Research On Vehicle-Road-Cloud Collaborative Resource Management Method For Intelligent And Connected Vehicles

In recent years,with the development of intelligent and connected vehicles(ICVs),the vehicle-road-cloud collaborative computing architecture sinks the computing resources from the cloud to the roadside network edge near the side of ICVs,which can meet the low-latency requirements of ICVs’ computing tasks and ensure the quality of service.Since the computing,storage and network resources of each roadside edge resource node are limited,there is a certain conflict in the use of limited resources for a large number of highly concurrent computing tasks in the ICVs traffic scenario.Meanwhile,the failure of any resource will affect the effectiveness of the allocation policy.Therefore,it is necessary to formulate a reasonable and reliable resource allocation strategy.At the same time,since the signal coverage of the base station at the roadside edge node is limited,it can only allocate resources and provide services for the ICVs within the range.However,the ICVs may enter the signal range of different base stations during the driving process.Continuing to access the previously allocated resources will lead to an increase in service delay or even service interruption,which cannot guarantee the continuity of the services received by the vehicles.Service migration technology can effectively solve this problem,and different migration options will correspond to different costs,so it is necessary to determine a reasonable and efficient service migration strategy to minimize migration overhead on the basis of ensuring the quality of service and user experience.In this paper,two aspects of vehicle-road-cloud collaborative resource management in the scenario of ICVs are investigated,including resource allocation and service migration.For the resource allocation problem,this paper takes the quality of service as the starting point,and introduces the model of obtaining high-reliability services through the cooperation of multiple low-cost and low-reliability resource nodes.A resource allocation optimization problem model is constructed,which takes the system revenue maximization as the optimization objective and considers resource capacity,task delay and other constraints.Then the problem is transformed into an unconstrained problem based on the penalty function method,and solved by genetic algorithm to obtain the optimal resource allocation decision scheme to achieve the best matching between tasks and resource nodes.Finally,the simulation experimental results show that the method outperforms the comparison method in terms of system revenue and task completion rate,which proves the effectiveness and practicality of the method.To address the service migration problem,this paper takes the service characteristics of the ICVs as the starting point,and divides the services into two types:immediate service and continuous service.And then,this paper provides different options such as migration,transit or rebuilding instances according to their service characteristics.A service migration optimization problem model is constructed,which minimizes system cost and considers constraints such as resource capacity and maximum connection distance.In the solution process,an algorithm is designed to determine the candidate connection nodes and cost based to the location of the ICVs.Then the problem is transformed into an unconstrained problem based on the penalty function method,and solved by genetic algorithm to obtain the service migration decision with minimum cost.Finally,the effectiveness of the proposed method is verified by simulation experiments,and compared with the existing methods,the results show that the proposed method in this paper can obtain lower system cost.