Research Of Multi-Aerial Base Stations Deployment Method For Burst Service And Weak Coverage Under 5G Network

In recent years,the rapid development of communication technology has accelerated the application of Unmanned Aerial Vehicle(UAV),UAV has the advantages of controllable mobility and flexible deployment.Equipped with base station functions to form an aerial base station,which can not only be deployed in hotspot areas to share the ground base station traffic,improve user communication capacity;at the same time,it can also be used as an emergency rescue and disaster relief communication platform to complete disaster relief and achieve coverage compensation.Due to the limitation of the battery of aerial base station,it cannot work continuously for a long time,so it is of great significance to improve the energy efficiency(EE)of aerial base station in the face of the diverse communication needs of ground users.Existing research does not consider the deployment cost of aerial base stations and user dynamics when improving EE,which cannot adapt to network dynamic changes,and lack an efficient user perception model.The traditional multi-aerial base stations trajectory optimization method has low execution efficiency and is difficult to obtain the optimal solution,while the Deep Reinforcement Learning(DRL)algorithm can solve complex models efficiently and intelligently.Therefore,this thesis proposes an intelligent deployment method of multi-aerial base stations based on DRL algorithm for two typical scenarios of sudden hotspot and disasters to meet user communication needs.For sudden hotspot scenarios,in order to ensure the communication quality of aggregated users,this thesis proposes an intelligent deployment method of multi-aerial base stations for capacity enhancement,which shares data traffic when ground base stations are overloaded and provides communication services for users in hotspot areas.First,in order to determine the number of aerial base stations deployed,the Affinity Propagation(AP)algorithm is used to complete the detection of hotspot.Secondly,considering the dynamic distribution of users,energy consumption and interference of aerial base stations,EE is defined as the ratio of user transmission throughput and aerial base station energy.Established optimization model of maximizing EE,since the model belongs to the mixed integer nonlinear programming problem,and aerial base station action decision problem has certain dynamics,so the DRL algorithm is applied to solve the problem of trajectory optimization and EE improvement of aerial base station in hotspot areas.The simulation results show that the proposed algorithm can effectively solve the threedimensional deployment problem of aerial base stations in sudden hotspot,and improve the user transmission throughput based on the full use of the energy of aerial base stations.For disaster scenarios,in order to ensure the recovery of communication for disaster-affected users,this thesis proposes a coverage compensation-oriented multi-aerial base stations intelligent deployment method,which jointly deploys Coverage-aimed Unmanned Aerial Vehicle Base Station(Co-UAV-BSs)and Capacity-aimed Unmanned Aerial Vehicle Base Stations(Ca-UAV-BSs)provide communication services for disaster-stricken users.After disaster occurs,the distribution of ground users is usually unknown.This method first deploys Co-UAV-BSs to complete ground user detection,and then applies Density-Based Spatial Clustering of Applications with Noise(DBSCAN)algorithm to complete the clustering of affected users and determine the number of aerial base stations deployed.Co-UAV-BSs are responsible for the full coverage of the disaster area,Ca-UAV-BSs transmit data based on Millimeter waves and are responsible for the capacity improvement of user gathering areas after disaster.EE is defined as the ratio of ground user transmission rate to aerial base station energy efficiency consumption and deployment cost.Finally,the DRL algorithm was applied to continuously adjust the deployment positions of Co-UAV-BSs and Ca-UAV-BSs.The simulation results show that the method solves the coverage problem of the interrupted area,improves transmission rate,and is more flexible and fast than other rescue methods.In summary,this thesis studies the hotspot area scenarios and postdisaster scenarios,solves the energy efficiency optimization problem of aerial base stations in different scenarios,effectively improves the quality of communication,and provides a feasible solution for fast and efficient network compensation in emergency scenarios.