1 Intelligent automation
Intelligent
ericsson.com/
intelligent
automation
platform
automation
How Intelligent RAN Automation is
creating key use cases for service providers
Intelligent Automation Guide Series
October 2022
2 Intelligent automation
Executive summary
The next-generation Radio Access In this new scenario, the main challenge This is unsustainable in terms of cost
Network (RAN) will need to support many for the service providers is optimizing and environmental impact. To break the
use cases with different requirements radio network resource utilization. growing trend of energy consumption,
at the same time, from simple voice and At Ericsson, we can help you solve this it’s necessary to address the different
text messaging to extremely data-hungry challenge by making sure you run a more parts of the network holistically and,
applications, such as the streaming of sustainable network, while saving time while doing so, to make sure that the
real-time virtual reality (VR) feeds. for your existing workforce, especially full 5G experience is delivered for
Intelligent automation is essential to the operations team. This will lead to the your consumers.
meet growing network demands, manage optimization of investment decisions and Furthermore, as 5G deployments are
the increasing complexity, operate reduce your capex and opex. evolving to the cloud in the RAN domain,
networks more efficiently to provide Service providers are looking for you should consider the new security
the best network performance, and solutions to improve RAN efficiency, risks. While the cloud introduces
deliver a great customer experience with performance, and user experience as advantages, it also expands the 5G
rationalized capex investments. a must for monetization use cases, attack surface that must be protected
Our experience has shown us such as extended reality applications. with a security-by-design approach.
that service providers have a unique The use case for improving energy Our intention with this guide is to
opportunity to monetize new 5G efficiency is also in high demand to show the intelligent automation solutions
applications. However, there are also meet sustainability and cost-control that address the main service provider
new challenges that must be solved. goals. Service providers have realized challenges. They provide the highest
Increased operational complexity and the that if 5G is managed the same way it ROI and shortest time to market, with a
demand for superior user experience must was in previous generations, energy future-proof architecture that can meet
both be satisfied to achieve successful consumption will increase dramatically to the ever-changing demands for each
monetization of 5G. meet rising traffic demands. service provider.
Figure 1: Drivers for automation
New possibilities drives complexity Winning ambition
Explosion of use cases Accelerate time to customer
Data-hungry applications Performance leadership
IoT Diverse devices Maximize return on investment
Security demands Reduce complexity
3 Intelligent automation
The four main domains of
Ericsson Intelligent RAN
Automation solutions
Exploring the typical use cases in the categories of evolution,
deployment, optimization, and healing of the network.
At Ericsson, we have identified the main Network optimization
areas where service providers can benefit • offers intelligent autonomous functions
from AI-based automated solutions. to optimize user experience and return
We have applied our vast expertise in on investments, such as RF shaping,
RAN automation spanning over two traffic and mobility management,
decades to provide a wide range of use and energy efficiency
cases that can fit in the following four
categories: network evolution, network Network healing
deployment, network optimization and • ensures service continuity and the
network healing. resolution of both basic and complex
These four categories are supported incidents, delivering high availability
by automation and the AI foundation that while keeping opex at a minimum
is applied in the software functionality,
services and platform. These domains serve four main
value drivers: network sustainability,
Network evolution trustworthiness and security, efficiency
• enhances network planning with more and user experience. The intelligent
efficient radio frequency (RF) planning, automation evolution journey improves
site selection and capacity management these benefits by creating networks that
• improves network and service are more proactive and data-driven.
performance and enables new revenues
thanks to data-driven and intent-based
insights and recommendations The AI and automation foundation for
Ericsson Intelligent RAN Automation
Network deployment enables high performance and a
• handles provisioning and life cycle shorter time to market for all the
management (LCM) of complex different use cases, while ensuring a
networks with optimal costs and high degree of flexibility.
speed to market
Figure 2: Use cases for Ericsson Intelligent RAN Automation
Network evolution
Network deployment Network optimization Network healing
Automation and AI foundation
4 Intelligent automation
Make smarter
decisions with AI
AI, automation and novel data-handling capabilities that support
service provider decision making meet in the Ericsson Intelligent
RAN Automation solutions.
