Virtual antenna array (VAA) formed by unmanned aerial vehicle (UAV) antenna units using collaborative beamforming (CB) technology plays an important role in the air communication system, and can be used in radar, military, disaster rescue and other places. However, there are still some issues with the beam pattern formed by this method, such as high sidelobe level (SLL), high cost and low efficiency. In this article, each UAV carries an omnidirectional antenna unit, and a large number of UAVs form a UAV virtual rectangular antenna array (UVRAA) to communicate with the ground base station (BS). We formulate an overhead minimization and efficient communication multi‐objective optimization problem (OMECMOP) which jointly optimize the excitation current weights of the UVRAA and reduce the number of UAVs in operation to improve the beam pattern, enhance the communication efficiency and decrease the overhead of UVRAA. In addition, we also propose an improved multi‐objective multi‐verse optimization algorithm based on the inverse decline curve type (ISDT‐MOMVO) which introduces a strategy optimization initialization solution with quasi‐opposition based learning (QBL) and a hybrid solution updating operators to solve the OMECMOP. The simulation results show that compared with other traditional swarm intelligence (SI) optimization algorithms the ISDT‐MOMVO algorithm produces better beam pattern and the thinning rate can reach 50%.

中文翻译：

---

无人机虚拟矩形天线阵波束方向图合成多目标优化方法

无人机天线单元采用协同波束形成（CB）技术形成的虚拟天线阵列（VAA）在空中通信系统中发挥着重要作用，可应用于雷达、军事、灾害救援等场所。然而，该方法形成的波束方向图仍存在旁瓣电平(SLL)高、成本高、效率低等问题。本文中，每架无人机携带一个全向天线单元，大量无人机组成无人机虚拟矩形天线阵列（UVRAA）与地面基站（BS）进行通信。我们提出了开销最小化和高效通信多目标优化问题（OMECMOP），联合优化 UVRAA 的激励电流权重并减少运行中的无人机数量，以改善波束方向图，提高通信效率并降低 UVRAA 的开销。此外，我们还提出了一种基于逆下降曲线类型的改进多目标多宇宙优化算法（ISDT-MOMVO），该算法引入了基于准对抗学习（QBL）的策略优化初始化解决方案和混合解决方案更新算子解决 OMECMOP。仿真结果表明，与其他传统群体智能（SI）优化算法相比，ISDT-MOMVO算法产生了更好的波束方向图，细化率可达50%。