Design Of Computation Offloadingoptimization Algorithm In Vehicle Fog Network

With the continuous development of Intelligent Traffic System(ITS),communication and computing needs are ubiquitous.Vehicle Fog Computing(VFC),as a distributed and marginalized computing mode,improves the efficiency of data processing and service response by offloading tasks to fog nodes,while achieving efficient utilization of communication and computing resources.Despite the long-term development of VFC,intelligent vehicles have the problem of high latency when dealing with computationally intensive tasks.Therefore,this project focuses on computation offloading in vehicle fog networks,and designs models and algorithms to address related issues during the offloading process,ensuring reliability while achieving low latency computing task offloading.The specific research content is as follows.(1)Aiming at the problem of link interruption in offloading scenarios,the relationship between link connectivity and vehicle motion in urban road scenarios is studied,and a V2 X link duration prediction model,LDPM-TFM,is designed that does not depend on vehicle speed distribution and link duration probability model.Firstly,the parameters affecting the duration of the V2 X link are analyzed to avoid underfitting caused by too few model parameters.Secondly,the following model in the micro traffic flow model is used to predict whether the vehicle can pass the intersection in the current green light cycle and whether the vehicle node pair can pass the intersection in the same green light cycle.Then,in order to improve the accuracy of the prediction model and reduce the complexity,the average speed and time of vehicles passing through the intersection area are predicted by the speed-density relationship in the macro traffic flow model.Finally,the V2 X link duration prediction model is established and solved by using the vehicle position relationship,vehicle speed and driving trajectory,and the connection time of fog node pairs in urban road scenes is simulated by SUMO,and the accuracy of the model is verified.(2)In the single-hop offloading scenario in the vehicle fog network,a computation offloading algorithm,SHCOA-GA,is proposed for the fine-grained task offloading problem to optimize the offloading decision and sequence.Firstly,the calculation offloading process is modeled based on the two-stage production plan,and the objective function is obtained to minimize the delay.Secondly,the application of traditional genetic algorithm in offloading decision-making is introduced,and its convergence and application deficiency in solving this problem are analyzed.In view of the above problems,the traditional genetic algorithm is improved,and Johnson Rules are introduced to ensure the optimal offloading order.Finally,the average offload delay and convergence are verified by simulation.(3)In the multi-hop offloading scenario in the vehicle fog network,a local multi-hop opportunistic routing offloading algorithm,MHORO-NS,is designed to solve the coarse-grained task offloading problem.Firstly,considering the link loss caused by link interruption and signal fading,the link model is established,and the link transmission rate is obtained as the single-hop link quality index.Secondly,the forwarding area is determined according to the vehicle density of the current road section,and the candidate set is screened by designing rules.Next,the node similarity indicator is designed,and the nodes in the candidate set are assigned priority according to the index.Finally,a comparative simulation experiment is carried out on the algorithm,and the results show that MHORO-NS has better performance in terms of delay,number of hops in the offloading path and stability.