Inhomogeneous Multi-UAV Aerial Base Stations Deployment: A Mean-Field-Type Game Approach.

In recent years, unmanned aerial vehicles (UAVs) are more widely applied due to low cost and high flexibility. Facing the suddenness of emergency events and the uncertainty of user service requests, the rational deployment of the UAV aerial base stations (ABSs) has become an effective solution. However, how to deploy multiple UAVs with the variety of properties (power, service radius, and so on) is a challenge problem. There is a mean-field-type game (MFTG) to obtian optimal startegies which has been applied in practical applications. The arbitrariness of the number, the distinguishability of agents and the non-negligible effect on system are considered in MFTG. Because the homogeneity of multiple UAVs cannot be guaranteed in the practical communication model, this paper formulates the deployment of multiple UAV ABSs with various properties as a MFTG. In this scenario, each UAV decides the flight strategy at the next moment to minimize its own cost function by analyzing the communication requests of the ground users. In addition, the existence of the Nash equilibrium of the MFTG problem is proved. And the direct square complement method is used to solve the problem to minimize the cost function of each UAV. The simulation results show the correctness of the proposed method and the rationality of the deployment.