Dynamic Beam Management for High–Low-Frequency UAV Networks

Terahertz (THz) communication with its wide bandwidth resource and large channel capacity offers significant potential for data-intensive Internet of Things (IoT) applications through unmanned aerial vehicle (UAV) networks. As THz waves have high directionality and severe propagation loss, massive MIMO system with beamforming is usually applied to achieve high gain and mitigate transmission interference. However, with a large number of antennas and fast varying channels, it will cause unacceptable overhead. To further address the challenges of THz propagation, this paper proposes a dynamic beam management method to stabilize the high frequency directional network and optimize the beamforming in UAV-assisted high-low frequency IoT communications. In particular, a high-low frequency collaborative network with beam training and tracking is employed, utilizing out-of-band lower frequency for auxiliary information transmission. In addition, a novel hierarchical codebook design is explored to enhance beam training performance, followed by a simplified beam training procedure that utilizes low frequency information. Moreover, an effective beam tracking method using previous information further reduces the re-training complexity. The communication framework based on this procedure can quickly complete beam alignment and dynamically track the beams according to the real-time quality of service. Numerical results validate superior performance of our proposed dynamic beam management method compared to the benchmark methods.