| With the gradual maturity and application of the vehicular networks,there will be a large number of vehicles in the future to access network resources and edge computing requirements.Scientific and intelligent network resource scheduling technology will enable the vehicular networks,improve the driving experience of connected vehicles,greatly improve traffic efficiency and enhance the safety of the traffic system.However,the limited and scarce spectrum and computing resources available in current wireless networks present a new challenge to meet the needs of connected vehicles.In order to meet the increasing resource demand of the development of the vehicular networks,more effective network resource organization and scheduling technology is needed.In order to solve the scheduling problem of wireless resources in the vehicular networks,this paper studies the intelligent scheduling technology of wireless resources that can adapt to the environment of the vehicular networks.The main research contents and innovations are as follows:（1）With cognitive radio as the technical background,Finder-MCTS,a spectrum scheduling scheme based on vehicle driving state priority,is proposed for urban vehicle network.In Finder-MCTS,the scheduling priorities of vehicle users are obtained according to the driving state of vehicles,and the deep neural network of offline training is used as the environment model.The dynamic scenarios simulation of main users is introduced to correct the actual network benefits of vehicle users,and the Monte Carlo search tree algorithm is used to provide the scheduling solution of spectrum resources.This approach has two advantages.On the one hand,this scheme can improve the utilization rate of spectrum resources and the quality of users.On the other hand,this method can improve the search efficiency by adding the interference constraint of cognitive radio resource allocation into the tree search process.（2）Aiming at the problem that the existing edge computing infrastructure has high deployment and maintenance costs and is not suitable for the vehicular networks environment in remote areas,this paper introduces the concept of integrated vehicular networks with space and ground,and proposes a low-orbit communication satellite as the edge computing relay.A task scheduling method for offloading excess computing tasks to urban cloud computing centers within the coverage of low-orbit communication satellites.In the scheduling method,based on the vehicle user computing task properties and computing resource pricing system,put forward delay cost minimization of vehicle users and computing resources supply side profit maximization goal of the collaborative optimization,using the combination of dobby slot-machine adaptive learning output optimal computing tasks,and cloud computing center matching solutions.At the same time,in view of the analysis of the prior assumptions in the optimization problem,this paper adopts the combined multi-arm gambling machine to independently learn and update the parameters of the computing service quality of the ground city cloud computing center within the coverage range of the operation track of the low-orbit communication satellite,thus optimizing the quality of the scheduling.（3）The feasibility and efficiency of spectrum allocation and task scheduling method proposed in this paper are verified by simulation.Experimental results show that compared with existing algorithms,Finder-MCTS can achieve better performance in channel utilization,average link capacity and average convergence time.At the same time,in the simulation scenario of the vehicular networks between heaven and earth,compared with the traditional task scheduling method,the proposed collaborative optimization method based on adaptive parameter learning has advantages in terms of average scheduling delay and matching system benefits.In conclusion,in this paper,through a spectrum allocation method based on vehicle behavior priority scheduling and scenarios simulation and an adaptive task scheduling method based on collaborative optimization about user delay cost and network benefits,we guarantee the quality of wireless resource management of vehicular networks,greatly improve the efficiency of resource management. |
