ericsson.

com/
intelligentautomationplatform

Ericsson Intelligent
Automation
Platform
Solution brief
2 Ericsson | Ericsson Intelligent Automation Platform

Bringing automation
to RAN at scale

Ericsson Intelligent Automation Platform implements Service

Management and Orchestration (SMO) for Open RAN, and
extends those to take openness and automation forward
with multi-vendor and multi-technology RAN support.

1. Managing network complexity Bringing openness and

Automation is no created by legacy and 5G disaggregation to RAN
2. Improving agility in managing legacy Service providers are looking at
longer optional services and launching new services reducing RAN opex, through Open RAN
Telecommunication networks are across existing physical 5G networks architecture and automation, with
becoming more powerful every day, and the newer cloudified 5G network the following main drivers:
but they are not getting any simpler. 3. Increasing productivity of resources • openness and disaggregation
The evolution of technology over the and accelerating upskilling for to ensure supply chain security
last few decades has made it possible the management of cloud-native • cloudification for leveraging commercial
to provide mobile communications telco networks off-the-shelf (COTS) cloud infrastructure
services at scale, performing in a 4. Securing new enterprise to improve return on capex and reduce
way that may have seemed impossible revenues requiring stringent RAN-related software operations and
a few years ago. As a result, the volume service level agreement maintenance (OAM) costs
of traffic and diversity of use cases management and customization • accelerate innovation to drive
running on top of mobile networks 5. Differentiating customer experience new 5G revenues
has grown exponentially. to reduce churn, improve net
Moreover, the increased disaggregation promoter score (NPS), as well This introduces additional complexity
of the Radio Access Network (RAN) nodes as overall customer satisfaction and costs from managing multiple
associated with adoption of cloud-native vendors across different RAN components.
RAN architectures multiply the elements According to Ericsson’s internal analysis Therefore, a common RAN automation
to manage. and estimates, doing nothing is not an platform is needed to efficiently manage
As communications service providers option as operational costs will more and operate the radio network along
around the globe take the next step in than double over the next five years. For with existing brownfield deployments.
the evolution of mobile technology and communications service providers, the Open RAN service management and
start launching 5G services, the existing biggest part of this opex is related to RAN orchestration (SMO) platform defines
challenges related to opex efficiency deployment, management and operations, the architecture and interfaces for
and revenue growth are intensifying: so automation of these activities is key. Open RAN automation platform.

Figure 1: Opex will rise unless network and service automation is introduced

New industries
Network
complexity Cloud RAN
and opex
Network slicing

NFV/SON/Cloud native
Network automation is essential
5G to address complexity
Unmanageable without automation and reduce costs

Manageable without automation

Target is to reach a level of

automation where the network can
be considered nearly zero touch

2G 3G 4G 5G
3 Ericsson | Ericsson Intelligent Automation Platform

Enter Open RAN SMO

Open RAN SMO enables flexibility and agility when

deploying RAN network functions. With the internal
and external SMO interfaces, managing multi-vendor
Open RAN networks is made possible.

Open RAN defines network functions There are various internal and external • A1 interface, between the near-RT
as central units (CU), distributed units (DU), SMO interfaces – R1, A1, O1, and O2 – that and the non-RT RIC, provides
near-real-time Radio Intelligent Controller allow SMO to manage multi-vendor fine-grained policy guidance and other
(near-RT RIC) and non-real-time Radio Open RAN network. data enrichments to RAN functions.
Intelligent Controller (non-RT RIC), • R1 interface, internal to SMO, is • O1 interface, between SMO and
as cloud-native virtualized functions motivated by the need for portability the Open RAN network functions,
running on “O-Cloud” infrastructure. of multi-vendor rApps and providing is the OAM interface towards
SMO is the entity that is responsible value-added services to rApp Open RAN functions to achieve FCAPS
for RAN domain management. SMO developers and solution providers. management, software management
oversees life cycle management of It is a collection of services including, and file management capabilities.
network functions as well as the but not limited to, service registration • O2 interface is the O-Cloud
O-Cloud. SMO embraces the principles and discovery services, authentication Management and deployment interface
of openness in creating an automation and authorization services, artificial between SMO and O-Cloud. It supports
platform based on cloud-native principles intelligence/machine learning (AI/ML) orchestration of O-Cloud infrastructure
that enable agility in deploying RAN workflow services, and A1, O1 and resource management (such as
network functions and applications O2-related services. In addition, the inventory, monitoring, provisioning,
through standardized and open interfaces. interface enables new open APIs to be software management and life cycle
integrated into the SMO framework. management) and deployment
of the Open RAN network functions.

