2021 Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit): Network Softwarisation

(NET)

Monitoring as a Service over a 5G Network Slice

2021 Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit) | 978-1-6654-1526-2/21/$31.00 ©2021 IEEE | DOI: 10.1109/EUCNC/6GSUMMIT51104.2021.9482534

Dimitris Giannopoulos Panagiotis Papaioannou

dept. of Electrical and Computer Engineering dept. of Electrical and Computer Engineering
University of Patras University of Patras
Patras, Greece Patras, Greece
[email protected] [email protected]

Christos Tranoris Spyros Denazis

dept. of Electrical and Computer Engineering dept. of Electrical and Computer Engineering
University of Patras University of Patras
Patras, Greece Patras, Greece
[email protected] [email protected]

Abstract—Network slicing, i.e. multiple virtual networks infrastructure. This infrastructure will be accessible by means
running on a common infrastructure, is considered a key of network slices. Network slices, namely logical networks,
mechanism to serve the multitude of tenants (e.g. vertical including transmission, network functions and applications,
industries) targeted by forthcoming fifth generation (5G) are layered on top of a physical network. More specifically,
systems, in a flexible and cost-efficient manner. It is predicted
that one of the most popular models for customers will be the
each network slice is a collection of core network (CN) and
Network Slice as a Service (NSaaS) model. This model allows a radio access network (RAN) functions whose settings are
Network Service Customer to order and configure a Network configured to meet the diverse requirements of given use
Slice and offered it as a service. However, ensuring that the cases in terms of functionalities and delivery performance.
agreed SLA for the delivered Network Slice is fulfilled, Many telecommunication operators have shown particular
monitoring and telemetry mechanisms need to be enabled at interest in this technology, in order to boost their revenues,
various resource levels: the Programmable Infrastructure, the allowing the deployment of new services on top of their 5G
5G Core and RAN and the resources of the NS. This work infrastructure [5].
presents an approach to deliver such monitoring and telemetry Following the Anything-as-a-Service (XaaS) paradigm [6],
mechanisms as a service for the emerging 5G NSaaS delivery
model.
strong attention has been drawn to the Network Slice-as-a-
Service (NSaaS) model. NSaaS is able to map service
demands from a customer to functionalities, topology,
Keywords—Network Slicing, Network Slice as a Service,
policies, and parameters of a network slice automatically, as
Monitoring as a Service, 5G
well as to provide component-based and auto-configured
I. INTRODUCTION network functionalities for operators to design and launch
network services more conveniently. Some of the examples
Forthcoming fifth generation (5G) systems are expected to be
of NSaaS models include the Service Blueprints as defined in
highly flexible and programmable end-to-end
[7], [8] or the Generic Slice Template as defined by GSMA
communication, networking, and computing infrastructures.
[9]. Under the NSaaS model, the network slices are intended
They are expected to able to meet the diversified
to be delivered to verticals as services. Following this model,
requirements from multiple services. Such services may
allows each customer to use the provided slice to set up and
require high multi-Gbps peak data speeds, ultra-low latency,
run one or more use case with complete independence.
massive device connectivity, reliability, increased network
Furthermore, an operator is allowed to compose and operate
capacity, increased availability, and data-driven insights. At
different network slices in parallel, such as the case of hosting
the same time, 5G systems are expected to operate in a
multiple enterprises, while guaranteeing slice. This means
flexible and cost-efficient manner, while also serving
that data communication in one slice does not negatively
multiple tenants (e.g., verticals). The International
affect services in other slices [2]. Network slicing realizes the
Telecommunication Union (ITU) has broken the services
service separation for multi-tenancy, so as to virtually build
provided by 5G into major use cases [1]. These include
an exclusive network for each tenancy.
enhanced mobile broadband (eMBB) including ultra-high
One of the main cornerstones of network slicing is NFV.
definition (UHD) TV, massive machine-type
ETSI NFV has defined an architectural framework [10],
communications (mMTC) for metering, logistics, smart
where the NFV technology contains the general-purpose
agriculture, and ultra-reliable and low latency
processor platform, the cloud operating system, the
communications (URLLC) for autonomous driving and
hypervisor, distributed computing, and the software of
automated factory.
network elements. It decouples hardware from software and
The Third Generation Partnership Project (3GPP) [2] and
shields the hardware details for virtual network functions
other organizations, such as the Next Generation Mobile
(VNFs). More specifically, there is a management and
Network Alliance [3] and the 5G Public-Private Partnership
orchestration (MANO) block, which plays the role of
(5G-PPP) [4], have agreed on the convergence of mobile
controlling the lifecycle of VNFs, as well as the connectivity
broadband and vertical sectors onto a common physical
among them. The idea is that a customer orders a network