We have considered how to provide support Investment decisions are driven by the insight into how they affect each other.
for decision making across the spectrum long-term evolution of the network and its Network performance is one factor
– from decision granularity to time frame, capabilities, such as the addition of new that impacts the user experience and,
and from network planning to deployment, radio equipment, site development and by extension, the service provider. But
parameter configuration and tuning, new enriched radio functionality, including energy efficiency also benefits service
and operations. Ericsson Intelligent RAN software upgrades. providers, their opex, the environment
Automation frees up time and provides The network evolution area in Ericsson and increasingly environmentally aware
actionable insights for making decisions to Intelligent RAN Automation deals directly consumers. By using data insight and AI,
create more end value for service providers with these use cases, such as capacity asymmetrical opportunities can be found.
and their consumers. planning. Our solutions reduce the time it This means that in some traffic scenarios,
Reliable and relevant data is the single takes to evaluate new scenario development, a small-to-negligible decrease in network
most important factor for making network a cumbersome process when done manually, performance can allow for significant
decisions. Nevertheless, there is a limit to which could limit the total number of energy savings. For example, this can be
the amount of data humans can handle, scenarios and, consequently, the information done by pushing user equipment from a
and this is where AI comes in. AI helps to available. Therefore, faster scenario capacity cell to a coverage cell, allowing
automate the data handling process to development gives both a more robust data the former to be put in sleep mode. Finding
provide actionable insights. set of multiple scenarios and frees up time for the scenarios where this has minimal
Properly coupling AI, data, and deciding what improvements can be made impact on user experience while granting
automation creates a virtuous cycle where and where. This intelligent process is further high energy savings will require increased
better data enhances the AI, which allows supported by radio frequency design and data insights and AI that considers traffic,
the organization to handle more data, site-selection features. time, location and services offered.
thus creating even further improvement. A future opportunity for leveraging AI Overall, the purpose of our AI-based
Our findings demonstrate that the power comes in the form of trade-off decisions. functionality and data-enabling
of an intelligent data approach includes One such trade-off example is the choice architecture is to create a virtuous cycle that
advanced microwave insights, more between network performance and improves insights and decision making to
sustainable products, intelligent services energy savings. Today, these decisions enhance service providers’ networks with
and network evolution capabilities. are taken independently, or with limited reduced effort from operations teams.
Figure 3: How AI delivers value in radio networks
18% Benefits 25%
Network load redistribution Better 5G coverage with
more efficient use of resources 5G-aware traffic management
Energy management and sustainability
Self-organising network
15 min 14%
Efficient operations
vs 1 week Energy savings with
to automatically classify 100,000 cells with Augmented MIMO sleep
Cell issue classifier
Customer experience
80% >70%
Reduced signaling with Incidents prevented with
Machine learning assisted paging Network performance RAN KPI degradation prediction
5 Intelligent automation
What is the holistic approach of
Intelligent RAN Automation?
Intelligent automation is embedded in our
solutions, from RAN to network management,
as well as our professional services offering.
Our intelligent automation is executed hundred milliseconds. In many cases,
where required to ensure the best network AI/machine learning (ML) makes
performance and consumer experience. it possible to enhance the network
When operating a RAN, deployed performance by up to 30 percent and also
resources must be managed efficiently. to minimize the amount of configuration
There are two control loops acting optimization that is needed in the high
together according to different time scales level control loops by up to 90 percent.
and scopes. Ericsson Intelligent RAN The high-level control loops are related
Automation solutions use AI algorithms to traditional network operations. Examples
integrated with existing processes and include RAN coordination and network
algorithms in these control loops. power management, which decreases
The two fastest control loops are energy consumption by up 20 percent.
related to the traditional Radio Resource These high-level control loops encompass
Management – that is, traffic mobility the bulk of the manual work that will
operations. Innovation examples include disappear as a result of intelligent RAN
the link adaptation feature that improves automation, which explains why AI/ML is
spectral efficiency by up to 15 percent in especially attractive in these control loops.
the fastest control loop. Functionality in This holistic take is applied in all these
these control loops is mostly autonomous, use cases, which act on both fast and slow
driven by engineered algorithms requiring control loops for distributed and centralized
complex configurations in a timeframe automation to make the improvements
ranging from milliseconds to several scalable across the entire network.