Figure 2: High-level architecture of Open RAN

External system providing

RAN SMO framework
Non-RT enrichment data to the SMO
management RIC (R1)
O2 A1 O1 Open fronthaul
M-plane

NG
Open RAN network functions NG-Core
RAN
O-RU
O-Cloud

O1 and M-plane – SMO FCAPS interfaces to RAN Legend

O2 – SMO cloud-native deployment interface Open RAN defined interface
A1 – Non-RT-RIC to near-RT-RIC interface 3GPP defined interface
R1 – rApps interface to Non-RT-RIC and/or SMO services Interface out of scope

Source: O-RAN Alliance architecture

4 Ericsson | Ericsson Intelligent Automation Platform

Going beyond SMO

Realizing the benefits of openness and automation Key benefits

in the RAN domain requires a multi-technology and Handles multi-vendor RAN
multi-vendor approach. complexity and scalability
• Provides management and
orchestration of multi-vendor RAN
networks over open interfaces
Ericsson Intelligent Automation • tutorials and how-tos • Automates third-party
Platform comprises the implementation • example applications purpose-built RAN via EMS
of Open RAN’s SMO function and takes • API examples and full documentation integration leveraging OSSii
this forward by extending it with • build utilities • Ericsson has deployed Network
the capabilities to manage not only • verification environment Design and Optimization (NDO)
Open RAN and Ericsson’s Cloud RAN • integration with CI/CD toolchain applications in 9 million
technology, but purpose-built RAN • helper libraries and modules cells worldwide, with half in
networks like Ericsson Radio System. multi-vendor RAN environments
Also, the platform provides both rApps
design and runtime environments Aligned to Open RAN principles Multi-technology, enabling
for RAN automation use cases and implementing use cases in adoption of Open RAN
across physical and Cloud RAN. all areas of RAN automation: • Supports automation
The runtime environment enables • network evolution, such of purpose-built and
multi-vendor applications to run on as capacity planning Cloud RAN/Open RAN networks
the platform, access data, and control • network deployment, such as the rApps • Ericsson is launching rApps
network elements. Ericsson Sector Carrier Orchestration to maximize network
Additionally, the design environment and Ericsson Canary Upgrade ROI across purpose-built
allows for rapid application development • network optimization, such as Ericsson and Cloud RAN/Open RAN
and life cycle management of applications. Performance Diagnostics rApp
It is supported by a software development • network healing, such as AI-powered Open SDK empowers rApp
kit (SDK) that is launched together with advanced cell supervision developers to maximize ROI
the platform. • Enables development and
Ericsson Intelligent Automation Services deployment of rApps from service
Platform supports all the interfaces A complete set of capabilities to providers and third-party vendors
in line with Open RAN principles – R1, accompany the service provider • The platform’s SDK has
A1, O1, and O2 – and provides a safe in the RAN automation journey: been used to create rApps
and feature-rich execution environment • transformation and implementation on a common platform
for rApps that implement diverse use plan – based on “as-is” and “to-be” • Ericsson strongly supports
cases in the area of RAN automation. scenarios, a detailed plan and design openness and innovation as a
of how to realize the to-be scenario major contributor to Open RAN
A comprehensive offering • consulting – Ericsson services working groups, multiple global
To provide value to service providers offer business qualification standards and industry bodies
from day one, Ericsson Intelligent on the possible efficiency gains including 3GPP, ONAP and TMF,
Automation Platform is launching in • training – all courses required as well as being a founding
conjunction with other components to kickstart platform operations, member of OSSii
and products. development and customization
• deployment and life cycle management Connects RAN to network
Software development kit – initial installation as well as slicing, management,
Much bigger than a set of APIs upgrades of the automation platform assurance and orchestration
for development and integration, • multi-vendor customizations – • Supports RAN and
an SDK provides developers with customer-specific multi-vendor and OSS evolution strategy
everything they need to build and multi-technology implementations • Handles disaggregated
take the application through the • design and build automation use RAN complexity
different stages from development cases – to extend provided rApps • Supports rapid innovation,
to production in the most effective way: and create new rApps when required agility and fast time-to-market,
decoupling development
and life cycle management
5 Ericsson | Ericsson Intelligent Automation Platform

