Research On Intelligent Vehicle Network Scheduling Strategy Based On Spray Calculation

With the continuous improvement of the economic level,people’s needs for a better life are increasing day by day,and the market for food,clothing,housing and transportation demand has a broad space for development.While the automobile industry is booming,which provides great convenience for people’s travel,it also brings many problems to the society,such as frequent traffic congestion and frequent traffic accidents.People hope to enjoy a better driving experience,have a safer driving environment,and look forward to the early arrival of the era of intelligent transportation.In the "Internet of Everything" information age,autonomous vehicles are envisioned as a promising technology to ensure road safety and traffic efficiency.In order to ensure that the vehicle can drive safely and maintain a high level of intelligent driving,the vehicle is loaded with a large number of sensors to collect information about the surrounding environment,which makes driving a car a lot of computing tasks to process.Vehicle fog computing（VFC）is envisioned as a promising solution for handling explosive tasks in autonomous vehicle networks.Therefore,solving the network scheduling problem of requesting vehicles in a vehicle fog computing environment is of great significance for efficiently handling vehicle unloading tasks.This paper studies the scheduling problem of autonomous vehicles requesting offloading tasks to access the network in a fog computing environment.Due to the limited computing resources of the fog computing network system,and the randomness of the arrival and departure of vehicles in the network,it is expected to comprehensively evaluate the performance of the fog computing network for decision-making reference.The scheduling strategy of entering the VFC network is a difficult challenge.Therefore,in this paper,the task offloading of the requested vehicle is treated as a very short period of time,and it is discretized and divided into multiple time slots with the same interval,and the problem is expressed as a Markov decision with finite state and finite discrete time.process（MDP）,and introduce a multi-attribute decision-making method,the distance method between superior and inferior solutions（TOPSIS）to construct the reward function of the model,the purpose is to find an optimal scheduling strategy to make the expected return of the requesting vehicle in a task unloading cycle to reach maximum.According to the above model and the actual situation,this paper sets the parameters for numerical analysis,and solves the optimal scheduling strategy of the request vehicle on the VFC network and the expected return of the system through the value iteration algorithm based on the Bellman equation.The influence of resource service rate on expected return is calculated,and the reason for the difference between the MDP scheme combined with the superior and inferior solution distance method and the general MDP scheme is explained,and the optimal strategy obtained in this paper is compared with the random strategy and the short-sighted strategy to demonstrate its effectiveness.This paper provides a reference for dynamically changing network scheduling problems,and provides ideas for practical problems that expect to make decisions based on multiple factors.