978-1-6654-1526-2/21/$31.00 ©2021 IEEE 329

Authorized licensed use limited to: UNIVERSIDADE FEDERAL DO RIO GRANDE DO NORTE. Downloaded on December 10,2021 at 13:52:49 UTC from IEEE Xplore. Restrictions apply.
 2021 Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit): Network Softwarisation
(NET)

slice from the OSS/BSS platform and that order is then passed APIs [22]. At the VNF layer, a built-in monitoring function
onto the MANO platform to enable the brokering of NFV (e.g., VNF Event Stream (VES) agent) or a VNF Manager
resources. (VNFM) takes responsibility for monitoring VNFs. At the NS
From the point of view of a vertical customer, the slice is seen and E2E-Slice layer, external virtual probes (vProbe) (aka.,
as an isolated and tailored experimentation platform. Each virtual test agents (VTA)) are deployed to measure the
vertical needs to have only their own services exposed to performance of a NS within a domain or the performance of
them. This exposure can be defined as the ability of the an E2E slice [23]. It is clear that the aforementioned solutions
facility service provider to securely expose the management offer enough monitoring data in layers, but they lack
capabilities of every slice instance towards the vertical cohesiveness, no single point of view is offered, through
customer. This requires proper isolation mechanisms to be which to aggregate all data and have a clear holistic image.
embedded in the slice instance in order to avoid security and There is a prevalent fragmentation caused by the component-
privacy breaches, say, between slice instances used by other specific features.
vertical customers. By regulating the service exposure, the One well established example of a vendor’s monitoring
facility service provider can define the degree of visibility architecture is described in [24]. The authors use Nagios [25]
and control the vertical can have of the respective slice as the core of their proposed monitoring solution, which
instance. In the 5G-VINNI white paper [11], the authors offers great flexibility and configurability. Such a solution
define different exposure levels in 5G-VINNI highlighting could easily be offered as a set of VNFs composing a
that higher the exposure level, the more advanced capabilities Network Service. However, there are specific drawbacks to
the vertical can get. this solution, such as the fact that the Nagios system
Network slices are highly complicated and both operators and configuration is based on complicated text files and most
verticals need to constantly monitor various points in the importantly, it does not have an integrated database of the
network. Building on the approach proposed by the authors past performance records [26]. Another approach is that of
in the 5G-VINNI white paper, in this paper, we propose a 5GTANGO [27]. The 5GTango monitoring framework
framework, which provides Monitoring-as-a-Service at gathers and processes data from many sources, enabling the
different access levels over a 5G network slice, which offers developer to activate measurements and thresholds to capture
a holistic telemetry image to the vertical, providing metrics at generic or service-specific behaviors. In addition, the
various points of the network using multiple telemetry developer can define rules based on metrics collected from
sensors. In addition, our purpose is to map the monitoring one or more VNFs in one or more Network Functions
access levels that our proposed framework provides, with the Virtualization Infrastructure (NFVIs) to receive run-time
exposure levels defined by the authors of the 5G-VINNI notifications. Another solution aimed at operators is proposed
white paper [11]. in [28], where the authors propose an architecture to obtain
and aggregate monitoring data for the aim of performing per-
II. RELATED WORK slice monitoring.
In order to contend with the growing technological In the proposed solution, we will be using Prometheus and
complexity, 3GPP has suggested that different vendors Netdata as our monitoring tools of choice, in order to collect
provide different service assurance (SA) functions [12]. data at the VNF level. Considering that Netdata is relatively
Accordingly, monitoring, as the basic SA function, has been simple to configure, the user interface is extremely user-
offered by many NFV vendors [13]. Monitoring tools have friendly, modular and interactive, as well as actively kept up
long been used for tracking resource utilization and the to date by the developers, it makes for a great choice for the
performance of systems and networks. General purpose core general-purpose monitoring tool of our proposed
infrastructure monitoring tools have, for a long time, been solution. Combining that with the integrated metrics database
utilizing a client–server model by installing an agent in every and widespread use of Prometheus, makes for a well-rounded
system to be monitored. These monitoring agents measure the monitoring solution.
metric values from monitored components and send them to
the monitoring server. The server then goes on to store the TABLE 1: THE DIFFERENT EXPOSURE LEVELS IN 5G-VINNI
HIGHLIGHTING THAT HIGHER THE EXPOSURE LEVEL, THE MORE
collected metrics into a database for further analysis and
ADVANCED CAPABILITIES THE VERTICAL CAN GET [11]
sends alerts. It may then generate graphs, trending reports and
SLA reports based on the monitored metrics retrieved from 5G-VINNI is able to consume Level Level Level Level
the database. After a monitoring agent is initiated for operations related to ... 1 2 3 4
collecting metrics, it measures the relevant metric values E2E network slice application
from the monitored components and passes them on to the ✓ ✓ ✓ ✓
layer config & management
monitoring server. Depending on the configuration of the Network slice subnet / network
system, the server may send alerts to interested parties on function application layer config ✕ ✓ ✓ ✓
occurrence of an event [13]. & management
Network slice subnet / network
At the Infrastructure layer, the Virtualized Infrastructure function virtualized resource layer
Manager (VIM) OpenStack [14] has Ceilometer [15], config & management → ETSI ✕ ✕ ✓ ✓
Gnocchi [16] or Aodh [17] to monitor the infrastructure [18]. NFV Network Service (and VNF)
At the orchestration layer, the ETSI compliant OSM [19] orchestration
collects data from integrated monitoring functions (e.g. Infrastructure resource control &
management → NFVI with
MON) and aggregates and presents the data per tenant ✕ ✕ ✕ ✓
optional EPA capabilities and
(Prometheus [20], Grafana [21]) through the appropriate infrastructural SDN control

