Research On Geographic Location Service Of UAV Cluster

With the development of unmanned aerial vehicle(UAV)technology,UAVs can accomplish increasingly complex tasks.However,a single UAV has disadvantages like the lacking of multiple functions and weak anti-strike capability.Compared with individual UAV,the UAV clusters have higher scalability and better survivability,and can achieve more complex tasks.UAV clusters have become a key area of current research.However,the fast movement speed and frequent topological changes of UAV clusters lead to UAV clusters’ communication networking problem,which has been a bottleneck of UAV clusters’ development.A variety of mobile ad hoc network(MANET)routing protocols has been proposed by researchers,among which,geolocation-based routing protocols are well adapted to the high mobility and frequent topology changes of UAV clusters and have good scalability.In these protocols,nodes need to obtain the location of target nodes to route and forward packets.Therefore the location services are the basis of location-based routing protocols.Currently,there are several problems of existing location service methods that need to be solved in the field of UAV clusters,such as poor scalability,high control overhead,and low query success rate.These problems directly lead to the poor performance of communication networking in UAV clusters.In this thesis,we propose a backbone ringbased location service framework to address poor scalability and the inability to adapt to the frequent topological changes of UAV clusters.The main research efforts and innovations are as follows.(1)A new location service framework is proposed for a network scenario in which the overall location of a UAV cluster moves and the nodes move with a certain regularity.The framework first groups the clusters based on the similarity of node movements,and chooses a group leader for each group as the location server,and then organizes the location servers into a backbone ring to provide location services for the network.The framework can adapt to the frequent topological changes of UAV clusters with a high rate of detection.(2)A cluster grouping strategy based on the minimum distribution set algorithm is proposed,which takes into account the movement model of UAV clusters in real application scenarios and groups UAV clusters in combination with the correlation between the movement of UAVs.The proposed grouping method is able to adapt to the changes of node topology in the cluster,reduce the node switching among groups,and finally reduce group maintenance overhead.(3)A backbone ring-based location service algorithm is also proposed.The algorithm firstly organizes all the Group Leader(GL)nodes in the cluster into a backbone ring by breadth-first search,and then selects a location server node on the ring for each node using a consistency hash algorithm to provide location service for it.In this algorithm,the location server node does not directly store the specific location of nodes,which avoids the location lookup error caused by the delay of location update and increases the accuracy of location lookup.And because the location server nodes move together with the grouping,the switching and loss of server nodes caused by the movement of nodes can be avoided,which improves the scalability and reliability of the location service algorithm.In this thesis,the node model and location service algorithm of the UAV are implemented in the simulation software,and several experimental scenarios are designed to verify the feasibility of the proposed method and analyze its performance.The results show that the proposed method improves the accuracy of the location service by 30%,and the location update delay of GL nodes is only 2 seconds.Besides,it can adapt to the changes of cluster topology with good scalability.