雾计算环境下车联网优化部署与规模测算

发布时间：2018-03-29 09:30

本文选题：车联网　切入点：雾计算　出处：《大连海事大学》2017年硕士论文

【摘要】：车联网作为智能交通的一部分,为解决城市交通问题提供了发展方向。近年来,随着车辆数目和车联网数据的爆炸式增长,云计算网络带宽资源将严重不足,而且无法满足车联网移动性和低时延的特性。雾计算部署在网络边缘,具有位置感知和低时延的特性,能够更好地应用到车联网中。虽然雾计算的概念由思科在2012年提出,但是在车联网方面的研究还处于起步阶段,仅限于理论方面。本文首次将雾计算与车联网联合考虑,就其部署与规模测算问题展开研究。首先,结合雾计算和车联网技术,本文设计了雾计算环境下的车联网架构(FC-IoV)。该架构通过在网络边缘引入雾设备,使得车载终端能够快速接入网络,就近选择雾设备为其提供计算和存储服务,节约了带宽资源,提升了数据传输速率,减少了时延,具有重要的研究意义。其次,针对车联网中基础设施增多带来的网络部署成本太高的问题,本文提出了一种以最小成本为目标的网络部署方法。该方法以城市道路为应用场景,主要对车联网中的V2R(Vehicle to Roadside)进行部署,通过考虑网络覆盖范围和设备容量等因素,建立了整型线性规划(IntegerLinearProgramming,ILP)模型,并运用优化求解器Gurobi对模型进行求解,得到网络的最终部署方案。最后,针对大规模场景下,Gurobi求解线性规划模型效率过低,甚至无法求解的问题,本文提出了快速启发式算法。该算法能够得到模型的近似最优解,并且使求解时间大大减少。通过仿真实验,对数学模型和算法求解的结果进行对比,显示了算法求解的有效性和高效性,体现了算法的优越性。本文通过对雾计算环境下车联网进行优化部署,降低了网络部署成本,使得车联网部署更加经济可靠,为车联网实际部署提供了理论依据和技术支持。
[Abstract]:Car networking as part of intelligent transportation, and provides the development direction to solve the city traffic problem. In recent years, with the explosion of the number of vehicles and car networking growth data, cloud computing, network bandwidth will be a serious shortage of resources, and can not meet the car networking mobility and low delay characteristics. Deployed at the network edge computing fog with position perception and low delay characteristics, can be better applied to the car networking. Although the concept of fog calculation proposed by CISCO in 2012, but the research in the car networking is still in the initial stage, only the theory. For the first time the fog calculation and car networking combined with its deployment, research and measurement of the scale of problem first, combined with fog computing and vehicle networking technology, this paper designs the fog computing environment of car networking architecture (FC-IoV). The architecture by introducing fog devices at the network edge, which makes the vehicle terminal Terminal can quickly access the network, choose the nearest spray equipment to provide services for its computation and storage, saving bandwidth resources, enhance the data transmission rate, reduces the time delay, and has important significance. Secondly, aiming at the car networking infrastructure increased in the high cost of the problem brought by the network deployment, this paper puts forward a method to the minimum cost for the target network deployment method. The method in city road is the main application scenarios of car networking in V2R (Vehicle to Roadside) were deployed by considering the network coverage and capacity and other factors, to establish the integer linear programming (IntegerLinearProgramming, ILP) model, and used to solve the model optimization Gurobi, get the final deployment scheme of the network. Finally, according to the scene, the Gurobi linear programming model of low efficiency, even no method to solve the problem, this paper. A fast heuristic algorithm. The algorithm can obtain the approximate optimal solution of the model, and the computation time is greatly reduced. Through simulation experiments, comparing model and algorithm of mathematical results show the effectiveness and efficiency of the algorithm, shows the algorithm superiority. Based on the computing environment to get off the fog network optimization deployment, reduce the cost of network deployment, making the car networking deployment is more economical and reliable, provide a theoretical basis and technical support for the actual deployment of car networking.

【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5;U495

【相似文献】