Research On 3D FANET Routing Algorithm Based On Geographic Location Information

As one of the important branches of MANET,the Flying Ad hoc Network(FANET)composed of multiple Unmanned Aerial Vehicles(UAV)has greater flexibility and mobility,and is particularly suitable for complex agricultural remote sensing,traffic monitoring,emergency rescue and military mission scenario.At the same time,the three dimensional characteristics and dynamic changes of FANET network topology also pose higher challenges to the design of routing algorithms.Existing GPSR routing algorithms only use two dimensional geographic location information and fail to fully consider the impact of outdated location information.Therefore,it is easy to cause routing decision errors,thereby increasing end to end delay and reducing packet delivery rate.In view of the above problems,this thesis designs an improved 3D GPSR routing algorithm suitable for FANET networks by in-depth study of the UAV physical layer channel and its motion model.The main contents are as follows:The existing GPSR algorithms in FANET use the ideal omnidirectional antenna model,which does not take into account the differences in antenna gain at different elevation angles,which does not meet the actual three-dimensional characteristics of FANET.Aiming at this problem,the actual threedimensional omnidirectional antenna model of the UAV node is first established to solve the packet retransmission probability.Then a link metric based on the number of retransmissions and the energy consumption model is designed for the greedy forwarding mode,which provides a basis for selecting a highly reliable next hop node.Aiming at the perimeter forwarding mode,a local flood forwarding method is used instead of the plane right-hand rule to bypass routing holes,further improving the reliability of the routing algorithm.Simulation results show that the three-dimensional omnidirectional GPSR(OGPSR)routing algorithm proposed in this thesis can improve the packet delivery rate,reduce end-to-end delay and energy consumption.In order to improve the applicability of the above OGPSR routing algorithm to highly dynamic scenes,a three-dimensional OGPSR routing algorithm based on location prediction is proposed.Considering the randomness and predictability of the actual UAV node,a Gaussian Motion model is first established for the UAV node.Aiming at the problem that the traditional kalman filter prediction results of random Gaussian Motion model node acceleration do not converge at a certain observation frequency,the system parameter matrix of the Gaussian Motion model is used to improve the processing noise matrix of the kalman filter to adaptively predict the three-dimensional position of the node at the next moment.On this basis,the process of the OGPSR routing algorithm based on adaptive kalman prediction is given by analyzing the optimal prediction moment.Simulation results show that the improved OGPSR routing algorithm improves the performance of end to end delay and packet delivery rate in high dynamic scenarios.