Research On The Optimal Selection Method Of Central Node Of UAV Cluster

Unmanned aerial vehicle（UAV）has made significant breakthroughs in target search,agricultural plant protection,disaster rescue and other fields relaying on friendly usage,quick response,and flexible.Compared with a single UAV,UAV cluster is effective and fast to accomplish tasks by using collaborative network,and the central node has the capability for data aggregation and decision making.The selection of central node directly affects the direction and data size of the whole network flow.The network is complex during operations,which is multi-hop and self-organized.It is greatly beneficial to choose the central node of UAV by using selection algorithm of complex network.The network of UAV cluster demands stringent control and coordination since the location of each node is dynamic and node join and exit at any time.Therefore,this work focuses on the optimal selection technique of UAV.The main researches of this work are as follows:1)For single evaluation index and high computational complexity of selection algorithm in complex network problems,this work designs a central node selection algorithm which integrates the importance of nodes.Specifically,first,the complex network is mapped as a graph and expresses its characteristics.Second,it is introducing laplacian centrality for node self-evaluation,and designing weakness factors and combining efficiency values to evaluate the influence of neighbor nodes.Finally,it builds an importance matrix which is used for evaluating the node importance and choosing central node.The comparative analysis of experimental results and simulation show that the proposed algorithm has the characters of low latency,high node importance distinctiveness,and accurate central node selection results.Besides,the selected central node has better propagation and robustness in network.2)For the optimal central mode and dynamically changes in UAV problems,this work designs an adaptive UAV cluster center selection method.Specifically,first,the method takes mobility similarity,quality of wireless communication,data processing capability,and residual energy as the inputs and making normalization.The network structure of UAV is built by considering the weights of links.Second,the central nodes of the UAV cluster are selected by incorporating the importance of nodes.Finally,the method proposes an adaptive time window reselection mechanism by taking reselection mechanism and node join and exit mechanism.The experimental results show that the proposed method is more suitable for dynamic central node selection and owns accurate result.The experimental results show that the designed central node selection algorithm and adaptive time window reselection mechanism achieve better performance for the selection of the optimal central node of UAV cluster and the replacement of central node,which verifies the effectiveness of the proposed method.