面向车联网的地理辅助多播路由协议研究

发布时间：2018-01-20 05:02

本文关键词： 车联网位置辅助多播拥塞避免拥塞控制　出处：《西安电子科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：在当今智能交通网络中,基于地理位置信息的服务(如地域性交通信息广播、区域性广告投放等)为人们提供了许多便利,有效地提高了智能交通网络的运行效率。地理辅助多播技术(Geocast)作为支撑地理位置信息服务的核心技术,在当今智能交通网络中得到了十分广泛的应用。因此,研究面向车联网的地理辅助多播路由算法具有重要的现实意义。车联网在城市环境中具有节点密度高、移动速度快、拓扑更新频繁和网络间歇性断开等特点,而传统的基于几何距离的地理辅助多播路由算法应用在城市车联网中存在网络过载大、数据包到达率低、端到端延迟急剧增加等缺点。因此,传统的地理辅助多播路由算法难以适应城市车联网这一应用环境。本文面向城市车联网环境,提出了一种具有拥塞控制能力的地理辅助多播路由协议。根据区域数据包吞吐能力与该区域交通流量正相关这一原理,本协议分析了大量车辆节点的真实轨迹数据并完成了城市交通宏观、微观层建模,得到了宏观流量分布模型和微观节点运动模型,并以此设计了路由协议的数据包转发策略。在数据包转发策略的基础之上,本文还设计了基于宏观模型的车联网拥塞控制机制,从而有效的解决了网络拥塞问题。本文的主要工作在于:1.根据大量车辆节点的真实轨迹数据建立了宏观交通流量模型、微观节点运动模型,并据此设计了数据包转发策略;2.拥塞避免机制:在宏观层模型的基础上建立了城市交通拥塞时变模型,使得最优路径在生成时就避免通过拥塞区;3.拥塞控制机制:提出“移动熵”的概念来表征节点运动的随机性,并将“移动熵”值大的节点设置为“摆渡船”节点,利用“摆渡船”完成对全网拥塞状态的监控和拥塞信息的扩散。仿真结果表明,本文所提出的GeoCon路由协议与传统的地理辅助多播路由协议相比,在数据包到达率,网络过载率,端到端平均延时,节点缓冲区平均等待时间等性能上,都有一定的进步。GeoCon路由协议能够适用于城市大规模车联网环境。
[Abstract]:In today's intelligent transportation network, the services based on geographical location information (such as regional traffic information broadcast, regional advertising, etc.) provide a lot of convenience for people. Geo-assisted multicast technology is the core technology to support geographic location information service. Nowadays, it has been widely used in intelligent transportation network. Therefore, it is very important to study the geo-assisted multicast routing algorithm for vehicle networking, which has high node density in urban environment. The traditional geo-assisted multicast routing algorithm based on geometric distance has many advantages such as fast moving speed frequent topology updating and intermittent disconnection. However the traditional geo-assisted multicast routing algorithm based on geometric distance has high network overload and low packet arrival rate in urban vehicle network. The end-to-end delay increases rapidly. Therefore, the traditional geo-assisted multicast routing algorithm is difficult to adapt to the application environment of urban vehicle networking. A geo-assisted multicast routing protocol with congestion control capability is proposed, which is based on the principle of positive correlation between regional packet throughput and traffic flow in the region. This protocol analyzes the real track data of a large number of vehicle nodes and completes the macro-level and micro-level modeling of urban traffic. The macro-traffic distribution model and the micro-node motion model are obtained. On the basis of the packet forwarding strategy, this paper also designs the congestion control mechanism based on macroscopic model. In order to effectively solve the network congestion problem. The main work of this paper is: 1. According to the real track data of a large number of vehicle nodes to establish a macro-traffic flow model, micro-node movement model. Based on this, the packet forwarding strategy is designed. 2. Congestion avoidance mechanism: based on the macro-level model, a time-varying model of urban traffic congestion is established, so that the optimal path is generated to avoid passing through the congestion zone; 3. Congestion control mechanism: the concept of "mobile entropy" is proposed to characterize the randomness of node movement, and the node with large mobile entropy is set as "ferry" node. The "ferryboat" is used to monitor the congestion state of the whole network and spread congestion information. The simulation results show that the proposed GeoCon routing protocol is compared with the traditional geo-assisted multicast routing protocol. In the performance of packet arrival rate, network overload rate, end-to-end average delay, node buffer average wait time and so on. Both have some progress. GeoCon routing protocol can be used in urban mass car networking environment.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN929.5;U495;TN915.04

【相似文献】