2024

Cloud-native network
transformation and the
5G edge

Kerem Arsal Roberto Kompany Pablo Tomasi

Senior Principal Analyst, Principal Analyst, Principal Analyst,
Networked Edge Mobile Infrastructure Private Networks
— Omdia — Omdia — Omdia
 Executive 5G slow to reach 5G edge computing now Essential role of network Challenges in the journey Keys to success and the Private 5G networks Conclusion Appendix
summary full potential so far set for strong growth cloudification for CSPs to cloud native networking role of open source and edge computing

Contents

Executive summary.............................................. 3 SECTION 04

Challenges in the journey to

cloud native networking..................................20
SECTION 01
Managing complexity and interoperability are the main
5G slow to reach full potential so far............ 5 challenges to 5G cloud core deployments; business case
This is tied to slow and delayed deployments of 5G SA, clarity is not a concern.................................................................................... 20
which enables new services that can translate into Integration challenges are the primary obstacles to the
new revenue streams........................................................................................6 adoption of open vRAN.................................................................................. 21
Without ‘true’ 5G networks, edge computing on
telecom networks has also struggled to flourish.....................................6
SECTION 05
CSPs count many challenges in their edge deployments.....................7
Keys to success and the role of
CSPs acknowledge their lack of infrastructure readiness
and technical challenges in IT and network domains as their open source...........................................................23
biggest hurdles….................................................................................................8 A common cloud layer can act as a platform of
…and view Opex and Capex as the two trailing obstacles....................8 interoperability to reduce vendor fragmentation................................ 23
Kubernetes is a powerful approach to container
orchestration, especially when supported by services
SECTION 02
that offer visibility, control and management of clusters.................. 24
5G edge computing now set for Telco clouds may run in confined environments due to strict
strong growth.......................................................10 requirements, but CaaS platforms will be agnostic to network
and third-party workloads............................................................................ 25
CSPs value the versatility of edge computing, which
brings both internal and external benefits............................................. 11 CSPs will receive very frequent updates to their applications
through CI/CD principles............................................................................... 25
Confidence in presence of demand breeds confidence
in the future….................................................................................................... 11 Platforms that can support the co-existence of virtualized
and containerized network functions will be critical in
…with applications and premium connectivity at the
managing the transition period.................................................................. 26
edge viewed as key revenue streams in the future............................. 12
Analytics will create network awareness ensuring new
To capture demand, edge infrastructures are set to
differentiated and guaranteed service delivery
expand toward highly distributed computing....................................... 13
and monetization............................................................................................. 27
Omdia expects networked edge uptake to grow over the
next several years, leading to an opportunity of more
than $26bn of IaaS and SaaS by 2032....................................................... 13 SECTION 06

Private 5G networks and

SECTION 03 edge computing...................................................29
Essential role of network Selected examples of spectrum liberalization initiatives................... 30
cloudification for CSPs......................................15 Different private 5G deployment options meet different
enterprise needs............................................................................................... 31
Becoming cloud native is not only to pursue the edge
computing opportunity; it ranks among the highest overall Private 5G and edge computing align well to serve
priorities for CSPs…......................................................................................... 16 the enterprise.................................................................................................... 32
… because it offers many benefits and suggests a paradigm Examples of enterprises leveraging private 5G and
shift for them..................................................................................................... 16 edge computing................................................................................................ 32
In the core, cloudification will bring benefits of agility,
efficiency and scalability................................................................................ 17
RAN is the largest capex spend for mobile operators; its
Conclusion..............................................................33
virtualization will enable both software and geolocation
disaggregation................................................................................................... 17
RAN disaggregation will then lead to centralized computing, Appendix.................................................................34
facilitate CUPS and drive new edge services’ business models....... 18 5G core network functions and definitions............................................ 35
Omdia................................................................................................................... 36
Canonical............................................................................................................. 36

2 Cloud-native network transformation and the 5G edge

 Executive 5G slow to reach 5G edge computing now Essential role of network Challenges in the journey Keys to success and the Private 5G networks Conclusion Appendix
summary full potential so far set for strong growth cloudification for CSPs to cloud native networking role of open source and edge computing