330
Authorized licensed use limited to: UNIVERSIDADE FEDERAL DO RIO GRANDE DO NORTE. Downloaded on December 10,2021 at 13:52:49 UTC from IEEE Xplore. Restrictions apply.
 2021 Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit): Network Softwarisation
(NET)

In the following sections the remaining of this paper is

organized as follows. We present the motivation of this work
and the approach we took, and we introduce our framework
and discuss its architecture and the various monitoring access
levels provided, as well as their relationship to the exposure
levels defined in [11] (Section III). Next, we describe our
framework’s architecture and the testbed we used to
implement the framework, as well as a use-case we chose to
showcase the framework in action (Section IV). Finally, we
conclude the paper by presenting our insight and our future
research directions (Section V).
III. APPROACH
Today, there is an emerging need from Vertical applications
of deployed Network Slices for enhanced telemetry Figure 1 Framework architecture and matching between
capabilities; especially in 5G systems, that not only involves monitoring access levels and 5G-VINNI exposure levels
multiple resources but also resources that reside on different
resource access levels. Depending on the level of access a expose to vertical applications access to the 5G core itself or
vertical has to the network slice, their telemetry capabilities the underlying orchestrators and cloud infrastructure.
should perfectly match their needs. In order to meet this To accomplish the above, one must adopt a multi-tenancy
requirement, we propose a framework, which uses “telemetry approach for each of the levels. 5G slicing concept is also
sensors” applied to multiple levels of the stack. In addition, making this isolation easier since at least the user plane is
we have designed our framework in such a way, as to match isolated per tenant, while for the 5G system, it implies a
the definition of the 5G-VINNI “exposure levels”. More requirement that the 3GPP defined NWDAF function [29]
specifically, the exposure levels are seen in Table 1. must support multitenancy access. Our aim is to provide the
Depending on the vertical’s needs, the proposed monitoring aforementioned framework as-a-Service in the spirit of
framework provides the appropriate telemetry access level. Monitoring-as-a-Service, as a core capability of the Network-
Furthermore, the so called “telemetry sensors” are provided Slice-as-a-Service.
on-demand to the vertical, when they need them to adjust their IV. ARCHITECTURE AND IMPLEMENTATION
overlay Network Slice during operation while observing the
underlying 5G infrastructure. These telemetry sensors are A. Architecture
provided in the form of monitoring agents or probes. In our In this section, we will be explaining some prototype technical
proposed architecture we provide monitoring data at four specifications and implementation of our framework. At
different Access Levels, as seen in Table 2. Measurements at access level 1 of our framework, the customer is given access
various network layers of different access levels are taken, to a portal, which provides monitoring data on their network
thus providing a holistic telemetry approach, thanks to the service. The monitoring data provided are collected by a
multiple telemetry sensors offered. Netdata module [30], [31], which offers a very high
The defined access levels imply the different requirements on granularity (up to 1s). The data is then pulled by a Prometheus
resource access a vertical might have. The reason is that for module [20], which is a widely known and used monitoring
most infrastructure providers, while it is easy to provide in- open-source tool. In case the customer’s network service
slice isolated access for monitoring, it is very difficult to consists of multiple VNFs, the total of the telemetry data
streams is aggregated into a single Prometheus node. At
access level 2, the customer is provided access to a portal
TABLE 2: THE DIFFERENT ACCESS LEVELS IN OUR PROPOSED similar to the access level 1 portal, with the addition of a set
FRAMEWORK HIGHLIGHTING THAT THE HIGHER THE of monitoring data streams that correspond to the monitoring
ACCESS LEVEL, THE MORE ADVANCED MONITORING
CAPABILITIES THE VERTICAL CAN GET of the physical infrastructure of the 5G system, namely the
gNodeBs. This is achieved through the use of a custom made
Monitoring Services Provided
Level Level Level Level Netdata plugin [32], which is installed onto the physical 5G
1 2 3 4 appliance. At access level 3, the customer is given access to
In-slice telemetry access to the
telemetry data provided by the MANO orchestrator. More
service data of their network slice ✓ ✓ ✓ ✓
during operation specifically, we used the ETSI NFV-compliant Open-Source
Telemetry access to the specific MANO (OSM) version SEVEN [33]. Therefore, at access
data analytics of each 5G system level 3, the customer is given access directly to the OSM’s
function, such as AMF, SMF, etc., ✕ ✓ ✓ ✓ integrated Prometheus. Finally, at access level 4, the
as well as 5G Radio Access
Network customer is provided access to a Prometheus module, which
Per tenant telemetry data from the is installed on the bare metal of the infrastructure, which itself
NFV Orchestrator while it
✕ ✕ ✓ ✓ pulls data from a Netdata module, also installed on the bare
monitors the underlying NFV metal, in order to provide insight into the behavior of the
infrastructure
Per tenant bare metal and VM and Facility. The architecture of our framework and its
containers telemetry data from the ✕ ✕ ✕ ✓
cloud facility

331
Authorized licensed use limited to: UNIVERSIDADE FEDERAL DO RIO GRANDE DO NORTE. Downloaded on December 10,2021 at 13:52:49 UTC from IEEE Xplore. Restrictions apply.
 2021 Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit): Network Softwarisation
(NET)

