ericsson.

com/
en/ai-and-automation

Employing
AI techniques to
enhance returns
on 5G network
investments

Ericsson AI & Automation

2 Ericsson  |  Employing AI techniques to enhance returns on 5G network investments

Contents

Methodology
03 Key findings This report is based on a research survey “Machine learning
performed during October and November
04 AI technology adoption
2018. Telephone interviews were conducted is assisting our
06

Impact of AI adoption
within networks
with 165 senior executives from 132 mobile engineers to make
communications service providers globally.
08 AI opportunities for
The respondents were predominantly our mobile networks
Chief Technology Officers (CTOs) and
service providers
Chief Operating Officers (COOs), but also
more proactive
09 Challenges for successful included Chief Information Officers (CIOs), and immediately
adoption of AI Chief Marketing Officers (CMOs) and
Chief Financial Officers (CFOs) as well as responsive to the needs
10 Successful implementation of AI some Line of Business (LOB) managers. of our customers.”
11 Recommendations To conduct the survey, Ericsson
commissioned Coleman Parkes Research, Head of Radio Products,
Tier-1 Operator, Europe
which specializes in recruiting and
interviewing senior-level responders
across multiple global markets, vertical
sectors and functional areas for a wide
range of clients.

AI will be essential for processing the vast volume of data required to provide enhanced, personalized services to customers
3 Ericsson  |  Employing AI techniques to enhance returns on 5G network investments

Key findings

AI is creating both benefits and challenges

for service providers at the advent of 5G.

At a time of increasing complexity for AI is being adopted in networks AI will help recoup network investments
mobile communications service providers, AI is being embraced by service At the advent of 5G, service providers are
driven by deployment of 5G and the providers around the world, making huge investments in their networks
Internet of Things (IoT), this report according to our research. to enable the new use cases that 5G offers.
examines the crucial role of artificial More than half of service providers AI will help operators recoup these
intelligence (AI) in helping to maximize – a total of 53 percent – expect to have investments more quickly. Service providers
returns on network investments. It adopted AI within their networks by the believe the highest potential return from
explores the extent to which service end of 2020. Some are working to an even AI adoption will be in network planning
providers have adopted AI, and looks shorter timescale, and expect to have (70 percent) whilst 64 percent intend to
at their plans for further deployment. adopted AI by the end of this year. A further maximize their returns by focusing their
The importance of AI in developing and 19 percent are looking at an adoption AI adoption efforts on network
underpinning customer relationships timescale of within 3 to 5 years. performance management.
is also examined. It is our belief that The main areas in which service providers
AI will open up exciting opportunities are already seeing value and return AI is creating network data challenges
for the mobile communications sector, from AI are in building new revenue Embracing AI in networks is a challenge
as it can be utilized to create a more and enjoying operational cost savings. for service providers. The main concern
personal approach for customers, identified by 71 percent was defining and
while helping to manage the costs of AI will be vital to customer service implementing standardized interfaces.
deploying and maintaining networks. Enhancing customer experience was Other difficulties highlighted included data
identified by 55 percent of service quality (65 percent); excess of data from
providers as a key area where AI is too many sources (59 percent); problems
presently having the greatest impact finding indicators of degradation or
within core network activities. In addition, compromised assets (59 percent); storage
68 percent highlighted enhancing of data in too many systems (56 percent);
customer service as a business and and lack of single ownership or oversight
operational objective over the next of the data (55 percent).
3 years. A further 72 percent agreed
strongly that AI will be important
in enabling monetization of new
network technologies and providing
a better service to customers.
4 Ericsson  |  Employing AI techniques to enhance returns on 5G network investments

AI technology adoption

Many service providers are already concluding

successful trials on using AI in their networks.
Figure 1: Progress by service providers in achieving benefits through AI Capital expenditure reduction
Operational cost savings
Building new revenue streams
Optimizing network performance, reliability or availability
Creating personalized and adaptive customer journeys

