5G enables multiple and diverse new use cases, such as: B. real-time streaming of virtual reality feeds from the metaverse, but also increases the complexity for CSPs. To be future-proof, CSPs need to simplify operations, increase sustainability, and improve user experiences. To do this successfully, we need the technology of tomorrow. In other words, networks need to evolve. The good news is that we don’t have to wait until tomorrow, we are already creating the networks of the future and offering an evolutionary path towards intelligent automated radio networks.
In my previous blog Who Are the Intelligent Agents in Networking and Why Do We Need Them? I presented the use of AI technology in networking and explained how this technology can help achieve better results in network optimization and performance. In this blog, I build on that information and provide tips—or should we say foundation—on how CSPs can succeed on their journey — and build the networks of the future.
As previously mentioned, 5G radio access networks (RAN) need to support a variety of use cases. The challenge is to overcome this complexity and ensure great customer experiences with efficient energy consumption for sustainability.
It will require evaluating multiple scenarios with the different network configurations to predict the optimal configuration for network performance, power consumption and radio resource efficiency. For this task we have to apply new technologies. We need Artificial Intelligence (AI) and Deep Reinforcement Learning (DRL) techniques applied everywhere, both in products and services.
However, the use of AI is not enough. We need AI everywhere, but not just anywhere. In other words, we need to run it where it makes sense to unleash all the potential benefits of AI. Below I share four pillars that will enable CSPs to thrive today and into the future based on this holistic approach.
1. Using AI to support CSPs’ decision making
Data is the most important factor in decision-making – good decisions cannot be made without relevant and reliable data. Still, the amount of data that one person or team can process is limited, and that’s where AI comes in.
AI also helps automate the data processing process, sifting through large amounts of data to distill it and ultimately generate actionable insights. By coupling AI, data and automation, a virtuous cycle is created, where data improves the AI algorithm and further improves the networks. AI data-driven algorithms enable networks to adapt to an ever-changing environment to meet ever-changing demands.
The power of an intelligent approach can be demonstrated by Ericsson’s Advanced Microwave Insights solution. This solution detects events on a live network, such as B. Weather events or temporary blocking events (blocking vehicles). It can detect different types of interference and could potentially be used to track weather in the future. With this information, CSPs can identify performance issues for faster resolution.
There are many more examples of this AI data-driven approach in the intelligent services and network development capabilities that are part of the intelligent RAN automation solutions.
In short, CSPs benefit from these AI data-driven automation solutions in network planning, deployment, optimization and repair. Some examples of these solutions available today are listed below.
Figure 1: How AI delivers value in networks
2. rApps to increase efficiency and promote innovation
Have you heard of the rApps? What are they and why are they needed? We mention them here because they are part of the next step in the evolution of smart grids. Imagine having the Networks Evolution Almanac, which tells you what new services your customers need and want. Well, you also want to be prepared for these new services as soon as possible to gain competitive advantage.
rApps provide emerging technologies needed to realize CSPs’ key ambitions: helping drive innovation with a short time to market, improve performance, and optimize CAPEX and OPEX investments.
The rApps are a new software architecture paradigm that brings the following advantages to the CSPs:
- Agility for new use cases Deployed in an open ecosystem where customers and partners can be developers.
- Simplified automated network operations with true closed loop automation with CSPs under control.
- Increase in network efficiency with AI applied in the network at scale to make the best use of the wireless network.
rApps are applications designed for automation and optimization. They are deployed and managed on the non-real-time RAN Intelligent Controller (RIC). If you’ve never heard of RIC, don’t worry, this is an industry term to talk about the platform the apps run on. We will talk more about this in the third pillar.
We can already provide examples of rApps that improve energy efficiency and performance to create more sustainable networks with great user experiences. You can reduce daily power consumption by 15 percent without impacting performance, reduce transmission power by 20 percent, and increase user throughput by up to 30 percent on a congested network.
3. SMO and intelligent automation are a must for future-proof architecture to ensure security and openness
Let’s talk about the development of the intelligent network and the new architecture in more detail.
There is a demand to streamline the number of platforms and make them more open and secure. This is realized in the industry-defined Service Management and Orchestration (SMO) platform. Ericsson Intelligent Automation Platform (EIAP) is the Ericsson SMO with a strong AI foundation. To be clearer, SMO is the industry standard and EIAP is the implementation of the SMO platform with added value and functionality.
Let’s look at the relationship between the rApps and the platform and why both are needed.
rApps are placed in the EIAP, where both rApps and AI functions are automatically supported. EIAP automates rApp lifecycle management and supports AI components such as deployment, model monitoring, and retraining. This functionality enables custom development and deployment with automated orchestration and conflict management. In summary, rApps orchestrate radio and the EIAP is the rApps orchestrator.
Ericsson is a leader within industry standards (O-RAN Alliance) to ensure the platform and its rApps and interfaces are secure. In addition, faster and deeper threat detection based on the developed platform and rApps functionality will improve RAN security.
4. Open ecosystem for innovation and monetization
The main goal for the open architecture of the platform and rApps is innovation and monetization. This goes beyond the standardization of interfaces; We need to enable CSPs to collaborate with partners and develop new use cases faster.
When it comes to intelligent RAN automation solutions and how they can support service providers with faster time to market for new use cases and new services, the Software Development Kit (SDK) appears as a key enabler. Along with the EIAP, the Ericsson SDK, frameworks and components enable and expand the developer ecosystem around rApps.
Figure 2: Evolution of network architecture and rApps ecosystem
In summary, this environment supports the ambitions of both CSPs and third parties to provide new functionalities covering network-wide automation use cases (e.g. for monetization).
Conclusion
The AI applied to network automation together with the architecture presented with these four main pillars supports an open ecosystem of constant innovation and collaboration for all parties: service providers, vendors and partners in the future development of the telecom industry.
AI technology will enable CSPs to win the 5G game by providing a competitive advantage with data-driven AI algorithms while enabling them to reduce CAPEX and OPEX, simplify operational complexity, increase sustainability and improve user experience. At Ericsson, some of these functionalities have already been tested with excellent results.
Figure 3: Benefits of intelligent RAN automation
To master the 5G challenges and successfully use the new possibilities, your “Back to the Future” almanac is here: the Intelligent RAN automation.
learn more
Read the Intelligent Automation Guide on how Intelligent RAN Automation creates key use cases for service providers
Dive deeper into intelligent RAN automation
Discover the intelligent automation platform
Learn more about rApps
Read more about Telecom SDKs
Learn more about AI in networks