Any mobile network.

Intelligently automated
Figure 3: Ericsson Intelligent Automation Platform components and interfaces
Automation System
Portal and external APIs

Design time Runtime

Design and onboarding API discovery and exposure

environment
App execution environment
Service design
Ericsson Ericsson Service Vendor C Vendor D
Policy design rApp management rApp rApp provider rApp rApp rApp

Workflow design RI interface Exposure framework

Assurance prof. design
Runtime platform service
rApp design
Policy Flow Service Analytics Topology Access control
Intent design framework framework orchestration and intents service
Runtime Logging
Design framework Data management and movement
catalog
License management
Design catalog Analytics data AI/ML exec
Controller framework
collection framework environment Cert. management

Vendor A1 A1 O1 A1 O1 O2
FCAPS
xApp

EMS Near-RT RIC

Cloud RAN CU O-CU

gNodeB Cloud RAN DU O-DU

O-Cloud

Analytics and intents OSS portal

Platform components Components and services providing Web-based single point of
analytic insights such as event-based entry to connect to the rest of
The main functional components counters and KPI calculation, as well as the operational support systems
of the platform are as follows: expression and management of intents. (OSS), supporting integration
and discovery of OSS capabilities.
Application manager Analytics data collection
Life cycle management and Set of all file and stream-based ML execution environment
monitoring of applications hosted data for fault, performance and Supports ML model training,
within and by the platform. log data. It will initially support execution of ML algorithms and
the O1 interface. life cycle management of models.
Service orchestrator
Manages network service life cycles using Controller framework Runtime catalog
a TOSCA-based orchestration engine. Supporting control access towards Supports all verified artifacts
multiple external entities, including and software that is accepted
Policy framework networks and other systems, and validated for deployment
Enables the model-driven life through individual controllers. in production and execution.
cycle management (deployment
and execution) of policies and Topology service Workflow execution environment
supports the addition of new Provides access towards and Enables users to execute and monitor
policy engines and formats. management of a near real-time pre-defined or user-defined workflows
standardized topology and and supports the definition of new tasks.
Data management and movement inventory model.
Infrastructure supporting the transit
or sharing of data in the platform, as
well as various persistence technologies.
6 Ericsson | Ericsson Intelligent Automation Platform

RAN automation use

cases and rApps

rApps provide realization of RAN automation use cases,

and the need for sharing resources and capabilities among
rApps demands the best from the supporting platform.

The rApp concept is defined as an for intent-based RAN optimization rApps are modular applications
application designed to run on the and automation. that leverage the functionality exposed
non-RT RIC to realize different RAN As rApps are running on top of by the non-RT RIC/SMO framework to
automation and management use cases, Ericsson Intelligent Automation perform multi-vendor RAN optimization
with control loops on a time scale of one Platform, external data sources can and assurance capabilities via the R1,
second and longer. rApps will be using be integrated for the rApps to add A1, O1 and O2 interfaces. rApps can
an open and standardized R1 interface additional data to its use cases. be developed and delivered by the
on the Ericsson Intelligent Automation More precisely, these data enrichment multi-vendor community and are
Platform. For the interaction with RAN, capabilities are one of the foundations standards-based and application
the A1 interface can be used. The A1 open for the differential value the rApp platform agnostic, meaning they
interface will interact with the non-RT-RIC concept will enable. can run on any vendor’s non-RT-RIC.

