| UAV(Unmanned Aerial Vehicle,UAV)routing technology is one of the difficulties in the field of UAV technology research.However,most of the classic self-organizing network routing protocols such as AODV(Ad-hoc On-demand Distance Vector,AODV)routing protocol and DSR(Dynamic Source Routing,DSR)routing protocol cannot be applied to routing in UAV networks.Some researches have been found that the GPSR(Greedy Perimeter Stateless Routing,GPSR)protocol is a more suitable method of UAV routing technology.However,GPSR technology has the defects of routing holes and inability to balance load.When there is a routing hole in the GPSR routing,the flood routing mechanism can be used to find a better relay node,but it will also increase the network delay of data packet transmission,and the traffic load in the network may be too high,which may cause the congestion of the network.In view of the above technical difficulties,this dissertation takes the routing protocol in the UAV self-organizing network as the research goal.Firstly,the GPSR routing protocol for the UAV selforganizing network is proposed.The routing protocol is based on the routing mechanism of the original GPSR routing protocol and can avoid the flooding pathfinding mechanism when routing holes appear.Then,an optimized routing protocol in the UAV self-organizing network is designed,based on the the GPSR routing protocol for the UAV self-organizing network,so as to achieve a higher data packet arrival rate and lower data packet transmission in the UAV self-organizing network routing scenario.The specific research work includes the following three aspects:(1)The GPSR routing protocol based on chaos simulated annealing algorithmThe GPSR routing protocol based on chaotic simulated annealing algorithm is designed,based on the proposed GPSR routing protocol for UAV self-organizing network.The protocol comprehensively considers various factors in path finding and calculates the optimal greedy selection function during path finding.It can effectively avoid routing holes and reduce the packet transmission delay of the network.The simulation experiment results show that the GPSR routing protocol based on chaos simulated annealing algorithm can effectively avoid the occurrence of routing holes,realize one-time routing and effectively reduce packet transmission delay of the network.(2)The routing protocol with load balancing performance based on a deep reinforcement learning network.The GPSR routing protocol based on chaos simulated annealing algorithm proposed does not have the ability to balance traffic,that is,the GPSR routing protocol based on chaos simulated annealing algorithm can only find a fixed routing path.When multiple routing conflicts appear,the routing strategy cannot be changed adaptively according to the load change of the network,which will cause the network to have a large degree of congestion and a large amount of queuing congestion and a large amount of packet loss.Aiming at this problem,the routing protocol with load balancing performance based on a deep reinforcement learning network is designed.Experimental results show that the routing protocol can effectively balance traffic under heavy traffic loads whicn will increase the packet arrival rate and reduce packet transmission delay of the network(3)The traffic adaptive routing protocolThe routing protocol with load balancing performance based on a deep reinforcement learning network still could be improved in packet arrival rate,and this routing protocol can only show excellent performance when the network congestion is large,while the GPSR routing protocol based on chaos simulated annealing algorithm has better performance of optimal path finding when the network congestion is small.Therefore,a traffic adaptive routing protocol is designed,which combines the advantages of the GPSR routing protocol based on chaos simulated annealing algorithm and the routing protocol with load balancing performance based on a deep reinforcement learning network.The greedy function changes with the traffic loads,and the chaos simulated annealing algorithm is used to search for the optimal parameters of this routing protocol,which greatly improves the network packet arrival rate and reduces the packet transmission delay of the network. |