48%
41%
35% 33%
28% 27% 29%
22% 24% 22% 21% 23%
21% 19%
13%

Testing AI applications Not doing but Adopting and implementing

and executing trials plan to initiate trials in networks

16%
12% 12% 14%
8% 7% 6% 6% 6%
3% 1% 1% 1% 1%

Already using successfully Not doing and no plans Used but stopped

AI assesses data and quickly delivers Only 12 percent feel they have a detailed
analytical outcomes to users. It encompasses knowledge of AI's application. However,
Machine Learning (ML), which analyzes 49 percent considered themselves to have a
"There has been an
raw data and autonomously looks for fairly detailed knowledge of AI application. increased complexity
patterns that can yield further insights. More than half expect AI to be adopted
Already, sectors such as manufacturing, in their networks before the end of 2020 in the management
healthcare and banking are capturing the (a total of 53 percent globally) and there is of a vast number of
benefits of AI. For service providers, AI a general expectation (55 percent globally)
offers opportunities that need to be built that the benefits will be evident within devices and huge
jointly with infrastructure providers on a
case-by-case basis, with a common goal
1 to 2 years.
The majority of service providers are
data. AI and ML
to more effectively manage complexity at the stage of testing AI, with 48 percent will help to reduce
and optimize network performance. focusing on AI to reduce capital expenditure.
Our research has revealed service A further 41 percent are focusing on using AI this complexity."
providers around the world are presently for optimizing network performance, and Machine Learning Scientist,
at various stages on their journey with AI. 35 percent for new revenue streams. Asia Pacific region

1
Ericsson Mobility Report (November 2018)
5 Ericsson  |  Employing AI techniques to enhance returns on 5G network investments

40%
40 percent of the world’s
population is expected to
benefit from 5G by 2024.

5G is forecast to cover more than 40 percent of the world's population by the end of 2024, according to the Ericsson Mobility Report (November 2018)

Figure 2: Focus of service providers testing AI

Capital expenditure
reduction
48%

Optimizing network
performance
41%

New revenue
streams
35%

The highest current use of AI among advent of 5G will make network topologies
service providers worldwide is in service relatively complex, with small cells and new
quality management (17 percent) and antennas making usage patterns more
operational cost savings (16 percent). difficult for humans alone to predict, and
The broader context for the interest being current radio propagation models becoming
shown in AI by service providers is that 5G is more complex to compute as a result of new
expected to cover more than 40 percent of the radio spectrum bands, denser topologies,
world’s population by 2024,1 and total mobile Massive MIMO and beamforming.
data traffic is predicted to have increased by
a factor of 5 by this date.
Early adopters of AI among service
providers will undoubtedly gain an
advantage, as they will be well placed to
deal with new challenges that result from the
proliferation of additional devices following
the introduction of 5G. This is because the
6 Ericsson  |  Employing AI techniques to enhance returns on 5G network investments

Impact of AI adoption
within networks

Service providers around the world

are already seeing the impact of AI.
Service providers around the world are
observing improved reliability for customers “AI and machine learning will be key to our customer
as the area in which AI is currently having
the greatest impact upon core network experience-centric operations management.”
activities. The research conducted for this
Chief Technical Officer, Tier-1 Operator, India
report has also revealed that AI is presently
facilitating improvements ranging from
simplifying network evolution to improving
performance across existing networks.

Figure 3: Biggest impact of AI currently within core network activities (worldwide)

Improving the
reliability of network 70%
for customers

Simplifying network
evolution to next “G“ 64%

Analyzing service
performance 58%
for customers

Optimizing capacity
of networks 58%

Enhancing the
customer experience 55%

Improving
performance across 52%
the existing network

The impact of AI is already visible in core network activities

7 Ericsson  |  Employing AI techniques to enhance returns on 5G network investments

The impact of AI upon core network

activities varies between regions.
Figure 4: Biggest impact of AI currently within core network activities (regional)

Improving the reliability Simplifying network Analyzing service

of networks for customers evolution to next “G“ performance for customers

56% 56% 56%

91% 55% 45%
67% 60% 51%
75% 58% 62%
69% 77% 69%
70% 73% 64%

North America
Europe
Latin America
Middle East and Africa
North East Asia
South East Asia, Oceania and India

Optimizing capacity Enhancing the Improving performance

of networks customer experience across the existing network

61% 50% 39%

73% 64% 55%
58% 60% 49%
79% 71% 46%
15% 62% 69%
52% 39% 57%
8 Ericsson  |  Employing AI techniques to enhance returns on 5G network investments

AI opportunities for
service providers

The improvements to efficiency that can be brought about

by AI are of particular interest to service providers.

