (what is the information that
we need to send, what is the format that we need to send etc.) and knowing it one will be able to work with any vendor. In
this way, it will need only to study one unique structure to configure any radio, and not learn about each specific vendor.
Additionally, within this project we will also obtain the radio metrics from each radio type to analyze them. A further
study, if time permits, will be to apply machine learning techniques to optimize some of the configuration of the different
radios automatically.

STRUCTURE:
1. Thorough analysis of existing communications standards, parameters as well as frameworks. This will also include
detailed background study on the 5G standards, Dockers, Containerized solutions and Kubernetes.
2. A unified framework definition will be created wherein the explanation for each of the building blocks of the framework
will be provisioned.
3. Creation of a unified framework which encompasses the Netconf/Yang model, TR-069 model and the API calls and radios from
Amarisoft, Accelleran and Sunwave (BTI wireless) will be integrated.
4. In the subsequent step an educational video will be created and posted on Youtube. Furthermore, a demo paper will be
written and submitted for potential publication and presentation in a related conference.
5. For the part of monitoring framework, the Prometheus exporter will be utilized and a scalability test will be executed
to analyze the load that the telemetry data will generate given the number of UEs, radios, etc. Results are expected to be
submitted to a journal.
 
Overview (resum en anglès):

Private networks based on the 5th generation (5G) of mobile telephony are gaining a
lot of importance in the Industry due to the use cases they support. Until now, private
networks were set up using Wireless Fidelity (WiFi) technology, a technology that was
easy to configure and use. However, due to technological progress in recent years, private
networks are forced to use more powerful technology such as 5G.
5G technology is more complex due to its architecture and the number of parameters to
configure. On the other hand, it is a very flexible technology, since most of the components
are pieces of software that can be installed in different parts of the network, depending on
the interests of the customer. This comes with the disadvantage that the number of possible
architectures increases, and with it the complexity of managing this type of network.
In addition, in the case of private networks, there are many providers that offer 5G networks,
or different parts of it, which, on the one hand, increases flexibility when choosing
the providers that will be used to set up the private network, but on the other hand, it makes
it necessary to study how each of them works to configure the different components. This
requires some study time to see how to deploy and configure the components of each
provider that will be used in the private network.
This project is developed in the context of the Neutroon company, whose main objective
is to create a platform that solves this type of problem. It allows, in a very simple way, to
manage private networks that use WiFi, Long Term Evolution (LTE) and 5G technologies.
Neutroon not only reduces the complexity of deploying and configuring private networks,
but also centralizes in a single platform the management of the different applications that
can be running in this same network. In addition, Neutroon has the flexibility to manage
the network regardless of the provider or providers chosen to set up the private network,
since it has several of them integrated into its platform so that it is not a limitation for the
client.
The objective of this document focuses on the radio configuration part of private networks,
essentially the configuration of the different components that make up a 5G private network.
For this, a study has been carried out of the current architecture that Neutroon uses
to integrate different providers and of the scalability and maintenance problems that it may
have as Neutroon integrates more components and more providers.
To solve this problem, a new component has been added within Neutroon¿s own architecture
that will greatly simplify the work when configuring the different parameters of a radio
from the different providers, whether 5G or WiFi, creating a model of standard data so that
it is not necessary to learn the data model of each provider and it is only necessary to
focus on the values that should be assigned in the radio devices. In addition, it is a very
easy component to integrate with other pieces of software, such as Neutroon, so that it
can also take advantage of this component and reduce the complexity of its architecture


© CBLTIC Campus del Baix Llobregat - UPC