Research On Resource Scheduling And Intelligent Optimization Methods For Internet Of Vehicles

In recent years,with the continuous evolution and development of wireless network and artificial intelligence technology,the application of the Internet of Vehicles(Io V)industry has been promoted and implemented,which is considered to be beneficial to the realization of the vision of intelligent transportation and smart city.At present,Io V has become a research field with important application value in academia and business.In Io V system,according to the established communication protocol and data interaction standard,the dynamic information exchange and sharing among vehicles,road side units,pedestrians and communication network can be realized in an intelligent way,which provides theoretical support for the massive data analysis and application service delivery of Io V.However,the computing capacity of mobile devices in the current Io V is relatively low,and there is a risk that the information will be leaked or tampered when the billing data generated in the process of functional business transaction is transmitted to mobile users.In addition,the extensive development of Io V has led to an exponential increase in the number of mobile devices.The large scale of the device group makes the available resources in the system unable to be reasonably allocated,resulting in excessive consumption of computing and cache resources,which is also the major concern of the current Io V system.In order to cope with the above problems,we first discuss and analyze the architecture of Io V.Combined with the high integration of the distributed structure of mobile edge computing(MEC)and blockchain,we propose a system architecture of charging billing data calculation offloading and resource intelligent scheduling for new energy vehicles in Io V,and the reinforcement learning method is used to solve the optimal calculation offloading and resource allocation strategy.Then,combined with the fusion computing theory of cloud computing and MEC,we design an Io V system framework for collaborative intelligent scheduling of computing and caching resources,and introduce a deep reinforcement learning(DRL)method to iteratively solve the optimal intelligent resource allocation policy of collaborative computing and caching.The specific research works are listed as follows.(1)Research on data computing offloading and resource intelligent scheduling strategy of Io V based on reinforcement learningAiming at the problems of high energy consumption and unreliable billing data transmission during the charging process of new energy vehicles in Io V system,integrating the application principles and technical characteristics of MEC and blockchain,we propose an architecture of data computing offloading and resource intelligent scheduling system for Io V.The reinforcement learning method is used to iteratively solve the decision optimization problem of joint data offloading and resource scheduling,which is used to efficiently realize the optimal computing offloading and resource allocation strategy,so as to effectively reduce the latency and energy consumption in the process of billing data uplink and transmission,and improve the transaction throughput of the blockchain.(2)Research on collaborative intelligent scheduling strategy of computing and cache resources for Io V based on DRLOn the basis of the data computing offloading and resource intelligent scheduling system architecture for Io V,we further consider the energy efficiency of massive data processing and the efficiency of edge application content delivery.According to the application characteristics of cloud computing and MEC,we propose an Io V system framework for collaborative intelligent scheduling of computing and caching resources.Based on the large-scale and dynamic characteristics of the framework system,DRL method is introduced to deal with the optimization problem of joint decision-making of computing offloading,content caching and resource allocation,and iteratively solves the optimal cooperative intelligent scheduling strategy for computing and caching resources to achieve the optimization objectives of minimizing energy consumption,computation overhead and maximizing blockchain transaction throughput.