Executive summary

Realizing the full potential of 5G To truly flourish, edge computing on telecom To tackle these challenges and help CSPs in
networks, whether to enable new networks ideally needs connectivity and their journeys toward cloud native networks
computing domains to meet at and build upon and 5G edge, there are six key elements of
services and revenue streams or to
a common denominator of containerization success:
achieve operational efficiencies and and cloud native principles. CIRCLE-1 A common cloud layer to cut across
scalability of applications, remains CSPs recognize this, and view vendor siloes and take advantage of a
one of the biggest challenges for cloudification of both the core and RAN to broad ecosystem
communication service providers (CSPs) be among the biggest technological drivers CIRCLE-2 Automation and orchestration
around the world. According to a recent of edge computing. that leverages Kubernetes as the
global analysis by Omdia, there has been no With 5G SA (standalone) core, CSPs can underlying force
statistical relationship between mobile ARPU localize and scale user plane function CIRCLE-3 Workload-agnostic cloud platforms /
changes and 5G uptake between 2020 and (UPF) instances according to demand, CaaS for network functions and IT
2023. Arguably, this is largely because most and efficiently steer and break traffic CIRCLE-4 Carrier-grade support for continuous
5G deployments have been non-standalone out where needed and depending on integration and deployment (CI/CD)
(NSA), which means the networks continue applications. CIRCLE-5 Transition management between
to leverage the 4G core. Further they are not In the RAN, developments around virtualization and containerization
built using cloud native cores, limiting the network disaggregation, cloudification, CIRCLE-6 Network awareness and intelligence
network’s agility, resiliency and scalability and open RAN boost the possibility of for better customer experience and
and therefore cannot natively support key using shared hardware and software for new services
capabilities, such as 5G edge computing, both network functions and IT workloads
which can stimulate innovative use cases on the same infrastructure. Open-source technologies can play a major role
and drive revenue growth. Private 5G networks are also ranked in achieving these elements.
highly by CSPs as a driver of edge By acting as vendor-neutral cloud layers,
Despite a slow initial uptake, CSPs report open source platforms can bridge different
great optimism about the potential of 5G
computing; these networks also rely on
edge computing and plan to expand their cloud cores and local UPFs in enterprise hardware environments and network
edge infrastructures toward highly distributed environments. applications. They can simplify automation
computing fabrics suitable for the era of and orchestration built on Kubernetes,
artificial intelligence. Moving to cloud native networks is a top which is inherently open source anyway.
When computing comes closer to end transformational priority for CSPs but there are With support for both virtualized and
users and devices, and it is integrated challenges to resolve. containerized workloads, they can provide
with network functions, workloads can be According to Omdia surveys, managing unified platforms to host and orchestrate
optimized for low latency and bandwidth multiple clouds and hybrid environments, legacy and cloud native applications. With a
savings (e.g., relative to public cloud) clashes between cloud and legacy vast ecosystem of developers and already
while applications can run near or within systems, and interoperability across strong adoption, open source can become
customer premises to satisfy privacy, vendors are key concerns both in the core a true engine of innovation for CSPs in their
control and performance requirements. and RAN domains. journeys to cloud native networking.

96% x12
CSPs* that will
Projected growth of
have launched
installed edge servers
fully commercial
on public telecom
edge computing
networks between
deployments within
2023 and 2032
the next two years

99% #2
Priority ranking of
CSPs* that expect
‘becoming a cloud
some or substantial
native organization’
revenue growth from
among CSPs (out of
edge computing
nine priorities asked)

* In the US, the UK and Germany

3 Cloud-native network transformation and the 5G edge

 Executive 5G slow to reach 5G edge computing now Essential role of network Challenges in the journey Keys to success and the Private 5G networks Conclusion Appendix
summary full potential so far set for strong growth cloudification for CSPs to cloud native networking role of open source and edge computing

SECTION 01

“5G monetization has been poor

due to a lack of 5G SA deployments,
which in turn limited the uptake of
edge computing and the arrival of
new revenue-generating use cases
and services”
 Executive 5G slow to reach 5G edge computing now Essential role of network Challenges in the journey Keys to success and the Private 5G networks Conclusion Appendix
summary full potential so far set for strong growth cloudification for CSPs to cloud native networking role of open source and edge computing

5G slow to reach full

potential so far
There has been no clear relationship between growth in 5G
uptake and changes in mobile ARPU.

