The unmanned aerial vehicle- (UAV-) assisted sub-6 GHz disaster relief networks cannot meet high-speed transmission requirements. In this paper, the millimeter wave (mmWave) frequency band is combined with the sub-6 GHz frequency band to build a high-speed UAV-assisted disaster relief network. However, the high propagation path loss of mmWave signals usually needs to be compensated by beamforming, where the ground-facing beam of each UAV is the desired receiving beam of ground user information. The different channels need to be allocated to a single UAV so that this kind of beam can be used simultaneously by different ground users to communicate with this UAV. Also, the other UAVs should reuse these channels as much as possible to save spectrum resources. In this paper, the beamforming training (BFT) mechanism is firstly used to obtain the signal-to-noise ratio (SNR) values of all possible links between ground terminals and UAVs, which are used to estimate these links’ energy efficiency. Then, an interference graph construction algorithm is proposed to identify the links that cannot be used simultaneously in the same channel according to the system energy efficiency. Finally, an iterative channel allocation algorithm is designed to allocate new channels to eliminate the edges of the interference graph, so that the links obtained by the BFT process can be used simultaneously as much as possible under the constraint of the number of channels. The simulation results show that our proposed scheme can achieve the shorter average convergence time, the higher data rate (or the lower data loss rate), and the higher energy efficiency.

中文翻译：

---

集成毫米波/低于 6 GHz 无人机辅助救灾网络中的高效无线电信道分配

无人机-（UAV-）辅助的6 GHz以下救灾网络无法满足高速传输要求。本文将毫米波（mmWave）频段与sub-6 GHz频段相结合，构建高速无人机辅助救灾网络。然而，毫米波信号的高传播路径损耗通常需要通过波束成形进行补偿，其中每架无人机的面向地面波束是所需的地面用户信息接收波束。需要为单个无人机分配不同的信道，以便不同的地面用户可以同时使用这种波束与这架无人机进行通信。此外，其他无人机应尽可能复用这些信道以节省频谱资源。在这篇论文中，波束成形训练（BFT）机制首先用于获取地面终端和无人机之间所有可能链路的信噪比（SNR）值，用于估计这些链路的能量效率。然后，提出一种干扰图构建算法，根据系统能效识别同一信道中不能同时使用的链路。最后，设计了一种迭代信道分配算法，分配新的信道以消除干扰图的边缘，使BFT过程得到的链路在信道数的约束下尽可能多地同时使用。仿真结果表明，我们提出的方案可以实现更短的平均收敛时间，更高的数据速率（或更低的数据丢失率），