“Communication-Computing” Integrated Resource Allocation Strategy For VFC

With the continuous development of society,future mobile network applications will also develop rapidly,such as the emergence and popularization of new applications,virtual/augmented reality(V/AR)and vehicle autonomous driving.The traditional network architecture mainly satisfies the user’s social networking and web browsing needs,which has been exhausted compared to new network applications that require high bandwidth and low latency.Therefore,in order to meet the requirements of new network applications,fog computing is used for solving the above problems.This thesis mainly studies the application of fog computing in the field of vehicular networking,which is called vehicular fog computing(VFC).In the process of VFC resource allocation,the stability and mobility of vehicle nodes will affect the data offloading process during VFC resource allocation,leading to more complex channel status and resource allocation process,and a series of urgent problems has to be solved.In VFC,the vehicle fog node can be parked on the side of the road,or it can be a moving vehicle.This thesis mainly considers the case of parking vehicles as fog nodes.According to different scenarios,this thesis considers different decisions and constraints.The problem of how to make effective resource allocation decisions in VFC has been studied in detail.The main work is as follows:1.Aiming at the problem of increased time delay due to poor vehicle stability when user vehicles are unloading tasks to fog clusters,a VFC resource allocation strategy based on stable time delay is proposed.In the problem modeling,the actual environmental factors are considered,including the location of the user’s vehicle,the rate of data uploading and downloading,and the computing resources in the system.And put forward the stability measurement method of the vehicle.Based on the above constraints,a resource allocation strategy is proposed to minimize the system’s stable delay.Since the proposed optimization problem is an NP-hard problem,the genetic algorithm is selected to solve it according to the variable type.The traditional genetic algorithm is improved to prevent the algorithm from falling into the local optimal solution too quickly.2.Aiming at the problem of vehicle mobility,a VFC resource allocation strategy based on system revenue is proposed.In the first study,we considered the stability of the vehicle and proposed a calculation method for the stability of the vehicle,but did not consider the impact of the mobility of the vehicle on the allocation of resources.Therefore,in this research,we consider the mobility of vehicles and propose a system revenue model.With the goal of maximizing system revenue,the resource allocation process is abstracted into a series of mathematical decisions,and the problem is transformed into Markov decision process and use value iteration algorithm to solve it.In summary,this thesis fully considers the characteristics of VFC technology,studies the resource allocation problem in this scenario,improves the corresponding algorithm,and effectively allocates the resources in the system while ensuring the stability of the vehicle.Finally,through a large number of numerical simulation experiments,the superiority of the proposed model and algorithm is verified.