In the past few years,unmanned aerial vehicles(UAVs)have continued to develop with the characteristics of high flexibility,low cost and strong concealment,and their application fields have gradually extended from military to civilian.With the continuous enrichment of UAV application scenarios,the single UAV has been unable to meet the increasingly complex requirements.Therefore,it is increasingly urgent to build a collaborative and interactive network among multiple UAVs.As the first step of networking,neighbor discovery directly affects the performance of UAV networks.Therefore,the main research goal of this thesis is to realize the fast topology construction of UAV networks.The research in this thesis solves the problems of slow neighbor discovery and low networking efficiency in large-scale UAV networks.The specific contents are as follows.Firstly,this thesis proposes a fast neighbor discovery algorithm based on successive interference cancellation and multi-packet reception by referring to 5G NOMA to distinguish users from power difference to solve the packet collision problem that the existing neighbor discovery algorithms have not solved.The main idea is to use the large power difference between the received signals to unpack the collision packets.When unpacking,the low-power signals are regarded as interference.After the highest-power signal successfully unpacked,it is eliminated from the received signals to unpack the lower-power signals,thus reducing the neighbor discovery time.The simulation shows that the application of successive interference cancellation technology reduces the neighbor discovery time by 27.92%and 26.88%on average compared to the classical scan-based algorithms(SBA)and completely random algorithms(CRA).In addition,on this basis,this thesis also applies the multi-packet reception technology to reduce the probability of packet collision and further accelerate the neighbor discovery speed.Compared with the SBA and CRA algorithms,the average neighbor discovery time is reduced by 69.02%and 66.03%.Then,this thesis focuses on the existence of multiple nodes as neighbors to each other in large-scale UAV networks,and proposes a neighbor discovery algorithm joint relative positioning.The main idea is to measure the distance between nodes based on the handshake signals during neighbor discovery using two-way ranging technology.The information of neighbors is spread through Gossip mechanism to speed up the discovery of potential neighbors.Moreover,the scheme achieves the relative positioning of the network by transmitting ranging information between nodes while discovering neighbors and obtains the approximate distribution of nodes in the network and the geometric shape of the network topology.The simulation results show that when the number of modulation modes is 6,the proposed algorithm further reduces the neighbor discovery time by about 19.14%,and not only completes the neighbor discovery but also finishes the relative positioning of the network at the same time. |