Optimal Deployment And Dimensioning Of Fog Computing Supported Internet Of Vehicles

As part of Intelligent Transportation System(ITS),Internet of Vehicles(IoV)provides promising solution to the problem of modern urban traffic.With the explosive growth of vehicle numbers and IoV data in recent years,cloud computing network bandwidth resource will be critically inadequate,moreover,it fails to meet the demand of mobility and low latency of IoV.Fog computing,implemented at the edge of network,can be implied in IoV for better performance owing to its features of location awareness and low latency.Although fog computing concept has been put forward by cisco in 2012,the research of IoV is still at an early stage and limited to theory.This thesis jointly considers fog computing and IoV for the first time and study on the problem of deployment and dimensioning.Firstly,combining with fog computing and IoV,we design the IoV architecture under the circumstance of fog computing in this thesis,named as FC-IoV.Taking advantages of the fog devices on the edge of the network,vehicle terminal can quickly access network and choose the nearby fog device for computing and storage services.This architecture provides IoV with abundant service which can save the bandwidth resources,enhance the data transmission rate and reduce the latency.Secondly,the cost of network deployment will extremely high with the increasing number of infrastructure.Aiming at this problem,we propose a kind of network architecture deployment method based on minimum cost in this thesis.This method takes the urban road as the application scene,mainly about V2R in the IoV.The method establish a model named Integer Linear Programming(ILP)under the consideration of the network coverage and device capacity.Then we use gurobi solver to calculate the model and get the final deployment scheme and minimum cost.Finally,in the large-scale network scenarios,we propose the heuristic algorithm to solve the issue of high time complexity when using gurobi to solve the ILP problem.Heuristic algorithm shows reduction in computing time under the premise of getting approximate optimal solution ILP.By comparing the solution of mathematic model and algorithm,we verify the efficiency and rationality of algorithm solution,which shows the superiority of the algorithm.This thesis mainly studies the deployment of FC-IoV,which reduces the network deployment cost,and makes network deployment more economical and reliable.This study provides the theoretical basis and technical support to IoV in practice.