Figure 2 UC at 5G VINNI Patras5G

correspondence with the 5G-VINNI definition of exposure quality UGC (smartphones). The facility, along with the
levels is shown in Figure 1. components is seen in Figure 2. More specifically, the LiveU
It is worth noting that Prometheus comprises a great solution equipment is connected to a CPE (Huawei 5G CPE Pro),
for a time-series database to keep persistent monitoring data, which communicates over 5G with the gNodeB (amarisoft
and that is our solution of choice in this framework. Once Callbox Classic). The gNodeB is directly connected to our
inside Prometheus, these metrics can be queried and displayed Cloud infrastructure, which runs OpenStack and hosts the
using any metric analytic tool capable of retrieving various network services. The monitoring service is seen
Prometheus data. Prometheus is responsible for both through Netdata’s Web interface and it is displayed on a
collecting the data and communicating with the time-series monitor, alongside the actual Video Stream.
database for retrieving data upon request. To provide 5G Standalone Access, the 5G System is based on
the open-source Open5GS solution. Open5GS [39] contains a
B. Testbed Infrastructure
series of software components and network functions that
In order to showcase our monitoring framework, we ran implement the 4G/ 5G NSA and 5G SA core functions. The
experiments on our testbed infrastructure Patras5G [34], [35], complete Open5GS documentation and installation procedure
located in Patras on the University of Patras (UoP) Campus. can be found in [40]. The virtualized core’s VNFs and their
Our VIM of choice is OpenStack [36] Rocky 3.16.1 and, as connectivity are seen in Figure 3.
we mentioned earlier, our NFV Orchestrator of choice is the In order to create our network service, we used Openslice [41].
ETSI OSM version SEVEN. The hardware used for the 5G Openslice is a prototype open source, operations support
capable network in these scenarios is the AMARI Callbox system. It supports VNF/NSD onboarding to OSM and NSD
Classic which acts as a 3GPP compliant eNodeB, gNodeB, deployment management. It also supports TMFORUM
EPC and 5GC also supporting 5G NR in both SA and NSA OpenAPIs regarding Service Catalog Management, Ordering,
mode [37]. This was used as a Physical Network Function Resource, etc. After onboarding our VNFs/NSDs onto
(PNF) for the purposes of the experiment.
C. Implementation
A first deployment of our approach applied on a Media
Vertical from the 5G-SOLUTIONS European project [38] and
specifically on a Use-Case of User Generated Content (UGC)
& Machine Generated Content (MGC). 5G-SOLUTIONS
[38] is the flagship ICT-19 RIA project supporting EC’s 5G
policy by implementing the last phase (Phase 3b) of the
5GPPP roadmap. The project aims to prove and validate that
5G provides prominent industry verticals with ubiquitous
access to a wide range of forward-looking services with orders
of magnitude of improvement over 4G, thus bringing the 5G
vision closer to realization. The proposed UC involved the
generation and sharing of UGC using custom applications
provided by partners, which allowed specific tests to be
executed on the 5G-VINNI Patras5G [35] facility, such as the Figure 3 5G Core architecture
contribution of high-quality live video (4K+) as well as lower

332
Authorized licensed use limited to: UNIVERSIDADE FEDERAL DO RIO GRANDE DO NORTE. Downloaded on December 10,2021 at 13:52:49 UTC from IEEE Xplore. Restrictions apply.
 2021 Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit): Network Softwarisation
(NET)

