A Dynamic Cellular Network Framework for Multi-UAV-BS Deployment

UAV-assisted base stations (UAV-BSs) to provide wireless access for terrestrial users is an alternative solution in emergency communication and disaster scenarios. Multi-UAV-BS networks have greater coverage and flexible networking capability, but face problems such as frequency resource conflicts, inter-base station relay links and deployment costs. It is an open topic for discussion whether multi-UAV-BS networks can adopt the cellular architecture applied in terrestrial base station networks. Due to the mobile deployment of UAV-BSs, static interference cancellation or coordination methods are not practical in dynamic scenarios, resulting in further deterioration of signal quality for the edge users. In this paper, a new dynamic cellular network framework is developed for UAV-BS networks with frequency reuse. First, the coverage radius of UAV-BS cells is maximized by solving the convex optimization problem. Second, a non-convex optimization model is proposed to minimize the number of UAV-BSs required, and quickly solved based on the movable location candidate set. The new solution can form a connected backbone network among UAV-BSs and adapt to the dynamic changes of ground users. Simulations are conducted to verify the effectiveness of the proposed method.