Research On D2D Mobile Intelligent Network Based On Particle Swarm Optimization Algorithm

As communications scenarios and needs become more and more diversified,communication networks are often required to be mobile as well,while traditional mobile communication network nodes are fixed after planned.Based on this,this paper propose a system using UAV cluster in the air intelligent networking providing communication services for users on the ground.The UAV(unmanned aerial vehicle)equipped with communications equipment as a network node,which communicate with each other in the form of D2D(device to device)communication,and move along with the user's mobile intelligently to form a new optimal mobile network to serve users.The work of this paper is mainly about the application design of the behavior decision algorithm of UAV,the building of UAV experimental platform,and the algorithm simulation,indoor simulation and partial outfield experiments.The main contents of this paper are as follows:Firstly,we combine the research system and multi-agent cooperative evolution to model the application scenario,and design a multi-agent mobile intelligent networking system based on PSO(particle swarm optimization)algorithm.The system is divided into user's environment model,user tracking model and user space environment adjustment model,network optimization scheme,multi-agent position adjustment scheme and particle swarm optimization algorithm.For particle swarm algorithm,we design the fitness function,particle structure and population structure and particle update strategy.Secondly,based on the design above,we build the system simulation and experiment platform,the platform consists of hardware and software.The hardware part is divided into the ground hardware and air hardware.Software is divided into user mobility simulation software,GPS distribution software,center control software,UAV control software,and particle swarm algorithm software.The algorithm simulation and result analysis are carried out on this platform,and the algorithm simulation includes two cases:one user with one UAV and multiple users with multiple UAVs.Based on the algorithm simulation and indoor simulation results of one UAV with one user,we analyze the trajectory of the user and the UAV and find UAV can track user.Then the outfield experiment is carried out,analyzing the trajectories of UAV and user,studying the effect of UAV inertia on the result,and then add the inertia processing in UAV flight.The multi-users with multi-UAVs simulation including two cases,one is the number of user and UAV is same,the other is the number of users are more than the number of UAVs.According to the simulation distance indicator between user and UAV in this scenario,the coverage of the network and algorithm performance are analyzed.