VANET中基于认知代理与回归聚集的低延迟数据处理算法  

Low latency critical data process algorithm based on cognitive agent and regression gathering in VANET

作　　者：范凌云[1] 文国琴[1] 张文涛[2] 

机构地区：[1]重庆城市职业学院,重庆402160 [2]重庆大学计算机科学与技术学院,重庆400030

出　　处：《计算机应用研究》2016年第6期1871-1876,共6页Application Research of Computers

基　　金：国家自然科学基金资助项目(60772146);重庆市教委科学技术研究基金资助项目(KJ134101);重庆市科委自然科学基金计划资助项目(2010BB2399)

摘　　要：车载自组织网络(VANET)中有关安全的信息无疑具有最高优先级,应为其设计低延迟、高准确率的信息传递算法,基于此,提出一种基于认知代理的紧急数据处理算法。为每个簇分配三种代理,数据采集代理与数据传递代理独立、并行地工作,提高了整体处理的效率;紧急数据的聚集过程中,基于回归模型,过滤其中的冗余信息与重复信息,提高了数据紧凑度。实验结果表明,相较于其他仅针对数据传输的VANET路由算法,算法的端到端延迟与控制信息的开销较低。In vehicular Ad hoc networks,the information about traffic safety is top priority no doubt,and a low latency and high accuracy route algorithm should be designed for it,based on that,this paper proposed a critical data package process algorithm based on cognitive agent. Three kinds of agent were assigned for each cluster,data collection agent and data deliver agent worked independently and parallelly,thus it improved the total efficiency. In the procedure of data aggregation,it filtered the redundancy data,enhanced the compact degree of the data. Experiments result show that the proposed algorithm has lower end-toend delay and control information overhead than the other algorithms which focus on the route algorithm of VANET.

关 键 词：车载自组织网络 道路安全 认知代理 回归分析 端到端延迟 

分 类 号：TP393.1[自动化与计算机技术—计算机应用技术] TP301.6[自动化与计算机技术—计算机科学与技术]