Title:The Coverage Overlapping Problem of Serving Arbitrary Crowds in 3D Drone Cellular Networks

Abstract:Providing coverage for flash crowds is an important application for drone base stations (DBSs). However, any arbitrary crowd is likely to be distributed at a high density. Under the condition for each DBS to serve the same number of ground users, multiple DBSs may be placed at the same horizontal location but different altitudes and will cause severe co-channel interference, to which we refer as the coverage overlapping problem. To solve this problem, we then proposed the data-driven 3D placement (DDP) and the enhanced DDP (eDDP) algorithms. The proposed DDP and eDDP can effectively find the appropriate number, altitude, location, and coverage of DBSs in the serving area in polynomial time to maximize the system sum rate and guarantee the minimum data rate requirement of the user equipment. The simulation results show that, compared with the balanced k-means approach, the proposed eDDP can increase the system sum rate by 200% and reduce the computation time by 50%. In particular, eDDP can effectively reduce the occurrence of the coverage overlapping problem and then outperform DDP by about 100% in terms of system sum rate.