An Operational Model for Unmanned Aerial Vehicle-assisted Device-to-Device Relaying

Abstract

In the aftermath of severe flooding, earthquakes, or warfare, the failure of ground-based cellular infrastructure often leaves affected populations and first responders without vital communications connectivity. While Unmanned Aerial Vehicles (UAVs) and 5G NR Device-to-Device (D2D) relaying offer a promising solution for rapid network restoration, a significant gap exists between terrestrial relaying standards and a functional, deployable aerial network. This thesis bridges this gap by designing a 3GPP-aligned operational model for UAV-based relaying.

By adapting mature 5G terrestrial relaying standards to the unique constraints of an aerial network, the proposed operational model introduces specific architecture enhancements to directly address the critical needs of these network-denied environments. For localised use cases through UE-to-UE (U2U) architectures, the operational model helps stranded individuals contact emergency services and supports local coordination. Additionally, for core-connected use cases through UE-to-Network (U2N) architectures, the operational model restores public network access, bridging the isolated disaster zone back to the core network.

The proposed operational model prescribes specific protocol procedures for synchronisation, discovery and connection establishment tailored for aerial networks. To overcome the high collision rates caused by standard decentralised Mode 2 protocols in high-density environments, the operational model introduces a critical architectural enhancement: UAV-governed scheduling, which leverages centralised coordination to ensure reliable out-of-coverage connectivity. Furthermore, the operational model incorporates the specific signalling exchanges required to retrieve real-time radio environment data, enabling the UAV to dynamically optimise its 3D position to maximise network throughput.

The effectiveness of the proposed operational model is systematically evaluated through system-level simulations. The simulations reveal that while standard Mode 2 resource allocation suffers from severe half-duplex and physical collisions under high network loads, the proposed UAV-governed scheduling eliminates steady-state collisions entirely. This coordination guarantees near-perfect packet delivery ratios and cuts end-to-end latency by more than half compared to the uncoordinated baseline. Ultimately, the results demonstrate that UAV-based relaying provides a highly effective 3GPP-aligned solution for rapid network restoration in an emergency scenario.

https://doi.org/10.4121/625a1cd3-63b8-40eb-895d-a20bdbdc0416 Repository link