Figure 4: Example of Ericsson’s holistic take on AI applied to RAN automation
• Execute where it makes sense
Weeks
• Ensure optimal efficiency
Centralized automation and performance
• Network-wide coordination
Days
• More time for complex Network benchmarking and evolution,
decisions AI/ML LCM, software upgrades
• Human interaction
Seconds Orchestration, optimization,
programmability, embedded AI LCM
Distributed automation
Handover decisions, quality of service,
Milliseconds dual connectivity control, spectrum
• Local
load balancing
• High volume of decisions
• Fully automatic
Beamforming, scheduling,
Microseconds coordinated multipoint,
fast spectrum management
6 Intelligent automation
Why choose rApps
for automation?
Deciding where to place the intelligence is not about deciding
on one place to fit all, but on leveraging the different capabilities
of the network layout to maximize overall performance.
The industry consensus is that increased That is the role of the Ericsson Intelligent • Ericsson 5G Centralized Neighbor
openness and the resulting wider Automation Platform, our realization of Relations rApp: The automatic
ecosystem will allow for more innovation, a service management and orchestration establishment of cell neighbor relations
improved network efficiency at scale, and (SMO) platform. Efficient handling of based on various measurements
better network orchestration. rApps will be crucial to support ecosystem and detection
rApps have oversight of the network, development efforts to foster innovation. • Ericsson Frequency Layer Manager
which makes them good candidates for From the telecom technological evolution rApp: The dynamic redistribution of
containing large portions of the policy and perspective, intelligence in the rApp users in the most effective way to
optimization. Intelligent deployment will enable network-wide performance, increase user experience
rApps act on the high-level control loops, controlled via intent-based automated • Ericsson Performance Diagnostics rApp:
half a second and above, which makes decisions in the higher-level control loops. The automatic cell-issue classification
them ideal for central placement. Later, this will be cascaded to the network in and root-cause analysis for the entire
Centralized decisions are made with fast control loops in terms of execution. RAN performance optimization
a holistic network view, and this will help The architecture of rApps will provide a
to improve actions locally at the RAN broader ecosystem to engage in the telecom The next step in the intelligent network
level, for example, by optimizing the best space to create novel products, design evolution will be intent-based management,
baseband partner on the network level. specific solutions based on certain niche where intents will cascade from a higher
Another example is an Ericsson rApp needs, and foster innovation to propel the abstraction level. Central orchestration and
which can increase data throughput industry forward. handling of these intents will be part of the
by up to 30 percent at the cell edge in a Ericsson solutions for energy saving RAN evolution, for which the rApps and
congested network. For these reasons, include radio features such as multiple SMO layer are well suited. Therefore, this
rApps can further enhance radio network input, multiple output (MIMO) Sleep mode, architecture prepares us for the next step in
functionality by providing a centralized Cell Sleep mode, Radio Deep Sleep, and the intelligent network evolution journey,
view of the network. rApps for network centralized energy where the main value drivers are network
Inter-rApp orchestration can be management, optimization and control. sustainability, trustworthiness and security,
executed to secure network performance. Some examples of existing rApps are: efficiency and user experience.
Figure 5: Evolution of intelligent automation
Automation of Realizing the vision of cognitive & autonomous
Boosting automation capabilities with AI
operations and optimization network operations
Reinforcement Intent-based
Scripting Process automation ML Machine reasoning
learning management
Value
Technology
journey
Efficiency
Customer experience
Sustainability
Trustworthiness and security
Manual and reactive Data-driven and proactive
7 Intelligent automation
Intelligent services
for tailored business
requirements
All networks are different, and all service providers deserve
dedicated use cases to match their requirements.
Ericsson provides customized use cases This AI software consists of an event team become more successful, providing
using AI technology, thanks to our or correlated set of events along with early notifications before alarms and
intelligent services. We offer a dedicated corresponding analysis, thresholds for complaints arise. Therefore, the O&M
local team comprising Ericsson network distinguishing between normal and team can act before issues arise,
experts and data scientists, synchronized abnormal conditions (where applicable), resulting in improved network
with Ericsson product R&D for use case an engagement plan for notification performance and revenues.
design and enhancement of network and/or execution of actions, and an Our collaborative approach allows
products and features. action plan for handling the condition. us to generate a wide range of use cases.
