On UAV Scheduling Approach For Maritime Spectrum Monitoring

With the rapid development of information technology,the radio service has been greatly expanded,resulting in low utilization and serious interference of scarce spectrum resources.The efficient completion of spectrum monitoring plays a key role in solving the above problems.However,due to the limitations of traditional means,maritime monitoring faces the disadvantages of high cost,low real-time and low precision.If UAVs are used to achieve spectrum monitoring with flexibility and high precision,they still face three main problems: data sparsity,sampling point selection,and rapid convergence of tasks.In order to solve the above problems and realize the spectrum monitoring in the complex electromagnetic environment over the sea,a framework for using the ship to place the drone into the target sea area for monitoring is proposed,and the detailed design and execution flow are given;according to the spatial distribution characteristics of the electromagnetic spectrum,the sampling point selection problem in the task initialization stage is studied to maximize the reconstruction benefit.The greedy strategy local optimal idea is used to design the UAV scheduling algorithm in the task execution process to achieve rapid task convergence.The simulation experiment method is adopted to verify the validity of the proposed algorithm and mechanism.Main works are listed as follows.Firstly,the UAVSense framework for maritime spectrum monitoring is designed.The framework design and execution flow are described in detail,and the two main functional modules of spectrum sensing and spectrum database are introduced.It is difficult for a small number of UAVs to complete large-scale spectrum monitoring tasks.For the sampling data sparsity problem caused by the insufficient number of UAVs,the spectrum interpolation algorithm for data reconstruction is studied in the framework.The interpolation experiment simulation experiment is set to verify the effectiveness of the UAVSense framework and find the OK algorithm with the highest interpolation accuracy.Secondly,aiming at the sampling point selection problem,a first-round task allocation mechanism CAMB for balanced coverage is proposed.Since the spectral data is a regionalized variable with spatial correlation,the distribution of sampling points will directly affect the interpolation precision.According to this feature,the granularity of the regional division is defined.The data coverage entropy is used to measure the reasonable degree of sampling point selection,so that the sampling point selection problem is solved in the absence of prior knowledge.The simulation comparison with the random selection method shows that the performance of the CAMB mechanism is more stable and more accurate.Finally,to solve the problem of rapid convergence,a multi-slot UAV scheduling algorithm SAMD for minimizing the difference is designed.A greedy strategy with minimal difference is proposed to quickly eliminate local interpolation errors.The iterative processing makes the overall interpolation accuracy reach the convergence criterion quickly.Compared with the classic UAV scheduling algorithm,the SAMD algorithm shows real-time performance and higher precision.It also shows that the cost of using the UAVSense system framework to complete the maritime spectrum monitoring task is lower.