基于隐马尔可夫模型的高速公路路由算法
Highway Routing Algorithm Based on Hidden Markov Model
周德宇 1袁学松2
作者信息
- 1. 昆明理工大学 昆明 650500
- 2. 安徽机电职业技术学院 芜湖 241000
- 折叠
摘要
在VANET高速公路场景中,由于车辆高速运动导致车辆和车辆、车辆和RSU之间所构成的网络拓扑频繁改变,使得绝大多数路由协议需要及时更新自己的邻居表来指定路由.针对邻居选择错误会导致数据不断重发、传输时延高且不可靠等现象,文章提出SDN-NDHM算法解决邻居选择错误的问题.算法利用典型的GPSR算法思想,运用HMM来预测高速运动节点的下一时刻位置信息,并且利用SDN来修正预测值、集中管理和调度网络资源.通过MATLAB的仿真实验表明,该算法与经典的GPSR改进算法相比能更好判断车辆节点的加入和离开,并且拥有更好的平均邻居正确率和更高的吞吐量.
Abstract
In the VANET highway scenario,due to the high-speed motion of vehicles,the topology of networks between vehicles,vehicles and RSU changes frequently.Most routing protocols need to update their neighbor tables in time to specify routes.Wrong neighbor selection will lead to continuous data retransmission,resulting in high transmission delay and unreliability.Therefore,SDN-NDHM is proposed to solve the problem of wrong neighbor selection.Method proposed by this paper uses the typical idea of GPSR algorithm,uses HMM in VANET to predict the next time position information of high-speed moving nodes,and uses SDN controller to modify the predicted value,centrally manage and schedule network resources.The simulation results of MATLAB show that this algorithm can judge the joining and leaving of nodes better than the improved GPSR algorithm,and has better neighbor accuracy and higher throughput.
关键词
车载自组网/GPSR协议/隐马尔科夫模型/软件定义网络/平均邻居发现错误率/吞吐量Key words
VANET/GPSR/HMM/Software Define Network/Average Neighbor Discovery Error Rate/Throughput引用本文复制引用
基金项目
安徽省高等学校优秀青年人才支持计划(2019)(2019gxyq332)
安徽省高等学校自然科学重点项目(2021)(KJ2021A1521)
出版年
2024
NETL
NSTL
万方数据