Figure 5: Areas where service providers will be focusing upon adopting AI in their networks

64% 61% 56% 53% 53% 52%

Network capacity Reduced time for Infrastructure Service quality Network Network trouble
planning/ network planning modeling to reduce management performance ticket management
management capital expenditure management
(CAPEX)

51% 50% 49% 47% 44%

Self-optimization Self-configuration Self-healing Virtual employee Network alarm

assistants management

Continuous improvements to efficiency The rapid expansion of advanced IoT new AI use cases. This was followed by
are essential to any business sector, devices will need real-time latency, reduction of time in network planning.
and the mobile communications and successful adoption of AI will The results of our research highlight the
sector is no exception. facilitate faster, more responsive, strong appetite among service providers for
The arrival of 5G will be characterized by available-on-demand networks. building AI reasoning-based systems that
improved speed, consistency, reliability and However, the benefits of AI extend can enable the self-healing, self-optimizing
capacity. Features such as advanced load beyond cost-management and and self-configuring networks of tomorrow.
management will enable service providers network-management efficiencies. Figure 5, above, shows the areas where
to effectively manage the network traffic, Enhancing user experiences is expected service providers indicated how they
ensuring network performance has minimal to be one of the many notable benefits of are focusing their AI adoption efforts.
to no impact with additional devices coming AI-managed services on the networks. It shows how increased reliance on smart
over the networks. Service providers will be In our research, network capacity automation enabled by AI will be an
able to achieve successful management planning/management emerged as the top investment for service providers, with
of network performance by developing area where operators see opportunities return on investment (ROI) visible in
cognitive and predictive AI algorithms. in terms of implementing and adopting significantly reduced operating costs.

“The business model the communication sector has to come up with will be
value-based. For voice, they were charging per minute. For data, it was per MB.
What they always missed in their business models was customer value.”
Assistant Professor, Middle East and Africa region
9 Ericsson  |  Employing AI techniques to enhance returns on 5G network investments

Challenges for successful

adoption of AI

Accessing network data, and using it to solve business

challenges, are hurdles facing many service providers.

In order to maximize the potential of AI, of respondents globally, and 89 percent 5G networks. The barriers highlighted in
there needs to be effective collection, in North America. our research (data quality, multiplicity of
structuring and analysis of data to ensure Other concerns expressed globally sources and systems) are indicative of the
clear and actionable insights for network included an excess of data from too many limitations of the traditional approach to
development and improvement. sources (59 percent); problems finding data analytics.
This is an area of concern among service indicators of degradation or compromised AI-enabled networks employ advanced
providers and infrastructure vendors across assets (59 percent); storage of data in too data analytics that make systems smart,
the globe, with 71 percent expressing many systems (56 percent); and lack of adaptive, self-aware, proactive and
concern about defining and implementing single ownership or oversight of the data prescriptive. These can play a key role in
standardized interfaces. (55 percent). reducing associated operating costs and in
As part of our research, we asked At the heart of these concerns is the addressing many of the barriers that service
respondents to provide their views on fact that traditional approaches to data providers have indicated are preventing
whether their company was collecting analytics have been based upon centralized insights from data being acted upon.
data effectively. and generic analysis tools. These are
It became apparent that data quality unsuitable for addressing the increasing
was an issue. This was cited by 65 percent complexity and vast volumes of data in

Figure 6: Challenges identified by service providers in sharing data to enable real AI benefits

Defining and understanding Understanding/identifying

standardized interfaces which data to share and use “The opportunity for
78% 56% Machine Learning, in
55% 91% this case, is the ability
75% 60% to identify patterns. It's
83% 75%
a fantastic way to learn
69% 69%
from experience, like
61% 68%
when we are children.”
Head of Radio products,
Tier-1 Operator, Europe.

