Build trust in telecommunications AI for the benefit of all
Artificial intelligence (AI) is without question the most important technology for the next generation of system automation. In addition to freeing up valuable resources, improving performance and increasing innovation, AI also offers a variety of opportunities for communication service providers (CSPs) to grow their business beyond simply providing connectivity.
However, one of the biggest challenges to the widespread adoption of AI is the need to fully address the trust issues related to technology, especially with regard to cultural and educational differences. At Ericsson, we believe that the best way to build trust is to share knowledge and engage with stakeholders about the specific AI techniques we research, test and use in our products and solutions. The critical infrastructure we provide for digital businesses makes it imperative that our customers are assured that we are using AI in a trustworthy manner on networks, using techniques that are robust, explainable, understandable, and unbiased.
The purpose of this special edition of Ericsson Technology Review is to provide our stakeholders with better insight into our view of the role of AI in future network development. In it, we present our latest insights into the potential of trusted AI techniques to help us achieve breakthroughs in six key areas: business support systems (BSS), operational support systems (OSS), data protection, data ingestion, RAN, and general network performance.
BSS and OSS both play a crucial role in the future development of our industry, especially 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 results with minimal manual intervention. AI-native BSS enables the various applications within the BSS to share business information with one another in an efficient and secure manner. Similarly, our AI-based OSS concept, which enables the use of industry-wide defined interfaces and open source modules, offers great advantages over conventional OSS – especially when it comes to supporting new 5G and IoT use cases.
Advanced AI techniques are also proving extremely helpful in solving privacy issues and reducing the network footprint. Our research shows that migrating from a traditional machine learning model to a federated learning model – a more advanced AI technique – dramatically reduces the amount of information exchanged between different parts of the network without negatively impacting QoE.
AI techniques are also opening up new opportunities to achieve the level of RAN automation required to solve the complex resource management challenges of 5G and future wireless systems. For example, our RAN experts have already integrated the most important software enablers for AI-based RAN automation into a comprehensive framework that provides a solid and flexible technological basis for the development of the AI-based RAN. To help CSPs ensure superior performance in their operations, we have also used AI techniques to develop a harmonized data collection architecture that ensures secure and efficient access to data.
In order to achieve our ultimate goal of creating completely cognitive networks without contact, an AI is required that corresponds to the human ability to argue and decide within complex dependencies and across wide areas. While AI is not yet able to offer this level, a group of Ericsson experts has proven that it is already possible to achieve a high level of practical autonomous operation in networks by using existing AI techniques in a flexible architecture can be combined 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 reflect on the role AI-based solutions will play in your company in the years to come. Please let your colleagues and business partners know so we can publicize the benefits of trustworthy AI for our industry.