Our specialist teams work with These solutions, known as These use cases have become part of
service providers to translate their “service continuity”, are part of the our global library and can be replicated to
business goals, pain points and data into Intelligent RAN Automation solution. solve similar issues for service providers
customized use cases for an optimized They detect anomalies, and predict worldwide. Bringing references from
result in terms of efficiency, consumer and prevent incidents. In essence, it networks around the world allows us
experience and network performance. is about actively working closely with to test a specific network for globally
Ericsson develops, per specific the service provider’s operations and known issues. These experiences help
service provider’s needs, customized maintenance (O&M) to prevent issues us to design a wide variety of use cases
AI algorithms for detecting, predicting, and have a positive impact with early that can be tailored to fit individual
and/or measuring specific operational detection, notification and actionable customers. Our intelligent services
characteristics of interest in the service recommendations to stay ahead of the apply advanced software functionality
provider’s operated network. issues. The service will help the O&M with AI technology.
Figure 6: Use case co-creation process for tailor-made service provider needs
Ideation and co-creation Develop and introduce
Engage with service providers to develop use cases Unify the latest technologies with data and telco domain
for prioritized network problems experts for fast development and introduction
Technology enhanced by people
Ecosystem of knowledge
Actionable insights
Expand and utilize globally the library with
Detect anomolies and provide actionable insights
wide range of use cases
8 Intelligent automation
Conclusion
The target for service providers is to not viable in the old paradigm, that of The foundation of this network evolution
achieve more sustainable networks, continuously training and adapting the is a future-proof architecture with Ericsson
better user experience, optimized ML algorithms. We are providing Intelligent Automation Platform, which
capex investment and reduced opex by AI-powered solutions that are more offers automation in the LCM, data
improving network performance. accurate and efficient for certain use access, and network element control. It
To achieve these objectives, intelligent cases or applications than rule-based also supports the execution of rApps from
network automation is necessary. Networks solutions, since they can dynamically Ericsson, customers, and third parties,
require predictive capabilities to provide adapt to network changes. In the long enabling an open ecosystem for innovation.
insights dynamically and in real time to term, AI-powered solutions will be Both rApps and Ericsson Intelligent
support service providers in making the ubiquitous in the RAN and will facilitate Automation platform are based on
best decisions for their specific challenges the creation of the best performing and cloud-native microservices architecture.
and goals. AI-powered radio features and most cost-effective networks. We are a leading player within the
rApps, together with intelligent services We have developed the following use O-RAN Alliance to ensure that the SMO
and a future-proof intelligent platform, cases that can be provided as an rApp: and its Non-RT RIC, rApps, and R1 and
are required to succeed. • reducing the overall transmitted power A1 interfaces will be secure. For more
Technology leadership in AI by 20 percent, with a 3.4 percent information, see: Ericsson in O-RAN alliance.
technologies (such as reinforcement saving on the electricity bill per base The telecom industry is in an
learning and digital twins), combined station (Ericsson Power Optimizers) evolutionary process where networks have
with network data expertise, is applied • saving daily radio energy consumption become increasingly autonomous. The
to RAN automation and makes the by 15 percent, with zero impact on target is to evolve current networks so
foundation of the Ericsson Intelligent user experience that they can follow the automation
RAN Automation ecosystem. We are • automating in-service software evolution path. Network automation
committed to reducing the complexity upgrades with no service disruption is a journey that started more than
of operations and empowering service • integrating new radio units in a shorter 20 years ago with Ericsson. Today, with
providers to make the best decisions time with optimized configuration and AI technologies applied in radio features,
according to their needs. with automated closed-loop assurance rApps, and services, we are in an advanced
The methodology used in Ericsson • improving network performance automation phase that enables service
Intelligent RAN Automation is leveraging increased user throughput and radio providers to overcome the complexity
knowledge from historical and current coverage at network scale without challenges and succeed in capitalizing on
real-time data to an extent that was manual intervention new 5G opportunities.
Figure 7: Intelligent RAN Automation measured benefits for service providers
20% 50% 40% 15%
Lower capex with AI planning Less time from site survey to Reduction on bad quality cells Increase spectral efficiency
design complete
15 min 17% 60% 30%
Integration Improved energy efficiency Reduction in network Improvement on
performance issues service availability
About Ericsson 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 16/22102-FGM101030
SE-164 80 Stockholm, Sweden to revision without notice due to © Ericsson 2022
Telephone +46 10 719 0000 continued progress in methodology,
www.ericsson.com design and manufacturing. Ericsson
shall have no liability for any error or
damage of any kind resulting from the
use of this document
