VANET中基于代理与回归聚集的低延迟紧急数据处理算法*

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、并行的工作,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。试验结果表明,相较于其他仅针对数据传输的VANET路由算法,本算法具有更低的端到端延迟与控制信息的开销。

Agent based low latency critical data process algorithm in VANET

In vehicular ad hoc networks, the information about traffic safety is top priority with no doubt, and a low latency and high accuracy route algorithm should be designed for it, based on that, a critical data package process algorithm based on agent is proposed. Three kinds of agent are assigned for each cluster, data collection agent and data deliver agent work independently and parallel, thus the total efficiency is improved. In the procedure of data aggregation, the redundancy data are filtered, the compact degree of the data is enhanced. Experiments result show that the proposed algorithm has lower end-to-end delay and control information overhead than the other algorithms which focus on the route algorithm of VANET.