Multiple Aerial Base Station Deployment and User Association Based on Binary Radio Map

Abstract

The utilization of the aerial base station (ABS) has been treated as a promising solution to the coverage, deployment and cost savings problems in the wireless networks. Different from the terrestrial networks, in the ABS assisted air-to-ground network, the service positions of ABSs and user association should be jointly optimized, which is challenging due to the unknown propagation environment in the task area. In this paper, a binary radio map (BRM) is constructed to grant the ABSs the location-specific channel knowledge within the entire task area. With the assistance of the BRM, a multiple ABS deployment and user association framework is proposed. The framework consists of an offline optimization stage during which the initial ABS deployment position and user association are jointly optimized based on the BRM, and an online refinement stage during which the ABS positions and user association are refined in real time according to both the observed quality of service (QoS) and virtual QoS emulated by the BRM. The performance of the proposed framework is examined in a complex urban scenario with mobile users. The results show that higher achievable rate performance and lower on-board energy consumption can be realized when compared with the reference schemes.