Rationalizing and validating Ensuring regulatory

data and sources compliance (GDPR etc.)

72% 44% North America

Europe
55% 45% Latin America
Middle East and Africa
58% 47%
North East Asia
54% 46% South East Asia, Oceania and India

46% 46%
64% 52%
10 Ericsson  |  Employing AI techniques to enhance returns on 5G network investments

Successful implementation of AI

Comprehensive experience in building and managing mobile

network products is vital to successful implementation of AI.

Service providers are keen to embed AI more Figure 7: Service providers wanting to embed AI into their networks
readily into their networks, according to our
research. This sentiment is shared among
77 percent globally, and is particularly
strong in South East Asia, Oceania and
India (91 percent) and North America
(83 percent).
77% Worldwide

There is broad agreement among service

providers that delivery of new services to
customers can be accelerated by adoption of
83% North America

AI technology across networks. Worldwide,

South East Asia,
this view is held by 76 percent of service 91% Oceania and India
providers, with 85 percent of respondents
in North East Asia expressing this view.
There is also widespread belief that AI
will enable service providers to scale network
operations and evolve into digital service
providers, while managing the associated
cost. This view is endorsed by 73 percent
globally and there is almost universal
support in North America at 94 percent.
In order to achieve this integration
and acceleration, AI technology needs
to be defined and implemented using
standardized interfaces. Therefore,
successful AI implementation is considered
to be dependent largely on telecom “We are keen to lead
expertise (63 percent worldwide) and
easy-to-adopt functionality that enables the way in the region
staged AI functionality to be easily Therefore, early adopters of the
implemented (70 percent), as well as AI-enabled networks would seem likely
when it comes to
configurable AI functionality (62 percent). to benefit the most in terms of their ROI. artificial intelligence,
AI will undoubtedly play a crucial role Ericsson is working in collaboration
as network complexity increases and with service providers at the heart of as it paves the road
service providers face the demands of new developments, with a focus upon for implementation
handling multiple technologies such as using AI to reduce costs and increase
4G, 5G and IoT, as well as growth in the network performance. With more than of new technologies
number of devices. This is because AI will 600 AI patents approved for filing,
enable service providers across the globe Ericsson has played a pioneering role in
across all our markets.“
to manage increasing network deployment transforming the engineered network CEO, Tier-1 Operator, Middle East
and network operation costs. into a continuous learning network.
11 Ericsson  |  Employing AI techniques to enhance returns on 5G network investments

Recommendations

AI will bring myriad opportunities to mobile communications

AI will bring significant benefits to mobile communications, notably cost

and network management efficiencies, and improvements to customer
experience. However, in order to make the most of the opportunities, service
providers will need to take a proactive approach which includes:

– building standardized interfaces to access relevant and actionable data

– exploring ways of using AI to optimize customer experience
– running early trials with new customer segments to identify AI opportunities
– examining use of AI and automation for network operations,
including planning and optimization
– ensuring early adoption of new solutions for AI and automation
to facilitate introduction of new use cases
 Ericsson enables communications service providers
to capture the full value of connectivity. The company’s
portfolio spans Networks, Digital Services, Managed
Services, and Emerging Business and is designed to
help our customers go digital, increase efficiency and
find new revenue streams. Ericsson’s investments in
innovation have delivered the benefits of telephony
and mobile broadband to billions of people around
the world. The Ericsson stock is listed on Nasdaq
Stockholm and on Nasdaq New York.
www.ericsson.com

Ericsson The content of this document is subject to BNEW-19:007926 Uen

SE-164 80 Stockholm, Sweden revision without notice due to continued © Ericsson AB 2019
Telephone +46 10 719 0000 progress in methodology, design and
www.ericsson.com manufacturing. Ericsson shall have no
liability for any error or damage of any kind
resulting from the use of this document