Handover-Aware Resource Allocation for Aerial RU-Assisted Open RAN Systems

Abstract

Open radio access network (Open RAN) disaggregation across the distributed unit (DU), centralized unit (CU), and near-real-time radio intelligent controller (near-RT RIC) enables programmable, hierarchical radio access control. This architectural split allows independent optimization of short-term resource management at the DU and long-term mobility decisions at the near-RT RIC, making it ideal for beyond fifth-generation (B5G) networks leveraging unmanned aerial vehicle (UAV)-mounted aerial radio units (ARUs). However, these ARUs must contend with heterogeneous user equipment (UE) velocities and varying service demands. To ensure seamless connectivity, this paper proposes a handover-aware mobility management framework incorporating an adaptive time-to-trigger (TTT) mechanism. The framework is formulated as a mixed-integer nonlinear programming (MINLP) problem using a weighted-sum scalarization approach to maximize aggregate data rate while minimizing handovers, subject to qualityof-service (QoS), power control, and fronthaul capacity constraints. To mitigate computational complexity, a two-timescale heuristic solution executes short-term resource and power allocation at the DU, and long-term user association and handover management at the near-RT RIC. Simulation results demonstrate that the proposed approach effectively controls handover frequency and improves throughput, reducing handovers by 60.37% and achieving up to a $3.3 \times$ data rate improvement over the analytical-TTT baseline.