Research On Location Service Of Large Scale UAV Cluster

UAV ad hoc network(UANET)is essentially composed of a large number of single function or multi-function UAVs.UANET is more suitable for cluster cooperation and it has important applications in military and civil fields.UANET has many characteristics,and it also faces many technical problems in the process of research and development.Among them,the factor that particularly affects the combat efficiency of UAV cluster is the communication networking problem.Due to the high-speed movement of UAV nodes,the topology of UANET will change frequently,which will lead to the sharp decline of the performance of UANET.Greedy perimeter stateless routing(GPSR)protocol is a typical location-based routing protocol,which has better scalability in UANET.GPSR routing protocol must be supported by location services.Therefore,location services suitable for large-scale UANET with frequent topology changes and high node mobility have become a research hotspot.Aiming at the phenomena of high location service overhead,low accuracy of location information and high packet loss rate in large-scale UANET,this paper proposes a hierarchical location service method based on regional ring and improves the dynamic location prediction method based on Kalman filter.They both can effectively reduce the overhead of location service,improve the accuracy of location information and reduce the packet loss rate in large-scale UANET.The main research work and innovations of this paper are as followed:Firstly,a hierarchical location service method based on regional ring is proposed in this paper.At the first of all,the UAV nodes are grouped based on the grouping strategy,and one node is selected in each group as the group leader node.Then,after using the strategy of dynamic merging and splitting,the group leader nodes are divided into several reasonable regional rings by constructing the regional rings.The nodes in the regional ring are closely connected,so the location information can be updated with low cost in a short time.The regional rings can maintain communication and information transmission through the group leader nodes that are neighbors to each other in different regional rings.Finally,all nodes in the network can use the hash server nodes to obtain the location information of the destination node through the regional rings.This method solves the problem that the location service cost increases greatly due to the gradual expansion of the scale of UAV cluster,and effectively improves the phenomenon that the accuracy of location query reduces greatly.Secondly,with the expansion of the scale of UAV cluster and considering the problem of query delay and end-to-end delay,in order to reduce the location information error of the destination node,reduce the packet loss rate and maintain the computational overhead at a reasonable level,a dynamic location prediction method based on Kalman filter is improved.The method uses the state transition equation to estimate the location of UAV nodes.At the same time,the correction matrix equation is used to correct the Kalman gain and other information.The method improves the state transition matrix by adding UAV flight angle parameters,which reduces the error of position prediction and improves the prediction accuracy.Finally,the OPNET Modeler network simulation software is used to build the experimental simulation platform of this paper.The experimental results show that the hierarchical location service method based on regional ring can better adapt to large-scale UAV clusters,reduce the overhead of node location service and improve the accuracy of location information.The dynamic position prediction algorithm based on Kalman filter can effectively reduce the position error and improve the position information hit rate,so as to reduce the packet loss rate.