For example, Ericsson Frequency Some more examples of the

rApps share Layer Manager rApp optimizes the capabilities enabled by these
automated load balancing features software building blocks are:
common services of the radio network between different • cell type classification
frequency layers. This rApp is composed • site and sector identification
Each rApp can also interwork with other of several smaller functions to provide • geographical neighbors
rApps by means of standardized interfaces the load balance recommendations, • mobility scoring
to build a more complex automation and some of these functions are also • coverage balance and contiguity
function, in which the insights from one shared with other centralized SON • device slicing
rApp serve as input to another rApp to take algorithms such as Ericsson 5G
more complex decisions. We can say that Centralized Neighbor Relations rApp, cell This sharing of services among
“small” rApps compose “large” use cases. type classifier, physical topology builder. the rApps has a number of benefits:
• “platform value” from reuse
Figure 4: Ericsson Frequency Layer Manager rApp example: by enabling even more rapid
sharing a number of common services with other rApps via the platform application development
• reduces overall complexity,
Ericsson Intelligent Automation Platform as the size of the rApps can
be smaller
Ericsson Frequency Layer Manager rApp
• data and insights can
be shared among rApps
Coverage • standard life cycle
balance estimator management procedures
across the rApp ecosystem
Cell/sector App coverage Inter-frequency
coverage estimator estimator load manager

Cell type Physical topology Geo-neighboring

classifier builder (site and sector) discover

Cell relation
manager

Ericsson 5G Centralized Neighbor Relations rApp

R1
7 Ericsson | Ericsson Intelligent Automation Platform

rApps work together

to fulfill use cases

In many cases, rApps are composed to provide the

functionality that more complex use cases require.

The different use cases in the wide Figure 5: Different areas of RAN automation empowered by
area of RAN automation, with some the Ericsson Intelligent Automation Platform and the rApps
examples, can be presented as follows:
• Network evolution: Secures
network investments aligned Network evolution
with business strategy, improving
energy and service performance,
as well as enabling new revenues
through evolution of network assets Network deployment
by data-driven and intent-based
insights and recommendations.
Example: Capacity planning.
• Network deployment: Handles Network optimization
provisioning and life cycle management
of complex networks with optimal
costs and speed to market. Examples:
the rApps Ericsson Canary Upgrade, Network healing
Ericsson Sector Carrier Orchestration,
Ericsson 5G Centralized Neighbor
Relations, Advanced Radio
Coordination (E-RAN management).
• Network optimization: Offers
intelligent autonomous functions
to optimize customer experiences
and return on investments. Examples:
Ericsson Frequency Layer Manager
rApp, user equipment prioritization
for service assurance (Ericsson QoS
Nudging rApp), Ericsson Performance
Diagnostics rApp.
• Network healing: Ensures
service continuity and resolution
of basic and complex incidents,
delivering high availability while
keeping the opex at a minimum.
8 Ericsson | Ericsson Intelligent Automation Platform

Empower the rApp ecosystem

with the best SDK

Ericsson Intelligent Automation Platform provides rApp developers

with an SDK, which has unique capabilities to streamline the
development and life cycle management process.

The platform SDK is much bigger • AI/ML training and execution: • Controller framework: Enables
than a set of development and Provides AI/ML model life cycle parameter configuration on managed
integration APIs. It offers everything management, execution and elements and uses an events-based
needed to efficiently see the application a training environment. API to provide notifications of
through each stage. • Analytics data collection: application changes to allow near
Built on a strong foundation of Offers raw analytical data in a real-time action. Abstracts complex
services, the platform is cloud native file or stream-based data format. details of the underlying system to allow
(deployed on Kubernetes) and provides Standard collection interfaces for developers to focus on their use case.
a defined architecture for ease of fault management (FM), performance • Policy design and execution:
application structuring. Standard management (PM) and others will be Design and execution for policy
interfaces to integrate with logging, provided and supported out of the box. handling using multiple engines.
authentication, certificate management • Analytics insights: Analytics • Service orchestration:
and data movement allow developers processing capabilities enable TOSCA-based orchestration engine
to focus on the business logic and valuable insights into network supporting declarative orchestration,
deliver the key value of the use case. behavior and performance. through ETSI’s open source
The automation platform offers • Base services: Offer everything an management and orchestration
developers a suite of APIs and capabilities: application needs such as logging, (ETSI-MANO) resource orchestrators,
• Data management and movement: user management, certificate of end-to-end services.
Includes access to key data bus of management and UI portal. • Workflow execution: Realization
the system. • Design and onboarding environment: of flows of activities that flexibly
• Topology service: Near-RT DevOps pipelines enable the automated orchestrate system functions and
source of truth for inventory onboarding, verification, promotion and capabilities, supporting new use cases.
and topology. publication of software and models.

