Ericsson Technology Review Magazine 2021 – AI special edition

Artificial intelligence (AI) is without question the primary enabling technology for the next generation of system automation. Aside from freeing up valuable resources, improving performance and boosting innovation, AI also opens up a wealth of opportunities for communication service providers (CSPs) to grow their businesses beyond simply providing connectivity.

Magazine

#ericssontechnologyreview

May 7, 2021

Download the magazine

Building trust in telecom AI for the benefit of all

One of the biggest challenges to the wide-scale adoption of AI, however, is the need to fully address the trust issues related to the technology, particularly with regard to cultural and educational differences. At Ericsson, we believe that the best way to build trust is through knowledge sharing and dialogue with stakeholders about the specific AI techniques that we are exploring, testing and using in our products and solutions. The critical infrastructure that we provide for digital enterprises makes it imperative that our customers are confident that we are deploying AI in a trustworthy way in networks, using techniques that are robust, explainable, traceable and unbiased.

The purpose of this special issue of Ericsson Technology Review is to offer our stakeholders greater insight into our view of the role of AI in future network development. In it, we present our latest findings about the potential for trustworthy AI techniques to help us make breakthroughs in six key areas: business support systems (BSS), operation support systems (OSS), privacy, data ingestion, RAN and overall network performance.

BSS and OSS both have pivotal roles to play in the future development of our industry, particularly when it comes to 5G and the Internet of Things (IoT). Many 5G and IoT use cases require BSS that can handle complex business situations and optimize outcomes with minimal manual intervention. AI-native BSS enables the various applications within the BSS to share business information with each other in an efficient and secure manner. Similarly, our AI-powered OSS concept, which enables the use of industry-defined interfaces and open-source modules, offers big advantages over traditional OSS – especially when it comes to supporting emerging 5G and IoT use cases.

Advanced AI techniques are also proving to be enormously helpful in addressing privacy issues and reducing network footprint. Our research indicates that migration from a conventional machine learning model to a federated learning model – a more advanced AI technique – dramatically reduces the amount of information that is exchanged between different parts of the network with no negative impact on QoE.

AI techniques also open up new possibilities to achieve the level of RAN automation required to solve the complex resource management challenges presented by 5G and future radio systems. For example, our RAN experts have already integrated the key software enablers for AI-based RAN automation into a comprehensive framework that provides a solid and flexible technological foundation for the development of AI-based RAN. To help CSPs ensure superior performance across their operations, we have also used AI techniques to develop a harmonized data ingestion architecture that ensures secure and efficient access to data.

Reaching our ultimate goal of creating fully cognitive, zero-touch networks will require AI that is on par with human capability to reason and decide within complex dependencies and across broad domains. While AI is not yet able to offer that level of sophistication, a group of Ericsson experts have proven that it is already possible to reach a high degree of practical autonomous operation in networks by combining existing AI techniques within a flexible architecture to create what we call a cognitive layer. It’s an exciting step in the right direction.

We hope this special issue of our magazine inspires you to think about the role AI-based solutions will play in your organization in the years ahead. Please share it with your colleagues and business partners to help us spread the word about the benefits that trustworthy AI will bring to our industry.

Figure 1: The business intent hierarchy mirrors the business rule hierarchy

BSS and artificial intelligence – time to go native

The growing need to support disruptive services emerging from the Internet of Things and 5G requires a fundamental transformation of business support systems (BSS). At Ericsson, we believe that the best way to achieve this is by forging BSS and artificial intelligence (AI) together to create truly AI-native BSS.

Figure 1: Analytics and policy-driven automation: closed control loop

Open, intelligent and model-driven: evolving OSS

Ericsson is leveraging the open-source approach to build open and intelligent operations support systems (OSS) that are designed to support autonomic networks and agile services. Our concept benefits from our ability to combine the open-source approach with our unique network and domain knowledge.

Figure 1: Overview of federated learning

Privacy-aware machine learning with low network footprint

Federated learning makes it possible to train machine learning models without transferring potentially sensitive user data from devices or local deployments to a central server. As such, it addresses security and privacy concerns at the same time that it improves functionality and performance.

Cognitive processes for adaptive intent-based networking

Extensive automation will be necessary to cope with the unprecedented flexibility and dynamic adaptation that 5G networks are introducing into service delivery and network resource utilization. Our intent-based networking approach combines several artificial intelligence (AI) techniques within a flexible architecture and uses intents to specify what the autonomous system is expected to do.

Figure 1: Data pipeline using the harmonized data ingestion architecture

Data ingestion architecture for telecom applications

Our harmonized data ingestion architecture is designed to help communication service providers ensure superior performance across their operations, including everything from network management to product serviceability and tailored service offerings. An additional benefit of our approach is that it frees up application development resources to focus on use-case realizations instead of data management.

Figure 1: Software-enabler framework for AI-based RAN automation

Artificial intelligence in RAN – a software framework for AI-driven RAN automation

When used correctly, AI techniques have tremendous potential to overcome complex cross-domain automation challenges in radio networks. Ericsson has identified the key software enablers for AI-based RAN automation and integrated them into a comprehensive framework that provides a solid and flexible technological foundation.

Receive the latest industry insights straight to your inbox

Subscribe to Ericsson Technology Review

Subscribe here