Performance Analysis And Modeling Of Navigation Cloud Based On The Internet Of Vehicles

With the development of automobile industry and the population of automobiles,the problem of road congestion and traffic accidents turns into an increasingly serious social issue.As one kind of key technologies in intelligent transportation system,navigation technology,however,still has not played its role in advancing traffic efficiency.In order to solve the problems above,this paper puts forword a model of navigation cloud based on the Internet of vehicles.The core technology of this navigation system is the central navigation cloud(CNC),which is a central dynamic navigation system deployed on a cloud computing platform.Compared to distributed navigation systems,CNC,as one kind of central dynamic navigation systems,has an ability to fuse real time global traffic information dynamically.And then it can simultaneously offer optimum routes to multiple vehicles in the target map,thus reducing road congestion and enhancing traffic efficiency.Nevertheless,it is quite a challenge for the navigation cloud system to offer central navigation service to all vehicles at the same time.It must be equipped with a large enough scale of virtual machines to meet the needs of computing power.In this paper,we establish a performance model of CNC using the average response time of navigation service requests as the performance metrics.Assuming that the interarrival times of navigation requests are in accordance with poisson distribution,the service times of CNC are in accordance with negative exponential distribution,the queuing rule is FCFS,the queuing space is infinite,and all the virtual machines are independent of each other,obviously this is an M/ M/ C/ ?/ m queueing system.When there are requests in queue,the response time is composed of waiting time and serving time.We focus on the queueing system when there are requests in queue in this paper and three groups simulation experiments are taken to check the validity of our performance model.Besides,in order to effectively illustrate the significance of our model,we take one set of real data validation based on traffic flow theory.The results of the simulation and real data experiments show that the proposed performance model of central navigation cloud is validated,which is instructive to build a navigation cloud.