Abstract
Unmanned aerial vehicle-based aerial base stations (BSs) can provide rapid communication services to ground users and are thus promising for future communication systems. In this paper, we consider a scenario where no functional terrestrial BSs are available and the aim is deploying multiple aerial BSs to cover a maximum number of users within a certain target area. To this end, we first propose a naive successive deployment method, which converts the non-convex constraints in the involved optimization into a combination of linear constraints through geometrical relaxation. Then, we investigate a deployment method based on $K$ -means clustering. The method divides the target area into $K$ convex subareas, where within each subarea, a mixed integer non-linear problem is solved. An iterative power efficient technique is further proposed to improve coverage probability with reduced power. Finally, we propose a robust technique for compensating the loss of coverage probability in the existence of inaccurate user location information. Our simulation results show that the proposed techniques achieve an up to 30% higher coverage probability when users are not distributed uniformly. In addition, the proposed simultaneous deployment techniques, especially the one using iterative algorithm, improve power-efficiency by up to 15% compared with the benchmark circle packing theory.
Highlights
L OW-ALTITUDE unmanned aerial vehicles (UAVs) have been increasingly appealing to future wireless communication systems
The non-convex constraints are addressed with a simple geometrical relaxation which converts each non-convex constraint into four linear constraints that can be solved
We have studied the efficient deployment of multiple aerial base stations (BSs) in order to maximize the number of covered users while avoiding inter-cell interference (ICI)
Summary
L OW-ALTITUDE unmanned aerial vehicles (UAVs) have been increasingly appealing to future wireless communication systems. We study the efficient deployment of multiple UAVs so the maximum user coverage probability is achieved, where we define the coverage probability as the ratio of number of covered users to the total number of users within a specific target area. Geometrical Relaxation: we propose a geometrical relaxation scheme where the optimal locations of UAVs are obtained in a ’step-by-step’ fashion, in which the UAV is always deployed in a position such that it covers the most number of remaining users in the target area until there is no space for accommodating more UAVs The formulation of such a problem includes an increased number of non-convex constraints to avoid interference between any two aerial BSs. The non-convex constraints are addressed with a simple geometrical relaxation which converts each non-convex constraint into four linear constraints that can be solved.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
