Deployment And Optimization Of UAV-assisted Stereo Wireless Sensor Network For Marine Ranch Monitoring

A new round of global science and technology is accelerating the development of marine fishery,and setting off a new wave of blue economic development.Since,the construction of marine ranching in China has entered a period of rapid development.In the development process of marine ranch,the whole process monitoring and evaluation,early warning and prediction after the establishment of the marine ranch are ignored,which will lead to the fact that the ecological environment quality of the marine ranch is still invisible and unknown,and the economic species resources are not statistically analyzed.The UAV-assisted Underwater Wireless Sensor Network is studied to realize the ecological monitoring and information transmission of marine ranch.The three-dimensional network of underwater,water,air and land ensures the real-time and effective monitoring of the environment,and provides a theoretical basis and support for the monitoring practice of marine ranch.In this thesis,the TOD problem of UWSN network under UAV-UWSN architecture is deeply studied.The sensor is responsible for the data acquisition of the ecological environment of the marine ranch,and the unmanned aerial vehicle is linked with the underwater sensor network,which is responsible for transmitting data to shore-based monitoring center.Firstly,we conduct the research on the network Topology Optimization and Dimensioning(TOD)under UAV-UWSN architecture.In order to minimize the total energy consumption of network,a mathematical formulation is provided for the TOD problem considering the constraints of network connectivity,communication distance and link hops.The Gurobi solver is used to verify the correctness and scalability of the mathematical formulation in small and medium-sized network scenarios.Aiming at solving the problem of the exponential growth of Gurobi computational complexity especially in large scenes,a heuristic algorithm based on artificial bee colony(ABC)is designed to solve the TOD problem.Through the simulation of different scene scales,the superiority of ABC algorithm in solving hierarchical deployment problem is verified.Because ABC algorithm is easy to fall into local optimum,we propose an improved bee colony algorithm(IABC)to increase the accuracy and speed of the algorithm and reduce network energy consumption.Secondly,we study the efficient data collection under UAV-UWSN framework,focusing on the problem of Global Path Planning and Local Hovering(GPPLH)of UAVs.In the study of UAV local flight planning,a cylindrical geometry is established to optimize flight path,hovering position and hovering time in continuous geometric space.In the study of global problems.In order to improve the working efficiency of UAV and reduce energy consumption,a integrated Brain Storm Cuckoo Search optimization(BS-CS)algorithm is proposed to optimize path and hover position set of UAV.Compared with the original Brain Storm Optimizatio algorithm,the time for a round of information collection is optimized.A large number of simulation results show the BS-CS algorithm is superior in task completion time,path and energy consumption in solving GPPLH problem.