Project
Integrated E2E network-wide management
Developing an end-to-end fault management system that provides a comprehensive view of the network.
Initial approach
This project takes a cross-layer, end-to-end (E2E) approach to transforming cellular network management, covering:
- The radio layer, including interactions between base stations and user equipment.
- Virtualised resources and the configuration of RAN elements.
The project uses advanced Machine Learning techniques to estimate indicators of end-user perceived quality under different network conditions, resource availability and contextual factors. These predictive capabilities provide the foundation for advanced AI mechanisms designed to optimise E2E quality, improving user experience while making more efficient use of network resources.
One of the project’s main objectives is to develop an integrated fault management system capable of providing a comprehensive view of the network. The system brings together information from both the radio link and virtualised RAN resources to improve network resilience.
To achieve this, the project will develop new AI algorithms and mechanisms capable of taking a cross-layer, E2E view of the network. Building on metrics and methodologies established in previous initiatives, these solutions will bring them together into a robust framework for optimisation and fault management in next-generation cellular networks.
Use case
As networks become increasingly decentralised and AI-driven network control grows in importance, identifying the source of problems within an O-RAN network has become even more complex.
Consider a group of users on a video call with friends in another country. Midway through the conversation, video quality suddenly deteriorates. At that point, it may not be clear whether the cause is network interference, incorrect radio resource configuration, disruption caused by poor handover management, high load on a gNB or even an issue with the media relay server.
There is therefore significant interest in giving operators deeper insight into what is happening across the Radio Access Network by detecting and mitigating faults specific to O-RAN environments.
Innovation focus
Our work on fault detection and mitigation focuses on gaining a detailed understanding of how different applications perform under real-world conditions, including their communication patterns with users and servers and their ability to adapt to changing network conditions.
By analysing the characteristics of the network traffic generated by these applications, we gain valuable, independent insight into their behaviour and performance. This knowledge enables us to develop advanced tools such as real-time Quality of Experience (QoE) estimators capable of identifying the root causes of network faults.
This end-to-end approach not only improves our ability to detect faults quickly, but also supports effective mitigation strategies, helping applications run smoothly while improving overall network reliability.
Let’s talk
We look at every challenge in its real-world context to determine what’s worth building — and the right way to build it.