Research on Improving Communication Capacity in mmWave Backhaul UAV Networks

Millimeter-wave (mmWave) unmanned aerial vehicle (UAV) networks are a promising solution for rapidly deployable backhaul systems in urban disaster scenarios, where terrestrial infrastructure may become unavailable. Although mmWave bands provide wide bandwidth for high-capacity transmission, their strong susceptibility to blockage and beam misalignment poses significant challenges in dense urban environments, particularly under UAV positional fluctuations caused by wind. This study investigates the optimization of multi-hop mmWave UAV backhaul networks with the objective of maximizing the bottleneck link capacity. A three-dimensional urban model of the Shinjuku area in Tokyo is employed, and radio propagation is evaluated using a ray-tracing-based approach considering line-of-sight (LoS) constraints and inter-link interference. Particle Swarm Optimization (PSO) is used to determine optimal UAV placements for two- to four-hop configurations. Numerical results demonstrate that multi-hop relaying combined with directional 2 × 2 patch antennas significantly improves the minimum link capacity, enabling the target backhaul capacity of approximately 9 Gbps per link under static conditions. However, capacity degradation is observed when UAV jitter is introduced due to LoS blockage and beam misalignment. To address this issue, a jitter-aware optimization method incorporating an expanded Fresnel-zone constraint is proposed. The proposed method substantially mitigates capacity degradation under realistic positional fluctuations, resulting in more robust backhaul performance. These findings demonstrate that jitter-aware placement design is essential for realizing reliable high-capacity mmWave UAV backhaul networks in dense urban disaster environments.