Research On Key Technologies Of Distributed Flocking Control In Flying Ad Hoc Networks

With the wide application of the unmanned aerial vehicle(UAV)in various fields in recent years,single UAV has gradually become inapplicable to meet the needs of increasingly complex scenarios.At the same time,problems such as low autonomy and high dependence on control station in centralized UAV swarm networking,have gradually become prominent.Therefore,it is necessary to investigate the distributed swarm networking technology,improve the coordination ability of UAV swarms,and establish a flexible and efficient Flying Ad hoc Network(FANET),to solve the above problems.Nevertheless,varoious problems remain to be settled in distributed UAV swarm networking,such as topology establishment,network optimization and connectivity analysis.This thesis takes the distributed swarming in the FANET as the background,uses mathematical tools such as algebraic graph theory and random geometry,analyzes the theoretical conditions of network topology construction and connected network,and designs a series of distributed UAV swarming framework.The main innovations and contributions of this thesis include the following four aspects:1)Focusing on the problems of the asymmetry of the generated network graph and the uncontrollable upper bound of the node degree in the traditional algorithm,a distributed topology generation mechanism is designed to realize the "selfconfiguration" of the FANET.Based on the traditional Acute Angle Test(AAT)algorithm,an improved topology establishment algorithm is proposed.Compared with the original one,the proposed algorithm ensures the symmetry of the network graph while making the upper bound of the node degree controllable at the same time.In addition,the theoretical derivation and corresponding proof of the upper bound of the corresponding vertex degree in the three dimensional space are given.2)Focusing on the problem of low-connectivity network topology during the convergence of the UAV swarm,a distributed connectivity optimization algorithm considering critical edges is designed to realize the "self-optimization" of the FANET.For the problem that the iterative vector eventually converges to the vector of all ones in traditional power iteration-based distributed Fiedler vector estimation algorithm,a periodic correction step is intrduced to obtain the correct value.In addition,the need for real-time acquisition of the global information of Fiedler vector is avoided,thus the algorithm can be implemented in a distributed manner.The analysis results show that the designed algorithm can increase the network connectivity probability by 55%and reduce the proportion of low-degree nodes in the network by 14.5%.3)Focusing on the scenario of the failure of the node in the swarm,the impact of the failed node on the connectivity is analyzed.After that,a distributed connectivity maintenance algorithm based on the importance of the node is designed to realize the "self-healing" of the FANET.Compared with the existing Distributed Spectrum based Estimation(DSE)algorithm,the proposed algorithm further analyzes the two steps caused by the failed node,including the removal of associated edges and the reconstruction of new associated edges between its neighboring nodes,thus making the theoretical results closer to the actual topology changing in a FANET.The analysis results show that the designed connectivity maintenance mechanism increases the minimum algebraic connectivity caused by node failure by 4.4 times.4)Considering the conditions of the UAV swarm to establish an initial connected network topology,the connected probability of the UAV in the 3Dimensional(3D)space is analyzed,and the corresponding closed-form expression is derived,to provide a theoretical basis for establishing an initially connected FANET.Compared with the existing research,the interference problems caused by other member UAVs in the swarm in the 3D space is further analyzed.In addition,the connected probability of the special case where the receiver UAV is located in the center of the swarm,and the general case where the receiver is located arbitrarily in the swarm,are given.