Research On Optimization Method Of 5G Network Air Base Station Deployment Based On UAV

The Internet of Things(IoT)has broad prospects as an application of the fifth generation(5G)mobile communications.However,insufficient terrestrial network infrastructure resources limit the deployment and integration of Io T devices.UAVs can be deployed on-demand as air-assisted base stations(AABS)for Io T,facilitating equipment integration and overcoming the limitations of insufficient ground infrastructure.In this thesis,the optimization method of UAV-based 5G network air base station deployment is studied,and the corresponding improved algorithm is adopted to improve the optimized performance for the two deployment problems presented.In the scenario of static deployment of AABS,UAVs need to cover Io T devices that cannot be served by ground base stations and provide communication and services.Considering energy consumption and communication performance,solving UAV deployment problems requires optimizing both its optimal number and communicationrelated parameters.Therefore,the problem is modeled as a joint deployment optimization problem of UAVs(JDOPUAV),which minimizes the number of deployed UAVs while maximizing the average throughput of UAVs and the minimum throughput of UAVdevice pairs.On this basis,an improved biogeographical-based optimization algorithm IBBOFCI is proposed to solve the constructed problem.The algorithm introduces a flexible local selection mechanism in the process of regenerating the solution,which avoids the waste of a large number of computing resources by reducing the number of loops in the algorithm iteration.In addition,both the chaotic mechanism in migration operations and the introduction of intrusion operators for high HIS habitats enhance population diversity,thereby expanding the search range of the algorithm.In simulation experiments with other comparative algorithms,the effectiveness of the improvement factor of the IBBOFCI algorithm is proved.In the scenario where data collection is completed within the wireless transmission field of the UAV flight to the Io T device,it is necessary to ensure the "freshness" of the data and limit the total time for the UAV to complete a single mission trajectory within the set threshold.The dynamic deployment problem of UAV data collection in the above scenario is constructed as a multi-target problem for UAV data collection trajectory optimization(MOPDCTO),which reduces the calculated number of UAV deployments and UAV flight time by optimizing the UAV’s hover target coordinates and access order when collecting data while maximizing the average throughput of UAV.To solve MOPDCTO,an improved non-dominated sorting genetic algorithm-Ⅲ(IMC-NSGA-Ⅲ)is proposed.The algorithm introduces a weight graph initialization strategy for generating the initial flight trajectory and generates the initial task flight trajectory through the calculation of the weight map,which further accelerates the convergence speed of the algorithm.In addition,the introduced path-finishing compression strategy solves the problem of unreasonable task allocation caused by a "fragmented" trajectory,while the proposed crossover operator in the group solves the problem of dimensional mismatch of the original cross-operation,and the performance of the original algorithm is improved through these improved mechanisms.The effectiveness of the proposed improved algorithm is verified in different Io T scale simulation experimental environments,and the results show that the IMC-NSGA-Ⅲ.The algorithm is superior to other comparison algorithms in solving the MOPDCTO problem.