A Potential Game Approach to Multiple UAVs 3D Placement in Iot Communication Networks

Abstract

In this paper, we study the three-dimensional deployment of UAV in Internet of Things communication, and propose a game theory based UAV self-organization deployment scheme. According to the dynamic characteristics of UAVs, the game model is designed as an extended game, and the optimal performance of the system is obtained by solving the sub-game refined Nash equilibrium solution at each stage. Aiming at the optimal objective function to maximize global coverage, the utility function and potential function satisfying the potential game are designed. In each sub-game stage, a UAV three-dimensional deployment algorithm based on automatic learning machine is proposed to obtain the pure Nash equilibrium in each sub-game stage. Finally, numerical simulation is carried out to verify the UAV self-organizing three-dimensional deployment scheme based on the game framework in this paper. The simulation results show that the Nash equilibrium solution that can be obtained by the learning algorithm proposed in each sub-game stage is almost optimal, which proves the effectiveness of the game framework and the automatic learning algorithm.