rApps – your best bet for innovation in 5G networks

Press play on photo to watch the video

Do you want to listen to this videoepisode instead? Play the audio only.

The rApps are a new software architecture paradigm that brings valuable differentiators to the CSPs:

• Agility for new use cases implementation
• Simplified and automated network operations (Truly Closed loop Automation with CSPs in control)
• Boosting network efficiency with AI applied in the network at scale

Ericsson is developing an industry-leading portfolio of rApps for both 5G and 4G radio access networks. In this blog post, we introduce the technical side of the rApps and elaborate on why they are needed to succeed in the new challenges and opportunities with 5G networks, especially for sustainability, network performance, and operational efficiency. We will explain the product principles: technological foundation, functionality, and execution environment.

So, keep reading to find the answers to the questions of what are rApp, why they are needed for advanced RAN automation, and what value they bring to the 5G networks. In other words - why rApps are the best bet on innovation for 5G networks.

The context – why we need rApps in 5G

There are several factors that are creating high complexity in networks’ operations with 5G introduction. Networks are evolving toward new cloud-based architecture. On the other hand, there is an increased variety of network services, from simple voice and text messages to extremely data-hungry applications, e.g., stream real-time XR feeds.

Together with these new services, there is the explosion of numerous types of devices. This complex environment requires RAN not only to cope with all diverse use cases and devices but also to be able to deliver an impeccable user experience for each of them.

And we are not done yet with elaborating on the networks’ complexity, as it further increases with the deployment of new hardware and software capabilities. More deployments mean increased network density, with added capabilities and new security threats that must be taken into account too.

Faced with such networks’ complexity, the operations teams really need good support to keep CAPEX and OPEX optimized. The necessity for advanced automation has never been greater, and rApps have emerged as an innovative element that helps solve this complexity.

But please relax and enjoy the journey because it’s not only about challenges. It’s also about joining the ride of the new possibilities, where rApps can provide CSPs with crucial competitive advantages for winning in the telecom market. If done right, accelerating time to customer and developing new use cases will create performance leadership and maximize the return on investment, and rApps are here to make it real!

What are rApps actually?

To call rApps your best bet for innovation in 5G networks is a bold statement, so let us elaborate.

We are living in times of networks technology evolution with decoupled hardware and software and new software functionalities deployed as microservices, so-called cloud-native applications. We are on the journey to the prevailing cloud deployments. Moreover, AI technology is applied everywhere in software and services to create more intelligent networks.

If your radio network functions were musicians, rApps are their conductors. Just like each musician in the orchestra can be a virtuoso per se, and a conductor is here to unite all the individual players into perfect harmony, in the same way, rApps are needed to harmonize all your radio access network functions.

Now, to a more technical explanation…

“rApps” are applications designed for automation and optimization. They are deployed and managed on the non-real time RAN Intelligent Controller (RIC). If you haven’t heard about non-RT RIC before, you will find more details in the rApps execution environment section. Essentially, what you need to know is that Ericsson is developing rApps as cloud-native software applications that can benefit from using AI to improve the performance of the radio network functions at scale.

The ultimate purpose of rApps is to automate the network’s operations in an orchestrated way to get the most out of the radio network.

A bit more on rApps… What is the rApps role?

To understand the rApps role, we need to introduce different control loops for automation.

When provided with network feedback data on its performance and execution, control loops enable further optimization and the establishment of a positive feedback loop, e.g., optimizing configuration, orchestration of radio resources, or handover decisions within the RAN. They can execute from start to end in a closed loop on a time scale varying from microseconds to weeks, depending on their scope.

Since rApps act on so-called “higher-level” control loops, with a response time of half a second and above, they are suited for handling the network at scale, including operations like orchestration, optimization, and many more. This ability to orchestrate across the network requires rApps to be placed centrally.

Have we made your head spin in loops? Stay with us as we explain further.

For those who want to know more - why are rApps placed centrally?

There are three considerations for how and where to execute a control loop (and justifications why rApps can be placed centrally):

1. Time scale: operations that require swift response times (e.g., handover decisions) will need to execute in the radio network functions, whereas the orchestration and other high-level operations (e.g., network optimization) benefit from a centralized location.
2. Compute and data resources: managing compute and data resources more efficiently is the following reason for having a centralized deployment. Aspects that count here are CAPEX, operational efficiency, and sustainability. By centralizing compute resources, CSPs can achieve higher margins and scale.
3. RAN management efficiency: some operations can benefit from distributed or local execution, making the best use of the equipment and radio resources. This is the case for the fast control loops that don’t require input from other areas of the network.

rApps can be deployed in a centralized platform connected to the radio network functions to have the network view. For this reason, rApps are better suited to control the higher-level control loops since they have insights into the broader network and its data. They can provide more efficient compute use, data handling, and the best decision-making process.

