UAV-BS Deployment Strategy Based On Intelligent Optimization Algorithm

5G mobile communication and Internet of Things are hot areas of research and application in the intelligent era.As one of the emergency communication infrastructures of 5G and Internet of Things,The unmanned aerial vehicles-base station(UAV-BS)has been rapidly developed and deeply studied in recent years.When the number of outdoor wireless users surges and the fixed base stations are difficult to adapt to high-traffic communication,the UAV-BS can provide communication services as an emergency communication scheme.However,there are many issues and challenges with the UAV-BS and its network.For example,the UAV-BS coverage capability is one of the crucial primary issues for improving the quality of network communication.At present,according to the application scenarios,the UAV-BS coverage problem can be divided into scanning and deployment coverages,in which the deployment coverage can be subdivided into coarse-grained and fine-grained deployments.Aiming at the problem of maximum coverage and maximum network throughput in the fine-grained deployment of UAVBSs,this thesis studies the UAV node deployment strategies of the single-objective optimization and multi-objective optimization with limited energy while improving the network performance and the duration of flight of UAV-BSs.As an NP-hard problem,the fine-grained deployment of multi-UAV-BS is difficult to solve by traditional methods,while the intelligent optimization algorithms based on computational intelligence mechanisms can usually provide approximate optimal solutions.Compared with other intelligent optimization algorithms,the Particle Swarm Optimization(PSO)has the advantages of easy implementation,the shared information among particles,and being less affected by the problem dimensionality.Therefore,this thesis solves the single-objective optimization and multi-objective optimization problems of UAV-BS deployment based on the improved PSO algorithm to enhance the performance of UAV-BS deployment.For the single-objective deployment of UAV-BS,this thesis aims to optimize the user coverage and throughput that determine the quality of service(Qo S)of the whole network.We regard both user coverage and network throughput as single-objective optimization problems,and propose an intelligent hybrid algorithm,named PSO-VFA.The algorithm is based on PSO and firefly algorithms to make up for the shortcomings of particle swarm optimization algorithm,such as low accuracy and insufficient local search capability.The PSO-VFA algorithm is divided into two stages with different intelligent algorithms: first,search the coverage area globally using the particle swarm optimization algorithm;then,the Virtual Repulsive Force-based Firefly Algorithm(VFA)proposed in this thesis performs a local search to maximize user coverage and network throughput.In the VFA algorithm,in addition,the user is taken as the target to attract the UAV,and the virtual repulsion force is introduced for the UAV position adjustment to achieve better fine-grained deployment.Simulation results show that,compared with a single intelligent algorithm,the intelligent hybrid algorithm proposed in this thesis has higher user coverage,network throughput,and faster convergence speed.In this thesis,the UAV-BS deployment problem in three-dimensional space is regarded as a multi-objective optimization problem,and the user coverage,throughput and mobile energy consumption in the network are simultaneously optimized.The battery capacity carried by a UAV node is limited,and a lot of energy will be consumed when flying and moving in the target area.Therefore,mobile energy consumption should be considered in the deployment of UAV nodes.This thesis proposes a Virtual Force-based Multi-Objective Particle Swarm Optimization(VF-MOPSO)algorithm,which uses the idea of the Pareto optimal solution set to simultaneously optimize user coverage,network throughput,and mobile energy consumption.Finally,we provide a deployment scheme for multiple UAV-BS-groups.To provide mutation operations for particles on the boundary and improve the solution accuracy of the algorithm,the VF-MOPSO algorithm introduces mutation operators on the basis of a multi-objective particle swarm optimization algorithm.Moreover,the virtual force is used to adjust the position between UAV nodes and enhance the convergence speed of the algorithm.The simulation results show that the proposed VF-MOPSO algorithm can get a desired deployment scheme in the multi-objective optimization problem.Compared with the MOPSO algorithm,it can reduce the mobile energy consumption of UAV while maintaining the coverage rate and network throughput.