一种面向车辆自组网的多人博弈数据融合算法  

Data Aggregation Algorithm Based on Multi-player Game Theory for VANET

作　　者：陈羽中[1,2] 翁诗宁 郭昆[1,2] 

机构地区：[1]福州大学数学与计算机科学学院,福州350108 [2]福州大学福建省网络计算与智能信息处理重点实验室,福州350108

出　　处：《小型微型计算机系统》2016年第8期1807-1811,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61103175;61300104;61300102;61571129)资助;福建省教育厅重点项目(JK2012003)资助;福建省科技创新平台项目(2009J1007)资助;福建省自然科学基金项目(2013J01232)资助;福建省科技厅产学重大项目(2014H6014)资助

摘　　要：针对车辆自组织网络中数据融合的研究,目前主要集中于大规模簇间数据融合分析,而对簇内数据融合的研究却很少涉及.为了评估簇内数据融合性能,研究了节点可信特征以及动态车辆自组网的网络结构变化特征,提出了一种基于纳什均衡的数据融合算法,用于分析簇内节点在数据融合过程中存在的竞争和合作关系,进而分析簇特征冗余度和网络结构变化程度的均衡关系.通过对算法进行扰动,可以得到效益最优传输策略.实验结果证明,最优效益解在数据融合精确度和网络结构稳定性方面具有优势.For studying data aggregation in VANET, current researches mainly focus on data aggregation in large scale environment, while the aggregation efforts on cluster level have rarely been discussed. For aggregating local data in VANET, node reliability characteristics and network structure variation characteristics are studied. In addition, we present a novel data aggregation algorithm based on the Nash equilibrium theory by studying the inherent correlation in sensor data to analyze the competition and cooperation relationship among vehicles. Furthermore, we make equilibrium analysis of cluster redundancy and cluster variation degree in VANET. Finally, we get the optional transmission strategy by interruption process. Simulations show that our algorithm has advantages in aggregation accu- racy and topology stability.

关 键 词：车辆自组织网络 数据融合 博弈论 可信模型 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]