Research On On-demand Coverage And Communication Optimization Technology In UAV Ad Hoc Network

Due to the flying nature of Unmanned aerial vehicles(UAVs),it is very attractive to use UAVs as aerial base stations to provide on-demand coverage for ground users at temporary events(such as disaster relief network),hotspot areas or military events.In many applications,the UAVs are connected to construct a UAV airborne network,where a challenging problem is how to deploy multiple UAVs for on-demand coverage while at the same time maintain the connectivity of UAV airborne networks.As a solution to this problem,in this thesis,we first study the on-demand coverage problem of UAV airborne networks,and then propose a user access strategy to achieve load balance among the UAV base stations.We also optimize the network traffic to reduce unnecessary communications among UAVs.Firstly,this thesis proposes an on-demand deployment method of UAV networks based on virtual forces field.The force field is composed by three kinds of virtual forces:(i)The attractive forces to drag all UAVs together towards the hotspots and that to connect the UAVs covering around different hotspots;(ii)The repulsive forces to push away UAVs that are closer than a desired distance,avoiding the overlap of UAVs’ coverage areas;(iii)The attractive force towards the connection line between two hotspots,so as to drag the UAVs to form a multi-hop connection between the flocks of UAVs covering different hotspots.And each UAV will follow the force field to move towards its proper position in a total distributed manner.The algorithm obtains ideal deployment results with short convergence time.The algorithm is validated to be fair to every UAV in the network and resistant to GPS error.Secondly,this thesis analyzes the user traffic in the airspace,and then further improve the deployment algorithm based on virtual forces.In the improved algorithm,the user’s distribution and the load balancing among UAV base stations are considered.And then a novel user access strategy is proposed.In the proposed access strategy,users are more like to access the UAV base station that are associated the least number of users in the available range.The effectiveness of the proposed access strategy is validated via simulation,comparing with the traditional maximum signal-to-noise-ratio access method.Finally,the optimization of network traffic is studied,and a method to reduce the communications among UAVs are studied based on the multi-agent decision model.Under the multi-agent decision model,we conduct experiments with two agents and multiple agents.Comparing with the non-communication strategy and the complete communication strategy,we find that the proposed method can significantly reduce the traffic only at the expense of very little performance degradation.This result provides a promising solution to reduce the network traffic.