Research On Safe And Efficient Fog-Based Resource Management Optimization In Internet Of Vehicles

With the continuous development of society,people have more and more diversified demands for on-board services,such as automatic driving,on-board augmented reality(AR)navigation,etc.,which can greatly improve people’s driving experience.However,in order to support the implementation of on-board services and improve service quality,existing Internet of Vehicles(Io V)resource management schemes will face severe challenges.On the one hand,on-board services have a large demand for computing resources,and the existing computing resources cannot meet the computing needs of onboard services.On the other hand,the mobility of vehicles makes the wireless channel environment and the distribution of computing resources change rapidly,and the original static resource allocation and task offloading scheme cannot solve the dynamic problem.Therefore,this paper will take the expansion of computing resources,optimization of resource allocation and task offloading scheme as the entry point,and take the system delay,throughput,energy consumption and other factors as the index points to measure the comprehensive performance of the system to carry out in-depth research on the optimization of resource management of the Io V.Firstly,a resource management and task offloading model based on vehicle fog calculation and contract theory is constructed in this paper.In order to expand computing resources of Io V,the concept of vehicle fog computing is introduced in this paper,in which vehicles are regarded as moving fog nodes with certain computing capacity and computing tasks can be offloaded to vehicle nodes for processing.However,due to the lack of incentive mechanism,the enthusiasm of vehicles to contribute computing resources for task processing is not high in reality.In view of this,this paper designed a contract theory-based Io V resource transaction strategy,in which vehicles can get certain rewards when they contribute their computing resources,so as to enhance the enthusiasm of vehicles to contribute resources.Secondly,based on the above system model,this paper proposes an optimization scheme of Io V resource transaction based on deep reinforcement learning algorithm.In the vehicle fog computing environment,due to the mobility of the vehicle fog nodes and the real-time changes of the resource distribution topology of the Io V,the traditional static resource management scheme cannot solve the dynamic problems.Therefore,reinforcement learning algorithm in machine learning is adopted in this paper to solve such decision problems.Considering the large state space of the system model and the slow convergence rate of reinforcement learning model,deep reinforcement learning algorithm was adopted in this paper to replace reinforcement learning agent model with deep neural network,which has improved the training efficiency of the model.Finally,based on the designed transaction strategy of Io V resource,this paper constructs a Io V resource management model based on blockchain technology.Due to the lack of authentication institutions,there is the problem of fake transaction information produced by dishonest nodes in the process of Io Vs resource transactions,which disturbs the security and fairness of resource transactions.Therefore,this paper introduces the concept of blockchain in resource transactions,and verifies the validity of resource transactions by using the security attributes of the blockchain.Considering that the establishment of blocks is the basis for blockchain real-time update,which requires a large amount of computing resources to complete the proof of work(Po W),this paper designed a scheduling scheme of computing resources based on contract theory to expand the computing capacity of the nodes of blockchain and improve the success rate of block establishment to realize the blockchain-based resource security transaction mechanism.There are 29 figures,5 tables and 71 references in this paper.