[2] 3GPP, Technical Specification Group Services and System

Aspects; Feasibility Study on New Services and Markets
Technology Enablers, Stage 1 (Release 14), TR 22.891, 2016.
[3] N. G. M. N. Alliance, 5G white paper." Next generation mobile
networks, white paper 1, 2015.
[4] 5GPPP, White Paper on Automotive Vertical Sector, 5G
Automotive Vision, 2015.
[5] L. M. Contreras, P. Doolan, H. Lønsethagen and D. R. López,
"‘‘Operational, organizational and business challenges for
network operators in the context of SDN and NFV," Comput.
Netw., vol. 92, p. 211–217, Dec. 2015.
[6] Y. Duan and e. al, "Everything as a service (XaaS) on the cloud:
origins, current and future trends," in 2015 IEEE 8th
International Conference on Cloud Computing, 2015.
[7] 5G-TRANSFORMER, "D5.4. Report Trials Results.," 2019.
[8] J. Ordonez-Lucena, C. Tranoris and J. Rodrigues, "Modeling
Network Slice as a Service in a Multi-Vendor 5G
Experimentation Ecosystem," Dublin, 2020.
[9] GSMA, PRD NG.116 v3.0, “Generic Network Slice Template”,
2020.
[10] ETSI, "Virtual Network Function Architectural Framework,"
Paris, France, Nov. 2014.
[11] C. Tranoris, A. Cattoni, D. Warren, K. Mahmood, M. Everett, M.
Figure 4 The Monitoring 5G service from Openslice at Ghassemian, M. Xie, P. Grønsund, A. Gavras, T. Ghosh, D. R.
https://patras5g.eu/services/services_marketplace Lopez, J. Ordonez-Lucena and J. A. Rodrigues, "Onboarding
Vertical Applications on 5G-VINNI Facility," March 2020.
Openslice, we ordered the network services, as seen in Figure [12] 3GPP, System architecture for the 5G System (5GS), TS 23.501.
4. [13] K. Fatema, V. C. Emeakaroha, P. D. Healy, J. P. Morrison and T.
Lynn, "A survey of cloud monitoring tools: Taxonomy,
V. CONCLUSIONS – FUTURE WORK capabilities and objectives," Journal of Parallel and Distributed
Computing, vol. 74, no. 10, pp. 2918-2933, 2014.
In this paper we proposed a framework, which offers a multi- [14] OpenStack Community , "OpenStack," 2020. [Online].
level telemetry solution for Cloud and 5G infrastructures, Available: http://www.openstack.org/. [Accessed 2021].
which closely matches the exposure levels as defined in the [15] "Ceilometer documetnation," [Online]. Available:
5G-VINNI white paper [11]. We examined what each https://docs.openstack.org/ceilometer/latest/. [Accessed Jan.
monitoring access levels provides to the customer and saw an 2021].
indicative application, where the monitoring framework was [16] The Gnocchi Developers, "Gnocchi Documentation," 2018.
