3D deployment of UAV-mounted base stations for heterogeneous access requirements

Recently, unmanned aerial vehicles (UAVs) have been reported a lot as aerial base stations (BSs) to assist wireless communication in Internet of Things (IoT). However, most results for UAV deployment require uniform access requirements and obstacle-free environment. This paper considers multi-UAV-mounted BSs deployment problem in obstacle-laden environment to achieve on-demand coverage for ground user equipments (UEs), with the goal of minimizing the number of required UAVs for full coverage of all UEs by optimizing the three-dimensional (3D) positions of UAVs as well as UE clustering. On one hand, the considered problem is challenging as the constraints on the non-uniform Quality of Service (QoS) requirements, the service ability of each UAV, the beamwidth of the directional antenna equipped by each UAV and obstacle avoidance are jointly taken into account. On the other hand, these practical considerations contribute to broadening the scope of applications. We formulate the problem as a mixed-integer programming problem and develop a three-step placement algorithm to achieve the goal. First, the maximum allowed service radius of each UE is derived based on the Karush-Kuhn-Tucker (KKT) condition to satisfy required signal-to-interference-plus-ratio (SINR) threshold. Second, a greed-prior chaotic search enabled adaptive artificial bee colony (GP-CA2BC) algorithm is proposed to group the UEs such that the minimum number of required UAVs can be obtained. Third, the 3D position of each UAV is optimized to further enhance the QoS. Finally, some simulation results are presented to demonstrate the superiority of the proposed scheme in decreasing the number of required UAVs as well as improving the QoS compared with the existing works.