If you were wondering why rApps are not deployed as radio features, now you may already have the answer. It’s because rApps need to be deployed in a centralized platform connected to the network functions in order to have the broader network view. Do you remember the orchestra conductor from the beginning of our blog post and their central position on the stage to have a view of all the musicians for achieving the perfect harmony? The rApps role is to be the conductor, that’s it!

The rApps execution environment: Ericsson Intelligent Automation Platform

By now, we have explained what rApps are, how they work, and which position they have in the network.

The next question to answer is what is their execution environment. Bear with us for some more technical (but important to understand rApps) explanation.

rApps are placed in the non-Real Time Radio Intelligent Controller (Non-RT RIC) within the Service Management and Orchestration (SMO) framework in the network hierarchy. This layer and more functionalities, including the non-RT RIC, which orchestrates and manages the different rApps, are implemented by the Ericsson Intelligent Automation Platform (EIAP). Another crucial detail to mention is that AI-supporting technology for automating the rApps lifecycle management and data management functionality also resides in the EIAP. This includes lifecycle management components like deployment, model monitoring, and retraining support.

Further, the EIAP can lifecycle manage rApps individually, allowing for rApps to be developed and deployed individually with automated orchestration, conflict management, etc.

Just like the rApps conduct the RAN, the Ericsson Intelligent Automation Platform conducts the rApps.

Along with the EIAP, Ericsson provides software development kit (SDK), frameworks, and components for enabling and expanding the developer ecosystem around rApps. This environment supports both CSPs’ and third-party ambitions to provide new functionalities, covering network-wide automation use cases.

The rApps ecosystem benefits everyone

The architecture that enables rApps ties in with the bigger picture of increasing openness and a wider ecosystem of industry developments.

Indeed, it is the rApps nature to provide an ecosystem in which multiple players can benefit. The open interface (R1) that they utilize allows the development of rApps by CSPs, vendors, and third parties.

As rApps communicate with the RAN over the ORAN-defined R1 interface, this allows CSPs and 3PPs to create innovation in the form of rApps, while the platform handles their deployment, orchestration, and conflict management.

That way, the lifecycle management of applications and platforms becomes streamlined, and faster innovation allows CSPs to address issues like complexity and achieve their winning ambitions. Further, the architecture and ecosystem prepare them for the future development of the telecom industry.

In which use cases can rApps do their best?

The opportunity created by rApps to easily deploy intelligence centrally can be leveraged in a multitude of ways. So, in which use cases can rApps be used at their best?

Some examples include energy efficiency, smarter deployment and configuration, and upgrades.

Let us elaborate here about the energy efficiency use case. When optimizing centrally for energy efficiency, rApps can leverage local functionality, such as cell/sector sleep mode, to save energy in specific areas of the radio network in low-traffic scenarios. There are also other methods to address this case, but an AI-based rApp approach allows for simple prioritization according to CSP strategy while reducing manual operations.

For network deployment, the ideal scenario is to have a plug-and-play capability for the new radio network functions so that the manual intervention is minimal and the configuration and integration in the existing network are fully automated.

Another important use case of rApps is autonomous optimization. Some performance features, such as Carrier Aggregation, require a high amount of configuration effort to work at their optimum. In other words, many configuration parameters must be set to optimize the network. Furthermore, with new hardware and software introduction, these configurations require updates. So, the ability to address this complexity in an automated manner saves OPEX and optimizes performance.

Last but not least, one more example of rApps use cases is the verification and assurance for the network upgrade. When upgrading the network, verification and assurance of introduced features can be done more efficiently by adding AI capabilities to the software functionality that handles new deployments. Moreover, having centralized control of a given update enables in-service software upgrades (ISSU).

This all makes rApps a must-have for your 5G network’s impeccable performance

When we say – impeccable, we do mean it. It’s like listening to the Vienna New Year’s concert, and when all the tones merge into the perfect harmony for enjoyment.

The rApps come to the stage to harmonize the overall network performance, bringing with them differentiating benefits:

• Allowing intelligent closed loop automation at the network scale
• The functionality developed with shorter TTM and adapted to each CSP´s network
• Innovation with easy-to-develop and deploy customized use cases
• Future proof architecture for intelligent networks

It is up to CSPs and partners to join and enjoy the result they influenced themselves. Just like in Radetzky March in Musikverein Golden Hall when the public takes part in the performance to create the perfect outcome.

Related content

Ericsson Intelligent RAN Automation

Intelligent Automation Platform

Intelligent Automation rApps

Intelligent Automation Platform SDK

Innovation potential of Non Real-time RIC