Modeling and Performance Analysis of UAV-Assisted Cellular Networks in Isolated Regions

Unmanned aerial vehicle (UAV) provides a quick and feasible solution to urgent communications in isolated region, in which reliable wireless communication link is unavailable through conventional terrestrial base stations (BSs). In this paper, we model the UAV- assisted cellular network with isolated region as a homogeneous Poisson point process in the infinite 2D plane with a circular nulling area. Within the circular area, users are served by a UAV-mounted BS hovering at the center of this area. We apply a probabilistic multi-slope path-loss large-scale model as well as Nakagami-m small-scale model to characterize the air- to-ground channel. With mathematical tools from stochastic geometry, we derive the coverage probability of the users located within the isolated region. Based on the derived expression, we show the existence of the optimal UAV altitude to achieve the maximal coverage probability. We further investigate the impacts of isolated region size and ground BS density on the optimal UAV altitude and the achievable maximal coverage probability.