used. It is worth noting that our framework, regardless of the [Online]. Available: https://gnocchi.xyz/. [Accessed Jan. 2021].
access level, always provides access to the telemetry data via [17] "Aodh Documentation," [Online]. Available:
https://docs.openstack.org/aodh/latest/. [Accessed Jan. 2021].
the widely used Prometheus module. The fact that the
[18] "Telemetry," Openstack, [Online]. Available:
monitoring data is in Prometheus format, means that it is https://wiki.openstack.org/wiki/Telemetry. [Accessed 2021].
easily machine consumable. [19] "OSM," ETSI, [Online]. Available: https://osm.etsi.org.
We intend to focus our future efforts on the transition of our [Accessed 2021].
framework’s virtual machines to containers for a more robust [20] "Prometheus," [Online]. Available: https://prometheus.io/.
and agile deployment. In addition, we aim to incorporate [Accessed 2021].
policies and mechanisms that check and avoid the violation [21] "Grafana," [Online]. Available: https://grafana.com/. [Accessed
of the Service Level Agreements (SLAs), by triggering 2021].
autoscaling procedures, which depend on the measurements [22] "OSM Usage - Monitoring and autoscaling," ETSI, OSM,
taken at the appropriate level. [Online]. Available: https://osm.etsi.org/docs/user-guide/05-osm-
usage.html#monitoring-and-autoscaling. [Accessed 2021].
ACKNOWLEDGMENT [23] M. Xie, Q. Zhang, P. Grønsund, P. Palacharla and A. J.
Gonzalez, "Joint Monitoring and Analytics for Service Assurance
This work is supported by the H2020 European Projects 5G- of Network Slicing," in IEEE Conference on Network Function
VINNI (grant agreement No. 815279), 5G-VICTORI (grant Virtualization and Software Defined Networks (NFV-SDN):
agreement No. 857201), 5G-SOLUTIONS (grant agreement O4SDI - 4th Orchestration for Software-Defined Infrastructures,
2018.
No. 856691), and HELNET (MIS 5002781) co-financed
project by Greece and the EU. [24] C. Issariyapat, P. Pongpaibool, S. Mongkolluksame and K.
Meesublak, "Using Nagios as a groundwork for developing a
better network monitoring system," in Proceedings of
REFERENCES PICMET'12: : Technology Management for Emerging
Technologies. IEEE, 2012.
[25] Nagios Enterprises, "Nagios," [Online]. Available:
[1] M. Series, "IMT Vision–Framework and overall objectives of the https://www.nagios.org/. [Accessed 2021].
future development of IMT for 2020 and beyond," [26] S. Mongkolluksamee, P. Pongpaibool and C. Issariyapat,
Recommendation ITU 2083, 2015. "Strengths and Limitations of Nagios as a Network Monitoring
Solution," in Proceedings of the 7th International Joint

