Three-dimensional drone-cell deployment for congestion mitigation in cellular networks

Future cellular networks are expected to offload flash crowd traffic in an area (e.g., a stadium with ongoing sporting events). Owing to the temporary characteristic of this traffic, investing in infrastructures may not be financially viable to provide service coverage for this traffic. Resorting to drone-cells (i.e., unmanned aerial vehicles acting as aerial base stations) to offer rapid but temporary service coverage is a promising solution. Although drone-cells provide quick coverage, efficient placement of drones is challenging. In this paper, we model heterogeneous traffic distribution and formulate a three-dimensional (3-D) drone-cell placement problem with a goal of mitigating congestion for future cellular networks under heterogeneous traffic. To make this problem easy to solve, we convert the 3-D drone-cell placement problem into a new 3-D drone-cell placement problem based on our derived theoretical results. We propose four algorithms to solve the new 3-D drone-cell placement problem. In these algorithms, we first develop four types of drone-cell altitude solvers to achieve a drone-cell altitude. With such an altitude, we further convert the new 3-D problem into a mixed-integer quadratic cone optimization problem in 2-D space, which is solvable by an optimization tool. Simulation results show that these algorithms can effectively help deploy a drone-cell to mitigate network congestion.