
5G represents a tipping point in the telecommunications industry where networks are becoming too complex for humans to operate profitably without the use of automation tools and technologies. The complexity is partly due to 5G itself, which uses a much wider set of frequency bands, can prioritize latency-based services and supports huge increases in the number of network elements and end-user devices. But there are a plethora of other changes that further increase the complexity.
These include the evolution of physical hardware to virtual and cloud-native networks, end-to-end network slicing, adoption of Open Radio Access Network (RAN) technologies, and the addition of new enterprise business services. . There are also multi-technology networks with some communications service providers (CSPs) running 2G, 3G, 4G/LTE and 5G networks in parallel, as well as multi-vendor networks with typically two to four different RAN providers deployed in the network.
Artificial intelligence (AI) and machine learning (ML) are becoming commonplace in the telecommunications industry and are often the only way to manage the complexity we see in today’s multi-vendor, multi-technology networks. today. This complexity becomes even more apparent in RAN which is one of the most difficult areas to tackle due to its distributed nature, number of network nodes and proximity to end users, making it, not surprisingly , a major consumer of OPEX. .
The evolution of RAN incorporates automation
Telecommunications industry automation is strongly tied to the ubiquitous use of AI – but where it makes more sense depending on the use case. For example, improving CAPEX/OPEX rationalization and performance requires large-scale network actions. The good news is that the latest network technologies – 5G and Open RAN – were designed for large-scale automation. In fact, the O-RAN Alliance defines a new Service Management and Orchestration (SMO) architecture specifically designed to enable better RAN automation.
The key to success then is for network providers to automate the right things and aim to continually improve performance, which requires applied intelligence. When it comes to the evolution of RAN automation, we can see AI and ML technologies used primarily in three specific areas.
- SMO platform – the SMO platform itself is designed to integrate AI technologies. At its core, it has a built-in AI/ML runtime environment. The platform is designed to connect to large external data sources as well as support north and south interfaces.
- Life cycle management – we see an urgent need to make greater use of AI in the lifecycle management of virtual and cloud native network software components. One of the primary goals of RAN automation is to replace the manual work of developing, installing, deploying, managing, optimizing, and retiring RAN functions. Because AI and ML have proven to be effective tools for developing RAN automation functionality, the use of AI and ML to drive lifecycle management and CI/CD tools is obvious. . It is expected that AI and ML will be widely used in the training and retraining of deployment models, from using a generic and global model to a much more self-contained integrated model with self-retraining. Data collection and management is one of the biggest challenges for scaling AI/ML software and tools in CSPs. It is quite relevant to know how the data is managed in the life cycle of the algorithm.
- RAN Automation rApps – In the O-RAN SMO paradigm, RAN automation use cases are implemented in applications called “rApps”. rApps rely heavily on the use of AI and ML technologies simply to manage the large number of variables within the network. For example, you may have an rApp designed to detect and compensate for cell failures on the network. In the event of an outage, the rApp automatically extends coverage to neighboring cells to minimize the impact of an out of service cell, while maintaining acceptable service levels. The actions are then undone once the cell returns from the failure. The ability to automatically compensate for cell failures eliminates manual labor and increases resolution speed, improving user satisfaction. But AI technologies are needed to make this possible.
AI and ML are essential in modern mobile networks and any service management and orchestration system must use and support the use of AI. AI is in everything we do.
About the Author

Peo Lehto, OSS Solutions Area Manager, Ericsson Digital Services. Ericsson Digital Services provides solutions that realize the digital transformation of customers, including software and services in the areas of monetization and management systems (OSS/BSS), telecommunications core (packet core and communication services) and cloud infrastructure and NFV (network function virtualization). Peo also led Ericsson’s IP and Transport practice in North East Asia, led Fixed Broadband Convergence for Ericsson in Japan, as well as the Node Development Organization EPG for Ericsson in Sweden. Peo was born in Sweden in 1968. He holds an Ms.Sc. degree in Electrical Engineering and an MBA in Industrial Marketing and Purchasing from Chalmers University of Technology in Gothenburg.
Sign up for free at insideBIGDATA newsletter.
Join us on Twitter: https://twitter.com/InsideBigData1
Join us on Facebook: https://www.facebook.com/insideBIGDATANOW