333
Authorized licensed use limited to: UNIVERSIDADE FEDERAL DO RIO GRANDE DO NORTE. Downloaded on December 10,2021 at 13:52:49 UTC from IEEE Xplore. Restrictions apply.
 2021 Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit): Network Softwarisation
(NET)

Conference on Computer Science and Software Engineering,

Bangkok, Thailand, May 2010.
[27] C. Parada, J. Bonnet, E. Fotopoulou, A. Zafeiropoulos, E.
Kapassa, M. Touloupou and a. et., "5GTANGO: A Beyond-
MANO Service Platform," in 2018 European Conference on
Networks and Communications (EuCNC): Network
Softwarisation (NET), 2018.
[28] R. Montero, N. Calabretta, F. Agraz, A. Pagès and S. Spadaro,
"SDN-controlled and orchestrated OPSquare DC architecture
enabling network slice deployment with QoS guarantees,"
Photonics in Switching and Computing (PSC), IEEE, pp. 1-3,
September 2018.
[29] 3rd Generation Partnership Project, "TS 29.520
V15.8.0;Technical Specification Group Core Network and
Terminals; 5G System; Network Data Analytics Services;Stage
3(Release 15)," 3GPP, Dec. 2020.
[30] Netdata, Inc., "Netdata," 2021. [Online]. Available:
https://www.netdata.cloud/. [Accessed Jan. 2021].
[31] Netdata, Inc, "Netdata on Github," Jan 2021. [Online]. Available:
https://github.com/netdata/netdata. [Accessed Jan. 2021].
[32] Netdata, Inc., "External plugins overview," 2021. [Online].
Available:
https://learn.netdata.cloud/docs/agent/collectors/plugins.d.
[Accessed Jan. 2021].
[33] "OSM," [Online]. Available: https://osm.etsi.org/. [Accessed Jan.
2021].
[34] "Patras5G wiki," [Online]. Available: https://wiki.patras5g.eu/.
[Accessed Jan. 2021].
[35] C. Tranoris and S. Denazis, Patras 5G: An open Source based
End-to-End facility for 5G trials, ERCIM NEWS, Special Theme
5G, Number 117, April.
[36] "Openstack," [Online]. Available: http://www.openstack.org/.
[Accessed Jan. 2021].
[37] amarisoft, "AMARI Callbox Classic," amarisoft, 5 Feb. 2020.
[Online]. Available:
https://www.amarisoft.com/app/uploads/2020/02/AMARI-
Callbox-Classic.pdf. [Accessed Jan. 2021].
[38] "5G-SOLUTIONS," [Online]. Available:
https://www.5gsolutionsproject.eu/. [Accessed Jan. 2021].
[39] "Open5GS," [Online]. Available: https://open5gs.org/. [Accessed
Jan. 2021].
[40] L. Sukchan, "Open5GS documentation," [Online]. Available:
https://open5gs.org/open5gs/docs/guide/01-quickstart/. [Accessed
Jan. 2021].
[41] C. Tranoris, "Openslice: An opensource OSS for Delivering
Network Slice as a Service," 2021. [Online]. Available:
https://openslice.readthedocs.io/en/stable/.

334
Authorized licensed use limited to: UNIVERSIDADE FEDERAL DO RIO GRANDE DO NORTE. Downloaded on December 10,2021 at 13:52:49 UTC from IEEE Xplore. Restrictions apply.