Research On Joint Communication And Computing Resources Allocation Method In Internet Of Vehicles Based On Edge Computing

With the development of Internet of Vehicles,computationally intensive and delaysensitive services continue to emerge,such as autonomous driving,virtual reality,and dynamic high-precision map navigation,which poses great challenges to vehicles with limited computing power.As a new computing mode,Mobile Edge Computing(MEC)provides computing and storage services for mobile users on the edge of the network,allowing mobile users to offload computing tasks to the edge nodes for execution,which effectively reduces the end-to-end delay of mobile service delivery.Based on the above advantages,related scholars have proposed the application of MEC in the Internet of Vehicles and the construction of Vehicular Edge Computing(VEC)to solve the problem of insufficient vehicle computing capacity,long access delay and backhaul network congestion in traditional cloud computing schemes.However,since VEC is deployed in a local network,the distributed deployment of edge nodes has relatively limited communication and computing resources.How to improve network performance by reasonable resource allocation under the premise of ensuring service quality is one of the key challenges in VEC network.Therefore,this paper studies the resource allocation problem in VEC network.The specific research contents are summarized as follows:For partial task offloading model,the paper discusses the communication and computing resources allocation in the VEC network to achieve efficient multi-vehicle collaboration to complete a computing task,and proposes a resource allocation scheme based on minimizing completion delay.Specifically,it is proved that the problem of solving the feasibility is a convex optimization problem when the completion delay is fixed,and a series of convex optimization feasibility problems are solved by the dichotomy method to obtain the global optimal solution.In particular,for cooperative vehicles with equal computing power,the power allocation with the maximum transmission rate is first obtained by the water injection algorithm,and then the closed-form solution of the optimal delay is derived based on the convex optimization theory.Simulation results show that the proposed scheme can significantly reduce task completion delay.For binary task offloading model,this paper studies the resource allocation of task offloading among multiple vehicles and a single MEC server in VEC network.In this paper,the total cost of completing the task is defined as the weighted sum of the delay and energy consumption required to complete the task.Besides,the optimization problem is established to minimize the total cost of completing all tasks,while the constraints of the communication and computing resources are jointly considered.Since the optimization problem is a mixed integer nonlinear programming problem,the traditional optimization method is difficult to solve.Therefore,this paper models the optimization problem as a Markov decision process and proposes a solution based on Q-Learning and Deep Q Network.Simulation results show that both schemes effectively reduce the total cost of completing all tasks.