| With the expansion of functions and extension of tasks,future flying vehicles will certainly require their ability to fly in formation and work together to achieve clustering and intelligence of flying vehicles.With the wide application of mobile ad hoc networks,Flying Ad hoc Network(FANET)has received more and more attention from the industry.Routing technology is a key technology in the network to improve network performance and ensure normal network communication,and has a decisive impact on routing quality,which has a direct impact on network performance.For FANET,the highly dynamic changing topology poses a challenge to the design of routing algorithms.In FANET,the fast movement of UAV nodes and the low density of nodes in the network mean that problems such as frequent routing interruptions cannot be solved in the network by increasing transmission power and providing long-distance communication.On the one hand,the feasible paths between UAV nodes in the network contain all loop-free network paths,and it is impossible to search the feasible solution space in an exhaustive way when solving;on the other hand,most of the current research on routing protocols for FANET is still based on the evolution and improvement of network routing of MANET and VANET,and for FANET,the design of special routing protocols becomes It is particularly important to design special routing protocols for FANET.Therefore,studying the routing mechanism and researching the design of comprehensive,stable,efficient,and high-quality network routing is essential to improve the quality of service(e.g.,low latency,high transmission rate,etc.)in FANET.Meanwhile,biotic swarming behavior is a highly coordinated swarming movement.For example,individuals in flocking biological groups such as birds and beasts in nature use simple rules and local interactions to cooperate with each other to efficiently accomplish complex tasks and have intelligent characteristics such as distributed,simple and self-organized,which fits well with the needs of autonomous UAV cluster flight systems.Therefore,by drawing on the characteristics of biological population intelligence in nature,this thesis uses intelligent algorithms such as ant colony optimization algorithms,improved genetic algorithms,continuous Hopfield neural networks and Boltzmann machines,and improves the basic algorithms,and applies them in the study of flying ad hoc network routing to form a relatively mature population intelligence routing technology,which effectively solves the routing problem of highly dynamic FANET,thus It can provide relevant theoretical and methodological support for the research and application of FANET routing protocols.A series of research works have been carried out in this thesis around the need to establish efficient routing problems in FANET as follows.1.To improve routing availability,a FANET low latency routing protocol(MOLSR)with optimal link evaluation is proposedTo overcome the additional routing overhead caused by link interruptions due to node movement,an improved OLSR optimized routing protocol for highly dynamic scenarios based on the OLSR routing protocol is proposed based on a combination of factors such as neighbour node link stability evaluation,node power,node bandwidth,transmission distance,etc.,based on the characteristics of FANET,proposing the algorithm’s link stability evaluation mechanism,routing The improved method such as link stability evaluation mechanism,routing entry establishment and energy function design of the algorithm are proposed.Simulation results show that the improved OLSR routing protocol can perform better than the traditional OLSR routing protocol in key performance indicators such as average end-to-end delay,packet delivery rate and routing control overhead.2.To achieve comprehensive routing discovery,a FANET routing mechanism based on ant colony optimization(ACA)is proposedCombined with the characteristics of fast and dynamic changes in routing links in FANET,an improved ant colony algorithm is used to solve the routing problem.Improvements are made in various aspects such as strategy selection,pheromone update,parameter selection and route discovery strategy of the ant colony to reduce the power consumption of the algorithm,which can be better applied to the highly dynamic network of FANET.Simulation experiments are carried out using NS3 software for network simulation and the results show that compared with several existing algorithms such as AODV and DSR,the search capability in ACA is stronger,the energy consumption of nodes is more balanced,and network lifetime is also relatively longer.3.To enhance the stability of routing,a genetic algorithm based FANET routing protocol(GAR)is designedThe strategy fully considers the stability of FANET routing,the bandwidth of the links and the energy of the nodes,improves the coding method,operator operation and control parameters of the basic genetic algorithm to form a GAR,and performs route search.Simulation results show that GAR can effectively improve the stability of routing compared to several other algorithms in the context of real-world scenarios and specific network performance requirements,helping to improve network performance and thus the lifetime of the network.4.To improve the overall efficiency of routing,a FANET routing method based on continuous Hopfield neural network algorithm(CHNNR)is proposedThe continuous neural network routing algorithm(CHNNR)is designed by improving route discovery and maintenance through intelligent adjustments such as energy function and matrix conversion,taking into account the characteristics of highly dynamic and high speed changes under flying ad hoc network topology.Experimental results show that the method can reduce network latency,improve normalized network throughput and enhance data transmission success rate,and the comprehensive performance of CHNNR is better than other routing CHNNR outperforms other routing algorithms and is more suitable for FANET networks with frequent topology changes.5.To improve the quality of the routing mechanism,a Boltzmann machine based FANET routing algorithm(BMR)was designedCombining the characteristics of FANET,the Boltzmann network structure for routing is constructed,and a routing optimization method is proposed to design the routing algorithm for FANET in Boltzmann machines(BMR).The routing algorithms such as BMR,AODV and DSR are compared and simulated in the experiments,and the results prove that Boltzmann machines outperform other algorithms in terms of average end-to-end delay,average routing lifetime and control overhead,improving the quality of network routing in FANET with better performance.6.Design a flight verification platform based on group intelligence for flying ad hoc network routing to verify the feasibility and effectiveness of the improved algorithm in this thesisAccording to the demand of UAV autonomous coordination and the function mapping of group intelligence on UAV,the intelligent cluster is formed based on the group behavior,and the flight verification platform is built from two aspects of hardware design and software system,and several types of algorithms such as optimized ant colony and traditional network routing algorithms improved in this thesis are transplanted to three quad-rotor UAV systems by designing two different types of experimental scenarios and processes of flight tests.Finally,we compare and analyze the network performance indexes such as packet delivery success rate of various algorithms,and the results show that the improved algorithms in this thesis are better and more effective.The research work in this thesis proposes a series of solutions for different scenarios,using bionic population intelligence techniques to solve key problems in the field of flying ad hoc network routing,and verifies the feasibility,effectiveness and practicality of the relevant techniques.These works can provide credible solutions for information transmission interaction in FANET and promote the development of UAV bionic cluster autonomy theory and technology applications. |
PDF Full Text Request