Research On Resource Allocation And Pricing In Vehicle Edge Computing Based On Double Auction Mechanism

With the advent of the fifth generation mobile communication system(5G),the internet of vehicles has begun to flourish,and the high latency characteristics of traditional cloud computing(CM)can no longer meet the requirements of various intelligent applications in the internet of vehicles,such as autonomous driving,intelligent entertainment,intelligent transportation and so on.In order to solve the new requirements of these smart applications for data computing,storage and communication,Mobile Edge Computing(MEC)came into being.MEC sinks computing,which can effectively relieve computing pressure.Therefore,in order to solve the large amount of delay-sensitive data generated in the internet of vehicles,this paper considers making computing nodes close to users through MEC to support the operation of the internet of vehicles system,and introduces the auction mechanism to motivate vehicles and edge nodes to participate in the information interaction of the internet of vehicles system,so as to improve the efficiency of the whole system.Aiming at the resource allocation problem in the internet of vehicles,this paper proposes two auction mechanisms to achieve the goal of maximizing the logarithm of successful matching between resource providers and resource demanders.The specific auction mechanism design scheme is as follows:The first scheme proposes an auction mechanism to maximize the successful matching logarithm of the internet of vehicles system in view of the situation that edge nodes as resource providers and vehicles only serve as resource demander;the second scheme is to proposes an auction mechanism to maximize the successful matching logarithm of the internet of vehicles system in view of the situation that edge nodes serve as resource providers and vehicles can serve as both resource demander and resource provider.In the problem of resource allocation,the former uses an approximation algorithm,while the latter uses a greedy strategy algorithm and the vehicle that priority to satisfy when it can not meet the RSU distant constrained(PMVnot-MRD)algorithm;in the problem of price payment,the former uses the lowest successful bidding algorithm,while the latter uses the algorithm that corresponding to the next seller’s asking price to calculate the price paid.At the same time,this paper proves that the two mechanisms satisfy the economic properties of individual rationality,and authenticity.Through experiments,this paper evaluates the successful matching pairs of the two auction mechanisms,the effectiveness of buyers and sellers,and the overall utility of the internet of vehicles system.Compared with the traditional algorithm ICAM,the experimental results show that the auction mechanism proposed in this paper can effectively maximize the successful matching pairs of the internet of vehicles system,and can provide an effective resource allocation strategy for the internet of vehicles system.