Research On Joint Neighbor Discovery And Positioning For Unmanned Aerial Vehicle Networks

Unmanned Aerial Vehicles(UAVs)have developed rapidly in the past few years and have become one of the important tools in military,commercial,consumer and other fields.By carrying various sensors,UAVs can complete various tasks in the air.Due to its characteristics of rapid deployment,lightness and flexibility,UAVs are widely used in agricultural plant protection,emergency communications,environmental monitoring,power inspection and other fields,playing huge economic benefits and technical advantages.At the same time,more and more applications have also spawned the development of UAV network technology.As UAVs and UAV networks are used in richer scenarios,many problems arise.This thesis focuses on the localization and neighbor discovery problems of UAV networks deployed in environments with poor satellite signals,such as cities and large indoor spaces.The challenges that such environments bring to localization and neighbor discovery problems are analyzed,and network localization methods that can be used to solve these problems are introduced.First of all,this thesis proposes a joint localization and neighbor discovery method based on Two-Way Ranging and Time Difference of Arrival(TWR/TDoA),which is used for extremely harsh environments with only one anchor node.Discovery and location speed.Then,this thesis theoretically analyze the localization accuracy of the proposed method.Finally,this thesis conduct simulation analysis on the localization accuracy and neighbor discovery convergence performance of our method.The results show that the method is convergent in neighbor discovery,and its localization error can approach the theoretical lower bound when the running time is long enough.Numerically,the simulation results show that it achieves meter-level positioning in the presence of poor Global Navigation Satellite System(GNSS)signals.Then,this thesis considers a relatively loose environment,that is,there are multiple nodes with known GNSS information in the network,and designs a joint positioning and neighbor discovery method based on the Gossip algorithm,which can take into account the convergence speed,Positioning accuracy and robustness.At the same time,this thesis also propose an improved extended Kalman filter algorithm for the positioning problem.This method is mainly aimed at the UAV network with incomplete network location information.It is used for fast networking,positioning and tracking of UAVs in motion.High positioning accuracy and high robustness are achieved.The method can complete positioning and neighbor discovery in one process.Compared with the traditional positioning method,the method utilizes neighbor information and improves the positioning speed.In addition,the method also retains the strong robustness of the Gossip algorithm,and uses the redundant information of the Gossip algorithm to improve the positioning accuracy.Finally,this thesis perform theoretical and simulation verifications on the performance of the proposed method.Our proposed joint localization and neighbor discovery method can provide support for location-based applications in UAV networks and provide a foundation for further research on UAV networking.