Figure 6: Ericsson Intelligent Automation Platform SDK example workflow and platform services

App designers and developers

Design your app App manager Use an app

Access Human or machine
Ericsson Intelligent Launch SDK Design Build and Onboard
Automation Platform SDK tools and
portal your app package Network
documentation
deployment
Developer ecosystem applications
Validate
Network
Consumer runtime services optimisation
Deploy applications
AI/ML Analytics Data Design and Base services
Analytics Network
training data management onboarding
insights User healing
and execution collection and movement environment
management, Configure applications
certificate
Controller Workflow Topology Service Policy design management, Network
framework execution service orchestration and execution UI portal, evolution
Monitor applications
logging, etc

Ericsson Intelligent Automation Platform

9 Ericsson | Ericsson Intelligent Automation Platform

Connecting the dots of

end-to-end automation

Ericsson brings its end-to-end experience on network slicing

to ensure that the Open RAN functions are integrated with the
rest of the network domains for the best customer experience.

High network performance and enhanced customer experience is dependent on Ericsson also understands how to connect
customer experiences are as important the coordination and performance of the RAN technologies to the rest of network
in the evolution to Cloud RAN as opex the different network domains, where domains to enable real end-to-end-service
improvements. Ericsson understands RAN is a fundamental part, but not management in all areas, from fulfillment
end-to-end services and how the overall isolated (see Figure 7 for example). and assurance to monetization.

Figure 7: Network slices require the coordination of different network domains

End-to-end network slices

RAN Transport 5G core

QoS1
Slice A
QoS2 Slice A
RAN partition
Cloud-native
QoS3 Slice B VPN Slice B applications
Transport path
VPN

RAN partition Slice C

QoS4 Slice C Cloud-native
applications

QoS5 Slice D RAN partition VPN Transport path Slice D Cloud-native

CN applications
instance

Cases where this multi-domain

understanding of the service architecture An ecosystem of benefits
is required include the ability to:
• create and manage end-to-end Ericsson Intelligent Automation • revenue increase linked,
network slicing, which is critical Platform enables RAN automation for instance, to 5G enterprise
for 5G enterprise services to be realized and unlocks significant services and driven from
• implement AI/ML driven journey benefits: improved time-to-market
closed-loop assurance for ● faster time-to-market with RAN • opex savings – specifically
multi-domain services where capability alignment and management RAN opex – with considerations of
RAN is an integral part, to maintain across the multi-vendor radio network benefits coming from improved RAN
service integrity and availability ● operational flexibility with operations, planning and optimization
• orchestrate consumer and enterprise automated RAN deployment, operation, • customer satisfaction and metrics
services and scale them up and optimization and energy savings such as the NPS. Improvement here
down in response to demand ● end-to-end integration of results from reduction of network
operations and service experience detractors by improving RAN network
These capabilities will fuel the adoption aspects that increase satisfaction experience, and this can drive revenues
of Open RAN. Better overall performance and higher lifetime value for itself, as well from new customers and
and customer experience for lower enterprise and residential customers reduced churn
cost is a true driver for RAN evolution.
Ericsson is deploying these capabilities An estimation of the benefits By realizing these benefits, the
today in purpose-built and Cloud RAN produced by these drivers is reflected Ericsson Intelligent Automation
networks, and the platform makes it in the three most important areas for Platform becomes a platform of
possible to connect these capabilities the service provider: growth for the service provider.
to the Open RAN domain.
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 1/22112-FGM 101 030

SE-164 80 Stockholm, Sweden to revision without notice due to © Ericsson AB 2021
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