FIGURE 1
Relationship between % of change in 5G penetration and % of change in ARPU, 2020-23

25%

Countries with
ARPU growth India
20%

15% Bulgaria

Israel
Qatar
10% Bahrain

Philippines

Kuwait
5% Croatia Australia

Hong Kong
New Zealand Taiwan
% of change in mobile ARPU

Canada
0%
Denmark
Portugal Singapore
Slovakia US Finland R2=0.022
Czech Republic Switzerland
-5% Hungary Greece Oman
UAE
UK

France Chile Austria

-10% Spain

Romania
South Korea Saudi Arabia
-15% Belgium Sweden
China

Poland Italy Norway

Germany
-20% Thailand
Countries with
ARPU drop Netherlands

-25%

-30% Ireland Japan

R2 denotes statistical significance of

relationship between variables;
between 0.5 and 0.99 is acceptable
-35%
0ppt 5ppt 10ppt 15ppt 20ppt 25ppt 30ppt 35ppt 40ppt 45ppt

Ppt increase in 5G penetration

SOURCE: OMDIA, WORLD CELLULAR INFORMATION SERIES DATABASE 2023

5 Cloud-native network transformation and the 5G edge

 Executive 5G slow to reach 5G edge computing now Essential role of network Challenges in the journey Keys to success and the Private 5G networks Conclusion Appendix
summary full potential so far set for strong growth cloudification for CSPs to cloud native networking role of open source and edge computing

This is tied to slow and delayed deployments of 5G SA, which

enables new services that can translate into new revenue streams
Since 2019, the number of 5G deployments 15% of all 5G deployments worldwide. use cases that 5G SA can, such as those
has grown around the world. Of the 700+ Surprisingly, this ratio has remained requiring ultrareliable low-latency or those
mobile operators covered in Omdia’s unchanged between 2022 and 2023. In powered by multi-access edge computing
database, those with commercial 5G 2023, more than 40 5G NSA networks were or network slicing.
offerings have increased from 75 to nearly launched vs. 12 5G SA additions. In short, 5G NSA, while a differentiating
300 by the end of 2023. Most of these While this approach has allowed CSPs label to market, only has limited actual
deployments have been non-standalone to pace their investments and to introduce end user appeal, and therefore limited
(NSA) versions of 5G, i.e., they continue to a new network generation into their potential for sustained ARPU uplift. In fact,
rely on 4G core (EPC) and are largely radio markets (even if by name), 5G NSA brings according to Omdia’s data, there has been
overlays on 4G infrastructures. only modest performance improvements no clear relationship between the rise of
According to Omdia’s Mobile Core over 4G, relative to what an end-to-end 5G 5G penetration and changes in mobile
Market Tracker, by the end of 2023, there network with a 5G core (SA) can achieve. ARPU around the world (see Figure 1).
were only 47 5G SA rollouts, comprising Moreover, 5G NSA does not enable many

FIGURE 2
Mobile operators and 5G deployments worldwide, 2019-2023

2019 75 631

2020 128 571

2021 178 506

2022 38 205 463

2023 47 247 412

0 100 200 300 400 500 600 700 800

No. of mobile operators

 Standalone (SA)  Non-standalone (NSA)  No 5G

SOURCE: OMDIA, MOBILE CORE MARKET TRACKER 2H 2023, GSA

Without ‘true’ 5G networks, edge computing on telecom

networks has also struggled to flourish
‘True’ 5G standalone (SA) networks telecom networks. This, in turn, limits the Omdia estimated only around 100K
should combine cloud native cores with new revenue opportunities for CSPs. networked edge / MEC servers installed
virtualized / containerized RAN (NR), Not surprisingly, without sufficient on public telecom networks by the end
enabling a digital fabric of applications and cloudification of mobile networks, of 2023.
services on highly distributed computing neither at the core nor the RAN, edge Only a negligible number of vRAN units
nodes at the network edge, supported by computing uptake on these networks has been used for anything but RAN
capabilities, such as slicing. has also been limited. Many CSPs have workloads in the same period. vRAN
With only 5G NSA, some performance launched commercial edge products, and Open RAN are both journeys to
improvements are achieved, but the but these are often served only on a few cloudification, but they are still too
paradigm does not shift to the “fusion centralized nodes. nascent to expand to edge computing
of connectivity and computing” that true For the last few years, networked edge for third parties.
5G networks can bring. Without this step / MEC has been a highlight expectation of Demand from enterprises has stayed
change, third-party cloud infrastructures the type of infrastructure and computing limited, with only 4% of large enterprises
and innovative applications have no architecture that 5G would enable. claiming some adoption for 5G edge
incentive to benefit from and flourish on However: computing / MEC.

6 Cloud-native network transformation and the 5G edge