Project
Radio Resource Management in O-RAN
Developing advanced optimisation and fault management algorithms that use predictive insights and contextual data to improve network performance.
Initial approach
This project explores advances in future B5G/6G networks. These networks will introduce significant innovations, including:
- Higher frequency bands (mmWave, THz).
- Cell-less deployment models.
- Device-free deployment models.
- Reconfigurable intelligent surfaces.
These emerging radio environments bring significant challenges and call for entirely new approaches to traditional tasks such as mobility optimisation and resource allocation.
We address these challenges by developing advanced optimisation and fault management algorithms that use predictions and contextual information — such as user location — to improve network performance.
Use case
Mobility management is critical to telecommunications networks, ensuring seamless connectivity and a consistent user experience. It involves the intelligent coordination of handovers, session continuity and resource allocation to serve users in highly dynamic environments, such as densely populated urban areas or high-speed transport networks.
Current L3 handover mechanisms are primarily reactive, relying on historical measurement data or triggering events to initiate and execute handovers. While effective for low-mobility user equipment (UE) in macrocell environments, this approach can struggle in scenarios involving highly mobile UEs, dense small-cell deployments or advanced services such as XR.
Reactive approaches can lead to handover failures, radio link failures, ping-pong effects, throughput degradation and poorly timed handovers. To improve robustness, 3GPP has introduced Conditional Handover (Rel-16) and LTE HO (Rel-18), aimed at reducing interruption times during frequent handovers between small cells. However, these methods remain inherently reactive by design.
Innovation focus
Our mobility management use case explores advanced approaches based on the regression and link prediction capabilities of Graph Neural Networks (GNNs).
Within cellular networks, GNNs provide a powerful framework for modelling complex relationships between network entities such as user equipment (UE), base stations and radio links.
By applying GNNs, we aim to accurately predict critical radio-link metrics — including signal strength, interference levels and handover success probabilities — as well as determine the most likely next cell for a given UE.
This proactive approach enables smarter, more efficient handover decisions, reducing latency, minimising handover failures and improving overall network performance.
By integrating these advanced AI techniques, we aim to rethink mobility management, enabling seamless connectivity and a better user experience across dynamic, high-density network environments.
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.